一篇nature immunology的方法，跪着读完_(:з」∠)_

[ixxmu]

https://mp.weixin.qq.com/s/ZCOoeQU6ysK9G8apTnfjvQ

[github-actions[bot]]

一篇nature immunology的方法，跪着读完(:з」∠) by 东林的扯淡小屋

Mouse strains. 

Female C57BL/6 mice were purchased from The Jackson Laboratory. Eight- to ten-week-old mice were used in all of the experiments. All animal experiments were conducted in accordance with the Animal Welfare Guidelines of the Children’s Hospital Boston. The Children’s Hospital Animal Care and Use Committee approved and monitored all of the procedures

小鼠种系。

雌性C57BL/6小鼠购自Jackson实验室。所有的实验都使用了8到10周大的老鼠。所有动物实验均按照波士顿儿童医院动物福利指南进行。儿童医院动物护理和使用委员会批准并监测了所有的手术。

Mouse peritonitis model. Wild-type mice were intraperitoneally injected with 1 × 107 E. coli (ATCC 19138) in 300 μl phosphate-buffered saline (PBS). At different time points after injection, mice were anesthetized with isoflurane, retro-orbital blood was collected, and then mice were sacrificed by euthanizing with CO2 (ref. 51). Cells from different organs such as BM, spleen, liver and peritoneal exudate were collected as detailed below.

小鼠腹膜炎模型。野生型小鼠腹腔注射1×10^7大肠杆菌(ATCC19138)(300μl磷酸盐缓冲盐水(PBS)溶解）。在注射后不同时间点，用异氟醚麻醉小鼠，取眶后血，CO2处死小鼠(参考文献51)。收集骨髓、脾脏、肝脏、腹膜渗出液等不同器官的细胞如下。

 51. Sakai, J. et al. Reactive oxygen species-induced actin glutathionylation controls actin dynamics in neutrophils. Immunity 37, 1037–1049 (2012).

Mouse neutrophil isolation. Neutrophils display circadian oscillations in number and phenotype, and neutrophil aging is an intrinsically driven bona fide circadian process52. Thus, all samples in this study were prepared from mice sacrificed at the same time in the morning (08:00). Eight-week-old female mice were used for scRNA-seq analysis. PB (600–800 µl) was collected by retro-orbital bleeding and diluted with 3 ml Hanks’ balanced salt solution (HBSS) containing 15 mM EDTA53. Cells were centrifuged for 10 min at 500g. Red blood cells were lysed by resuspension in 5 ml ammonium-chloride-potassium (ACK) lysis buffer (Thermo Fisher Scientific) for 5 min at room temperature. Next, 10 ml RPMI + 2% fetal bovine serum were added to stop lysis followed by centrifugation at 500g for 5 min. Cells were washed twice with 10 ml HBSS + 2 mM EDTA + 1% bovine serum albumin (BSA) before being resuspended in 500 μl PBS + 1% BSA. For BM neutrophil isolation, whole BM cells were flushed from the femur, tibia and ilia leg bones with 5 ml HBSS + 2 mM EDTA + 1% BSA and filtered through a 70 μm cell strainer. Cells were centrifuged for 10 min at 500g. Red blood cells were lysed with 1 ml ACK lysis buffer for 2 min at room temperature and washed twice with HBSS + 2 mM EDTA + 1% BSA and resuspended in 200 μl PBS + 1% BSA. c-kit+ BM cells were first enriched by positive selection using c-kit (CD117) microbeads (Miltenyi Biotec) and further purified by FACS sorting c-kit+ cells. To isolate spleen neutrophils, spleens were dissected, placed in 3 ml PBS + 1 mM EDTA and then gently disaggregated through a 70 μm cell strainer using a 1 ml syringe plunger. Whole spleen cells were collected, centrifugated and resuspended in 1 ml PBS + 1 mM EDTA. The red blood cells were then lysed with 5 ml ACK lysis buffer for 2 min at room temperature. After centrifugation, cells were washed twice with PBS + 1 mM EDTA and resuspended in 500 μl PBS + 1% BSA. Finally, peritoneal cavity exudate cells were harvested by three successive washes with 10 ml HBSS + 15 mM EDTA + 1% BSA. After centrifugation, cells were washed twice with the same solution and resuspended in 100 μl PBS + 1% BSA.

