It is well known that NMJ nuclei are located beneath the endplate and produce essential components of the synaptic apparatus. In the 众所周知，NMJ 细胞核位于终板下方，并产生突触装置的基本组成部分。在
Fig. 2 | Mechanistic insights into human MuSC aging. a, UMAP visualization of MuSC subpopulations identified from scRNA-seq. b, Tree visualization of the GO terms enriched among marker genes for every MuSC subpopulation. Top 10 clusters of GO terms defined based on semantic similarity are shown. c, Beeswarm Milo plot showing the distribution of 图 2 | 人类 MuSC 衰老的机制洞察。a，UMAP 可视化 MuSC 单细胞 RNA 测序鉴定出的亚群。b，GO 术语的树状可视化显示出每个 MuSC 亚群标记基因富集的 GO 术语。根据语义相似性定义的前 10 个 GO 术语群集被展示。c，蜜蜂图 Mlio 绘制显示了对不同批次和一次批次间 MuSC 对数变化的分布情况。 in cell abundance with age across neighborhoods of MuSC subtypes; significantly differentially abundant neighborhoods are colored. d, Ribosome biogenesis enrichment score of MuSC subpopulations in young (five donors) versus aged (seven donors) individuals. value: two-tailed Mann-Whitney-Wilcoxon test. . e, Dot plot of ribosome biogenesis and RNA polymerase I complex genes in MuSC subpopulations. Dot size represents the proportion of cells expressing the gene in aged group, color represents in young versus aged. Significantly upregulated and downregulated genes were defined using the direction of (FC), the proportion of cells and LTSR (significance value, ranging from 0 to 1 , where 1 is confident estimate). See Source Data. , Expression of senescence-associated and ribosome assembly (h) genes in cultured human primary myoblasts (f) by both qPCR (three biological repeats per group) (g,h) and western blot . Three independent experiments were performed for western blot with similar results. value: unpaired two-tailed -test. ; . Illustration in was created with BioRender.com. , qPCR (three donors for both panels) of genes in FACS-sorted MuSC subpopulations. value in k: one-way ANOVA test; value in I: unpaired two-tailed -test. ; , Violin plots of CCL2, TNFAIP 3 and NFKBIZ in ICA MuSCs from scRNA-seq data. P value: unpaired two-tailed -test. n, qPCR of CHUK, NFKBIZ and CCL2 in FACS-sorted ICA MuSCs (three young versus three or four aged donors). value: unpaired two-tailed -test. . All data presented in and are mean s.e.m. with individual data points shown. The exact 值显示在源数据中。在本研究中，我们发现了一个以前未报道的 NMJ 辅助群体，该群体表达了与突触形成相关的标记基因，这些标记基因与 NMJ 细胞核标记不同（图 3j 和补充 表 4）。如GO分析所示，NMJ辅助物在“突触组织”和“轴突发育”方面富集，可能促进突触形成（扩展数据图4g）。在 NMJ 配件中 a
k Main MuSC - ICA MUSC - MUSC k 主要 MuSC - ICA MUSC - MUSC
m
n
n Young ICA Aged ICA n 年轻 ICA 年老 ICA
marker genes were GRIA2, encoding the key subunit of ionotropic glutamate receptor; EFNA5, encoding an essential ligand involved in axon guidance to the myotube during limb development ; and SORBS2, encoding an adapter protein involved in AChR cluster formation in mouse . pySCENIC transcription factor activity inference further confirmed the distinction between NMJ and NMJ accessory, highlighting that the ETV4 and ETV5 (ref.55) Transcription factors known to induce synapse formation were almost inactive in NMJ accessory (Extended Data Fig. 4i,j). Interestingly, NMJ accessory increased with age, both in our dataset and in publicly available human quadriceps nuclei data (Fig. 3e and Extended Data Fig. 5b,c). By co-staining CHRNE (NMJ nuclei) and GRIA2 (NMJ accessory) with RNAscope probes, we observed groups of nuclei with co-localization of these transcripts in aged donor tissue sections, which were rare in young ones (Fig. 3k). NMJ accessory also expressed more slow-twitch rather than fast-twitch MF markers (Extended Data Fig. 5d), which was also evident from RNAscope staining (Fig. 3k and Extended Data Fig. 5e; NMJ accessory in fast-twitch MFs was rarely present-data not shown). Using immunofluorescence, we identified SORBS2 nuclei directly beneath the postsynaptic endplate (as marked by -bungarotoxin staining) at the NMJ (Fig. 3l). 标记基因是 GRIA2，编码离子型谷氨酸受体的关键亚基；EFNA5，编码在肢体发育期间对肌管进行轴突引导的重要配体 ；以及 SORBS2，编码在小鼠 AChR 簇形成中涉及的适配器蛋白。pySCENIC 转录因子活性推断进一步确认了 NMJ 和 NMJ 辅助之间的区别，突出显示 ETV4 和 ETV5（参考文献 55）已知能诱导突触形成的转录因子在 NMJ 辅助中几乎不活跃（扩展数据图 4i，j）。有趣的是，NMJ 辅助随着年龄增长而增加，无论是在我们的数据集中还是在公开可用的人类股四头肌细胞核数据中（图 3e 和扩展数据图 5b，c）。通过共染色 CHRNE（NMJ 细胞核）和 GRIA2（NMJ 辅助）与 RNAscope 探针，我们观察到在年长供体组织切片中这些转录本的共定位的细胞核群，而在年轻供体中很少见（图 3k）。NMJ 辅助还表达更多的慢肌而不是快肌 MF 标记物（扩展数据图 5d），这也可以从 RNAscope 染色中看出（图 3k 和扩展数据图）。 5e; NMJ 的快收缩 MFs 中的附件很少出现-未显示数据）。使用免疫荧光，我们在 NMJ（图 3l）下直接确定了 SORBS2 核，位于肌肉运动柄过后的区域（由 -曼巴毒蛋白标记）。
To better understand the functional importance of the NMJ accessory population, we cultured human myotubes in vitro, where they are able to mimic different stages of AChR cluster formation even without axonal stimulation . We then used this myotube culture to perform knockdown of two NMJ accessory markers, EFNA5 and SORBS2. Both knockdowns led to a marked decrease of AChR clusters at all stages of aggregate assembly (Fig. and Extended Data Fig. ), whereas overexpression of EFNA5 on its own was sufficient to promote AChR cluster formation (Fig. 3n). Overall, this suggests that NMJ accessory increases with age to support NMJ re-innervation in aged MF (Fig. 30). We also observed a subset of denervation signature genes in NMJ accessory nuclei; however, they were not exclusive to NMJ accessory (Extended Data Fig. 5g). 为了更好地理解 NMJ 附件群体的功能重要性，我们在体外培养了人类肌肉束，它能在没有轴突刺激的情况下模拟不同阿 ChR 聚集形成阶段 。然后，我们使用这种肌束培养进行了两种 NMJ 附件标记物 EFNA5 和 SORBS2 的基因沉默。这两种沉默导致了所有聚集形成阶段的 AChR 集群显著减少（图 和扩展数据图 ），而单独过表达 EFNA5 就足以促进 AChR 集群的形成（图 3n）。总的来说，这表明 NMJ 附件在年龄增长时增加以支持老化 MF 的 NMJ 再神经化（图 30）。我们还观察到 NMJ 附件中一部分解神经标记基因 ; 但是，它们并不是 NMJ 附件的专有（扩展数据图 5g）。
Mechanisms countering fast-twitch MF loss in aging 缓解快收缩 MF 老化的机制
MFs have differential susceptibility to aging depending on their type: fast-twitch MFs are more vulnerable than slow-twitch ones . Here, we combined information about MFs and nuclei comprising them to compare their dynamics with age (Fig. 4a). For the MF, we used immunofluorescence staining of myosin heavy chain proteins (Fig. 4b) followed by automatic image analysis (Extended Data Fig. 6a,b and Methods) to distinguish slow-twitch (type I, MYH7'), fast-twitch (type IIA, MYH2 , and type IIX, ) and hybrid (type IIA-IIX and ) MFs. We then scored the expression of the same genes in the nuclei and were able to separate three pure nuclei types, with exclusive expression of or , as well as four hybrid types, , and (Fig. and Supplementary Table 5). 肌纤维根据其类型对衰老的敏感性有差异：快收缩型肌纤维比慢收缩型更容易受损。在这里，我们结合了有关肌纤维和构成它们的细胞核的信息，比较它们随着年龄的动态变化。对于肌纤维，我们使用肌球蛋白重链蛋白的免疫荧光染色，然后进行自动图像分析，以区分慢收缩型（I 型，MYH7'）、快收缩型（IIA 型，MYH2）和混合型（IIA-IIX 和 IIX 型）肌纤维。然后我们评分核中相同基因的表达，并能够区分三种纯核类型，具有独特的表达，以及四种混合类型。
As expected, fast-twitch MFs displayed reduced heterogeneity in aged compared to young intercostal muscles as determined by immunofluorescence (Fig. 4b). Both slow-twitch and fast-twitch MFs decreased in cross-sectional area, with type IIA MFs exhibiting the greatest reduction in size (Extended Data Fig. 6c,d). Detailed MF typing revealed that type IIX MFs almost completely disappeared in aged individuals; type IIA and hybrid IIA-IIX did not significantly change; and type I increased in proportion (Fig. 4c, Extended Data Fig. 6e and Supplementary Table 5). 预期的是，快收缩MFs在老年人肋间肌中显示出较年轻时减少的异质性，这是通过免疫荧光确定的（图4b）。慢收缩和快收缩MFs的横截面积均减少，其中IIA型MFs的尺寸减小最为显著（扩展数据图6c，d）。详细的MF分类显示，IIX型MFs在老年人中几乎完全消失；IIA型和混合型IIA-IIX型没有显著变化；而I型的比例增加（图4c，扩展数据图6e和附表5）。
At the nuclei level, nuclei (type IIX) had a tendency to increase (Fig. 4e and Extended Data Fig. 6f), even though the type IIX MFs that are expected to contain these nuclei almost disappeared with age. This may point to initiation or increase of expression in other MF types (such as type I ) and type II (MYH2 ) and acquisition of an early hybrid phenotype. By combining staining of MYH7 protein together with and RNA (Extended Data Fig. ), we observed a number of nuclei in aged slow-twitch MFs expressing fast-type mRNAs and (Fig. 4 f and Extended Data Fig. 7a). These fast type mRNAs were even found in the cytoplasm (Fig. and Extended Data Fig. 7a), a sign of hybrid MF, which was not observed in young skeletal muscle. Notably, slow-type nuclei were located only in slow-twitch MFs (Extended Data Fig. 6h). Together, this points to a 'slow-to-fast' myonuclear shift in aged skeletal muscle, which can be an intermediate stage toward a hybrid MF phenotype. Interestingly, changes within slow-twitch MF were accompanied by an increase in glycolytic enzyme expression in the cytoplasm (as estimated using MF fragments) and a decrease in the nuclear expression of PPARGC1A, a key mitochondrial biogenesis gene (Extended Data Fig. 7b). This agrees with previous proteomics data but is unexpected given the oxidative nature of slow-twitch MF metabolism. We also observed an increase in RNA inside the nuclei and cytoplasm of (Fig. and Extended Data Fig. 7c), pointing to an additional 'fast IIA-to-fast IIX' nuclear shift. The appearance of such hybrid states may be a response to a loss of fast-twitch MF in aging. 在细胞核水平， 细胞核（IIX型）有增加的趋势（图4e和扩展数据图6f），尽管预期包含这些细胞核的IIX型肌纤维随着年龄增长几乎消失。这可能指向其他肌纤维类型（如I型 ）和II型（MYH2 ）中 表达的启动或增加，以及早期混合表型的获得。通过结合MYH7蛋白的染色以及 和 RNA（扩展数据图 ），我们观察到在老年慢肌 肌纤维中表达快速型mRNA 和 的许多细胞核（图4f和扩展数据图7a）。这些快速型mRNA甚至在细胞质中发现（图 和扩展数据图7a），这是混合肌纤维的迹象，而在年轻骨骼肌中没有观察到。值得注意的是，慢型 细胞核仅位于慢肌肌纤维中（扩展数据图6h）。总的来说，这表明在老年骨骼肌中存在“慢到快”的肌核转变，这可能是通向混合肌纤维表型的中间阶段。 有趣的是，慢肌纤维内的变化伴随着细胞质中糖酵解酶表达的增加（使用MF碎片估计）和线粒体生物发生关键基因PPARGC1A的核表达减少（扩展数据图7b）。这与先前的蛋白质组学数据一致，但考虑到慢肌纤维的氧化代谢特性，这种情况是意外的。我们还观察到核内和细胞质中的RNA增加（图15和扩展数据图7c），指向额外的“快IIA到快IIX”的核转变。这种混合状态的出现可能是对老化中快肌纤维丢失的一种响应。
MYH8 myocyte-mediated regeneration may represent another mechanism countering fast-twitch MF loss. In particular, myocytes were an intermediate state in the trajectory from MuSC to MF, predominantly connecting to fast-twitch MF (Extended Data Fig. 7d,e). This is consistent with fetal MYH8 being described as a marker of muscle regeneration . MYH8 myocytes also expressed a much higher level of fast-twitch rather than slow-twitch MF structural genes and increased in proportion with age (Fig. ,i and Supplementary Table 5), as recently reported . Immunofluorescence staining confirmed that MYH8 expression significantly increased with age (Fig. 4j) and predominantly occurred in fast-twitch MFs (Fig. 4k). Moreover, nearly of fast-twitch MFs had centralized nuclei (Fig. 41), a sign of regeneration. This is similar to a process identified in Duchenne muscular dystrophy where atrophic fast-twitch MFs regenerate by de novo expression of embryonic myosin heavy chain 3 (MYH3) . MYH8肌细胞介导的再生可能代表了另一种对抗快收缩MF损失的机制。特别是，MYH8肌细胞是从MuSC到MF的轨迹中的中间状态，主要连接到快收缩MF。这与胎儿期的MYH8被描述为肌肉再生标记是一致的。MYH8肌细胞也表达了比慢收缩MF结构基因更高水平的快收缩MF，并且随着年龄增加而比例增加。免疫荧光染色证实，MYH8表达随着年龄显著增加，并主要发生在快收缩MF中。此外，近一半的快收缩MF具有核心核（中心核），这是再生的迹象。这类似于杜兴氏肌肉萎缩症中鉴定出的过程，其中萎缩的快收缩MF通过新表达胚胎肌球蛋白重链3（MYH3）进行再生。
In summary, our data suggest two putative mechanisms countering fast-twitch MF loss: a 'slow-to-fast' myonuclei shift and an increase in fast-twitch MF regeneration via MYH8 myocytes (Fig. ). 总的来说，我们的数据表明有两种可能的机制来对抗快速收缩肌纤维的流失：一种是“慢到快”的肌核转变，另一种是通过MYH8肌细胞增加快速收缩肌纤维再生（图1）。
staining of NMJ accessory (in yellow circle) on intercostal muscle FFPE sections (two young versus three aged donors). Scale bar, , Immunofluorescence staining of -bungarotoxin ( -BTX) and SORBS2 on teased human intercostal muscles (one young versus two aged donors). Scale bar, . , Immunofluorescence of AChRs on cultured human myotubes after siRNA knockdown of EFNA5 ( , left, 13 si-EFNA5 versus eight Scramble control fields) and overexpression of EFNA5 (n, left, eight OE-EFNA5 versus 11 control fields). AChRs on different stages of cluster formation (dotted to plaque to branched) were quantified by Fiji. value: unpaired two-tailed -test. Scale bar, . Both experiments in and were performed twice with similar results. , Schematic diagram showing NMJ accessory-mediated pro-survival mechanism against NMJ aging. All data presented in bar plots are mean s.e.m. with individual data points shown. ; . The exact values are shown in the Source Data. NMJ 附件（黄色圆圈内）在肋间肌 FFPE 切片上的染色（两个年轻对比三个年长的供体）。比例尺， ，人类肋间肌上的拉开式免疫荧光染色（一个年轻对比两个年长的供体）-β-蝮蛇毒素（ -BTX）和 SORBS2。比例尺， 。 ，EFNA5 的 siRNA 敲除后培养的人类肌管上的 AChRs 的免疫荧光（ ，左，13 个 si-EFNA5 对比八个 Scramble 对照区域）和 EFNA5 的过表达（n，左，八个 OE-EFNA5 对比 11 个对照区域）。 Fiji 对不同阶段的簇形成的 AChRs 进行了定量。 值：未配对的双尾 -test。比例尺， 。 和 中的两次实验结果相似。 ，示意图显示 NMJ 附件介导的 NMJ 抗衰老机制。所有条形图中呈现的数据 均为均值 s.e.m.，显示个体数据点。 ； 。源数据中显示确切的 值。
The human skeletal muscle microenvironment in aging 老化中的人类骨骼肌微环境
To further investigate the aging muscle microenvironment, we separated and finely annotated major cell populations, including immune cells, fibroblasts, Schwann cells, endothelial cells and SMCs (Supplementary Note 4, Extended Data Figs. 8 and 9a and Supplementary Table 6), and used Milo to study changes in cellular neighborhoods with age. 为进一步研究老化肌肉微环境，我们分离并精细注释了主要细胞群，包括免疫细胞、成纤维细胞、施万细胞、内皮细胞和 SMC（附注 4、扩展数据图 8 和 9a，以及补充表 6），并使用 Milo 来研究细胞邻域随着年龄的变化。
We found that B cells, 我们发现 B 细胞， cells and NK cells accumulated with age, whereas M2 macrophages decreased (Fig. 5a). Immunofluorescence confirmed that and cells increased with age across individuals (Fig. 5d-f and Extended Data Fig. 9b). In parallel, we also performed 15-plex immunofluorescence staining on young and aged muscle sections using the RareCyte commercial panel of ArgoFluor-conjugated antibodies to visualize subtypes of immune cells, vascular cells and a 和 NK 细胞随着年龄积累，而 M2 巨噬细胞减少（图 5a）。免疫荧光证实了 和 细胞在个体间随年龄增加（图 5d-f 和扩展数据图 9b）。与此同时，我们还使用 RareCyte 商业 ArgoFluor 偶联抗体的 15-plex 免疫荧光染色在年轻和老化的肌肉切片上进行，以可视化免疫细胞、血管细胞和
b
e
f
c
d
i
Fraction of cells in group (%) - 000 20406080100 Mean expression in group 组中细胞的分数 (%) - 000 20406080100 组中的平均表达
with age. Scale bar, . , Violin plot showing specific expression of fast-twitch MF structural genes in myocytes. , Bar plot showing proportion of myocytes, relative to the total MF cells in scRNA-seq (five young versus seven aged donors). value: unpaired two-tailed -test. j, Immunofluorescence (left) and area quantification (right) of MYH8 on teased human intercostal muscles (six young versus six aged donors). value: unpaired two-tailed -test. . Scale bar, , Co-immunofluorescence of MYH7, MYH2 and MYH8 on skeletal muscle cross-sections with lower (k) and higher (I) magnification. Bar plots illustrate proportion of MFs with centralized nuclei relative to all MYH2 MFs (five young versus four aged donors). Arrows point to MYH MFs. Scale bar in . Scale bar in . value: unpaired two-tailed -test. , Diagram illustrating different putative mechanisms of MF aging. All data in c,e,i,j and I are mean s.e.m. with individual data points shown. The exact values are shown in the Source Data. 随着年龄增长。比例尺，。小提琴图显示快速肌纤维结构基因在肌细胞中的特定表达。条形图显示相对于scRNA-seq中总MF细胞的比例（五个年轻对比七个老年供体）。值：未配对的双尾t检验。j，MYH8在拉开的人肋间肌上的免疫荧光（左）和面积定量（右）。值：未配对的双尾t检验。。比例尺，，MYH7、MYH2和MYH8在骨骼肌横截面上的共免疫荧光，放大（k）和放大（l）。条形图说明相对于所有MYH2 MF的中心化核的MF比例（五个年轻对比四个老年供体）。箭头指向MYH MF。比例尺在。值：未配对的双尾t检验。，图示MF老化的不同假设机制。c、e、i、j和I中的所有数据均为均值s.e.m.，显示各个数据点。 精确的 值显示在源数据中。
d
e
Young 年轻
Aged 年长
I
Young larger vasculature Aged larger vasculature 年轻时较大的血管 结老时较大的血管
Fig. 5 | The human skeletal muscle microenvironment in aging. a-c, Beeswarm Milo plots showing the distribution of -transformed fold change in cell abundance with age across neighborhoods of cells in the microenvironment. AdvFB, adventitial fibroblasts; EnFB, endoneurial fibroblasts; PnFB, perineural fibroblasts.d, Co-immunofluorescence of CD3 and laminin on fresh-frozen sections. Bar plot showing number of cells per field (four young versus six aged donors). Scale bar, . : unpaired two-tailed -test. . e, Subset of four markers from a 15-plex RareCyte protein panel indicating proximity between cells and vessels (two young versus two aged donors). Scale bar, . , Co-immunofluorescence of NKG7 and laminin on fresh-frozen sections. Bar plots show the number of cells per (three young versus four aged donors). Scale bar, . : unpaired two-tailed t-test. . , RNAscope (g) and bar plot (h) showing number of LYVE1 cells per field on FFPE sections (two young versus two aged donors). Scale bar, . : one-way 图5 | 老化中的人类骨骼肌微环境。a-c，蜂群Milo图显示细胞丰度的 -转换倍数随年龄在微环境中细胞邻域的分布。AdvFB，外膜成纤维细胞；EnFB，内神经鞘成纤维细胞；PnFB，周神经鞘成纤维细胞。d，新鲜冰冻切片上CD3和层粘蛋白的共免疫荧光。条形图显示每个视野中的 细胞数量（四个年轻对比六个老年供体）。比例尺， 。 ：未配对的双尾 -检验。 。e，来自15-plex RareCyte蛋白质面板的四个标记子集，指示 细胞与 血管之间的接近程度（两个年轻对比两个老年供体）。比例尺， 。 ，新鲜冰冻切片上NKG7和层粘蛋白的共免疫荧光。条形图显示每个 细胞数量（三个年轻对比四个老年供体）。比例尺， 。 ：未配对的双尾t检验。 。 ，RNAscope（g）和条形图（h）显示FFPE切片上每个视野中LYVE1细胞的数量（两个年轻对比两个老年供体）。比例尺， 。 : 单向
h
ANOVA test. , Co-immunofluorescence of ACTA2 and laminin on fresh-frozen sections (i) and bar plot illustrating proportion of MFs with 0 (none), or more cells surrounding them (j) (three young versus three aged donors). Scale bar, . : unpaired two-tailed -test. . , Dot plot illustrating aging changes of chemokine and interleukin genes. Significant genes were defined based on direction of change, proportion of cells and LTSR (significance) , Co-immunofluorescence of ACTA2 and CCL2 on FFPE sections. Bar plot shows percentage of cells (two young versus two aged donors). Scale bar, . value: one-way ANOVA test. . , CellPhoneDB analysis of cell-cell interactions mediated via CCL 2 produced by various cell types in the microenvironment. Emitter (ligand) cells: leftmost; receiver (receptor) cells: rightmost. FB, fibroblast. All data in and are mean s.e.m. with individual data points shown. See the Source Data for exact values. ANOVA 测试。 ，ACTA2 和层粘连蛋白在新鲜冰冻切片上的共免疫荧光（i）和条形图显示周围有 0（无）、 或更多 细胞的 MFs 比例（j）（三个年轻对比三个年老的供体）。比例尺， 。 ：未配对的双尾 -检验。 。 ，点图显示趋化因子和白细胞介素基因的老化变化。基因的显著性基于变化方向、细胞比例 和 LTSR（显著性） ，ACTA2 和 CCL2 在 FFPE 切片上的共免疫荧光。条形图显示百分比的 细胞（两个年轻对比两个年老的供体）。比例尺， 。 值：单向 ANOVA 测试。 。 ，CellPhoneDB 分析细胞间通过各种细胞类型产生的 CCL 2 介导的相互作用。发射细胞（配体）：最左侧；接收细胞（受体）：最右侧。FB，成纤维细胞。所有数据在 和 中均为均值 标准误差，显示各个数据点。请参阅源数据获取确切的 值。