小鼠中性粒细胞分离。

中性粒细胞在数量和表型上显示昼夜节律振荡，而中性粒细胞衰老是一个内在驱动的真实的昼夜节律过程52。因此，本研究中所有样本均取材于早晨(08:00)同一时间处死的小鼠。使用8周龄雌性小鼠进行scRNA-seq分析。PB(600- 800µl)收集了retro-orbital出血和稀释3毫升汉克斯平衡盐溶液(hbs)包含EDTA53 15ml。500g离心10min。红细胞在5ml氨-氯-钾(ACK)溶解缓冲液(Thermo Fisher Scientific)中重悬5分钟，室温。接下来,10毫升RPMI +2%胎牛血清添加停止溶解在500 g离心5分钟紧随其后。细胞被洗两次10毫升hbs + 2毫米EDTA +1%牛血清白蛋白(BSA)在500g被resuspended BSAμl PBS + 1%。对于骨髓中性粒细胞分离，用5 ml HBSS + 2 mM EDTA+ 1% BSA从股骨、胫骨和ilia腿骨中冲洗整个骨髓细胞，并通过70μm细胞过滤器过滤。500g离心10min。1 ml ACK裂解液室温裂解2min，用HBSS + 2 mM EDTA + 1% BSA洗涤2次，再用200μl PBS + 1% BSA重悬。c-kit+ BM细胞首先用c-kit(CD117)微珠(Miltenyi Biotec)阳性筛选富集，然后用FACS分选c-kit+细胞进一步纯化。分离脾脏中性粒细胞，将脾脏解剖，置于3mlPBS + 1mm EDTA中，用1ml注射器柱塞用70μm细胞过滤器轻轻分离。收集脾细胞，离心，用1ml PBS + 1mm EDTA重悬。然后用5 mlACK裂解液在室温下裂解红细胞2分钟。离心后用PBS + 1 mM EDTA洗涤2次，500μl PBS + 1% BSA重悬。最后，用10 ml HBSS +15 mM EDTA + 1% BSA连续三次洗涤收集腹腔渗出细胞。离心后，用相同的溶液洗涤2次，再用100μl PBS + 1% BSA重悬。

Human sample collection. PB (10 ml) was collected from healthy donors into heparin anticoagulant tubes54. An equal volume (10 ml) of 6% hydroxyethyl solution was added into the heparinized blood and inverted gently several times for adequate mixing. The blood was kept at room temperature for 20–30 min before the supernatant was pipetted into a 50 ml Falcon tube followed by centrifugation at 290g for 5 min without braking. Cells were washed twice and lysed with ACK to completely remove red blood cells. Samples were stained with Percp-cy5.5-conjugated anti-human CD33 antibody for 20 min and DAPI was added to cells before sorting by FACS with a FACSAria III cell sorter (BD Biosciences). The Ethics Committee of Tianjin Blood Disease Hospital approved the study protocol, and the donor provided written informed consent for sample collection and data analysis.

人类的样本收集。

从健康供体收集PB(10ml)放入肝素抗凝管54。将等体积(10毫升)的6%羟乙基溶液加入肝素化血液中，轻轻翻转几次以充分混合。室温保存20-30 min，取上清液于50mlFalcon管中，290g离心5min，不制动。细胞洗涤两次，用ACK溶解，完全清除红细胞。样品用percp-cy5.5结合的抗人CD33抗体染色20分钟，在FACS与FACSAria III细胞分选仪(BDBiosciences)进行分选前，将DAPI添加到细胞中。天津市血液病医院伦理委员会批准了该研究方案，供体提供了书面知情同意书进行样本采集和数据分析。