fibroblasts (Extended Data Fig. 9c). This confirmed an increase in both 成纤维细胞（扩展数据图 9c）。这证实了老化骨骼肌中两者增加 cells and cells in the aged skeletal muscle, which were often concentrated around blood vessels (Fig. 5e). At the same time, anti-inflammatory M2-like LYVE1 macrophages detected with LYVE1 RNAscope probes decreased with age (Fig. ), consistent with a recent mouse study showing a decrease in anti-inflammatory signals in the aged muscle. 细胞和 细胞，通常集中在血管周围（图 5e）。同时，使用 LYVE1 RNAscope 探针检测到的抗炎 M2 样 LYVE1 巨噬细胞随着年龄的增长而减少（图 ），与最近的一项小鼠研究 一致，显示老化肌肉中抗炎信号的减少。
Among the stromal populations, adventitial fibroblasts and perineural fibroblasts increased most strongly with age (Fig. 5b). At the same time, myelinating and non-myelinating cell, ParFB and InterFB, tenocyte and endoneurial fibroblast populations decreased (Fig. 5b). Loss of both types of Schwann cells is detrimental and can contribute to axonal and NMJ deterioration . Finally, in the vasculature, VenEC (VenEC_CCL2 ) showed the largest increase with age, followed by CCL26 pericytes (Pericyte_CCL26 ). Most of the SMC populations, pericytes and some arterial and capillary endothelial cells tended to decrease with age (Fig. 5c). This was confirmed by a reduction of capillaries surrounding MFs using both immunofluorescence staining of ACTA2 (Fig. 5i,j) and RareCyte staining of CD31 (Extended Data Fig. 9d). Although we confirm a general decline in tissue vascularization as previously described , our findings suggest that SMCs and pericytes are the most affected as opposed to endothelial subsets. 在基质细胞群中，随着年龄增长，外膜成纤维细胞和周神经成纤维细胞增加最为显著（图5b）。与此同时，髓鞘化和非髓鞘化细胞、ParFB和InterFB、腱细胞和内神经成纤维细胞群减少（图5b）。两种类型的施万细胞的丧失是有害的，可能导致轴突和NMJ的恶化。最后，在血管系统中，VenEC（VenEC_CCL2）随年龄增加最为显著，其次是CCL26（Pericyte_CCL26）的周细胞。大多数SMC群、周细胞和一些动脉和毛细血管内皮细胞倾向于随年龄减少（图5c）。通过ACTA2的免疫荧光染色和CD31的RareCyte染色证实了周围微纤维的毛细血管减少（图5i，j和扩展数据图9d）。尽管我们确认了组织血管化的普遍下降，与之前描述的情况相反，我们的发现表明SMCs和周细胞受到的影响最大，而不是内皮细胞亚群。
To further clarify the influence of aging muscle microenvironment, we performed aging DEG analysis on cell types comprising the muscle microenvironment (Supplementary Table 3), paying particular attention to chemokines and cytokines. Strikingly, several cell populations in the aging muscle either significantly upregulated or had a tendency toward increased expression of CCL2 (Fig. 5k).CCL2 is the major pro-inflammatory and monocyte/macrophage-attracting cytokine known to be activated in muscle injury . Indeed, we found that CCL2 expression was greatly increased in the small capillaries and in the SMCs of the large blood vessels with age (Fig. 51). CellPhoneDB analysis predicted that several microenvironment populations (fibroblasts, MuSCs, arterial endothelial cells, SMCs and pericytes) could produce CCL2 and attract monocytes, CDC2 and plasma cells to the aging muscle via both CCR2 and CCR10 receptors (Fig. ). In addition to the pan-microenvironment upregulation of , we also noted an increase in CCL3, CCL4 and CXCL8, which was restricted to immune cells (Fig. 5k). For instance, and , increasingly produced with age by monocytes, macrophages and NK cells, were predicted to attract a range of immune cells, including monocytes, macrophages, different types of DC, plasma and B cells as well as eosinophils and neutrophils (Extended Data Fig. 9e). In contrast, CXCL8 was predicted to exclusively attract neutrophils (Extended Data Fig. 9e). Finally, we noted an increase in the expression of the pro-inflammatory cytokine IL6 in several microenvironment cell types, coupled with a decrease in the anti-inflammatory IL1O in immune cells (Fig. 5k). 为了进一步阐明老化肌肉微环境的影响，我们对构成肌肉微环境的细胞类型进行了老化DEG分析（补充表3），特别关注趋化因子和细胞因子。值得注意的是，老化肌肉中的几种细胞群体要么显著上调，要么具有向CCL2的表达增加的趋势（图5k）。CCL2是已知在肌肉损伤中被激活的主要促炎因子和单核细胞/巨噬细胞引诱因子。 的确，我们发现CCL2表达在小毛细血管和老年大血管的平滑肌细胞中大幅增加（图51）。CellPhoneDB 分析预测几种微环境群体（成纤维细胞、MuSCs、动脉内皮细胞、平滑肌细胞和周细胞）可以产生CCL2，并通过CCR2和CCR10受体吸引单核细胞、CDC2和浆细胞到老化肌肉（图 ）。除了全面上调CCL2外，我们还注意到CCL3、CCL4和CXCL8的增加，这些趋化因子限制于免疫细胞（图5k）。 例如，通过年龄逐渐由单核细胞、巨噬细胞和 NK 细胞产生的 和 ，被预测会吸引一系列免疫细胞，包括单核细胞、巨噬细胞、不同类型的 DC、浆细胞和 B 细胞，以及嗜酸性粒细胞和中性粒细胞（扩展数据图 9e）。相比之下，CXCL8 被预测仅吸引中性粒细胞（扩展数据图 9e）。最后，我们注意到多种微环境细胞类型中促炎细胞因子 IL6 的表达增加，同时伴随着免疫细胞中抗炎因子 IL1O 的减少（图 5k）。
In summary, we observed an inflammatory state of aged muscle, exemplified by immune cell infiltration and increased production of pro-inflammatory cytokines. The increase in cytokine expression across multiple stromal cell types with age may be partially responsible for immune cell invasion. 总的来说，我们观察到老化肌肉呈现炎症状态，表现为免疫细胞浸润和促炎细胞因子的增加产生。随着年龄的增长，多种基质细胞类型中细胞因子表达的增加可能部分导致免疫细胞入侵。
Common skeletal muscle aging changes in human and mouse 人类和小鼠中骨骼肌老化的常见变化
To identify the common aging hallmarks across different species and muscle types, we integrated our in-house generated human and mouse skeletal muscle scRNA-seq data with two previously published human 为了识别不同物种和肌肉类型之间的共同衰老标志，我们将我们内部生成的人类和小鼠骨骼肌单细胞 RNA 测序数据与两个先前发表的人类和四个小鼠健康、非干扰的单细胞资源进行整合 and four mouse healthy, non-perturbed single-cell resources 。整合数据集包括 346,296 个细胞，包含来自 33 名人类供体（19-84 岁）和 31 只小鼠（1-30 个月）的样本，并涵盖了超过 13 种不同类型的肌肉（图 6a，扩展数据图 10a-d 和补充表 7）。. The integrated dataset comprised 346,296 cells, contained samples from 33 human donors (19-84 years old) and 31 mice (1-30 months old) and covered over 13 different types of muscles (Fig. 6a, Extended Data Fig. 10a-d and Supplementary Table 7).
Due to the large difference in cell type abundance between human and mouse datasets, we focused on investigating cell-type-specific aging DEGs in both species for common signals (Methods). We found on average a larger number of aging DEGs in human compared to mouse skeletal muscle (Fig. 6b and Supplementary Table 7). Most cell types in both species displayed more downregulated genes than upregulated ones. The consistency in aging DEGs between human and mouse ranged from to according to Jaccard similarity index (Fig. and Methods). This range includes overlap in aging DEGs between bulk human and mouse skeletal muscle aging datasets (calculated based on the data from Zhuang et al. ). Downregulated aging DEGs were more consistent between species than the upregulated ones (Fig. 6c), emphasizing that downregulation is a conserved aging mechanism across species, as noted previously . At the same time, immune cells tended to have larger consistency in the upregulated genes compared to other cell types (Fig. 6c; immune cells are highlighted in red). This may reflect activation of gene expression programs that contribute to age-related inflammation. 由于人类和小鼠数据集中细胞类型丰度的巨大差异，我们专注于调查两种物种中特定于细胞类型的衰老DEGs的共同信号（方法）。我们发现，与小鼠骨骼肌相比，人类平均具有更多的衰老DEGs（图6b和附表7）。两种物种中的大多数细胞类型显示出更多的下调基因而不是上调基因。根据Jaccard相似性指数，人类和小鼠之间的衰老DEGs的一致性范围从 到 （图 和方法）。这个范围包括人类和小鼠骨骼肌整体衰老数据集之间的 重叠（根据庄等人的数据计算得出 ）。下调的衰老DEGs在物种之间的一致性比上调的更高（图6c），强调了下调是跨物种保守的衰老机制，正如先前所指出的 。与此同时，免疫细胞倾向于在上调基因方面比其他细胞类型具有更大的一致性（图6c；免疫细胞以红色突出显示）。 这可能反映了激活导致与年龄相关的炎症的基因表达程序。
Next, we performed the Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway enrichment analysis on aging DEGs from both species to explore pathways consistently upregulated or downregulated with age in the same cell types (Fig. 6d). Immune-related pathways, such as phagosome synthesis, antigen processing and presentation, complement cascade and coagulation, were consistently enriched in several cell types of both species. This further emphasizes increased inflammation during muscle aging. In contrast, pathways involved in muscle growth, structural integrity and innervation, such as PI3K-Akt pathway, focal adhesion and axon guidance, were depleted with age (Fig. 6d and Supplementary Table 7). 接下来，我们对两种物种的衰老 DEGs 进行了京都基因和基因组百科全书（KEGG）通路富集分析，以探索在相同细胞类型中随着年龄上调或下调的通路（图 6d）。 免疫相关通路，如吞噬体合成，抗原处理和呈递，补体级联和凝血，在两种物种的几种细胞类型中一致富集。 这进一步强调了肌肉衰老期间炎症的增加。 相反，与肌肉生长，结构完整性和神经支配有关的通路，如 PI3K-Akt 通路，焦点粘附和轴突引导，随年龄减少（图 6d 和补充表 7）。
We next focused on pro-inflammatory genes (Fig. 6e). Of note, the increase in the immune-attracting cytokine CCL2 in the human skeletal muscle was not replicated in the mouse single-cell data in the resting aged state, which differs from the senescent state induced upon injury (Fig. 6e). At the same time, other pro-inflammatory molecules, such as the chemokines CXCL3, CCL17 and interleukin IL1B, together with inflammasome NLRP3, showed a general increase in monocytes, macrophages, and B cells in both species (Fig. 6e and Supplementary Table 7), further providing evidence for an increase in inflammation with age. Interestingly, we observed an increase in pro-inflammatory IL6 within multiple vascular (SMCs, pericytes and a trend in arterial endothelial cells) and stromal cells (tenocytes and fibroblasts) of both species (Fig. 6e). The increased production of IL-6 by fibroblasts was reported to inhibit IGF-1 (ref. 71), an important pro-growth factor that facilitates muscle regeneration. Indeed, IGF-1 expression was significantly decreased in fibroblasts in both species (Fig. 6e), indicating a reduction of muscle repair in aging. 接下来，我们专注于促炎基因（图6e）。值得注意的是，在人类骨骼肌中免疫吸引因子CCL2的增加在老年状态下的小鼠单细胞数据中没有得到复制，这与受伤后诱发的衰老状态不同（图6e）。与此同时，其他促炎分子，如趋化因子CXCL3、CCL17和白细胞介素IL1B，以及炎症小体NLRP3，在两种物种的单核细胞、巨噬细胞和B细胞中呈普遍增加（图6e和附表7），进一步证明了随着年龄的增长，炎症程度的增加。有趣的是，我们观察到在两种物种的多种血管细胞（SMCs、包细胞和动脉内皮细胞中趋势）和基质细胞（肌腱细胞和成纤维细胞）中促炎性IL6的增加（图6e）。据报道，成纤维细胞产生的IL-6的增加会抑制IGF-1（参考文献71），IGF-1是一种促进肌肉再生的重要生长因子。事实上，两种物种中成纤维细胞中的IGF-1表达显著减少（图6e），表明肌肉修复在衰老中受到了减少。
Together, the above pathway and gene-level analyses suggest that an increase in inflammation and decrease of pro-growth, repair and muscle innervation are common features of skeletal muscle aging of both species. However, inflammation can be orchestrated via different cell types or cytokines, emphasizing the need for studying the muscle aging process in humans. 以上途径和基因水平分析表明，炎症的增加和促生长、修复和肌肉内神经支配的减少是两种物种骨骼肌老化的共同特征。然而，炎症可以通过不同的细胞类型或细胞因子来协调，强调有必要研究人类肌肉老化过程。
Discussion 讨论
Although single-cell genomics studies have provided many insights into aging in rodent tissues, studies of human tissues are still limited. The small number of human studies have yielded interesting insights into, for instance, aging of pancreas , skin , retina and bone mar. The bottleneck for human studies is the limited access to healthy human tissue across the lifespan. There is also an additional challenge for skeletal muscle tissue, as it requires different types of single-cell processing for optimal capture of all cell types. In the present study, we combined scRNA-seq and snRNA-seq to build a human skeletal muscle aging atlas that includes both MuSCs and MF nuclei as well as cells from the microenvironment. We annotated 40 major and 82 fine-grained cell and nuclei states, providing deep and detailed insights that go beyond previous studies . 尽管单细胞基因组学研究为啮齿动物组织的衰老提供了许多见解，但对人类组织的研究仍然有限。少数人类研究已经为胰腺、皮肤、视网膜和骨髓等器官的衰老提供了有趣的见解。人类研究的瓶颈在于有限的跨寿命健康人类组织获取。对于骨骼肌组织还存在额外的挑战，因为它需要不同类型的单细胞处理才能最佳捕获所有细胞类型。在本研究中，我们结合了 scRNA-seq 和 snRNA-seq，构建了一个包括 MuSCs 和 MF 细胞核以及来自微环境的细胞的人类骨骼肌衰老图谱。我们注释了 40 个主要和 82 个细粒度的细胞和细胞核状态，提供了深入和详细的见解，超越了以往的研究。
d
e
Fig. 6 | Common skeletal muscle aging changes in human and mouse. 图 6 | 人类和小鼠骨骼肌衰老的共同变化。
a, UMAP plot showing main cell populations in the integrated human and mouse skeletal muscle dataset of 346,296 cells, including our muscle aging atlas as well as six other publicly available resources. b, Bar plots showing the number of significantly upregulated and downregulated DEGs in mouse and human across different cell types. c, Heatmap showing consistency of the DEGs within the same cell type in human and mouse for upregulated (left) and downregulated (right) genes. Consistency was calculated using Jaccard similarity index. Immune cells showcase the highest similarity among cell type groups and are highlighted in a, UMAP 图显示整合的人类和小鼠骨骼肌数据集中的主要细胞群，包括我们的肌肉老化图谱以及其他六个公开资源。 b, 条形图显示小鼠和人类在不同细胞类型中显著上调和下调的DEGs数量。 c, 热图显示人类和小鼠相同细胞类型中上调（左）和下调（右）基因的DEGs的一致性。一致性使用Jaccard相似性指数计算。免疫细胞展示了细胞类型群中最高的相似性，并在其中突出显示。
From our in-depth analysis, we identified aging mechanisms acting in parallel across different cell compartments. In the MuSC compartment, we found downregulation of ribosome assembly resulting in decreased MuSC activation as well as upregulation of pro-inflammatory pathways, such as NF-KB, and increased expression of cytokines, such as CCL2. In the MF microenvironment, we found several cell types that expressed pro-inflammatory chemokines, such as CCL2, CCL3 and CCL4. These cytokines may mediate the recruitment of lymphoid cells into muscle and the pro-inflammatory environment of aged muscle. Moreover, our cross-species and cross-muscle integrated aging atlas highlights an overall downregulation in gene expression, an increase in inflammation and a decrease in pro-growth, repair and innervation pathways. Pan-microenvironment upregulation of CCL2 with age was not recapitulated in mice, suggesting an interesting human-mouse distinction in orchestration of inflammation. However, although we have done our best to validate our findings on an independent patient cohort (which did not undergo ventilation), it is still possible that CCL2 upregulation is a result of biological confounder (comorbidity or ventilation effect). 根据我们的深度分析，我们确定了在不同细胞组分之间同时发生的衰老机制。在MuSC组分中，我们发现核糖体组装的下调导致MuSC激活减少，同时促炎途径（如NF-KB）的上调以及细胞因子（如CCL2）的增加表达。在MF微环境中，我们发现了几种表达促炎趋化因子（如CCL2、CCL3和CCL4）的细胞类型。这些细胞因子可能介导淋巴细胞进入肌肉并形成老年肌肉的促炎环境。此外，我们的跨物种和跨肌肉集成的衰老图谱强调了基因表达的总体下调，炎症的增加以及生长、修复和内神经的通路的减少。年龄相关的CCL2的全微环境上调在小鼠中没有重现，这表明在炎症协调中出现了有趣的人-老鼠区别。 然而，尽管我们已经尽力验证我们在一个独立的患者队列上的发现（该队列没有接受通气），但 CCL2 的上调仍可能是生物混杂因素（合并症或通气效应）的结果。
Fast-twitch MFs are more susceptible to atrophy in aging compared to slow-twitch ones, and several mechanisms have been proposed as 快速肌纤维在老化过程中比慢速肌纤维更容易发生萎缩，已经提出了几种机制作为解释。
explanations for this phenomenon 对这一现象的解释有几种机制被提出。. Using immunofluorescence-based MF typing, we confirmed almost complete loss of fast-twitch IIX MF in the intercostal muscle with age, but this loss was not replicated on the nuclei level. This apparent contradiction is due to the appearance of fast-type IIX nuclei within both slow-twitch and fast-twitch IIA MF with age. We also observed increased expression of fetal MYH8 in fast-twitch MF with age, which is a sign of a regenerative process. Both mechanisms are likely to represent compensatory changes in response to the loss of fast-twitch MF (type IIX) with age and can be a potential therapeutic target for reducing muscle aging. 通过免疫荧光基于MF分型，我们确认在肋间肌中快速收缩IIX型肌纤维几乎完全丧失随着年龄增长，但这种丧失在细胞核水平上并未复制。这种明显的矛盾是由于随着年龄增长，快速型IIX细胞核在慢速收缩和快速收缩IIA型肌纤维中出现。我们还观察到随着年龄增长，快速收缩肌纤维中胎儿MYH8的表达增加，这是再生过程的迹象。这两种机制可能代表了对年龄增长导致的快速收缩型IIXMF（肌型）丧失的补偿性变化，并可能作为降低肌肉老化的潜在治疗靶点。
Another important mechanism for MF degeneration and atrophy with age is loss of innervation. Higher efficiency of slow-twitch as opposed to fast-twitch MF re-innervation was suggested to contribute to the differential aging susceptibility of the two MF types. We observed a degeneration of NMJ in intercostal muscles, as judged by the decreased number of AChR clusters and a marked decline in protective terminal Schwann cells. Interestingly, we also describe a previously unreported type of NMJ accessory nuclei, which tend to co-localize with NMJ and increase considerably with age both in our intercostal and the published quadriceps muscle datasets. Our in vitro functional experiments knocking down and overexpressing NMJ accessory-specific genes show that these genes aid in the formation of AChR clusters, which form an essential part of the postsynaptic membrane. Hence, we propose that NMJ accessory may contribute to the re-innervation process. This is especially interesting in light of higher expression of slow-twitch MF markers within the NMJ accessory population. Another possibility is that NMJ accessory- nuclei represent a denervation state, as NMJ often undergoes repeated cycles of denervation-re-innervation with age . Overall, we cannot exclude that NMJ accessory both responds to denervation and stimulates the re-innervation of NMJ. Moreover, due to the structural complexity of the NMJ area-that is, densely packed nuclei-it is not always easy to distinguish NMJ and NMJ accessory in the same postsynaptic endplate. More precise spatial mapping of NMJ accessory will be a key task for future investigations. MF衰变和肌纤维萎缩的另一个重要机制是失去神经支配。相对于快收缩肌纤维，慢收缩肌纤维的再神经支配效率更高，可能有助于解释这两种肌纤维类型的不同老化易感性。我们观察到了肋间肌中神经肌肉连接的退化，这表现为AChR簇的数量减少和末梢Schwann细胞的显著下降。有趣的是，我们还描述了一种以前未报道的类型的NMJ附属核，它倾向于与NMJ共定位，并且在我们的肋间肌和已发布的股四头肌肌肉数据集中随着年龄的增长而明显增加。我们通过体外功能实验 knocking down 和过表达 NMJ 附属特异基因显示，这些基因有助于AChR簇的形成，这是 postsynaptic 膜的重要部分。因此，我们提出NMJ附属可能有助于再神经支配过程。这在NMJ附属人口中慢肌纤维标记的高表达的情况下尤为有趣。 另一种可能性是NMJ附属核代表一种脱神经状态，因为NMJ通常随着年龄的增长经历反复的脱神经-再神经化循环。总的来说，我们不能排除NMJ附属核既对脱神经作出反应又促进NMJ的再神经化。此外，由于NMJ区域的结构复杂性，即核密集地区，不总是容易区分同一 postsynaptic endplate 中的NMJ和NMJ附属核。对NMJ附属核的更精确空间映射将是未来研究的关键任务。
Our study design has some limitations, such as exposure of organ donors to ventilation (which did not vary between age groups) and a relatively small sample size, which precluded detailed investigation of confounding biological covariates, such as sex, BMI or exercise. Moreover, due to limitations in sample availability, we included a few specimens from middle-aged donors (approximately 50-60 years old) and mice at 19 months of age, which do not qualify as geriatric but already showed some aging features. Future efforts to incorporate data from a broader range of sources will help clarify biological changes between the extremes of young compared to aged humans and mice. However, the use of large biopsies from organ donors ensured minimal ischemic time and enabled multiple assays (scRNA-seq/snRNA-seq and imaging) on the same tissue piece, generating high-quality data. This allowed us to generate an integrated transcriptomics and imaging dataset that provides a global overview of muscle aging biology and lays strong foundations for future studies of this process. 我们的研究设计存在一些局限性，比如器官捐赠者暴露于通气（在不同年龄组之间没有变化）以及相对较小的样本量，这妨碍了对混杂的生物学协变量（如性别、BMI 或锻炼）进行详细调查。此外，由于样本可用性的限制，我们包括了一些来自中年捐赠者（大约 50-60 岁）和 19 个月大的老鼠的标本，这些标本虽然不符合老年人的标准，但已经显示出一些衰老特征。未来努力将包括来自更广泛来源的数据，有助于澄清年轻人和老年人以及老鼠之间生物学变化的差异。然而，使用器官捐赠者的大量活检确保了最小的缺血时间，并在同一组织片上进行了多次分析（scRNA-seq/snRNA-seq 和成像），生成了高质量的数据。这使我们能够生成一个集成的转录组学和成像数据集，提供了肌肉衰老生物学的全局概述，并为未来研究这一过程奠定了坚实的基础。
Methods 方法
Experimental methods 实验方法
Access to human and mouse tissue and ethics. Single-cell transcriptomics. Human intercostal muscle samples (inner part between the second and third ribs) for scRNA-seq and snRNA-seq were collected with consent from deceased transplant organ donors by the Collaborative Biorepository for Translational Medicine (CBTM), immediately placed in HypoThermosol FRS preservation solution and shipped to the Sanger Institute for processing. Ethical approval was granted by the Research Ethics Committee (REC) East of England-Cambridge South (REC ref.15/EE/0152), and written informed consent was obtained from the donor families. Full metadata information for the organ donors is provided in Supplementary Table 1. Three 19-month-old and five 3-month-old male mice of the C57BL/6JRj strain were obtained from Janvier Labs. All mice were housed in micro-isolator cages in standard housing conditions (ambient temperature of and humidity of 40-60%), illuminated from 07:00 to 19:00 with ad libitum access to diet and water, under establishment licence number X3AOED725 provided by the Home Office. They were used to dissect hindlimb muscles for the single-cell and single-nucleus isolation. 人类和小鼠组织以及伦理准入。单细胞转录组学。人类肋间肌样本（第二和第三肋骨之间的内部部分）用于scRNA-seq和snRNA-seq，经过合作生物库转化医学（CBTM）获得已故移植器官捐赠者的同意后收集，立即放入HypoThermosol FRS保存溶液并运送到桑格研究所进行处理。研究伦理委员会（REC）东英格兰-剑桥南（REC参考15/EE/0152）已批准伦理审批，并已从捐赠者家属处获得书面知情同意。器官捐赠者的完整元数据信息在附表1中提供。从Janvier Labs获得了C57BL/6JRj品系的三只19个月大和五只3个月大的雄性小鼠。所有小鼠都被放置在微隔离笼中，处于标准饲养条件下（环境温度 和湿度40-60%），从07:00到19:00照明，可随时获得饮食和水，根据内政部提供的X3AOED725号成立许可证号。它们被用于解剖后肢肌肉进行单细胞和单核分离。
Adult tissue from the UK for validation experiments. The same intercostal muscle samples collected with consent from deceased transplant organ donors (partially overlapping with the set of donors used for scRNA-seq/snRNA-seq) by the CBTM were used for experimental validations. 英国成人组织用于验证实验。由CBTM收集的已故移植器官捐赠者同意收集的相同肋间肌肉样本（部分与用于scRNA-seq/snRNA-seq的捐赠者集合重叠）用于实验验证。
Fetal and adult tissue from China for validation experiments. Adult human intercostal muscle biopsies were collected during thoracic surgeries at Sun Yat-sen Memorial Hospital under approval of the REC of Sun Yat-sen University (no. 2018-048). For isolation of human primary myoblasts, lower limb muscles were collected from one medically aborted embryo at post-conceptional week19 at Guangzhou Women and Children's Medical Center with ethical approval licence granted by both the REC of Sun Yat-sen University (no. 2019-075) and Guangzhou Women and Children's Medical Center (no. 2022-050A01). Both materials were registered at the China National Center for Bioinformation (PRJCA014979) and were approved by the Chinese Ministry of Science and Technology for the Review and the Approval of Human Genetic Resources (2023BAT0735). Appropriate written informed consent was obtained from each adult patient to retrieve a muscle biopsy together with resected tissue (usually tumor). Informed consent was also obtained from the mother after her voluntary decision to legally terminate pregnancy but before the abortion. Before terminating pregnancy, both the mother and the embryo were diagnosed as healthy with no underlying diseases. Participants were not financially compensated. The detailed metadata for the eight organ donors (UK), 40 patients (China) and one embryo used for validation experiments are provided in Supplementary Table 8. 中国的胎儿和成人组织用于验证实验。在中山大学附属第一医院进行胸部手术期间收集了成年人的肋间肌肌活检组织，获得了中山大学研究伦理委员会的批准（批准号2018-048）。为了分离人类原代肌母细胞，从广州市妇女儿童医疗中心在怀孕后第19周进行了一次医学流产手术的胚胎的下肢肌肉进行了收集，获得了中山大学研究伦理委员会（批准号2019-075）和广州市妇女儿童医疗中心（批准号2022-050A01）的道德许可证。这两种材料已在中国国家生物信息中心注册（PRJCA014979），并获得了中国科学技术部对人类遗传资源审查和批准的批准（2023BAT0735）。从每位成年患者获得了适当的书面知情同意，以取回肌肉活检组织以及切除的组织（通常是肿瘤）。在母亲自愿决定合法终止妊娠但在流产前也获得了知情同意。 在终止怀孕之前，母亲和胚胎均被诊断为健康，没有潜在疾病。参与者没有得到经济补偿。提供了八名器官捐赠者（英国）、40名患者（中国）和一个用于验证实验的胚胎的详细元数据，见附表8。
Single-cell/single-nucleus sample processing. Skeletal muscle tissue was processed according to the following protocols for single-cell and single-nucleus isolation from skeletal muscle deposited at https:// www.protocols.io/ (refs. 77,78). In brief, muscle tissue was minced, digested in a solution of Collagenase II (Worthington Biochemical, LS004176) and Dispase (Gibco, 17105041). The lysate was centrifuged in the gradient of Percoll to recover a fraction with single cells. For single-nucleus isolation, muscle tissue was ground using a dounce homogenizer and lysed in the nuclei lysis buffer, and, finally, Percoll gradient was used to separate intact nuclei and cell debris. For the scRNA-seq experiments, either 8,000 live cells or 8,000 intact nuclei were loaded per sample into a Chromium Controller (10x Genomics), and a Single Cell 3' v2 or v3 Reagent Kit was used to create GEM, perform cDNA synthesis and generate sequencing libraries. The libraries were sequenced on an Illumina HiSeq 4000 or a NovaSeq 6000 platform. 单细胞/单核样品处理。骨骼肌组织根据以下方案进行处理，以从骨骼肌中分离单细胞和单核，方案详见https://www.protocols.io/（参考文献77,78）。简而言之，肌肉组织被切碎，用Collagenase II（Worthington Biochemical, LS004176）和Dispase（Gibco, 17105041）的溶液消化。溶液在Percoll梯度中离心，以回收含有单细胞的部分。对于单核的分离，肌肉组织使用dounce均质器研磨，并在细胞核裂解缓冲液中裂解，最后使用Percoll梯度分离完整的细胞核和细胞碎片。对于scRNA-seq实验，每个样品加载8,000个活细胞或8,000个完整细胞核到Chromium Controller（10x Genomics）中，使用Single Cell 3' v2或v3试剂盒创建GEM，进行cDNA合成并生成测序文库。文库在Illumina HiSeq 4000或NovaSeq 6000平台上测序。
Skeletal muscle biopsy processing for FACS. Freshly obtained intercostal muscle biopsies were minced with fine scissors and digested at for with gentle shaking in (per gram of tissue) of solution containing Dispase II (Roche, 4942078001) and Collagenase (Roche, 11088815001) supplemented with and penicillin-streptavidin (Gibco, 15140122). Digestion was stopped with FBS and 2 mM EDTA solution in PBS, and tissue suspension was sequentially filtered through (Falcon, 352360) and 40- (Falcon, 352340) strainers to get the single-cell suspensions. After centrifugation and reconstitution, cells were adjusted to cells per milliliter with FACS buffer ( FBS in PBS) and incubated with fluorophore-antibody cocktails (Supplementary Table 9) for to sort MuSCs (CD31-CD82 TNFRSF ) and MuSCs (CD31-CD82 ). Cells were sorted 用于 FACS 的骨骼肌活检处理。新鲜获取的肋间肌活检用细剪开切并在 中与缓慢摇动的 （每克组织）溶液中进行 处理。溶液包含 二肽酶 II （Roche, 4942078001）和 胶原酶 （Roche, 11088815001），并添加 青霉素-链霉素 （Gibco, 15140122）。使用 FBS 和 2mM EDTA 溶液停止消化，组织悬液经顺序通过 （Falcon, 352360）和 40- （Falcon, 352340）的过滤器过滤以获得单细胞悬液。离心和重悬后，用 FACS 缓冲液（PBS 中的 FBS）将细胞调整至每毫升 个细胞，并与荧光素-抗体混合物（附录表 9）孵育 以分选 MuSCs（CD31-CD82 TNFRSF ）和 MuSCs（CD31-CD82 ）。分选后，细胞排序
and analyzed with a BD Influx cell sorter, and data were analyzed with FlowJo (version 10.4) software. 并用 BD Influx 细胞分选仪进行分析，数据使用 FlowJo (版本 10.4) 软件分析。
Dissociation of human primary myoblasts and cell culture. Human intercostal muscle biopsies were digested as described in FACS to obtain single-cell suspension. After centrifugation and repeated washing, cells were pre-plated in a 10-cm gelatin-coated (gelatin, STEMCELL Technologies, 07903) cell culture dish for to get rid of fibroblasts. After pre-plating, the cell supernatant was gently transferred to a new cell culture dish to enrich primary myoblasts and kept in the incubator at and . 人类原代肌母细胞的分离和细胞培养。人类肋间肌肌活检经 FACS 消化，获得单细胞悬浮液。离心和重复洗涤后，细胞预培养在一只 10 厘米明胶涂层（明胶，STEMCELL Technologies，07903）的细胞培养皿中，以去除成纤维细胞。预培养后，细胞上清液轻轻转移到新的细胞培养皿中，富集原代肌母细胞，并在孵育箱中保持在 和 。
Immunfluorescence. The detailed antibody information for immunofluorescence is provided in Supplementary Table 9, and quantifications were performed using either Fiji software or custom image analysis pipeline (for MF). 免疫荧光。免疫荧光的详细抗体信息请参见附表 9，并使用 Fiji 软件或自定义图像分析流程（用于 MF）进行定量分析。
For MF, immune and vasculature cell type stainings, fresh-frozen blocks of muscle biopsies (obtained from Chinese patients) were used. Tissue sections were fixed with paraformaldehyde (PFA) and incubated in citrate buffer ( ) in a pressure cooker to perform heat-activated antigen retrieval. Sections were then blocked with 10% AffiniPure Fab goat anti-mouse IgG (Jackson ImmunoResearch, 115-007-003) for and 5% normal goat serum (Jackson ImmunoResearch, 005-000-121) for , respectively. Next, they were incubated with primary antibodies at overnight followed by incubation with secondary antibodies for at room temperature. After staining with DAPI, sections were mounted with fluorescence-saving mounting medium (Millipore,345789) and imaged with a DMi8 inverted microscope (Leica Microsystems) or scanned with a digital pathology slide scanner (KFBIO, KF-FL-400). 对于MF、免疫和血管细胞类型染色，使用了来自中国患者的肌肉活检的新鲜冷冻块。组织切片用 对甲醛（PFA）固定，并在压力锅中孵育在柠檬酸盐缓冲液（ ）中进行热激活抗原检索。然后，分别使用10%的AffiniPure Fab山羊抗小鼠IgG（Jackson ImmunoResearch, 115-007-003）和5%的正常山羊血清（Jackson ImmunoResearch, 005-000-121）进行阻断。接下来，用 过夜孵育主抗体，然后在室温下用二抗孵育 。用DAPI染色后，切片用荧光保存载玻片（Millipore,345789）封装，并用DMi8倒置显微镜（Leica Microsystems）拍摄或用数字病理学切片扫描仪（KFBIO, KF-FL-400）扫描。
Co-staining of CCL2 and ACTA2 was performed on formalin-fixed paraffin-embedded (FFPE) blocks of muscle biopsies obtained from the organ donors in the UK. The stainings were performed using automated Leica Biosystems BOND RX, and all sections were baked and dewaxed and subjected to heat-induced epitope retrieval enzyme 2 for at . After incubation with primary antibodies, sections were first incubated with HRP-conjugated goat anti-mouse IgG and visualized with fluorophore Opal 570 for ACTA2 and stained with DAPI. After blocking HRP activities, sections were then incubated with HRP-conjugated goat anti-rabbit IgG and visualized with fluorophore Opal 650 for CCL2. Slides were imaged using a Hamamatsu S60 slide scanner at magnification, and images were visualized with OMERO Plus (Glencoe Software). 在英国的器官捐赠者的肌肉活检标本中，使用自动化的Leica Biosystems BOND RX进行了CCL2和ACTA2的共染色。所有切片均经过烘烤、脱蜡处理，并在 下进行热诱导表位检索酶2处理。与初级抗体孵育后，切片首先与HRP偶联的山羊抗小鼠IgG孵育，并使用荧光染料Opal 570可视化ACTA2，并用DAPI染色。阻断HRP活性后，切片随后与HRP偶联的山羊抗兔IgG孵育，并使用荧光染料Opal 650可视化CCL2。使用Hamamatsu S60幻灯片扫描仪在 放大倍率下拍摄幻灯片图像，并使用OMERO Plus（Glencoe Software）可视化图像。
The same FFPE blocks from organ donors were used to perform 15-plex RareCyte immunofluorescence staining (two young versus two aged donors). The blocks were sectioned at thickness, mounted on Superfrost slides and dried at for to adhere them to the slides. Sections were incubated with a 15 -plex cocktail of custom-formulated ArgoFluor-conjugated antibodies (RareCyte) according to the manufacturer's instructions, and Hoechst was used to stain nuclei. The stained slides were imaged using the RareCyte Orion platform with seven lasers and pre-processed using RareCyte Artemis 4.0 software, which compensates for channel crosstalk and autofluorescence. 器官捐赠者的相同FFPE块用于进行15-plex RareCyte免疫荧光染色（两名年轻捐赠者与两名年长捐赠者）。块被切割成 厚度，装在Superfrost玻片上并在 处干燥 ，以将它们粘附到玻片上。根据制造商的说明书，切片使用定制配制的ArgoFluor偶联抗体（RareCyte）的15-plex混合物孵育，Hoechst用于染色细胞核。染色玻片使用RareCyte Orion平台进行成像，配备七个激光器，并使用RareCyte Artemis 4.0软件进行预处理，该软件可以补偿通道串扰和自发荧光。
RNAscope. RNAscope staining for markers of MF nuclei populations and macrophage marker LYVE1 was performed on FFPE sections from organ donors acquired in the UK. RNAscope LS multiplex fluorescent reagent kit (ACD, Bio-Techne) and automated Leica Biosystems BOND RX were used for the staining, as per the manufacturer's instructions. All sections were baked and dewaxed and subjected to heat-induced epitope retrieval enzyme 2 for at and of protease III before staining. The detailed probe information can be found in Supplementary Table 9. For dual RNAscope and immunofluorescence staining, the sections were then stained with anti-MYH7 antibody RNAscope。在英国获得的器官捐赠者的 FFPE 切片上进行了 MF 细胞核群体和巨噬细胞标记物 LYVE1 的 RNAscope 染色。使用 RNAscope LS 多重荧光试剂盒（ACD，Bio-Techne）和自动化的 Leica Biosystems BOND RX 进行染色，按照制造商的说明进行。所有切片均经过烘烤、脱蜡并在染色前经过热诱导表位重现酶 2 处理 在 和 的蛋白酶 III。详细的探针信息可以在附录表 9 中找到。对于双重 RNAscope 和免疫荧光染色，然后用抗 MYH7 抗体染色
(Developmental Studies Hybridoma Bank, BA-F8, 1:14). Confocal imaging was performed on a PerkinElmer Operetta CLS High Content Analysis System using a （发育研究杂交瘤库，BA-F8，1:14）。在 PerkinElmer Operetta CLS 高内容分析系统上进行共聚焦成像 (numerical aperture per pixel) water immersion objective with a 9-11 z-stacks step. Confocal image stacks were stitched as individual z-stacks using proprietary Acapella scripts provided by PerkinElmer and visualized using OMERO Plus. The quantifications of cells were analyzed using Fiji. （数值孔径 每像素）水浸入式物镜，具有 9-11 个 z 层 步进。使用 PerkinElmer 提供的专有 Acapella 脚本将共聚焦图像堆叠为单独的 z 层，并使用 OMERO Plus 进行可视化。使用 Fiji 分析 细胞的数量。
NMJ accessory nuclei identification. NMJ endplates were defined based on the characteristic clustering of the nuclei reminiscent of NMJ, which also had an expression of CHRNE. To be noted, while performing RNAscope with GRIA2 and CHRNE probes, we found some RNA punctate (as compared to negative control staining) in MF cytoplasm and non-synaptic nuclei. This can be due to either biological mechanisms that have not been identified or non-specific staining that was more frequent on FFPE sections. NMJ 附件核识别。NMJ 末板基于与 NMJ 相似的核团聚集特征定义，该核团还表达 CHRNE。需要注意的是，在使用 GRIA2 和 CHRNE 探针进行 RNAscope 时，我们发现一些 RNA 点状（与阴性对照染色相比）在 MF 细胞质和非突触核中。这可能是由于尚未确定的生物机制或更频繁地出现在 FFPE 切片上的非特异性染色。
Differentiation of human primary myoblasts and induction of AChR aggregation. Purified embryonic myoblasts were grown in DMEM/F-12 cell culture medium containing human basic fibroblast growth factor and penicillin-streptavidin. For myogenic differentiation, cells at confluence were changed to DMEM/F-12 medium containing house serum and penicillin-streptavidin, and myoblasts were differentiated to myotubes within 2-3 d. 人类原代肌母细胞的分化和 AChR 聚集的诱导。纯化的胚胎肌母细胞在含有 人类基本成纤维细胞生长因子和 青霉素-链球菌的 DMEM/F-12 细胞培养基中生长。对于肌肌细胞分化，达到 且细胞的培养基切换到含有 家庭血清和 青霉素-链球菌的 DMEM/F-12 培养基中，肌母细胞在 2-3 天内分化为肌管。
For differentiation and AChR aggregation induction, cells were seeded in six-well cell culture plates pre-coated with natural mouse laminin (Gibco, 23017015) in DMEM/F-12 at for at least . Upon reaching confluence, cells were switched to differentiation media. On day 2 of myogenic differentiation, cells were first incubated with of laminin for and later supplemented with of differentiation medium to induce formation of mature AChR clusters. Once myoblasts got differentiated into myotubes on day 3, siRNAs targeting SORBS2 and EFNA5 or plasmids expressing EFNA5 were transfected into the myotubes. Forty-eight hours after transfection, myotubes were fixed with PFA, washed with PBST and stained with -BTX followed by DAPI. Topological AChR aggregates were visualized using the DMi8 inverted microscope and quantified using Fiji. 为了区分和 AChR 聚集诱导，细胞被播种在预涂有 天然小鼠层粘蛋白（Gibco，23017015）的六孔细胞培养板中，在 DMEM/F-12 中培养至少 。当细胞达到 的浓度时，细胞被转移到分化培养基中。在肌源性分化的第 2 天，细胞首先与 的层粘蛋白一起孵育 ，然后再补充 的分化培养基，诱导成熟 AChR 簇的形成。一旦肌母细胞在第 3 天分化为肌管，siRNA 靶向 SORBS2 和 EFNA5 或表达 EFNA5 的质粒被转染到肌管中。转染 48 小时后，肌管用 PFA 固定，用 PBST 洗涤，并用 -BTX 染色，然后用 DAPI 染色。通过 DMi8 倒置显微镜可视化拓扑 AChR 聚集，并使用 Fiji 进行定量。
Computational methods 计算方法
Single-cell data pre-processing and integration. The and v3 10x Genomics skeletal muscle sequencing data were aligned and quantified using Cell Ranger version 3.1.0 with GRCh38-3.0.0 human and mm10-1.2.0 mouse reference genomes. Pre-mRNA version of reference genomes was used for alignment of nuclei datasets. STARsolo pipeline (based on STAR 2.7.3) mimicking Cell Ranger 2.x.x with options '-soloFeatures Gene GeneFull Velocyto' was employed to separate spliced and unspliced counts, which were used to differentiate MF fragments. The following single-cell data analysis and visualization were mostly performed in Python (version 3 ) with some analysis done in (version 3.6.3 and version 4.0.4) with data.table (version 1.14.0), ggplot2 (version 3.3.2) and ggpubr (version 0.4.0). 单细胞数据预处理和整合。使用 Cell Ranger 版本 3.1.0 对 v3 10x Genomics 骨骼肌测序数据进行了对齐和量化，使用了 GRCh38-3.0.0 人类和 mm10-1.2.0 小鼠参考基因组。核基因组数据集的对齐使用了前体 mRNA 版本的参考基因组。使用了 STARsolo 流程（基于 STAR 2.7.3）来模拟 Cell Ranger 2.x.x，使用选项 '-soloFeatures Gene GeneFull Velocyto' 分离剪接和未剪接计数，用于区分MF片段。以下单细胞数据分析和可视化主要在 Python 中进行（版本 3），其中一些分析在 （版本 3.6.3 和版本 4.0.4）中进行，使用了 data.table（版本 1.14.0）、ggplot2（版本 3.3.2）和ggpubr（版本 0.4.0）。