Single cell collection, library construction and sequencing. Single-cell suspensions were stained for 30 min at 4 °C with fluorophore-conjugated antibodies (APC/CY7-conjugated anti-Gr1 and FITC-conjugated-anti-CD45), filtered through 40 μm cell strainers, and DAPI was added before sorting by FACS with a FACSAria III cell sorter (BD Biosciences). For mouse cells, designated cells were sorted into PBS containing 0.05% BSA following the 10× Genomics protocol. The cell preparation time before loading onto the 10× Chromium controller was <2 h. Cell viability and counting were evaluated with trypan blue by microscopy, and samples with viabilities >85% were used for sequencing. Libraries were constructed using the Single Cell 3′ Library Kit V2 (10× Genomics). Transcriptome profiles of individual cells were determined by 10× Genomics-based droplet sequencing. Once prepared, indexed complementary DNA (cDNA) libraries were sequenced with paired-end reads on an Illumina NovaSeq 6000 (Illumina). For human cells, the sample preparation, library construction and single-cell sequencing were performed using a BD Rhapsody Single-Cell Analysis System following a standard protocol provided by the manufacturer (BD Biosciences).

单细胞收集、文库构建和测序。

单细胞悬浮液用荧光标记抗体(APC/cy7标记抗gr1和fitc标记抗cd45)在4℃下染色30分钟，通过40μm的细胞过滤器过滤，加入DAPI，然后用FACS与FACSAriaIII细胞分选器进行分选(BD Biosciences)。对于小鼠细胞，按照10×Genomics protocol将指定的细胞分类成含有0.05%BSA的PBS。将细胞装入10×铬控制器前的细胞制备时间<2h。显微镜下用台虫蓝评估细胞活力和计数，取存活率为>85%的样品进行测序。使用单细胞3’Library KitV2(10×Genomics)构建文库。通过10×基因组液滴测序确定单个细胞的转录组谱。一旦准备好，索引的互补DNA (cDNA)库在IlluminaNovaSeq 6000 (Illumina)上进行配对端测序。对于人类细胞，样品制备、文库构建和单细胞测序使用BDRhapsody单细胞分析系统，遵循制造商(BD生物科学公司)提供的标准方案。

Bulk RNA isolation and sequencing. BM cells were prepared as previously described55. BM cells were first stained with the following antibodies for 20 min: biotin-conjugated anti-CD4; biotin-conjugated anti-CD8a; biotin-conjugated anti-Ter119; and biotin-conjugated anti-B220/CD45R. Then, the cells were stained with the following antibodies for 90 min at 4 °C: PE/cy7-conjugated streptavidin; APC-conjugated anti-c-kit; PE-conjugated anti-Ly6G; and FITC-conjugated anti-CD34. Myeloblasts, promyelocytes, metamyelocytes, myelocytes, mature band cells and segmented neutrophils were sorted with a MoFlo cell sorter (Beckman Coulter). Total RNA was extracted from those populations using the Qiagen RNeasy Mini Kit (Qiagen). RNA quality was evaluated spectrophotometrically, and the quality was assessed with the Agilent 2100 Bioanalyzer (Agilent Technologies). All samples showed RNA integrity of >7.5. RNA-seq libraries were prepared using the KAPA mRNA HyperPrep Kit (Illumina). Once prepared, indexed cDNA libraries were pooled in equimolar amounts and sequenced with paired-end reads on an Illumina HiSeq 2500