CellBender was used to remove ambient RNA contamination from both single-cell and single-nucleus data with the following parameters: and learning rate (for some samples, these were adjusted to 250 epochs and 0.00005 learning rate). Scrublet was used to identify potential doublets in each sample, and, after that, cells with scrublet score were filtered out as doublets. Next, additional filtering was performed to discard potential empty droplets and doublet cells using the custom thresholds for number of genes and unique molecular identifiers (UMI) counts: for cells ( and max 5,000 genes, min 700 counts and max 50,000 counts), for nuclei ( and max 5,000 genes, min 500 counts and counts). Cells with more than and nuclei with more than mitochondrial genes expressed were removed as potential low-quality cells. CellBender 用于从单细胞和单核数据中去除环境RNA污染，使用以下参数： 和学习率 （对于某些样本，这些参数被调整为250个epochs和0.00005的学习率）。Scrublet 用于识别每个样本中的潜在双细胞体，之后，具有Scrublet分数 的细胞被过滤为双细胞体。接下来，通过使用基因数和唯一分子标识符（UMI）计数的自定义阈值进行额外过滤，以丢弃潜在的空液滴和双细胞体细胞：对于细胞（ 和最多5,000个基因，最少700个计数和最多50,000个计数），对于细胞核（ 和最多5,000个基因，最少500个计数和 计数）。具有超过 个细胞和具有超过 个核基因表达的线粒体基因的细胞被移除为潜在低质量细胞。
Scanpy Python package (version 1.7.2) was used to load the cell-by-gene count matrix and perform processing according to the standard pipeline with modifications. Marker genes were identified using different approaches. In most cases, -test was applied to identify DEGs in the given cluster as compared to the rest using sc.tl.rank_gene_groups (method = 't-test_overestim_var', corr method = 'benjamini-hochberg'); obtained values were corrected using the Benjamini-Hochberg method; and the top 100 genes were considered. Alternatively, the scvi.model.SCVI.differential_expression function was used to calculate DEGs between a particular cluster and the reference using the Bayesian approach. This was used to better call markers for specialized nuclei populations, such as I-FAM, I-OTU, II-FAM, II-OTU and II-TNF. Specifically, DEGs were called by comparing every specialized population to the conventional type I or type II MF cluster, respectively. Later, DEGs were further pre-filtered to have -transformed fold change above 1 , to be expressed in at least of the cells and to have Bayes factor above 2. Finally, gene set overrepresentation analysis was performed on the marker genes using scanpy. queries.enrich, a wrapper for gprofiler or Metascape web interface . Scanpy Python软件包（版本1.7.2） 被用来加载细胞-基因计数矩阵，并根据带有修改的标准流程进行处理。使用不同方法识别标记基因。在大多数情况下，使用 -test来识别给定群集中的DEGs，与其余部分进行比较，使用sc.tl.rank_gene_groups（method='t-test_overestim_var'，corr method='benjamini-hochberg'）；获得的 值使用Benjamini-Hochberg方法进行校正；并考虑前100个基因。另外，使用scvi.model.SCVI.differential_expression函数使用贝叶斯方法计算特定群集与参考之间的DEGs。这用于更好地调用专门核团群体的标记物，如I-FAM、I-OTU、II-FAM、II-OTU和II-TNF。具体而言，通过将每个专门群体与传统的I型或II型MF群集进行比较来调用DEGs。随后，进一步预过滤DEGs，使 -转换的折叠变化大于1，在至少 的细胞中表达，并且贝叶斯因子大于2。 最后，使用 scanpy 对标记基因进行基因集过表示分析。queries.enrich 是 gprofiler 或 Metascape web 界面的包装器。
Myonuclei typing was performed using single-nucleus expression measurements of and , which were scaled by library size and log transformed; a minimum threshold of 0.5 was used to classify myonuclei as expressing a particular gene. Myonuclei with expression of all three genes below the 0.5 threshold and myonuclei belonging to the subtypes of fragments (MF-Isn and MF-IIsn) or rare Hyb type were deemed as 'unclassified'. 利用单核表达测量的 和 进行肌细胞核分型，通过库大小和对数转换进行缩放；将表达特定基因的肌细胞核分类的最低阈值为 0.5。所有三种基因表达低于 0.5 阈值的肌细胞核以及属于片段（MF-Isn 和 MF-IIsn）或罕见 Hyb 类型的肌细胞核被视为“未分类”。
Trajectory analysis to uncover intermediate stages between MuSC and MF was performed using the Monocle package (version 2.9.0) . The dataset for trajectory analysis was limited to MuSCs and MF cells (coming from scRNA-seq) and the known genes important for muscle differentiation: PAX7, MYF5, PDGFRA, MYOG, TPM1, MYH2, MYH3, NCAM1, TNNT1, TNNT2, TNNT3, TNNC1, CDK2, CCND1, CCNA1 and ID1. 使用 Monocle 软件包（版本 2.9.0）进行揭示 MuSC 和 MF 之间中间阶段的轨迹分析。轨迹分析的数据集仅限于来自 scRNA-seq 的 MuSC 和 MF 细胞以及对肌肉分化重要的已知基因：PAX7、MYF5、PDGFRA、MYOG、TPM1、MYH2、MYH3、NCAM1、TNNT1、TNNT2、TNNT3、TNNC1、CDK2、CCND1、CCNA1 和 ID1。
Integration of publicly available datasets. Throughout the study, we integrated various public skeletal muscle datasets to strengthen our findings, namely: 公开可用数据集的整合。在整个研究过程中，我们整合了各种公开的骨骼肌数据集来加强我们的发现，即：
We downloaded pre-processed mouse MuSCs from the following Gene Expression Omnibus datasets: GSE110878, GSE143476, GSE134540, GSE138707 and GSE149590. We used these to compare quiescent and activated human and mouse MuSC subtypes. 我们从以下基因表达 Omnibus 数据集下载了经过预处理的小鼠 MuSCs：GSE110878、GSE143476、GSE134540、GSE138707 和 GSE149590。我们使用这些数据来比较静止和激活的人类和小鼠 MuSC 亚型。
We obtained raw files and pre-processed and annotated the de novo quadriceps muscle dataset from Perez et al. (following the same pipeline that we used for in-house single-nuclei data). We used it to compare the change in myonuclei populations with age between intercostal and leg muscle types. 我们获取了 Perez 等人的 de novo 股四头肌肌肉数据集的原始文件，并对其进行了预处理和注释（遵循我们用于内部单核数据的相同流程）。我们用它来比较随着年龄变化而发生的肌核群体在肋间和腿部肌肉类型之间的变化。
We obtained pre-processed data from all human and mouse skeletal muscle scRNA-seq studies available at the moment of the analysis (human: GSE143704 and DRYAD (https://doi. org/10.7272/Q65X273X) and mouse: GSE110878, GSE138707, GSE134540 and GSE149590) to create a human-mouse integrated aging atlas. Next, human and mouse datasets underwent quality control filtering as described previously and concatenated into one object, retaining only homologous genes. The scVI model was used to produce an integrated embedding for the data, using SampleID as a batch and species, 10x chemistry, muscle type and sex as categorical covariates. 我们从所有人类和小鼠骨骼肌单细胞 RNA 测序研究中获取了预处理数据（人类：GSE143704 和 DRYAD（https://doi.org/10.7272/Q65X273X）和小鼠：GSE110878、GSE138707、GSE134540 和 GSE149590），以创建人-小鼠整合老化图谱。接下来，人类和小鼠数据集经过质量控制过滤，如前所述连接成一个对象，仅保留同源基因。使用 scVI 模型为数据生成集成嵌入，使用 SampleID 作为批次和物种、10x 化学、肌肉类型和性别作为分类协变量。
Aging cell type composition analysis. With the aim to identify the age effect on cell composition, we used two different approaches depending on the size of the dataset and distinctness of the clusters. 老化细胞类型组成分析。为了确定年龄对细胞组成的影响，我们根据数据集的大小和聚类的独特性采用了两种不同的方法。
Mixed-effect Poisson regression model was used to assess the effect of age on cell type abundance in human. 使用混合效应泊松回归模型评估年龄对人类细胞类型丰度的影响。
Specifically, the effect of age on cell-type-specific counts was modeled using the Poisson linear mixed-effect model accounting for the possible biological and technical covariates using the 'glmer' function from the 'Ime4' package in R. We provided all of the factors as mixed terms (for instance, 具体来说，年龄对细胞类型特异计数的影响是通过使用泊松线性混合效应模型进行建模的，考虑了可能的生物学和技术协变量，使用R中的'Ime4'包中的'glmer'函数。我们将所有因素作为混合项提供（例如， ) as these allow estimation of the coefficients despite the collinearity of covariates. The effect of age and most of the covariates were estimated as an interaction term with the cell type. The log-transformed FC for every covariate was calculated relatively to the grand mean and adjusted so its value is 0 when there is no effect. Local true sign rate (LTSR) was used to estimate statistical significance, which denotes the probability that the estimated direction of the effect is true (see details on its calculation for cell type composition analysis in ref. 85). LTSR ranges from 0 to 1 , where the higher value denotes higher probability. We used LTSR as a cutoff to call significant age effect on cell type compositions. It is worth noting that our dataset contains two technical replicates (that is, libraries) for a substantial number of the single-cell and single-nucleus donor libraries as well as one additional biological replicate for two donors' single-nucleus libraries (Supplementary Table 1). However, it is not currently possible to construct a model that hierarchically decomposes donor effect into the effect of samples and also models additional covariates. Hence, we chose to use the library as an individual replicate for the model fitting process. We acknowledge that this can artificially increase the significance for some of the results, and, therefore, we also provide plots illustrating the raw cell type proportion data showing the average proportion for the samples in each donor in Extended Data Figs. and . 由于这些问题，可以估算系数，尽管协变量存在共线性。年龄和大多数协变量的影响被作为与细胞类型的交互项进行估算。每个协变量的对数转换后的折叠变化（FC）相对于总均值进行计算，并调整使其值为0，以表示没有影响。本地真实符号率（LTSR）用于估算统计显著性，表示估计效应方向为真的概率。LTSR范围从0到1，数值越高表示概率越高。我们将LTSR 用作年龄对细胞类型组成的显著影响的截断值。值得注意的是，我们的数据集中包含数量可观的单细胞和单核桥接库的两个技术重复（即 库），以及两位捐赠者的单核桥接库的一个额外生物学重复（见附录表1）。 然而，目前尚无法构建一个模型，将供体效应分层分解为样本效应，并且还模拟了额外的协变量。因此，我们选择使用 库作为模型拟合过程中的单独复制品。我们意识到这可能会人为地增加某些结果的显着性，因此我们还提供了绘制图表，显示每个供体样本中细胞类型比例数据的原始数据，显示每个供体中样本的平均比例在扩展数据图中。 和 。
The model fitted for major human cell types is (Fig. 1c): Ncs Age + (1|Celltype) Sample batch (Age-1|Celltype) Celltype batch::Celltype (1|Sample::Celltype), where Ncs denotes the cell count of cell type in sample ; Age denotes age in years, scaled and centered; Sample denotes 10x library ID; and batch denotes cells or nuclei. 适用于主要人类细胞类型的模型为（图 1c）：Ncs 年龄 +（1|细胞类型） 样本 批次 （年龄-1|细胞类型） 细胞类型 批次::细胞类型 （1|样本::细胞类型），其中 Ncs 表示样本中细胞类型 的细胞计数；年龄表示以年为单位的年龄，经过缩放和居中处理；样本表示 10x 库 ID；批次表示细胞或细胞核。
MiloR (version 1.2.0) framework was used to detect aging changes in the states within one or several cell types (for MuSC, MF, immune, fibroblast and vascular cell types). MiloR（版本 1.2.0） 框架用于检测一个或多个细胞类型（对于 MuSC、MF、免疫、成纤维细胞和血管细胞类型）内状态的衰老变化。
We preferred the Milo approach for smaller-scale datasets as it provides more resolution for the change and is not limited by the size of clusters and cell types. Specifically, we first constructed a -nearest neighbor (KNN) graph of cells ( was adjusted depending on cell type, ) using embedding obtained after the application of the scVI model on a particular cell type(s). Next, we sampled a representative group of cell neighborhoods across the KNN graph and obtained a count matrix with neighborhoods in rows and samples (that is, individual libraries) in columns. To be noted, our dataset contains two technical replicates (that is, libraries) for a substantial number of the single-cell and single-nucleus donor libraries as well as one additional biological replicate for two donors' single-nucleus libraries (Supplementary Table 1). After that, we applied a negative binomial linear regression model to assess the effect of Age on the number of cells in each neighborhood from each sample (that is, 10x library) accounting for the chemistry and Sex effect. The significance was controlled for multiple testing using weighted Benjamini-Hochberg correc. We later assigned cell type labels to the neighborhoods based on the cell type that contributed the majority of cells to this neighborhood. If the most abundant cell type contributed less than of cells, the neighborhood was labeled as mixed and filtered out. Additionally, if the neighborhood contained more than of cells from one donor, it was labeled as 'Donor-specific' and was filtered out. 我们更倾向于在小规模数据集上采用Milo方法，因为它对变化提供了更高的分辨率，并且不受聚类和细胞类型大小的限制。具体来说，我们首先在应用scVI模型后，根据特定细胞类型构建了一个K近邻（KNN）图（ 根据细胞类型进行调整， ）。接下来，我们在KNN图中对一组细胞邻域进行抽样，并获得一个包含邻域在行和样本（即个别 文库）在列的计数矩阵。值得注意的是，我们的数据集包含两个技术复制品（即 文库），用于大量单细胞和单核供体文库，以及两个供体的单核文库的一个额外生物复制品（附录表1）。之后，我们应用负二项式线性回归模型来评估年龄对每个样本（即10x文库）中每个邻域中细胞数量的影响，考虑到 化学和性别影响。 通过加权Benjamini-Hochberg校正控制了多重检验的显著性。然后，根据向该邻域贡献大多数细胞的细胞类型，为邻域分配了细胞类型标签。如果最丰富的细胞类型贡献的细胞少于 ，则将该邻域标记为混合并将其过滤掉。此外，如果邻域中来自一个供体的细胞超过 ，则将其标记为“供体特异性”并将其过滤掉。
Aging differential gene expression analysis. We performed aging differential gene expression analysis twice to identify age-associated genes changing in every major cell type as well as in every subtype. With that aim, we employed the linear mixed model as proposed , which allowed us to account for various technical (10x chemistry and data modality: scRNA-seq or snRNA-seq) and biological covariates (Donor, Sex, Donor type (DBD or DCD), BMI and length of ventilation) to disentangle the true effect of age. After fitting the model, a Bayes factor of each covariate was calculated for every gene as described in the previous report . Later, Bayes factors were used to compute the posterior probability and significance measure, LTSR (see section 1.3 of the Supplementary Note for more details), for the influence of the factor on every gene. The same model was also applied to test for age-associated DEGs separately within each species (human and mouse) while using muscle type and donor/mouse sex as biological covariates and 10x chemistry and data modality (scRNA-seq or snRNA-seq) as technical covariates. 进行老化差异基因表达分析。我们进行了两次老化差异基因表达分析，以确定在每种主要细胞类型以及每个亚型中发生变化的与年龄相关的基因。为此，我们采用了提出的线性混合模型，该模型允许我们考虑各种技术（10x化学和数据模态：scRNA-seq或snRNA-seq）和生物学协变量（供体、性别、供体类型（DBD或DCD）、BMI和通气时间）以解开年龄的真实影响。在拟合模型后，为每个基因计算了每个协变量的贝叶斯因子，如前一份报告中所述。随后，贝叶斯因子用于计算后验概率和显著性测量LTSR（有关更多详细信息，请参见补充说明的1.3节），以评估因素对每个基因的影响。同样的模型也被应用于在每个物种（人类和小鼠）内分别测试与年龄相关的DEGs，同时使用肌肉类型和供体/小鼠性别作为生物学协变量，以及10x化学和数据模态（scRNA-seq或snRNA-seq）作为技术协变量。
To compare aging DEGs between species, we calculated the Jaccard index, which denotes the ratio between the size of the intersection of two sets and their union. Next, we used the clusterProfiler package to identify which KEGG pathways are enriched among aging DEGs in every species. We used genes that have above 0 or below 0 , LTSR greater than 0.9 and were expressed in at least in of cells in the relevant cell type to perform enrichment analysis. After selecting significantly enriched pathways, we scored them based on the number of cell types that showed simultaneous human and mouse age-related enrichment (for ) and depletion (for in them. 为了比较不同物种之间的衰老DEGs，我们计算了Jaccard指数，该指数表示两个集合的交集与并集之间的比率。接下来，我们使用clusterProfiler包来确定每个物种中富集在衰老DEGs中的KEGG通路。我们使用具有大于0或小于0的基因，LTSR大于0.9，并且在相关细胞类型的至少 个细胞中表达的基因进行富集分析。在选择显著富集的通路后，我们根据同时显示人类和小鼠年龄相关富集（对于 ）和贫化（对于 ）的细胞类型数量对它们进行评分。
Cell-cell communication analysis. We used the CellPhoneDB algorithm (version 3 and database to obtain the list of all possible cell type pairs and receptors that can interact through the following ligands: CCL2, CCL3, CCL4 and CXCL8. Normalized count data and broad cell type assignment were used as input. Only receptors and ligands expressed in more than of the cells in the specific cluster were considered to indicate relevant interactions. Next, we used the 'igraph' R package to visualize all possible emitter cell types, producing the ligands and receptor cell types, expressing the receptor. We also indicated if the expression of ligand or receptor has changed with age (according to aging DEG analysis). 细胞间通讯分析。我们使用了CellPhoneDB算法（版本3）和数据库，以获取所有可能的细胞类型对和能够通过以下配体相互作用的受体列表：CCL2、CCL3、CCL4和CXCL8。标准化计数数据和广泛的细胞类型分配作为输入。仅考虑在特定簇中超过 细胞中表达的受体和配体，以指示相关的相互作用。接下来，我们使用“igraph” R软件包来可视化所有可能的发射细胞类型，显示配体和表达受体的受体细胞类型。我们还指出了配体或受体的表达是否随年龄变化（根据衰老DEG分析）。
Automatic image segmentation and analyses. For unbiased MF subtyping and age-associated comparisons (Fig. 4b,c, Extended Data Fig. 6c-e and Supplementary Table 5), we developed an automated MF segmentation and image analysis workflow. In brief, for each multi-channel image exported from the microscope, all channels targeting non-nuclei channels were first max-projected and then sequentially processed through Gaussian blurring, gamma adjustment and . After that, MF segmentation was performed using a deep-learning-based object segmentation algorithm, Cellpose . To classify MFs into different fast and slow subtypes, we manually trained an object classifier using the object classification workflow in ilastik . This classifier also returns metrics that describe the properties of MFs that were later exported for downstream analysis. To ensure trivial MF quantification, all MFs located at the image border were discarded. Nuclei surrounding all MFs were also detected with Cellpose using a different set of parameters in both pre-processing and segmentation steps (see Extended Data Fig. 6a,b for details). 自动图像分割和分析。为了进行无偏的MF亚型划分和与年龄相关的比较（图4b、c，扩展数据图6c-e和补充表5），我们开发了一个自动化的MF分割和图像分析工作流程。简而言之，对于从显微镜导出的每个多通道图像，首先对所有针对非核通道的通道进行最大投影，然后依次通过高斯模糊、伽马调整和 进行处理。之后，使用基于深度学习的对象分割算法Cellpose 执行MF分割。为了将MF分类为不同的快速和慢速亚型，我们使用ilastik中的对象分类工作流程手动训练了一个对象分类器。该分类器还返回描述后续分析所需的MF属性的度量标准。为了确保微不足道的MF定量，所有位于图像边界的MF都被丢弃。还使用不同的参数在预处理和分割步骤中使用Cellpose检测所有MF周围的细胞核（详见扩展数据图6a、b）。
Statistics and reproducibility. This study was designed as a case-control to study the effect of age on the abundance and gene expression across different populations in human skeletal muscle. No statistical methods were used to pre-determine sample sizes, but our sample sizes are similar to those reported in previous aging single-cell works . Randomization was not performed as it is not applicable to this study, and data collection and analysis were not blinded to the conditions of the experiments. 统计学和可重复性。本研究设计为一项病例对照研究，旨在研究年龄对人类骨骼肌不同人群中丰度和基因表达的影响。未使用统计方法预先确定样本量，但我们的样本量与先前关于衰老单细胞研究中报告的相似。由于不适用于本研究，因此未进行随机化，数据收集和分析未对实验条件进行盲化处理。
Cell type abundance was modeled as counts, and either linear mixed-effect Poisson model assuming Poisson distribution (Fig. 1d) or Milo assuming negative binomial distribution was used to quantify change over age (Figs. 2c, 3e and 5a-c). For comparison, cell type abundances were also modeled as proportions. The difference between young and aged was tested using non-parametric Mann-WhitneyWilcoxon test (Extended Data Figs. 1e-g, 2a-c, , f and 5c). Single-cell differential expression was performed using Bayesian linear mixed-effect model, assuming negative binomial distribution of gene expression counts in the cells, and significance was defined using LTSR. Group comparison in validation experiments was performed using either unpaired two-tailed -test or one-way ANOVA, and data distribution was assumed to be normal, but this was not formally tested. 细胞类型丰度被建模为计数，使用线性混合效应泊松模型假设泊松分布（图1d），或使用Milo假设负二项分布来量化随年龄变化（图2c，3e和5a-c）。为了比较，细胞类型丰度也被建模为比例。年轻和年老之间的差异使用非参数Mann-Whitney-Wilcoxon检验进行测试（扩展数据图1e-g，2a-c，f和5c）。单细胞差异表达使用贝叶斯线性混合效应模型进行，假设细胞中基因表达计数的负二项分布，并使用LTSR定义显著性。验证实验中的组比较使用未配对的双尾t检验或单因素方差分析进行，数据分布被假定为正态，但这并没有得到正式检验。
Some data points were excluded from the following analyses. Rare or donor-biased cell types were excluded from cell type abundance analysis in Fig. 1d (see legend for more details); donor 343B was considered an outlier for myonuclei analyses in Fig. 3e due to demonstrating abnormally high abundance of II-OTU population; and donors 502B, and were excluded from analysis in Fig. due to the fact that more than of myonuclei in those donors could not be confidently assigned to a particular subtype (see Fig. 4 c legend). 一些数据点被排除在以下分析之外。 在图 1d 中排除了罕见或供体偏向的细胞类型以进行细胞类型丰度分析（有关详情，请参阅图例）; 由于显示出异常高的 II-OTU 人口丰度，捐赠者 343B 被认为是图 3e 中肌核分析的异常值; 捐赠者 502B，
Reporting summary 报告摘要
Further information on research design is available in the Nature Portfolio Reporting Summary linked to this article. 关于研究设计的更多信息，请参阅与本文相关的自然出版社报告摘要。
Data availability 数据可用性
The processed data objects generated within this study are available for browsing at https://www.muscleageingcellatlas.org. Raw sequencing data for the newly generated libraries have been deposited to ArrayExpress (E-MTAB-13874). The publicly available human skeletal muscle single-nuclei and single-cell datasets were downloaded from GSE167186, GSE143704 and DRYAD (https://doi.org/10.7272/ Q65X273X), and mouse datasets were obtained from GSE110878, GSE138707, GSE134540, GSE143476, GSE149590 and GSE142480 repositories. All other data supporting the findings of this study are available from the corresponding authors upon reasonable request. 本研究生成的处理过的数据对象可在 https://www.muscleageingcellatlas.org 进行浏览。新生成的文库的原始测序数据已存储在 ArrayExpress（E-MTAB-13874）中。公开可用的人类骨骼肌单核和单细胞数据集已从 GSE167186、GSE143704 和 DRYAD（https://doi.org/10.7272/Q65X273X）下载，而小鼠数据集则来自 GSE110878、GSE138707、GSE134540、GSE143476、GSE149590 和 GSE142480 存储库。本研究结果支持的所有其他数据可根据合理请求从相应作者处获得。
Code availability 代码可用性
All custom notebooks and scripts used in this study have been deposited to the https://github.com/Teichlab/SKM_ageing_atlas repository. 本研究中使用的所有自定义笔记本和脚本已存储在 https://github.com/Teichlab/SKM_ageing_atlas 存储库中。
References 参考文献
Tidball, J. G. Regulation of muscle growth and regeneration by the immune system. Nat. Rev. Immunol. 17, 165-178 (2017). Tidball, J. G. 免疫系统对肌肉生长和再生的调节。Nat. Rev. Immunol. 17, 165-178 (2017)。
Pedersen, B. K. & Febbraio, M. A. Muscles, exercise and obesity: skeletal muscle as a secretory organ. Nat. Rev. Endocrinol. 8, 457-465 (2012). Pedersen, B. K. & Febbraio, M. A. 肌肉，运动和肥胖：骨骼肌作为一个分泌器官。自然。内分泌学评论。8，457-465（2012 年）。
Hargreaves, M. & Spriet, L. L. Skeletal muscle energy metabolism during exercise. Nat. Metab. 2, 817-828 (2020). Hargreaves, M. & Spriet, L. L. 运动期间骨骼肌能量代谢。自然。代谢。2，817-828（2020 年）。
Siparsky, P. N., Kirkendall, D. T. & Garrett, W. E. Jr. Muscle changes in aging: understanding sarcopenia. Sports Health 6, 36-40 (2014). Siparsky, P. N., Kirkendall, D. T. & Garrett, W. E. Jr. 老化中的肌肉变化：理解肌肉衰退。运动健康。6，36-40（2014 年）。
Nilwik, R. et al. The decline in skeletal muscle mass with aging is mainly attributed to a reduction in type II muscle fiber size. Exp. Gerontol. 48, 492-498 (2013). Nilwik, R. 等。随着年龄增长骨骼肌质量下降主要归因于 II 型肌纤维尺寸的减少。Exp. Gerontol. 48, 492-498 (2013)。
Gopinath, S. D. & Rando, T. A. Stem cell review series: aging of the skeletal muscle stem cell niche. Aging Cell 7, 590-598 (2008). Gopinath, S. D. & Rando, T. A. 干细胞评估系列：骨骼肌干细胞微环境的老化。Aging Cell 7, 590-598 (2008)。
Chini, C. C. S. et al. CD38 ecto-enzyme in immune cells is induced during aging and regulates NAD and NMN levels. Nat. Metab. 2, 1284-1304 (2020). Chini, C. C. S. 等人。免疫细胞中的 CD38 外刊酶在衰老过程中被诱导，并调节 NAD 和 NMN 水平。自然代谢。2, 1284-1304 (2020)。
Kuswanto, W. et al. Poor repair of skeletal muscle in aging mice reflects a defect in local, interleukin-33-dependent accumulation of regulatory T cells. Immunity 44, 355-367 (2016). Kuswanto, W. 等人。老年小鼠骨骼肌修复不良反映出局部、依赖白细胞介素-33 的调节性 T 细胞积累缺陷。免疫。44, 355-367 (2016)。
Larsson, L. et al. Sarcopenia: aging-related loss of muscle mass and function. Physiol. Rev. 99, 427-511 (2019). Larsson, L. 等人。肌肉衰老：与年龄相关的肌肉质量和功能丧失。生理学评论。99, 427-511 (2019)。
Zhang, H. et al. repletion improves mitochondrial and stem cell function and enhances life span in mice. Science 1436-1443 (2016). 张，H. 等。 膳食< b0>重复改善线粒体和干细胞功能，并延长小鼠寿命。 《科学》 1436-1443（2016 年）。
Barruet, E. et al. Functionally heterogeneous human satellite cells identified by single cell RNA sequencing. eLife 9, e51576 (2020). Barruet, E. 等人。 通过单细胞 RNA 测序确定功能多元的人类卫星细胞。 《eLife》9，e51576（2020 年）。
De Micheli, A. J. et al. Single-cell analysis of the muscle stem cell hierarchy identifies heterotypic communication signals involved in skeletal muscle regeneration. Cell Rep. 30, 3583-3595 (2020), De Micheli, A. J. 等人。 肌肉干细胞层级的单细胞分析确定了参与骨骼肌再生的异质型通信信号。 《细胞报》30，3583-3595（2020 年）。
De Micheli, A. J., Spector, J. A., Elemento, O. & Cosgrove, B. D. A reference single-cell transcriptomic atlas of human skeletal muscle tissue reveals bifurcated muscle stem cell populations. Skelet. Muscle 10, 19 (2020). De Micheli, A. J., Spector, J. A., Elemento, O. & Cosgrove, B. D. 人类骨骼肌组织的参考单细胞转录组图谱揭示了分叉的肌肉干细胞群体。骨骼肌 10, 19 (2020)。
McKellar, D. W. et al. Large-scale integration of single-cell transcriptomic data captures transitional progenitor states in mouse skeletal muscle regeneration. Commun. Biol. 4, 1280 (2021). McKellar, D. W. 等人。单细胞转录组数据的大规模整合捕获了小鼠骨骼肌再生中的过渡祖细胞状态。Commun. Biol. 4, 1280 (2021)。
Rubenstein, A. B. et al. Single-cell transcriptional profiles in human skeletal muscle. Sci. Rep. 10, 229 (2020). Rubenstein, A. B. 等人。人类骨骼肌的单细胞转录谱。Sci. Rep. 10, 229 (2020)。
Dos Santos, M. et al. Single-nucleus RNA-seq and FISH identify coordinated transcriptional activity in mammalian myofibers. Nat. Commun. 11, 5102 (2020) Dos Santos, M. 等。单核 RNA 测序和 FISH 鉴定哺乳动物肌纤维中的协调转录活动。Nat. Commun. 11, 5102 (2020)
Kim, M. et al. Single-nucleus transcriptomics reveals functional compartmentalization in syncytial skeletal muscle cells. Nat. Commun. 11, 6375 (2020) Kim, M. 等。单核转录组学揭示合胞体骨骼肌细胞中的功能区隔。Nat. Commun. 11, 6375 (2020)
Orchard, P. et al. Human and rat skeletal muscle single-nuclei multi-omic integrative analyses nominate causal cell types, regulatory elements, and SNPs for complex traits. Genome Res. 31, 2258-2275 (2021). Orchard, P. 等。人类和大鼠骨骼肌单核多组学综合分析提名复杂特征的因果细胞类型、调节元件和 SNPs。Genome Res. 31, 2258-2275 (2021)
Perez, K. et al. Single nuclei profiling identifies cell specific markers of skeletal muscle aging, frailty, and senescence. Aging (Albany NY) 14, 9393-9422 (2022). Perez, K. 等。单核分析确定骨骼肌衰老、衰弱和衰老的细胞特异性标记。老化（纽约州奥尔巴尼）14，9393-9422（2022 年）。
Petrany, M. J. et al. Single-nucleus RNA-seq identifies transcriptional heterogeneity in multinucleated skeletal myofibers. Nat. Commun. 11, 6374 (2020). Petrany, M. J. 等。单核 RNA 测序确定多核骨骼肌肌纤维的转录异质性。自然通讯。11，6374（2020 年）。
Gayoso, A. et al. A Python library for probabilistic analysis of single-cell omics data. Nat. Biotechnol. 40, 163-166 (2022). Gayoso, A. 等。用于单细胞组学数据概率分析的 Python 库。自然生物技术。40，163-166（2022 年）。
Dulken, B. W. et al. Single-cell analysis reveals cell infiltration in old neurogenic niches. Nature 571, 205-210 (2019). Dulken, B. W.等人。单细胞分析揭示老年神经发生区的免疫细胞浸润。自然 571, 205-210 (2019)。
Schaum, N. et al. Ageing hallmarks exhibit organ-specific temporal signatures. Nature 583, 596-602 (2020). Schaum, N.等人。衰老标志表现出器官特异性的时间特征。自然 583, 596-602 (2020)。
Mangiola, S. et al. Whole body cell map tracks tissue-specific immune cell accumulation and plasticity loss through ageing. Preprint at bioRxiv https://doi.org/10.1101/2023.06.08.542671 (2023). Mangiola, S.等人。全身细胞图追踪组织特异性免疫细胞在衰老过程中的积累和可塑性丧失。在 bioRxiv 上的预印本 https://doi.org/10.1101/2023.06.08.542671 (2023)。
Tintignac, L. A., Brenner, H. R. & Ruegg, M. A. Mechanisms regulating neuromuscular junction development and function and causes of muscle wasting. Physiol. Rev. 95, 809-852 (2015). Tintignac, L. A., Brenner, H. R. & Ruegg, M. A. 调节神经肌肉连接发育和功能以及肌肉消耗原因的机制。生理学评论。95, 809-852 (2015)。
Ungvari, Z., Tarantini, S., Donato, A. J., Galvan, V. & Csiszar, A. Mechanisms of vascular aging. Circ. Res. 123, 849-867 (2018). Ungvari, Z., Tarantini, S., Donato, A. J., Galvan, V. & Csiszar, A. 血管老化机制。环。研究。123, 849-867 (2018)。
Tabula Sapiens Consortium et al. The Tabula Sapiens: a multiple-organ, single-cell transcriptomic atlas of humans. Science 376, eabl4896 (2022). Tabula Sapiens Consortium 等人。 Tabula Sapiens: 人类多器官单细胞转录图谱。科学 376, eabl4896 (2022)。
Dell'Orso, S. et al. Single cell analysis of adult mouse skeletal muscle stem cells in homeostatic and regenerative conditions. Development 146, dev174177 (2019). Dell'Orso, S.等。成年小鼠骨骼肌干细胞在稳态和再生条件下的单细胞分析。Development 146, dev174177 (2019)。
Machado, L. et al. In situ fixation redefines quiescence and early activation of skeletal muscle stem cells. Cell Rep. 21, 1982-1993 (2017). Machado, L.等。原位固定重新定义了骨骼肌干细胞的静止和早期激活。Cell Rep. 21, 1982-1993 (2017)。
Sharifi, S., da Costa, H. F. R. & Bierhoff, H. The circuitry between ribosome biogenesis and translation in stem cell function and ageing. Mech. Ageing Dev. 189, 111282 (2020). Sharifi, S.，da Costa, H. F. R.和 Bierhoff, H.核糖体生物合成与干细胞功能和衰老中的翻译之间的电路。Mech. Ageing Dev. 189, 111282 (2020)。
Giordani, L. et al. High-dimensional single-cell cartography reveals novel skeletal muscle-resident cell populations. Mol. Cell 74, 609-621 (2019). Giordani, L. 等。高维度单细胞制图揭示了新的骨骼肌驻留细胞群体。 Mol. Cell 74, 609-621 (2019)。
Kimmel, J. C., Hwang, A. B., Scaramozza, A., Marshall, W. F. & Brack, A. S. Aging induces aberrant state transition kinetics in murine muscle stem cells. Development 147, dev183855 (2020). Kimmel, J. C.，Hwang, A. B.，Scaramozza, A.，Marshall, W. F.和 Brack, A. S.老化导致小鼠肌肉干细胞状态转变动力学异常。 Development 147, dev183855 (2020)。
Li, H. et al. Muscle-secreted granulocyte colony-stimulating factor functions as metabolic niche factor ameliorating loss of muscle stem cells in aged mice. EMBO J. 38, e102154 (2019). Li, H. 等。肌肉分泌的粒细胞集落刺激因子作为代谢生态位因子，改善老年小鼠肌肉干细胞的丧失。 EMBO J. 38, e102154 (2019)。
Tabula Muris Consortium. A single-cell transcriptomic atlas characterizes ageing tissues in the mouse. Nature (2020). Tabula Muris Consortium。一项单细胞转录组图谱描绘了老化小鼠组织。自然 (2020)。
Dann, E., Henderson, N. C., Teichmann, S. A., Morgan, M. D. & Marioni, J. C. Differential abundance testing on single-cell data using -nearest neighbor graphs. Nat. Biotechnol. 40, 245-253 (2022). Dann, E.，Henderson, N. C.，Teichmann, S. A.，Morgan, M. D.和 Marioni, J. C.。使用 -最近邻图对单细胞数据进行差异丰度测试。自然生物技术。40, 245-253 (2022)。
Alexander, M. S. et al. CD82 is a marker for prospective isolation of human muscle satellite cells and is linked to muscular dystrophies. Cell Stem Cell 19, 800-807 (2016). Alexander, M. S.等。CD82 是人类肌肉卫星细胞潜在分离的标记物，并与肌肉萎缩症相关。细胞干细胞。19, 800-807 (2016)。
Relaix, F. et al. Perspectives on skeletal muscle stem cells. Nat. Commun. 12, 692 (2021). Relaix, F.等。骨骼肌干细胞的展望。Nat. Commun. 12, 692 (2021)。
Lessard, F. et al. Senescence-associated ribosome biogenesis defects contributes to cell cycle arrest through the Rb pathway. Nat. Cell Biol. 20, 789-799 (2018). Lessard, F.等。细胞衰老相关核糖体生物合成缺陷通过 Rb 途径导致细胞周期停滞。Nat. Cell Biol. 20, 789-799 (2018)。
Saul, D. et al. A new gene set identifies senescent cells and predicts senescence-associated pathways across tissues. Nat. Commun. 13, 4827 (2022). Saul, D.等。一个新的基因集识别衰老细胞并预测跨组织的衰老相关途径。Nat. Commun. 13, 4827 (2022)。
Bengal, E., Perdiguero, E., Serrano, A. L. & Munoz-Canoves, P. Rejuvenating stem cells to restore muscle regeneration in aging. F1000Res. 6, 76 (2017). 孟加拉，E.，佩迪格罗，E.，塞拉诺，A. L.和穆尼奥斯-卡诺维斯，P. 使干细胞恢复肌肉再生以恢复老化。F1000Res. 6，76（2017 年）。
Thompson, W. L. & Van Eldik, L. J. Inflammatory cytokines stimulate the chemokines CCL2/MCP-1 and CCL7/MCP-3 through NFKB and MAPK dependent pathways in rat astrocytes [corrected]. Brain Res. 1287, 47-57 (2009). 汤普森，W. L.和范埃尔迪克，L. J. 炎性细胞因子通过 NFKB 和 MAPK 依赖途径刺激大鼠星形胶质细胞的趋化因子 CCL2/MCP-1 和 CCL7/MCP-3 [更正]。脑研究。1287，47-57（2009 年）。
Aibar, S. et al. SCENIC: single-cell regulatory network inference and clustering. Nat. Methods 14, 1083-1086 (2017). Aibar，S.等。 SCENIC：单细胞调控网络推断和聚类。自然。方法 14，1083-1086（2017 年）。
Van de Sande, B. et al. A scalable SCENIC workflow for single-cell gene regulatory network analysis. Nat. Protoc. 15, 2247-2276 (2020). Van de Sande, B. 等人。单细胞基因调控网络分析的可扩展 SCENIC 工作流程。自然。Protoc. 15, 2247-2276 (2020)。
Petit, C. S., Roczniak-Ferguson, A. & Ferguson, S. M. Recruitment of folliculin to lysosomes supports the amino acid-dependent activation of Rag GTPases. J. Cell Biol. 202, 1107-1122 (2013) Petit, C. S., Roczniak-Ferguson, A. & Ferguson, S. M. 将 folliculin 招募到溶酶体，支持氨基酸依赖的 Rag GTPases 激活。J. 细胞生物。202, 1107-1122 (2013)。
Tsun, Z. Y. et al. The folliculin tumor suppressor is a GAP for the RagC/D GTPases that signal amino acid levels to mTORC1. Mol. Cell 52, 495-505 (2013). Tsun, Z. Y. 等人。folliculin 肿瘤抑制因子是 RagC/D GTPases 的 GAP，用于向 mTORC1 信号氨基酸水平。分子细胞 52, 495-505 (2013)。
Ratnayake, D. et al. Macrophages provide a transient muscle stem cell niche via NAMPT secretion. Nature 591, 281-287 (2021). Ratnayake, D. 等人。巨噬细胞通过 NAMPT 分泌提供短暂的肌肉干细胞生长环境。自然 591, 281-287 (2021)。
Moresi, V., Adamo, S. & Berghella, L. The JAK/STAT pathway in skeletal muscle pathophysiology. Front. Physiol. 10, 500 (2019). Moresi, V.，Adamo, S.和 Berghella, L. 骨骼肌病生理学中的 JAK/STAT 途径。生理学前沿 10, 500 (2019)。
Caldow, M. K., Steinberg, G. R. & Cameron-Smith, D. Impact of SOCS3 overexpression on human skeletal muscle development in vitro. Cytokine 55, 104-109 (2011). Caldow, M. K.，Steinberg, G. R.和 Cameron-Smith, D. SOCS3 过表达对体外人类骨骼肌发育的影响。细胞因子 55, 104-109 (2011)。
Sun, L. et al. JAK1-STAT1-STAT3, a key pathway promoting proliferation and preventing premature differentiation of myoblasts. J. Cell Biol. 179, 129-138 (2007). Sun, L.等。JAK1-STAT1-STAT3，一个促进肌母细胞增殖和防止过早分化的关键途径。J. Cell Biol. 179，129-138（2007 年）。
Sato, S., Ogura, Y. & Kumar, A. TWEAK/Fn14 signaling axis mediates skeletal muscle atrophy and metabolic dysfunction. Front. Immunol. 5, 18 (2014). Sato, S.，Ogura, Y.和 Kumar, A. TWEAK/Fn14 信号轴介导骨骼肌萎缩和代谢功能障碍。Front. Immunol. 5，18（2014 年）。
Li, L., Xiong, W. C. & Mei, L. Neuromuscular junction formation, aging, and disorders. Annu. Rev. Physiol. 80, 159-188 (2018). Li, L.，Xiong, W. C.和 Mei, L. 神经肌肉连接形成，衰老和疾病。Annu. Rev. Physiol. 80，159-188（2018 年）。
Bonanomi, D. et al. Ret is a multifunctional coreceptor that integrates diffusible- and contact-axon guidance signals. Cell 148, 568-582 (2012) Bonanomi, D. 等。Ret 是一个集成可扩散和接触轴突引导信号的多功能辅助受体。Cell 148, 568-582 (2012)
Hallock, P. T., Chin, S., Blais, S., Neubert, T. A. & Glass, D. J. Sorbs1 and -2 interact with and are required for acetylcholine receptor cluster formation. Mol. Cell. Biol. 36, 262-270 (2016). Hallock, P. T.，Chin, S.，Blais, S.，Neubert, T. A. 和 Glass, D. J. Sorbs1 和 -2 与 交互，并且对乙酰胆碱受体簇形成是必需的。Mol. Cell. Biol. 36, 262-270 (2016)。
Hippenmeyer, S., Huber, R. M., Ladle, D. R., Murphy, K. & Arber, S. ETS transcription factor Erm controls subsynaptic gene expression in skeletal muscles. Neuron (2007). Hippenmeyer, S.，Huber, R. M.，Ladle, D. R.，Murphy, K. 和 Arber, S. ETS 转录因子 Erm 控制骨骼肌中亚突触基因表达。Neuron (2007)。
Kummer, T. T., Misgeld, T., Lichtman, J. W. & Sanes, J. R. Nerve-independent formation of a topologically complex postsynaptic apparatus. J. Cell Biol. 164, 1077-1087 (2004). Kummer, T. T., Misgeld, T., Lichtman, J. W. & Sanes, J. R. 神经独立形成拓扑复杂的突触后装置。J. Cell Biol. 164, 1077-1087 (2004)。
Lin, H. et al. Decoding the transcriptome of denervated muscle at single-nucleus resolution. J. Cachexia Sarcopenia Muscle 13, 2102-2117 (2022) Lin, H. 等。解码单核分辨率下失神经肌肉的转录组。J. Cachexia Sarcopenia Muscle 13, 2102-2117 (2022)。
Murgia, M. et al. Single muscle fiber proteomics reveals fiber-type-specific features of human muscle aging. Cell Rep. 19, 2396-2409 (2017) Murgia, M. 等。单肌纤维蛋白质组学揭示人类肌肉老化的纤维类型特异特征。Cell Rep. 19, 2396-2409 (2017)。
Schiaffino, S., Rossi, A. C., Smerdu, V., Leinwand, L. A. & Reggiani, C. Developmental myosins: expression patterns and functional significance. Skelet. Muscle 5, 22 (2015). Schiaffino, S., Rossi, A. C., Smerdu, V., Leinwand, L. A. & Reggiani, C. 发育型肌球蛋白：表达模式和功能意义。骨骼肌 5, 22 (2015)。
Webster, C., Silberstein, L., Hays, A. P. & Blau, H. M. Fast muscle fibers are preferentially affected in Duchenne muscular dystrophy. Cell 52, 503-513 (1988). Webster, C., Silberstein, L., Hays, A. P. & Blau, H. M. 杜兴氏肌营养不良患者快速肌纤维受到优先影响。细胞 52, 503-513 (1988)。
Lin, J. R. et al. High-plex immunofluorescence imaging and traditional histology of the same tissue section for discovering image-based biomarkers. Nat. Cancer 4, 1036-1052 (2023). Lin, J. R. 等。同一组织切片的高复合免疫荧光成像和传统组织学，用于发现基于图像的生物标志物。自然癌症 4, 1036-1052 (2023)。
Krasniewski, L. K. et al. Single-cell analysis of skeletal muscle macrophages reveals age-associated functional subpopulations. elife 11, e77974 (2022). Krasniewski, L. K. 等。骨骼肌巨噬细胞的单细胞分析揭示了与年龄相关的功能亚群。elife 11, e77974（2022 年）。
Fuertes-Alvarez, S. & Izeta, A. Terminal Schwann cell aging: implications for age-associated neuromuscular dysfunction. Aging Dis. 12, 494-514 (2021). Fuertes-Alvarez, S. & Izeta, A. 末梢施万细胞老化：对与年龄相关的神经肌肉功能障碍的影响。老化与疾病。12, 494-514（2021 年）。
Fukada, K. & Kajiya, K. Age-related structural alterations of skeletal muscles and associated capillaries. Angiogenesis 23, 79-82 (2020) Fukada, K. & Kajiya, K. 骨骼肌及相关毛细血管的年龄相关结构改变。血管生成。23, 79-82（2020 年）.