bulk RNA分离和测序。

BM细胞如前所述制备55。BM细胞首先用以下抗体染色20分钟:生物素偶联抗cd4;biotin-conjugatedanti-CD8a;biotin-conjugated anti-Ter119;和biotin-conjugated anti-B220 /CD45R。然后，用以下抗体在4℃下染色90分钟:PE/cy7偶联链霉亲和素;APC-conjugated anti-c-kit;PE-conjugatedanti-Ly6G;和FITC-conjugated anti-CD34。用MoFlo细胞分选器(BeckmanCoulter)对髓母细胞、早髓细胞、变髓细胞、髓细胞、成熟带细胞和节段性中性粒细胞进行分选。使用QiagenRNeasy迷你试剂盒(Qiagen)从这些群体中提取总RNA。采用分光光度法评价RNA质量，采用Agilent 2100生物分析仪(AgilentTechnologies)评价RNA质量。所有样本均显示>7.5的RNA完整性。使用KAPA mRNA HyperPrep Kit(Illumina)制备RNA-seq文库。一旦准备好，索引的cDNA文库将以等分子量的量汇集起来，并在Illumina HiSeq 2500上进行配对端测序

Wright–Giemsa staining and examination of morphology-defined neutrophil populations. The first recognizable cells of neutrophil lineage in BM are myeloblasts, which are characterized by a high nuclear-to-cytoplasmic ratio and dispersed chromatin. Myeloblasts then irreversibly differentiate into promyelocytes, which are characterized by a round nucleus and azurophil granules, followed by myelocytes characterized by a round nucleus and specific granules. Metamyelocytes are characterized by nuclear indentations (kidney-shaped nuclei) and the emergence of secretary vesicles. Finally, metamyelocytes are divided into band cells with a band-shaped nucleus and segmented cells (segmented neutrophils; also known as polymorphonuclear granulocytes) with a segmented nucleus. Cells were sorted (Fig. 2d) and concentrated onto microscope slides by cytospinning. Slides were dried and stained using the Diff-Quick Stain Set (Siemens). Stained slides were rinsed under running tap water and air-dried for 10 min. Images were obtained under a microscope with a 63× objective.

Wright-Giemsa染色及形态学明确的中性粒细胞群检测。

骨髓中最早被识别的中性粒细胞系细胞是髓母细胞，其特征是核质比高，染色质分散。然后髓母细胞不可逆地分化为以圆形核和含azurophil颗粒为特征的早幼粒细胞，再分化为以圆形核和特定颗粒为特征的髓细胞。偏髓细胞的特征是核凹痕(肾形核)和出现分泌囊泡。最后，变髓细胞分为带状细胞核的带状细胞和分叶细胞(分叶中性粒细胞;也称为多形核粒细胞)，核分节段。细胞被分选(图2d)，并通过细胞纺丝将细胞集中到显微镜载玻片上。玻片用Diff-Quick染色仪(Siemens)进行干燥和染色。染色载玻片在自来水下冲洗，风干10分钟，在63×物镜显微镜下成像。

EdU incorporation assay. EdU—a thymidine analog—can track cells post-mitotically in BM and PB (Fig. 7e). EdU is incorporated into DNA in the S phase of the cell cycle, and the half-life of EdU is only about 30 min, so incorporation can only occur in the first 1–2 h after EdU intraperitoneal injection. After 1 h of intraperitoneal injection with 0.5 mg EdU, mice were injected with E. coli as above to induce peritonitis. Mice were sacrificed at designated time points, and BM, blood and spleen cells were harvested followed by staining with the following fluorescent-conjugated antibodies: APC-conjugated anti-CD11b; APC/ cy7-conjugated anti-Ly6G; and PE-conjugated anti-CXCR4. Labeled cells were fixed, permeabilized and stained with azide dye using an EdU Proliferation Kit (BD Biosciences). Cells were further washed and analyzed using a BD FACSCanto II (BD Biosciences). Data were analyzed using FlowJo software (BD Biosciences).