Hirata, A. et al. Expression profiling of cytokines and related genes in regenerating skeletal muscle after cardiotoxin injection: a role for osteopontin. Am. J. Pathol. 163, 203-215 (2003). Hirata, A. 等。在注射心肌毒素后再生骨骼肌中细胞因子和相关基因的表达分析：骨桥蛋白的作用。美国病理学杂志。163，203-215（2003 年）。
Warren, G. L. et al. Role of CC chemokines in skeletal muscle functional restoration after injury. Am. J. Physiol. Cell Physiol. 286, C1031-C1036 (2004). Warren, G. L. 等。CC 趋化因子在骨骼肌受伤后功能恢复中的作用。美国生理学杂志。286，C1031-C1036（2004 年）。
Efremova, M., Vento-Tormo, M., Teichmann, S. A. & Vento-Tormo, R. CellPhoneDB: inferring cell-cell communication from combined expression of multi-subunit ligand-receptor complexes. Nat. Protoc. 15, 1484-1506 (2020). Efremova, M.，Vento-Tormo, M.，Teichmann, S. A. 和 Vento-Tormo, R. CellPhoneDB：从多亚单位配体-受体复合物的联合表达推断细胞间通信。自然协议。15，1484-1506（2020 年）。
Zhuang, J. et al. Comparison of multi-tissue aging between human and mouse. Sci. Rep. 9, 6220 (2019). Zhuang, J.等。人类和小鼠多组织衰老的比较。Sci. Rep. 9, 6220 (2019)。
Zhang, M. J., Pisco, A. O., Darmanis, S. & Zou, J. Mouse aging cell atlas analysis reveals global and cell type-specific aging signatures. eLife 10, e62293 (2021). Zhang, M. J.，Pisco, A. O.，Darmanis, S.和 Zou, J. 小鼠衰老细胞图谱分析揭示了全球和细胞类型特异性的衰老特征。eLife 10, e62293 (2021)。
Moiseeva, V. et al. Senescence atlas reveals an aged-like inflamed niche that blunts muscle regeneration. Nature 613, 169-178 (2023). Moiseeva, V.等。衰老图谱揭示了一个类似老年的炎症环境，削弱了肌肉再生。Nature 613, 169-178 (2023)。
Forcina, L., Miano, C., Scicchitano, B. M. & Musaro, A. Signals from the niche: insights into the role of IGF-1 and IL-6 in modulating skeletal muscle fibrosis. Cells 8, 232 (2019). 来源
Enge, M. et al. Single-cell analysis of human pancreas reveals transcriptional signatures of aging and somatic mutation patterns. Cell 171, 321-330 (2017). Enge, M.等。对人类胰腺的单细胞分析揭示了衰老的转录特征和体细胞突变模式。《细胞》171，321-330（2017 年）。
Zou, Z. et al. A single-cell transcriptomic atlas of human skin aging. Dev. Cell 56, 383-397 (2021). Zou, Z.等。人类皮肤衰老的单细胞转录图谱。《发育细胞》56，383-397（2021 年）。
Yi, W. et al. A single-cell transcriptome atlas of the aging human and macaque retina. Natl Sci. Rev. 8, nwaa179 (2021). Yi, W. 等人。人类和猕猴视网膜衰老的单细胞转录组图谱。国家科学评论。8, nwaa179 (2021)。
Lee, N. Y. S., Li, M., Ang, K. S. & Chen, J. Establishing a human bone marrow single cell reference atlas to study ageing and diseases. Front. Immunol. 14, 1127879 (2023). Lee, N. Y. S., Li, M., Ang, K. S. & Chen, J. 建立人类骨髓单细胞参考图谱以研究衰老和疾病。免疫学前沿。14, 1127879 (2023)。
Hepple, R. T. & Rice, C. L. Innervation and neuromuscular control in ageing skeletal muscle. J. Physiol. 594, 1965-1978 (2016). Hepple, R. T. & Rice, C. L. 衰老骨骼肌的神经支配和神经肌肉控制。生理学杂志。594, 1965-1978 (2016)。
Fleming, S. J. et al. Unsupervised removal of systematic background noise from droplet-based single-cell experiments using CellBender. Nat. Methods 20, 1323-1335 (2023). 弗莱明，S. J. 等。使用 CellBender 从液滴单细胞实验中无监督地去除系统性背景噪音。Nat. Methods 20, 1323-1335 (2023)。
Wolock, S. L., Lopez, R. & Klein, A. M. Scrublet: computational identification of cell doublets in single-cell transcriptomic data. Cell Syst. 8, 281-291 (2019). Wolock, S. L., Lopez, R. & Klein, A. M. Scrublet: 单细胞转录组数据中细胞双体的计算识别。Cell Syst. 8, 281-291 (2019)。
Wolf, F. A., Angerer, P. & Theis, F. J. SCANPY: large-scale single-cell gene expression data analysis. Genome Biol. 19, 15 (2018). Wolf, F. A., Angerer, P. & Theis, F. J. SCANPY: 大规模单细胞基因表达数据分析。Genome Biol. 19, 15 (2018)。
Raudvere, U. et al. g:Profiler: a web server for functional enrichment analysis and conversions of gene lists (2019 update). Nucleic Acids Res. 47, W191-W198 (2019). Raudvere, U. 等。g:Profiler: 用于功能富集分析和基因列表转换的网络服务器（2019 更新）。Nucleic Acids Res. 47, W191-W198 (2019)。
Zhou, Y. et al. Metascape provides a biologist-oriented resource for the analysis of systems-level datasets. Nat. Commun. 10, 1523 (2019). 周，Y.等。Metascape 为生物学家提供了一个面向系统级数据集分析的资源。Nat. Commun. 10, 1523 (2019)。
Trapnell, C. et al. The dynamics and regulators of cell fate decisions are revealed by pseudotemporal ordering of single cells. Nat. Biotechnol. 32, 381-386 (2014). Trapnell，C.等。通过单细胞的伪时间排序揭示了细胞命运决策的动态和调节因子。Nat. Biotechnol. 32, 381-386 (2014)。
Yoshida, M. et al. Local and systemic responses to SARS-CoV-2 infection in children and adults. Nature 602, 321-327 (2022). 吉田，M.等。儿童和成人对 SARS-CoV-2 感染的局部和全身反应。Nature 602, 321-327 (2022)。
Young, A. M. H. et al. A map of transcriptional heterogeneity and regulatory variation in human microglia. Nat. Genet. 53, 861-868 (2021). Young, A. M. H. 等。人类小胶质细胞转录异质性和调控变异图谱。自然遗传学。53，861-868（2021 年）。
van der Walt, S. et al. scikit-image: image processing in Python PeerJ 2, e453 (2014). van der Walt, S. 等。scikit-image：Python 中的图像处理。PeerJ 2，e453（2014 年）。
Stringer, C., Wang, T., Michaelos, M. & Pachitariu, M. Cellpose: a generalist algorithm for cellular segmentation. Nat. Methods 18, 100-106 (2021). Stringer, C.，Wang, T.，Michaelos, M. & Pachitariu, M. Cellpose：细胞分割的通用算法。自然方法。18，100-106（2021 年）。
Berg, S. et al. ilastik: interactive machine learning for (bio)image analysis. Nat. Methods 16, 1226-1232 (2019). Berg, S. 等。ilastik: 交互式机器学习用于（生物）图像分析。Nat。Methods 16，1226-1232（2019）。
Acknowledgements 致谢
The human tissue material was provided by the Collaborative Biorepository for Translational Medicine (https://www.cbtm.group. cam.ac.uk/). We are grateful to the donors and their families. We acknowledge . Suo for assistance with muscle single-cell processing; C. Domínguez Conde and K. Tuong as well as the immune subgroup of Teichlab for advice on immune cells; A. Maartens for proofreading this manuscript; . Usher for graphical illustrations of Fig. 30; and members of the sequencing pipelines at Sanger for sequencing samples and aligning the data. We acknowledge that part of the illustrations in Figs. 1, 2 and 4 were created with BioRender. com. We also thank all members of Teichlab and the Zhang group for discussions. We are thankful to RareCyte employees (T. George and D. Crawton) for their support with experiments and image processing of 15-plex staining data. 人类组织材料由转化医学协作生物资料库提供（https://www.cbtm.group. cam.ac.uk/）。我们感谢捐赠者及其家人。我们感谢 .苏某在肌肉单细胞处理方面的协助；C.多明格斯·孔德和 K.村作为 Teichlab 免疫细胞小组的成员对免疫细胞提供建议；A.马滕斯为校对本文手稿； .乌舍为第 30 图的图形插图；以及 Sanger 测序管道的成员用于测序样本并对齐数据。我们承认图 1、2 和 4 部分插图采用 BioRender 创建。我们还感谢 Teichlab 和 Zhang 小组的所有成员进行讨论。我们感谢 RareCyte 员工（T.乔治和 D.克劳顿）对 15-plex 染色数据的实验和图像处理的支持。
This publication is part of the Human Cell Atlas. This work was made possible by a partnership between the Wellcome Sanger Institute and Sun Yat-sen University. The research in H.Z.'s laboratory is supported by the National Key R&D Program (grant number 2019YFA0801703) (to H.Z.); the National Natural Science Foundation of China (grant numbers 31871370 and 32000840 ) (to H.Z. and Y.W.); the Advanced Medical Technology Center of Sun Yat-sen University (K0507007) (to H.Z.); the Natural Science Foundation of Guangdong Province (2021A1515012065) (to Y.W.); and the EMBO Long-Term Fellowship (ALTF744-2017) (to H.Z.). The single-cell atlas was funded in full by Wellcome Human Cell Atlas Strategic Science Support (WT211276/Z/18/Z) (to S.A.T). S.A.T.'s research is funded by the Wellcome Trust (220540/Z/20/A and WT206194) (to S.A.T). This project has received funding from the European Union's Horizon 2020 Research and Innovation Program under Marie-Skłodowska-Curie grant agreement number 101026233 (to J.P.P.). 本出版物是人类细胞图谱项目的一部分。这项工作得以实现，要归功于惠康桑格研究所与中山大学之间的合作伙伴关系。H.Z.实验室的研究得到了国家重点研发计划（批准号 2019YFA0801703）（H.Z.获得）；中国国家自然科学基金（批准号 31871370 和 32000840）（H.Z.和 Y.W.获得）；中山大学先进医疗技术中心（K0507007）（H.Z.获得）；广东省自然科学基金（2021A1515012065）（Y.W.获得）；以及 EMBO 长期奖学金（ALTF744-2017）（H.Z.获得）的支持。单细胞图谱项目得到了惠康人类细胞图谱战略科学支持的全额资助（WT211276/Z/18/Z）（S.A.T.获得）。S.A.T.的研究得到了惠康信托基金（220540/Z/20/A 和 WT206194）（S.A.T.获得）的资助。该项目获得了欧盟的 Horizon 2020 研究与创新计划的资助，Marie-Skłodowska-Curie 授予协议编号 101026233（J.P.P.获得）。
Author contributions 作者贡献
H.Z. and S.A.T. conceived the study and supervised the work. K.T.M. and K.S.P. acquired organ donor tissue, and Z.S. and A.P.X. collected biopsies from patients. Y.W., H.Z., M.D., V.R.K., C.T. and M.P. preserved collected tissue. L.B., H.Z., E.S.F., E.P. and V.R.K. performed single-cell and nuclei processing of the tissues. Y.W. devised in vitro and in vivo validation strategies, established experimental workflows, performed wet lab experiments and interpreted data. X.C. helped Y.W. with tissue pre-processing. Q.G. assisted Y.W. with immunohistochemistry. X.Z. helped with primary cell culture and FACS. X.L. provided FACS instrumental support. K.T. performed RNAscope and immunofluorescence staining experiments. S.P. and N.-J.C. performed multiplex RareCyte staining following the plan from V.R.K. V.R.K., H.Z. and S.A.T. coordinated data analysis across two sites and shaped study direction, with input from Y.W. and T. Liu. V.R.K., V.K., N.K. and N.H. performed statistical method development for the project. Data processing: K.P. and V.R.K. Global analysis: V.R.K., T. Liu and J.P.P. MuSC analysis: T. Liu, Y.W. and V.R.K. Myofiber analysis: V.R.K., Y.W. and T. Liu. Immune analysis: V.R.K. and T. Liu. Vascular analysis: V.R.K., L.Y. and T. Liu. Fibroblast analysis: V.R.K. and T. Liu. Cross-species integration and mouse data analysis: T. Liu and V.R.K. Cross-species aging analysis: V.R.K. Myofiber segmentation and feature extraction: T. Li and O.A.B. Manual image analysis and myofiber type quantification: Y.W. Website and imaging data stitching and upload: M.P. Interpretation of the results: V.R.K., Y.W., H.Z., S.A.T., K.B.M., T. Liu and J.S.L. Manuscript writing: V.R.K., Y.W., H.Z., K.B.M., T. Liu, J.S.L. and S.A.T. H.Z.和S.A.T.构思了这项研究并监督了工作。K.T.M.和K.S.P.获取了器官捐赠者的组织，Z.S.和A.P.X.从患者身上采集了活检组织。Y.W.、H.Z.、M.D.、V.R.K.、C.T.和M.P.保存了采集的组织。L.B.、H.Z.、E.S.F.、E.P.和V.R.K.对组织进行了单细胞和细胞核处理。Y.W.设计了体外和体内验证策略，建立了实验工作流程，进行了湿实验室实验并解释了数据。X.C.帮助Y.W.进行组织预处理。Q.G.协助Y.W.进行免疫组织化学。X.Z.协助进行原代细胞培养和FACS。X.L.提供了FACS仪器支持。K.T.进行了RNAscope和免疫荧光染色实验。S.P.和N.-J.C.根据V.R.K.的计划进行了多重RareCyte染色。V.R.K.、H.Z.和S.A.T.协调了两个地点的数据分析，并塑造了研究方向，得到了Y.W.和T. Liu的意见。V.R.K.、V.K.、N.K.和N.H.为该项目开发了统计方法。数据处理：K.P.和V.R.K. 全球分析：V.R.K.、T. Liu和J.P.P. MuSC分析：T. Liu、Y.W.和V.R.K. 肌纤维分析：V.R.K.、Y.W.和T. Liu。免疫分析：V.R.K. 和 T. 刘。 血管分析：V.R.K.，L.Y. 和 T. 刘。 成纤维细胞分析：V.R.K. 和 T. 刘。 跨物种整合和小鼠数据分析：T. 刘 和 V.R.K. 跨物种衰老分析：V.R.K. 我的纤维分割和特征提取：T. 李 和 O.A.B. 手动图像分析和肌纤维类型定量：Y.W. 网站和成像数据拼接和上传：M.P. 结果解释：V.R.K.，Y.W.，H.Z.，S.A.T.，K.B.M.，T. 刘 和 J.S.L. 手稿撰写：V.R.K.，Y.W.，H.Z.，K.B.M.，T. 刘，J.S.L. 和 S.A.T.
Competing interests 竞争利益
In the past 3 years, S.A.T. has consulted for or been a member of scientific advisory boards at Qiagen, Sanofi, GlaxoSmithKline and ForeSite Labs. She is a co-founder and an equity holder of TransitionBio and EnsoCell and a SAB member of Element Biosciences. She is a part-time employee at GlaxoSmithKline. The remaining authors declare no competing interests. 在过去的 3 年中，S.A.T. 曾就职于或是齐亚根、赛诺菲、葛兰素史克和 ForeSite Labs 的科学咨询委员会成员。她是 TransitionBio 和 EnsoCell 的联合创始人和股权持有人，以及 Element Biosciences 的 SAB 成员。她是葛兰素史克的兼职员工。其余作者声明没有竞争利益。
Supplementary information The online version contains supplementary material available at https://doi.org/10.1038/s43587-024-00613-3. 补充信息 在线版本包含可在 https://doi.org/10.1038/s43587-024-00613-3 获取的补充资料。
Correspondence and requests for materials should be addressed to Kourosh Saeb-Parsy, Sarah A. Teichmann or Hongbo Zhang. 信函和材料请求应寄至 Kourosh Saeb-Parsy、Sarah A. Teichmann 或 Hongbo Zhang。
Peer review information Nature Aging thanks Vittorio Sartorelli and the other, anonymous, reviewer(s) for their contribution to the peer review of this work. 同行评审信息《自然-老化》感谢 Vittorio Sartorelli 和其他匿名审稿人对本研究的同行评审所做的贡献。
Reprints and permissions information is available at www.nature.com/reprints. 重印和权限信息可在 www.nature.com/reprints 获取。
Publisher's note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, 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 changes were made. 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/4.0/. 开放存取 本文章根据知识共享署名 4.0 国际许可协议进行许可，只要您正确给出原作者和来源、提供指向许可协议的链接，同时说明是否有进行了修改，您就可以自由地使用、共享、改编、分发本作品。本文章中的图片或其他第三方资料如不另行说明均包含在文内的知识共享许可协议下，如果在文内标注，除外。如果在文内没包含在知识共享许可协议下的内容且您拟使用的内容并非受法令规定的许可范围或超出法令规定的许可范围，您必须直接向版权拥有者获取许可。要查阅本许可协议的正本，请访问 http://creativecommons.org/licenses/by/4.0/。
(c) The Author(s) 2024 (c) 作者(们) 2024
a
d
Extended Data Fig. See next page for caption. 扩展数据图。请参阅下一页的说明。
f
Extended Data Fig. 1 | Single-cell and single-nucleus skeletal muscle aging atlas. a, Dot plot showing marker genes for major cell types in human skeletal muscle aging atlas. The size of the dot represents the proportion of cells expressing a gene. Colour denotes the scaled expression level. b, c, UMAP visualisation of human aging cell atlas coloured according to age (b) and batch: cells or nuclei (c). d, Full version of the plot in Fig. 1d, taking into account 10x chemistry (Methods). e-g, Box plots illustrating proportions of each cell type in every biological replicate (tissue piece) for scRNA-seq (Cells) vs. snRNA-seq (Nuclei) data (averaged across different technical replicates, 15 nuclei 扩展数据图1｜单细胞和单细胞核骨骼肌老化图谱。a，点图显示人类骨骼肌老化图谱中主要细胞类型的标记基因。点的大小代表表达基因的细胞比例。颜色表示经过缩放的表达水平。b、c，UMAP可视化表明根据年龄（b）和批次：细胞或细胞核（c）对人类老化细胞图谱进行了着色。d，图1d中的完整版本，考虑了10x化学方法（方法）。e-g，箱形图说明单细胞RNA测序（细胞）与单核RNA测序（细胞核）数据中每个生物复制品（组织片段）中每种细胞类型的比例（在不同技术复制品中平均，15个细胞核）。 cells replicates). Samples containing less than 1000 cells were excluded. The box boundary extends from the 1st quantile ( 25 percentile) to the 3 rd quantile ( 75 percentile), horizontal line represents median, 'whiskers' extend to points that lie within 1.5 IQRs of the lower and upper quartile, observations outside this range are considered 'outliers' and marked with a cross. Mann-Whitney-Wilcoxon twosided test with Benjamini-Hochberg correction was used to quantify the change between cells and nuclei, 细胞复制）。排除含有少于 1000 个细胞的样本。箱线图边界从第 1 四分位数（25 百分位数）延伸到第 3 四分位数（75 百分位数），水平线代表中位数，“须”延伸到位于下四分位数和上四分位数的 1.5 IQR 之内的点，超出此范围的观测值被视为“异常值”并用十字标记。使用 Benjamini-Hochberg 校正的 Mann-Whitney-Wilcoxon 双侧检验用于量化细胞和细胞核之间的变化，, see Source Data for exact ，请参阅确切的源数据 values. 值。
Extended Data Fig. 2 |Human cell type dynamics with age and a reference mouse skeletal muscle aging atlas. a-c, Box plots illustrating proportions of each cell type in every biological replicate (averaged across different technical replicates) in young (in violet) . aged (in pink) samples from scRNA-seq ( 5 young aged donors, a) or from snRNA-seq ( 7 young aged biological replicates from 6 young aged donors, . Samples containing less than 1000 cells were excluded. The box boundary extends from the 1st quantile ( 25 percentile) to the 3 rd quantile ( 75 percentile), horizontal line represents median, 'whiskers' extend to points that lie within of the lower and upper quartile, observations outside this range are considered 'outliers' and marked with a cross. Mann-Whitney-Wilcoxon two-sided test with Benjamini-Hochberg correction was used to quantify the change between aged and young, ; see Source Data for exact values. d-g, UMAP plot illustrating 96,529 cells/nuclei from mouse skeletal muscle across age with major cell types (d), age group (young . aged, f) and data type (scRNA-seq . snRNA-seq, g) shown. e, Dot plot of marker genes for each cell type in mouse skeletal muscle aging atlas. The size of the dot represents the proportion of cells expressing a gene. Colour denotes the scaled expression level. 扩展数据图2 | 年龄与参考小鼠骨骼肌老化图谱中的人类细胞类型动态。a-c，箱线图显示每个生物复制品中每种细胞类型的比例（在不同技术复制品中平均），来自scRNA-seq的年轻（紫色）和老化（粉色）样本（5个年轻和老化供体，a），或来自snRNA-seq的年轻和老化生物复制品（6个年轻和老化供体的7个年轻和老化生物复制品，样本中包含少于1000个细胞被排除。箱线图边界从第1四分位数（25百分位数）延伸到第3四分位数（75百分位数），水平线代表中位数，“whiskers”延伸到位于下四分位数和上四分位数之间的点，超出此范围的观察被视为“异常值”并用十字标记。使用Benjamini-Hochberg校正的Mann-Whitney-Wilcoxon双侧检验用于量化老化和年轻之间的变化；请参阅源数据以获取确切值。d-g，UMAP图显示小鼠骨骼肌中的96,529个细胞/核，跨年龄显示主要细胞类型（d），年龄组（年轻和老化，f）和数据类型（scRNA-seq）。 snRNA-seq, g) shown. e, Dot plot of marker genes for each cell type in mouse skeletal muscle aging atlas. The size of the dot represents the proportion of cells expressing a gene. Colour denotes the scaled expression level.
a
h
j
k
Extended Data Fig. 3| Mechanistic insights into human MuSC aging. Extended Data Fig. 3| 人类 MuSC 老化的机制洞察。
, UMAP and marker plots of in-house human and mouse MuSCs (b) integrated together with other publicly available resources (see Methods). The size of the dot represents the proportion of cells expressing a gene. Colour denotes the expression level. d, Gating strategy for FACS-based human MuSC sorting. e, Proportional changes of FACS-sorted human MuSC (4 young aged) in aging. value: unpaired two-tailed -test. , Complete membrane images corresponding to the blots shown in Fig. 2i.g, -Galactosidase ( -Gal) staining (g) and qPCR (h, 3 biological repeats per group) of SASP genes in cultured human primary myoblasts. Experiments in were performed twice with similar results. Scale bar in g: . value: unpaired e , UMAP and marker plots of in-house human and mouse MuSCs (b) integrated together with other publicly available resources (see Methods). The size of the dot represents the proportion of cells expressing a gene. Colour denotes the expression level. d, FACS-based 人类 MuSC 分选的门控策略。e, FACS 分选的人类 MuSC（4 个年轻 老化）在老化过程中的比例变化。 值：未配对的双侧 检验。 , 对培养的人类原代肌母细胞中 SASP 基因的β-半乳糖苷酶（ -Gal）染色（g）和 qPCR（h，每组 3 个生物学重复）。在 中的实验进行了两次，结果相似。g 中的比例尺： 。 值：未配对的 e
C
f
i
two-tailed -test. . i, FACS-based scatter plots ( 4 donors) of TNF+, and Main MuSC. j, FACS-based scatter plots ( 4 young aged donors) of MuSC and their proportion changes in aging. value: unpaired two-tailed -test. , Set of transcription factors which were inferred to regulate CCL2 expression in MuSCs using pySCENIC algorithm. I, Schematic diagram illustrating change in positive (CHUK) and negative (TNFAIP3 and NFKBIZ) regulators in during aging and their putative influence on CCL2 expression. All data presented in bar plots are expressed as mean s.e.m. with individual data points shown. The exact values were shown in the Source Data. 双尾 -测试。 。i，基于 FACS 的散点图（4 个供体）显示 TNF+， 和 Main MuSC。j，基于 FACS 的散点图（4 个年轻 年龄供体）显示 MuSC 以及其在衰老中的比例变化。 值：未配对双尾 -测试。 ，推断调控 CCL2 在 MuSCs 中的表达的转录因子组，使用 pySCENIC 算法。I，示意图说明（CHUK）正调控和（TNFAIP3 和 NFKBIZ）负调控在衰老中的变化及其对 CCL2 表达的可能影响。所有显示在条形图中的数据 表示为均值 标准误差，显示出个体数据点。精确的 值显示在来源数据中。
a
e
h
d
i
Regulon
g
j
to points that lie within 1.5 IQRs of the lower and upper quartile, observations outside this range are considered 'outliers' and marked with a cross. MannWhitney-Wilcoxon two-sided test with Benjamini-Hochberg correction was used to quantify the change between aged and young. , Dot plot showing gene sets enriched in different nuclei populations based on gProfiler over-representation analysis. Colour denotes score, the size of the dot represents of adjusted (adj.) value, significant values highlighted with a red edge. h, Joint immunofluorescence (IF) for MYH7 and RNAscope for TNFRSF12A and OTUD1 (left) as well as RNAscope for MYH2, TNFRSF12A and OTUD1 (right) highlighting I-OTU and II-OTU nuclei populations. i, pySCENIC regulon specificity scores for the NMJ and NMJ accessory population, ordered from highest to lowest (top 10 regulons are labeled).j, Dot plot illustrating regulon activity (accessed using AUCell) for transcription factors specific to NMJ accessory (acc.) us. NMJ populations (relative to their activity in baseline MF-I and MF-II states). 位于下四分位数和上四分位数的 1.5 IQR 范围内的点，被视为“异常值”，并用十字标记。使用 Benjamini-Hochberg 校正的 MannWhitney-Wilcoxon 双侧检验用于量化老年人和年轻人之间的变化。 ，点图显示基于 gProfiler 过表示分析在不同细胞核群体中富集的基因集。颜色表示 分数，点的大小表示已调整（adj.） 值的 ，显著值用红色边框突出显示。h，MYH7 和 TNFRSF12A 以及 OTUD1 的联合免疫荧光（IF）（左）以及 MYH2、TNFRSF12A 和 OTUD1 的 RNAscope（右）突出显示 I-OTU 和 II-OTU 细胞核群体。i，NMJ 和 NMJ 附属群体的 pySCENIC 调控子特异性分数，从高到低排序（前 10 个调控子已标记）。j，点图说明与 NMJ 附属（acc.）相对于基线 MF-I 和 MF-II 状态中的 NMJ 群体的活动相关的转录因子的调控子活动（使用 AUCell 访问）。
a
d
f
g Classical g 经典
denervation Denervation signature from markers Lin et al, 2022 除神经 Denervation signature from markers Lin et al, 2022
Extended Data Fig. 5 | NMJ accessory nuclei and their putative role in promoting AChRs cluster formation. a, IF staining (left) and quantification (right, 26 young aged fields) of human neuromuscular junction structures with age. AChRs: -BTX; motor neuron axon: anti-NEFH; Schwann cell: antiS100B. value: unpaired two-tailed -test. Scale bar: m. . 扩展数据图 5 | NMJ 附属细胞核及其在促进 AChRs 簇形成中的潜在作用。a，年轻和年老人类神经肌肉接头结构的 IF 染色（左）和定量（右，26 个年轻 年老领域）。AChRs： -BTX；运动神经元轴突：抗 NEFH；Schwann 细胞：抗 S100B。 值：未配对的双侧 -检验。比例尺： m。 。
b, UMAP visualisation of quadriceps single-nuclei data pre-processed and re-annotated in-house.c, Boxplots illustrating the proportion of each cell type in 11 aged young patients in b. value: two-sided Mann-Whitney-Wilcoxon test with FDR correction. . The box boundary extends from the 1st quantile ( 25 percentile) to the 3 rd quantile ( 75 percentile), horizontal line represents median, 'whiskers' extend to points that lie within 1.5 IQRs of the lower and upper quartile, observations outside this range are b Perez et al, 2022 (quadriceps aging) b，UMAP 可视化四头肌单核数据 在内部预处理和重新注释。c，箱线图显示 11 个年老 年轻患者中每种细胞类型的比例。 值：双侧 Mann-Whitney-Wilcoxon 检验与 FDR 校正。 。箱线图边界从第 1 四分位数（25 百分位数）延伸到第 3 四分位数（75 百分位数），水平线代表中位数，“须”延伸到位于下四分位数和上四分位数的 1.5 IQR 之内的点，超出此范围的观测值为 b Perez 等人，2022 年（四头肌老化）
e
a
Parameter
Model
(Cellpose)
Object size
(Cellpose)
Cell
probability
(Cellpose)
Gaussian
blurring
Gamma
adjustment
Unsharp
mask
(radius)
Unsharp
mask
(amount)
Myofiber
cyto
90
default
3
0.4
60
1
Nuclei
cyto
10
default
3
0.8
Not
applied
Not
applied
C
d
Extended Data Fig. General ageing changes in myofibers and myonuclei. , Schematic (a) and Exemplary images (b) illustrating automatic image analysis workflow and segmentation parameters. c,d, Beeswarm plots (c) and histograms (d) showing distribution of myofiber cross-sectional area in young ( 7 donors) . aged (4 donors) patients for and myofiber types. Gaussian curve fits for histograms were obtained using the nonlinear regression test. value: unpaired two-tailed -test. . , Paired bar plots showing proportion of rare or unclassified myofiber (e) and 扩展数据图 ：肌纤维和肌核的一般老化变化。 ，示意图（a）和示例图像（b）说明自动图像分析工作流程和分割参数。c，d，蜂群图（c）和直方图（d）显示年轻（7位捐赠者） 和老年（4位捐赠者）患者的肌纤维横截面积分布 和 肌纤维类型。直方图的高斯曲线拟合是通过非线性回归测试获得的。 值：未配对的双尾 -检验。 。 ，成对的条形图显示罕见或未分类的肌纤维比例（e）和
myonuclei types in young . aged individuals. Three snRNA-seq samples that have a high proportion (>75%) of unclassified populations were discarded. , Scanned images of joint IF for MYH7 protein and RNAscope either targeting MYH1 and (g, 1 young vs. 2 aged donors) or MYH7 and MYH2 genes (h, one donor section) on FFPE sections. Insets in highlight exclusive staining of gene and MYH7 gene and protein. Scale bar in g: ; Scale bar in h: . All data in (c-f) are presented as mean s.e.m. with individual data points shown in c, e and . The exact values are shown in the Source Data. 年轻个体中的 myonuclei 类型。丢弃了具有高比例（>75%）未分类人口的三个 snRNA-seq 样本。 ，MYH7 蛋白和 RNAscope 的联合 IF 的扫描图像，目标分别为 MYH1 和 （g，1 个年轻对比 2 个年老供体）或 MYH7 和 MYH2 基因（h，一个供体切片）在 FFPE 切片上。 中的插图突出显示了 基因和 MYH7 基因和蛋白的独特染色。 g 中的比例尺： ；h 中的比例尺： 。 （c-f）中的所有数据均表示为均值 s.e.m.，其中 c、e 和 中显示了各个数据点。源数据中显示了确切的 值。
a
Cases 病例
15 (nucleus) 15（细胞核）
b
C
d
e
Extended Data Fig. 7 | Mechanisms countering fast-twitch myofiber loss in aging. a, Additional examples to Fig. 4f. Joint RNAscope ( genes) and IF (MYH7 protein) highlighting emerging expression of fast-type mRNAs (MYH1 and ) in slow-twitch (MYH7') myofiber nuclei (middle) and cytoplasm (right) in ageing. Scale bar: .b, Dot plot illustrating age-associated changes in the glycolysis and mitochondrial biogenesis gene, PPARGC1A, in myofiber fragments and two main types of myonuclei. Dot size indicates proportion of cells expressing the gene in aged group, colour denotes in gene expression. Significantly up- and down-regulated genes are highlighted with red and blue edges, respectively c. Additional examples to Fig. 4g.Joint RNAscope 在老化过程中对抗快速肌纤维流失的机制。a，与图 4f 的其他示例。联合 RNAscope（ 基因）和 IF（MYH7 蛋白）突出了在老化过程中慢肌（MYH7'）肌纤维细胞核（中）和细胞质（右）中新出现的快速型 mRNA（MYH1 和 ）的表达。比例尺： 。b，点图说明了糖酵解和线粒体生物发生基因 PPARGC1A 在肌纤维碎片和两种主要类型的肌核中的年龄相关变化。点的大小表示在老年组中表达该基因的细胞比例，颜色表示基因表达中的 。显著上调和下调基因分别用红色和蓝色边缘突出显示。c，与图 4g 的其他示例。联合 RNAscope
(MYH1, MYH2 genes) and IF (MYH7 protein) highlighting expression of fasttype IIX mRNAs (MYH1) in fast type Ila （MYH1、MYH2 基因）和 IF（MYH7 蛋白）突出了快速型 IIX mRNA（MYH1）在快速型 Ila 中的表达myofiber nuclei (middle) and cytoplasm (on example on the right) in ageing. Scale bar: . d, UMAP plot (left) shows MuSCs and myofiber populations from scRNA-seq. Dot plot (right) shows their marker genes, which are presented in the order of their appearance in the myogenesis trajectory. Dot size represents the proportion of cells expressing a gene, colour indicates the expression level. e, Reduced dimensional space showing cellular trajectory between MuSC and myofiber inferred by Monocle 2 algorithm coloured according to populations in . 肌纤维核（中间）和细胞质（右侧示例）在衰老中。比例尺： 。d，UMAP 图（左侧）显示来自 scRNA-seq 的 MuSCs 和肌纤维细胞群体。点图（右侧）显示它们的标记基因，这些基因按照它们在肌肉发生轨迹中出现的顺序呈现。点的大小表示表达基因的细胞比例，颜色表示表达水平。e，通过 Monocle 2 算法推断的 MuSC 和肌纤维细胞之间的细胞轨迹的降维空间，根据 中的群体着色。
a Immune cells 免疫细胞
b
Fibroblasts & Schwann cells 成纤维细胞和施万细胞
C
d
f
Extended Data Fig. 8 | Cell type composition of human skeletal muscle microenvironment. a-c, UMAP plots showing annotated subpopulations of immune cells (a), fibroblasts and Schwann cells (b), as well as endothelial and smooth muscle cells (c). Cell type abbreviations are explained in Supplementary Table 10. Cell populations marked in grey contained very few cells or (and) were represented in 1-2 individuals, thus were excluded from further analyses. d-f, Dot plots illustrating marker genes specific for subpopulations of immune cells (d), fibroblasts and Schwann cells (e) as well as vasculature cells (f). Dot size represents the proportion of cells expressing a gene, colour indicates the scaled expression level. 人类骨骼肌微环境的细胞类型组成。a-c，UMAP 图显示免疫细胞（a）、成纤维细胞和施万细胞（b）以及内皮细胞和平滑肌细胞（c）的注释亚群。细胞类型缩写详见附表 10。灰色标记的细胞群中包含极少数细胞或（和）仅在 1-2 个个体中出现，因此被排除在进一步分析之外。d-f，点图展示特定于免疫细胞亚群（d）、成纤维细胞和施万细胞（e）以及血管细胞（f）的标记基因。点的大小表示表达该基因的细胞比例，颜色表示标度化的表达水平。
a
UMAP1
PDGFA
UMAP1
FOXD3
UMAP1
NCAM1
b
d
e
CCL4 interactions CCL4 互动
CCL3 interactions CCL3 互动
Cell type group 细胞类型组
Adipocyte Immune 脂肪细胞免疫
FB MuSC
Myofiber SMC 肌纤维 平滑肌细胞
Schwann Vessel 施万血管
Mesothelium 间皮
Regulation with age 随年龄变化的规定
down - up - none 下 - 上 - 无
Mean expression 平均表达
Extended Data Fig. 9 |Age-associated changes in the cell types within muscle microenvironment. a, UMAP plots highlighting expression of terminal Schwann cell markers in non-myelinating Schwann cell cluster. b, Scanned whole sections showing co-IF of CD3 and Laminin on fresh-frozen human young vs. aged skeletal muscles. Staining was performed on 6 young and 4 aged donors. Scale bar: . c, An exemplary field of view showing 15-plex RareCyte protein staining on aged FFPE skeletal muscle section. Staining is shown for 5 channels at a time together with Hoechst nuclei staining as well as for each channel separately highlighting various cell types and states. Antibodies used and the corresponding cell types they recognize are provided in Supplementary Table 9. Scale bar: , RareCyte staining of CD31 endothelial cells (left) and bar plots (right) illustrating number of cells/field in young (10 fields) . aged (10 fields) donors. Scale bar: . value: unpaired two-tailed -test. . Data were presented as mean s.e.m. with individual data points shown. e, Putative cell-cell interactions in the aged skeletal muscle mediated via CCL3, CCL4 and CXCL8 chemokines produced by microenvironment cells. Emitter (leftmost) and receiver (rightmost) cell types are marked with circles, which are coloured according to a broad cell type group; ligands and receptors are marked with square nodes. Solid edges connect cell types and ligands, or receptors, which they express; thickness of the line is proportional to the mean expression level of the gene in each cell type. Dotted edges connect putative receptors and their ligands. 扩展数据图9 | 肌肉微环境中细胞类型随年龄变化。a，UMAP图突出非髓鞘Schwann细胞群中终末Schwann细胞标记物的表达。b，扫描的整个切片显示新鲜冷冻人类年轻与老年骨骼肌上CD3和Laminin的共IF。染色在6名年轻和4名老年供体上进行。比例尺： 。c，展示15-plex RareCyte蛋白染色在老年FFPE骨骼肌切片上的示例视野。染色每次显示5个通道，同时显示Hoechst核染色以及每个通道单独突出显示各种细胞类型和状态。使用的抗体及其识别的相应细胞类型详见附表9。比例尺： ，RareCyte染色CD31 内皮细胞（左）和条形图（右）说明年轻（10个视野） 。老年（10个视野）供体中每个视野中的细胞数。比例尺： 。 值：未配对的双尾 -检验。 。数据以均值 s.e.m.呈现，显示各个数据点。 在老化的骨骼肌中，通过微环境细胞产生的 CCL3、CCL4 和 CXCL8 趋化因子介导的假定细胞间相互作用。发射器（最左侧）和接收器（最右侧）细胞类型用圆圈标记，根据广义细胞类型组着色；配体和受体用方形节点标记。实线连接细胞类型和它们表达的配体或受体；线的粗细与每种细胞类型中基因的平均表达水平成比例。虚线连接假定的受体和它们的配体。