EdU-a胸腺嘧啶类似物可以在BM和PB中跟踪有丝分裂后的细胞(图7e)。EdU是在细胞周期的S期并入DNA的，而EdU的半衰期仅为30min左右，所以只能在腹腔注射EdU后的前1-2 h内加入。以0.5 mg EdU腹腔注射1h后，同样注射大肠杆菌诱发腹膜炎。在指定时间点处死小鼠，取BM、血、脾细胞，用以下荧光偶联抗体进行染色:apc偶联抗cd11b;APC /cy7-conjugated anti-Ly6G;和PE-conjugated anti-CXCR4。使用EdU增殖试剂盒(BDBiosciences)对标记细胞进行固定、透化和叠氮染料染色。细胞进一步清洗和分析使用BD FACSCanto II (BDBiosciences)。使用FlowJo软件(BD Biosciences)分析数据。

Spleen cryosection preparation. Spleens were fixed in 1% formaldehyde (StatLab) for 4–8 h, rehydrated in 30% sucrose solution for 72 h, and snap frozen in O.C.T. (Sakura Finetek Japan). Single-cell-thick (5 μm) spleen cryosections were obtained using a Leica Cryostat and the CryoJane tape transfer system (Leica Microsystems). For immunofluorescent staining, slides were rehydrated in PBS for 10 min followed by rinsing in PBST (PBS + 0.1% Tween 20). Blocking was performed with PBS + 10% donkey serum for 20 min. The diluted primary rat anti-S100a8 (Thermo Fisher Scientific; 335806) and rabbit anti-IFIT1 (Abcam; ab236256) antibodies were added and incubated for 1 h at room temperature. After three washes with PBST, Alexa Fluor 488-conjugated donkey anti-rat antibody (Jackson ImmunoResearch; 141697) and Cy3-conjugated donkey anti-rabbit antibody (Jackson ImmunoResearch; 143460) were added and incubated for 30 min at room temperature. Slides were washed 3× with PBST and then stained with DAPI (0.5 μM) for 3 min. Slides were rinsed in PBS and were covered with mounting solution (Vectashield; Vector Laboratories).

脾脏cryosection准备。脾脏在1%甲醛(StatLab)中固定4-8 h，在30%蔗糖溶液中再水合72 h，在O.C.T.中快速冷冻(SakuraFinetekJapan)。使用徕卡低温恒温器和CryoJane带传输系统(徕卡微系统)获得单细胞厚(5μm)脾冰冻切片。免疫荧光染色时，载玻片在PBS中再水化10分钟，然后在PBST中冲洗(PBS+ 0.1%吐温20)。PBS + 10%驴血清阻断20分钟。稀释的大鼠抗s100a8 (Thermo FisherScientific;335806)和兔抗ifit1 (Abcam;加入ab236256抗体，室温孵育1 h。经过三次洗涤PBST, Alexa Fluor488偶联的驴抗鼠抗体(杰克逊免疫研究;141697)和cy3标记的驴抗兔抗体(Jackson免疫研究;加入143460)，室温孵育30min。PBS冲洗3×PBST，然后用DAPI(0.5μM)染色3分钟。PBS冲洗，覆盖载体液Vectashield;向量的实验室)。

Laser scanning cytometry (LSC). LSC is an emerging technology that images and quantitatively analyzes cellular and subcellular criteria within tissues, re-interrogating identified cell subpopulation(s) for in situ characterization of the molecular and cellular events associated with those cells. LSC was performed with an iCys Research Imaging Cytometer four-laser system (Thorlabs)43. Each section was first scanned with a 10× objective using the 405 nm laser to generate low-resolution images of the DAPI-stained nuclei and obtain a general view of the spleen. Subsequently, the sections were divided into small regions and scanned with a 40× dry objective lens to create high-resolution field images. Data were analyzed using iCys Cytometric Analysis Software (Thorlabs).

激光扫描细胞仪(LSC)。LSC是一种新兴的技术，可以对组织内的细胞和亚细胞标准进行图像和定量分析，重新询问已识别的细胞亚群，以原位表征与这些细胞相关的分子和细胞事件。LSC采用iCys研究成像细胞仪四激光系统(Thorlabs)43。每个切片首先用10×objective扫描，使用405nm激光生成dapi染色核的低分辨率图像，获得脾脏的一般视图。随后，切片被划分成小区域，并用40×干物镜扫描，以生成高分辨率的现场图像。数据采用iCys细胞分析软件(Thorlabs)进行分析。

Confocal imaging. Sections with a thickness of 20 μm were prepared and stained as described above. Confocal images were obtained using the Zeiss LSM 700 laser scanning confocal microscope (Carl Zeiss AG). Data were analyzed using Imaris Software (Oxford Instruments).

共焦成像。如上所述制备并染色厚度为20μm的切片。采用蔡司LSM 700激光扫描共焦显微镜(Carl ZeissAG)获得共焦图像。数据采用Imaris软件(Oxford Instruments)进行分析。

Intracellular protein staining for FACS analysis. PB and spleen cell suspensions were prepared as described above. After being washed with PBS twice, cells were blocked with rat anti-mouse CD16/CD32 antibody on ice for 10 min. APC-conjugated anti-CD45, APC/cy7-conjugated anti-CD11b, PE-conjugated anti-Ly6G and BV711-conjugated anti-CXCR4 antibodies were added and incubated for 20 min at 4 °C in the dark. Cells were washed with PBS twice, fixed and permeabilized with 1 ml PBS containing 4% paraformaldehyde (Electron Microscopy Sciences) and 0.1% saponin (Sigma–Aldrich) at 4 °C for 30 min, and pelleted by centrifugation at 3,000g for 3 min at 4 °C. After being washed with 1 ml Wash Buffer (PBS containing 0.2% BSA and 0.1% saponin), cells were blocked for 30 min with blocking buffer (PBS containing 5% goat serum and 5% BSA) and then stained with anti-IFIT1/p56 antibody (Sigma–Aldrich) for 30 min at 4 °C in 200 µl staining buffer (PBS containing 5% goat serum, 5% BSA and 0.1% saponin). Cells were washed twice with 1 ml Wash Buffer followed by incubation with the secondary goat anti-rabbit-Alexa Fluor 488 antibody (Invitrogen) in staining buffer for 30 min. Cells were then washed with 1 ml Wash buffer and resuspended in 1 ml PBS containing 0.5% BSA. The stained cells were either sorted with a FACSAria III cell sorter (BD Biosciences) or analyzed on an Attune NxT Flow Cytometer (Thermo Fisher Scientific). Fixation, washing, staining and sorting were performed at a concentration of 5–10 × 106 cells per ml

细胞内蛋白染色用于FACS分析。按照上述方法制备PB和脾脏细胞悬液。PBS冲洗两次后，冰上用大鼠抗小鼠CD16/CD32抗体阻断细胞10分钟。加入APC-偶联抗cd45、APC/cy7偶联抗cd11b、pe-偶联抗ly6g和bv711偶联抗cxcr4抗体，4℃黑暗孵育20分钟。用PBS洗涤细胞两次，用含有4%多聚甲醛(电子显微镜科学)和0.1%皂素(Sigma-Aldrich)的1mlPBS固定和透化，4℃30分钟，3000 g离心制粒，4℃3分钟。洗后用1毫升洗缓冲区包含0.2%BSA和0.1%皂苷(PBS),细胞被封锁的30分钟阻断缓冲区(PBS山羊血清含5%和5% BSA),然后沾anti-IFIT1 /p56抗体(Sigma-Aldrich) 30分钟在4°C 200µl染色缓冲区(PBS山羊血清含5%,5% BSA和皂苷0.1%)。细胞用1ml洗涤缓冲液洗涤2次，然后用山羊抗兔alexa Fluor 488抗体(Invitrogen)在染色缓冲液中孵育30分钟。然后用1ml洗涤缓冲液洗涤，再用含有0.5% BSA的1 ml PBS重悬。染色的细胞要么用FACSAria III细胞分选器(BDBiosciences)分类，要么用调好的NxT流式细胞仪(Thermo Fisher Scientific)分析。以5-10×106个细胞/ml的浓度进行固定、洗涤、染色和分选