a
Citation
Age
(year or month) （年份或月份）
Muscle type
Chemistry
Cells
Species
Kedlian et al. Kedlian 等人
yo
Intercostal muscles 肋间肌
v2/v3
91278
Human
Barruet et al., Barruet 等人，
elife, 2020
yo
Vastus lateralis/ 大腿外侧肌
Rectus abdominus/ 腹直肌
Pectoralis major 胸大肌
v3
106558
Human
Micheli et al., Micheli 等人,
Skelet Muscle, 骨骼肌
2020
43-81уo
Vastus lateralis/ 大腿外侧肌
Rectus abdominus/ 腹直肌
Eye lid/ Serratus/ 眼睑/ 锯齿肌/
Trapezius/
Orbicularis oris/ 唇周肌/
External oblique/ 外斜肌/
Flexor hallucis longus 趾长屈肌
v2/v3
22058
Human
Kedlian et al. Kedlian 等人
Hindlimb muscle 后肢肌肉
v2
68246
Mouse
Giordani et al., Giordani 等人，
Mol Cell, 2019 Mol Cell，2019 年
Hindlimb muscle 后肢肌肉
v2
12441
Mouse
Li et al., The Li 等人，《
EMBO Journal, EMBO 杂志，
2019
Tibialis anterior 胫骨前肌
v2
9988
Mouse
Tabula Muris
Senis, Nature, 感官，自然，
2020
Tibialis anterior 胫骨前肌
v2
28867
Mouse
Rubenstein et 鲁本斯坦等
al., Scientific 等人，科学
Reports, 2020 报告，2020 年
Quadriceps/
Diaphragm
v3
6890
Mouse
C
Extended Data Fig. 10 | Integrated human-mouse skeletal muscle aging atlas. a, Overview table showcasing metadata (age composition, muscle type, 10x chemistry, number of cells, species) for the datasets included into human-mouse skeletal muscle aging atlas. b, UMAP visualisation of human-mouse aging cell 扩展数据图 10 | 人-小鼠骨骼肌老化图谱。a，概述表展示了包括在人-小鼠骨骼肌老化图谱中的数据集的元数据（年龄构成、肌肉类型、10x 化学、细胞数量、物种）。b，人-小鼠老化细胞的 UMAP 可视化。
d
Fraction of cells Mean expression 细胞平均表达的部分
STRA6
LRP4
K1II
BXO2
DN11
DN10
CRG
atlas coloured according to species. c, d, Dot plot showing species-common (c) and -specific (d) marker genes for each major cell type annotated in the humanmouse skeletal muscle aging atlas. Dot size represents the proportion of cells expressing a gene, colour indicates its expression level. 根据物种着色的图谱。c，d，点图显示人类和小鼠骨骼肌老化图谱中每种主要细胞类型的物种共同（c）和特异性（d）标记基因。点的大小表示表达基因的细胞比例，颜色表示其表达水平。
Reporting Summary 汇报摘要
Nature Portfolio wishes to improve the reproducibility of the work that we publish. This form provides structure for consistency and transparency in reporting. For further information on Nature Portfolio policies, see our Editorial Policies and the Editorial Policy Checklist. Nature Portfolio 希望提高我们发表的作品的可重复性。此表格提供了一致性和透明度的结构以供报告。有关 Nature Portfolio 政策的更多信息，请参阅我们的编辑政策和编辑政策清单。
Statistics 统计数据
For all statistical analyses, confirm that the following items are present in the figure legend, table legend, main text, or Methods section. 对于所有统计分析，请确认以下项目是否出现在图例、表图例、主文本或方法部分中。
n/a Confirmed n/a 确认
The exact sample size for each experimental group/condition, given as a discrete number and unit of measurement 每个实验组/条件的确切样本大小 ，以离散的数字和计量单位给出
A statement on whether measurements were taken from distinct samples or whether the same sample was measured repeatedly 关于是否从不同样本进行了测量，还是重复测量了同一样本的说明
The statistical test(s) used AND whether they are one- or two-sided 使用的统计检验以及它们是单侧还是双侧的
Only common tests should be described solely by name; describe more complex techniques in the Methods section. 只有常见的检验应该仅通过名称描述；将更复杂的技术描述在方法部分。
A description of all covariates tested 所有协变量测试的描述
A description of any assumptions or corrections, such as tests of normality and adjustment for multiple comparisons 任何假设或更正的描述，例如正态性检验和多重比较的调整
A full description of the statistical parameters including central tendency (e.g. means) or other basic estimates (e.g. regression coefficient) AND variation (e.g. standard deviation) or associated estimates of uncertainty (e.g. confidence intervals) 包括中心趋势（例如均值）或其他基本估计（例如回归系数）和变异性（例如标准偏差）或相关不确定性估计（例如置信区间）的统计参数的详细描述
For null hypothesis testing, the test statistic (e.g. ) with confidence intervals, effect sizes, degrees of freedom and value noted Give values as exact values whenever suitable. 对于零假设检验，带有置信区间、效应大小、自由度和 值的检验统计量（例如 ）请在适当时给出 值作为精确值。
For Bayesian analysis, information on the choice of priors and Markov chain Monte Carlo settings 对于贝叶斯分析，有关先验选择和马尔可夫链蒙特卡洛设置的信息
For hierarchical and complex designs, identification of the appropriate level for tests and full reporting of outcomes 对于分层和复杂设计，确定测试的适当水平并全面报告结果
Estimates of effect sizes (e.g. Cohen's , Pearson's ), indicating how they were calculated 效应大小的估计（例如 Cohen's 的 ，Pearson's 的 ），说明它们是如何计算的
Our web collection on statistics for biologists contains articles on many of the points above. 我们针对生物学家的统计数据网站包含了许多上述要点的文章。
Software and code 软件和代码
Policy information about availability of computer code 有关计算机代码可用性的政策信息
Data collection RareCyte Orion system (no version available) with 7 lasers was used to acquire 15-plex IF staining data. 使用具有 7 个激光器的 RareCyte Orion 系统（无版本可用）进行 15-plex IF 染色数据采集。
Data analysis Software used for data alignment and mapping include: 10x Genomics Cell Ranger (v3.1.0), Starsolo (based on STAR 2.7.3). 用于数据对齐和映射的软件包括：10x Genomics Cell Ranger（v3.1.0）、Starsolo（基于 STAR 2.7.3）。
Single cell data analysis was mostly performed using Python (version 3) and scanpy (v1.7.2) or R (v3.6.3 and v4.0.4) with data.table (v1.14.0), ggplot2 (v3.3.2) and ggpubr (0.4.0). 大多数单细胞数据分析使用 Python（版本 3）和 scanpy（v1.7.2）或 R（v3.6.3 和 v4.0.4）进行，配合 data.table（v1.14.0）、ggplot2（v3.3.2）和 ggpubr（0.4.0）。
Ambient mRNA was removed using Cellbender (v0.2.0). 使用 Cellbender (v0.2.0) 移除环境 mRNA。
Doublets were removed using Scrublet (version 0.2.3). 使用 Scrublet (版本 0.2.3) 移除双倍体。
Batch correction was performed with scVI-tools (version 0.14.5). 使用 scVI-tools (版本 0.14.5) 进行批次校正。
Trajectory analysis was performed with Monocle 2 (v2.9.0). 使用 Monocle 2 (v2.9.0) 进行轨迹分析。
CellPhoneDB (v3.0.0) was used for ligand-receptor interaction analysis. 使用 CellPhoneDB (v3.0.0) 进行配体-受体相互作用分析。
Cell type enrichment analysis was performed using miloR (v0.99.18). 使用 miloR (v0.99.18) 进行细胞类型富集分析。
Ageing differential expression analysis was perfomed using approach, published in ref 108. 使用方法进行了衰老差异表达分析，发表在参考文献 108 中。
Gene set over-representation analysis was performed with clusterProfiler (v3.14.3) or gprofiler-official (v1.0.0). PYSCENIC version 0.11 .2 was used to infer transcription factor activity. 使用 clusterProfiler（v3.14.3）或 gprofiler-official（v1.0.0）进行基因集过表示分析。使用 PYSCENIC 版本 0.11.2 推断转录因子活性。
Software used for image analysis: Cellpose (Original version after release), ilastik (v.1.3.1), Fiji (2.1.0/1.53c) 用于图像分析的软件：Cellpose（发布后的原始版本），ilastik（v.1.3.1），Fiji（2.1.0/1.53c）
Exported raw data from BD InfluxTM Cell Sorter were processed with FlowJo (10.4). 从 BD InfluxTM Cell Sorter 导出的原始数据已经使用 FlowJo (10.4) 进行处理。
RareCyte Artemis 4.0 software was used to load 15-plex image data, compensate for channel crosstalk and autofluorescence. OMERO Plus (Glencoe software) was used to visualize IF and RNAscope imaging data. 使用 RareCyte Artemis 4.0 软件加载 15-plex 图像数据，进行通道串扰和自体荧光的补偿。使用 OMERO Plus (Glencoe 软件) 可视化 IF 和 RNAscope 成像数据。
All custom notebooks and scripts used in this study have been deposited to https://github.com/Teichlab/SKM_ageing_atlas repository. 本研究中使用的所有自定义笔记本和脚本已存储在 https://github.com/Teichlab/SKM_ageing_atlas 代码库中。
Policy information about availability of data 数据可用性政策信息
All manuscripts must include a data availability statement. This statement should provide the following information, where applicable: 所有手稿必须包含数据可用性声明。该声明应提供以下信息（如适用）：
Accession codes, unique identifiers, or web links for publicly available datasets 公开可用数据集的存取代码、唯一标识符或网页链接
A description of any restrictions on data availability 数据可用性限制的描述
For clinical datasets or third party data, please ensure that the statement adheres to our policy 对于临床数据集或第三方数据，请确保陈述符合我们的政策
The processed data objects generated within this study are available for browsing at https://www.muscleageingcellatlas.org. Raw sequencing data for the newly generated libraries has been deposited to ArrayExpress. The publicly available human skeletal muscle single-nuclei and single-cell datasets were downloaded from GSE167186, GSE143704 and DRYAD (https://doi.org/10.7272/Q65X273X), while mouse datasets were obtained from GSE110878, GSE138707, GSE134540, 本研究生成的处理过的数据对象可在 https://www.muscleageingcellatlas.org 上浏览。新生成的文库的原始测序数据已存储在 ArrayExpress 中。公开可用的人类骨骼肌单核和单细胞数据集从 GSE167186、GSE143704 和 DRYAD (https://doi.org/10.7272/Q65X273X) 下载，而小鼠数据集则来自 GSE110878、GSE138707、GSE134540。
GSE143476, GSE149590, GSE142480 repositories. All other data supporting the findings of this study are available from the corresponding authors upon request. All custom notebooks and scripts used in this study have been deposited to https://github.com/Teichlab/SKM_ageing_atlas repository. GSE143476、GSE149590、GSE142480 存储库。本研究发现的所有其他数据可根据请求从相应作者处获得。本研究中使用的所有自定义笔记本和脚本已存储到 https://github.com/Teichlab/SKM_ageing_atlas 存储库。
Field-specific reporting 领域特定报告
Please select the one below that is the best fit for your research. If you are not sure, read the appropriate sections before making your selection. 请选择以下最适合您研究的选项。如果不确定，请在选择之前阅读相应部分。
Life sciences Behavioural & social sciences Ecological, evolutionary & environmental sciences 生命科学 行为与社会科学 生态、进化与环境科学
All studies must disclose on these points even when the disclosure is negative. 所有研究必须在这些方面披露，即使披露是负面的。
Sample size 样本量
Sample size calculations were not performed as (I) there was no available human skeletal muscle single-cell RNA sequencing dataset at the beginning of this study and (II) there are also no widely accepted recommendations available to perform power calculations for single-cell studies. Sample size for single-cell and single-nuclei experiments was determined by availability of the donors within sampling timeframe. 由于在本研究开始时没有可用的人类骨骼肌单细胞 RNA 测序数据集，并且也没有广泛接受的建议可用于进行单细胞研究的功效计算，因此未进行样本量计算。单细胞和单核实验的样本量是由在采样时间范围内可用的捐赠者确定的。
我们用于不同实验的样品数量如下：2 个（1 个年轻 vs. 1 个年龄）用于核糖体组装基因的 QPCR、Western blot 和 b-Gal 染色（图 2 和扩展数据图 3），12 个（4 个年轻 vs. 8 年龄）用于 FACS（图 2 和扩展数据图 3），7 个（3 个年轻 vs. 4 年龄）用于 CCL2 的 qPCR（扩展数据图 3）， 3 个（1 个年轻 vs. 2 个年龄）用于挑逗肌纤维的 NMJ 染色（图 3，扩展数据图 5），11 个（7 个年轻 vs. 4 年龄）用于肌纤维分型（图 4 和扩展数据图），12 个（6 年轻 vs. 6 年龄）用于 MYH8 IF 和 MYH8 IHC（图 4 和扩展数据图 6），6 个（3 个年轻 vs. 3 个年龄）用于 FAM189A2+ 细胞核的 RNAscope 验证， 3 个（2 个年轻 vs. 1 个年龄）用于 OTUD1+ 细胞核的 RNAscope 验证，5 个（2 个年轻 vs. 3 个年龄）用于 NMJ 辅助核的 RNAscope 验证，3 个（1 个年轻 vs. 2 个年龄）用于双 RNAscope 和 IF 用于肌纤维分型， 10 个（6 年轻 vs. 4 年龄）用于 CD3 和层粘连蛋白的共 IF，7 个（4 年轻 vs. 3 个年龄）用于 NKG7 和层粘连蛋白的共 IF， 6 个（3 个年轻 vs. 3 岁）用于 ACTA2 的共 IF 和层粘连蛋白 2 （1 个年轻 vs. 1 个年龄）用于稀有细胞染色，4 个（2 个年轻 vs. 2 个年龄）用于 ACTA2 和 CCL2 的共 IF，4 个（2 年轻 vs. 2 岁）用于 LYVE1 研究的 RNAscope。由于对人体肌肉衰老研究的研究有限，样本量是根据先前发表的人体肌纤维研究（Murgia等人，Cell Reports，2017）以及我们在特定实验中可以获得多少样本来确定的。
Intercostal muscle biopsies and fetal muscle sample used for validation experiments were obtained from patients in China. Embryo was aged between post conceptional weeks. Adult donors were aged and years old. Samplespecific donor metadata is available in Supplementary table 8. 用于验证实验的肋间肌活检和胎儿肌肉样本来自于中国患者。胚胎年龄在 周孕龄。成年捐赠者年龄在 和 岁之间。样本特定的捐赠者元数据可在附表 8 中获取。
Recruitment 招募
Human intercostal muscle samples for single-cell and nuclei processing were collected from deceased. transplant organ donors by the Collaborative Biorepository for Translational Medicine, Cambridge, UK (CBTM). Written informed consent was obtained from the donor families. 人类肋间肌样本用于单细胞和细胞核处理，由剑桥，英国(CBTM)的转化医学协作生物库从已故的器官捐赠者处收集。从捐赠者家属获得了书面知情同意。
Tissue donors were recruited in China by collaborating doctors in the Sun Yat-sen Memorial Hospital and Guangzhou Women 通过与中山纪念医院和广州妇女儿童医疗中心的合作医生在中国招募组织捐赠者
retrieve muscle biopsy together with resected tissue (usually tumour). Informed consent was also obtained from the mother after her voluntary decision to legally terminate pregnancy but before the abortion. 一起取肌肉活检和切除组织（通常是肿瘤）。在母亲自愿决定合法终止妊娠但在流产前，也从母亲那里获得了知情同意。
Single-cell transcriptomics: Human intercostal muscle samples (inner part between the second and third ribs) for sc- and snRNA-seq were collected with consent from deceased transplant organ donors by the Cambridge Biorepository for Translational Medicine, Cambridge, UK (CBTM), immediately placed in HypoThermosol FRS preservation solution and shipped to Sanger Institute for processing. Ethical approval was granted by the Research Ethics Committee East of England Cambridge South (REC Ref 15/EE/0152) and the written informed consent was obtained from the donor families. Full metadata information for the organ donors is provided in the Supplementary Table 1. Three 19 months old and five 3 months old male mice of C57BL/6JRj strain were obtained from Janvier labs, France. All mice were housed in micro-isolator cages in standard housing conditions (ambient temperature of and humidity of ), illuminated from 7:00am-7:00pm with ad libitum access to diet and water, under Establishment licence number X3AOED725 provided by the Home Office. They were used to dissect hindlimb muscles for the single-cell and single-nucleus isolation. 单细胞转录组学：人体肋间肌肌肉样本（第二和第三肋骨之间的内部部分）用于单细胞和单核糖核酸测序，经剑桥转化医学生物库（CBTM）征得已故器官捐赠者同意后收集，立即置于HypoThermosol FRS保存液中并运送至桑格研究所进行处理。研究伦理委员会东英格兰剑桥南部（REC Ref 15/EE/0152）批准了该研究，捐赠者家属已签署知情同意书。器官捐赠者的完整元数据信息详见附表1。来自法国Janvier实验室的C57BL/6JRj品系的三只19个月大和五只3个月大雄性小鼠被获取。所有小鼠均被放置在微隔离笼中，处于标准饲养条件下（环境温度 和湿度 ），照明时间为上午7:00至下午7:00，可自由进食和饮水，根据英国内政部提供的X3AOED725号许可证进行饲养。它们被用于解剖后肢肌肉进行单细胞和单核糖核酸的分离。
Adult tissue from the UK for validation experiments: The same intercostal muscle samples collected with consent from deceased transplant organ donors (partially overlapping with the set of donors used for sc/snRNA-seq) by CBTM were used for experimental validations. 英国成人组织用于验证实验：CBTM使用同意收集的已故移植器官捐赠者的肋间肌样本（部分与用于sc/snRNA-seq的捐赠者集合重叠）进行实验验证。
Fetal and adult tissue from China for validation experiments: Adult human intercostal muscle biopsies were collected during the thoracic surgeries at Sun Yat-sen Memorial Hospital (Guangzhou, China) under approval of the Research Ethics Committee of Sun Yat-sen University (REC 2018-048). For isolation of human primary myoblasts, lower limb muscles were collected from one medically aborted embryo at post conceptional week (Pcw) 19 at Guangzhou Women and Children's Medical Center (Guangzhou, China) with ethical approval licence granted by both the Research Ethics Committee of Sun Yatsen University (REC 2019-075) and Guangzhou Women and Children's Medical Center (REC 2022-050A01). Both materials were registered at China National Center for Bioinformation (PRJCA014979) and have been approved by the Chinese Ministry of Science and Technology for the Review and the Approval of Human Genetic Resources (2023BAT0735). Appropriate written informed consent was obtained from each adult patient to retrieve muscle biopsy together with resected tissue (usually tumour). Informed consent was also obtained from the mother after her voluntary decision to legally terminate pregnancy but before the abortion. Before terminating pregnancy, both the mother and the embryo were diagnosed as healthy with no underlying diseases. Participants were not financially compensated. The detailed metadata for the 8 organ donors (UK), 40 patients (China) and 1 embryo used for validation experiments is provided in Supplementary Table 8. 中国胎儿和成人组织用于验证实验：在中山大学附属第一医院（中国广州）进行胸部手术期间收集了成人人类肋间肌肌活检组织，获得了中山大学研究伦理委员会（REC 2018-048）的批准。为了分离人类原代肌母细胞，从广州市妇女儿童医疗中心（中国广州）在胚胎受孕第19周进行医学流产的一个胚胎收集了下肢肌肉组织，获得了中山大学研究伦理委员会（REC 2019-075）和广州市妇女儿童医疗中心（REC 2022-050A01）的道德批准许可。这两种材料已在中国国家生物信息中心（PRJCA014979）注册，并已获得中国科学技术部对人类遗传资源审查和批准的批准（2023BAT0735）。从每位成人患者获得了适当的书面知情同意，以检索肌肉活检组织以及切除的组织（通常是肿瘤）。 在合法终止怀孕前，母亲自愿决定并在堕胎前获得知情同意。在终止怀孕前，母亲和胚胎均被诊断为健康，没有潜在疾病。参与者没有得到经济补偿。提供了 8 名器官捐赠者（英国）、40 名患者（中国）和 1 名用于验证实验的胚胎的详细元数据，见附表 8。
Note that full information on the approval of the study protocol must also be provided in the manuscript. 请注意，研究方案获得批准的完整信息也必须在手稿中提供。
Flow Cytometry 流式细胞术
Plots 绘图
Confirm that: 确认：
The axis labels state the marker and fluorochrome used (e.g. CD4-FITC). 轴标签说明使用的标记物和荧光染料（例如 CD4-FITC）。
The axis scales are clearly visible. Include numbers along axes only for bottom left plot of group (a 'group' is an analysis of identical markers). 轴标尺清晰可见。仅在组的左下角绘图中包含沿轴的数字（'group'是相同标记的分析）。
All plots are contour plots with outliers or pseudocolor plots. 所有图都是带有异常值或伪彩色图的等高线图。
A numerical value for number of cells or percentage (with statistics) is provided. 提供了细胞数量或百分比的数值（带有统计数据）。