RNA purification, cDNA synthesis and preamplification and quantitative PCR (qPCR). Total RNA was extracted from sorted fixed cells using a RecoverAll Total Nucleic Acid Isolation Kit (Thermo Fisher Scientific), starting at the protease digestion stage of the manufacturer-recommended protocol. The following modification to the isolation procedure was made: instead of incubating cells in digestion buffer for 15 min at 50 °C and 15 min at 80 °C, we carried out the incubation for 3 h at 50 °C. The cDNA was subsequently generated using a PrimeScript RT Master Mix (Perfect Real Time) (Takara Bio), and then subjected to 14 cycles of preamplification using a Prelude PreAmp Master Mix (Takara Bio) according to the manufacturer’s recommendations. The pre-amplified cDNA was subjected to qPCR in which the amplified product was detected using TB Green Premix Ex Taq (Tli RNase H Plus) (Takara Bio) on a CFX96 Real-Time PCR Detection System (Bio-Rad). ΔCt was calculated using GAPDH as a normalizer. The sequences of real-time qPCR primers are listed in Supplementary Table 8.

RNA纯化，cDNA合成，预扩增和定量PCR (qPCR)。从制造商推荐方案的蛋白酶消化阶段开始，使用RecoverAll总核酸分离试剂盒(ThermoFisherScientific)从分类固定细胞中提取总RNA。对分离过程进行了如下修改:我们不再将细胞在消化缓冲液中50℃培养15分钟，80℃培养15分钟，而是在50℃培养3小时。cDNA随后使用PrimeScriptRT主混合(完美实时)(Takara Bio)生成，然后根据制造商的建议，使用Prelude前置放大器主混合(TakaraBio)进行14个循环的预扩增。将预扩增的cDNA进行qPCR，扩增产物在CFX96实时荧光定量PCR检测系统(Bio- rad)上使用TB GreenPremix Ex Taq (Tli RNase H Plus) (Takara Bio)进行检测。δCt以GAPDH为归一化器计算。real-timeqPCR引物序列见补充表8。

scRNA-seq data processing. The quality of sequencing reads was evaluated using FastQC and MultiQC. Cell Ranger version 2.2.0 was used to align the sequencing reads (fastq) to the mm10 mouse transcriptome and quantify the expression of transcripts in each cell. This pipeline resulted in a gene expression matrix for each sample, which records the number of UMIs for each gene associated with each cell barcode. For human data, sequenced reads were aligned to the hg38 human transcriptome, then the expression of transcripts in each cell was quantified using the BD Rhapsody Whole Transcriptome Assay Analysis Pipeline. Unless otherwise stated, all downstream analyses were implemented using R version 3.5.2 and the package Seurat version 2.3.4 (ref. 56). Due to dissimilar data qualities, low-quality cells were filtered using sample-specific cutoffs (Supplementary Table 1). The NormalizeData function was performed using default parameters to remove the differences in sequencing depth across cells. For the experiment described in Fig. 1, cells from four samples were pooled and analyzed together. After rigorous quality control, we obtained 19,582 high-quality cells with an average of 1,241 genes per cell profiled, resulting in a total of 18,269 mouse genes detected in all cells (Extended Data Fig. 1e). For the experiment described in Fig. 5, after excluding low-quality cells, a total of 25,897 cells, including 4,421 cells from BM (eBM_Gr1), 6,232 cells from PB (ePB_Gr1), 5,989 cells from spleen (eSP_Gr1), 4,435 cells from liver (eLV_Gr1) and 4,169 cells from peritoneal cavity (ePC_Gr1), were available for analysis (Extended Data Fig. 6e,f).

scRNA-seq数据处理。使用FastQC和MultiQC评估测序reads的质量。使用CellRanger2.2.0版本将测序读数(fastq)与mm10小鼠转录组对齐，并量化转录本在每个细胞中的表达。这个流水线为每个样本生成了一个基因表达矩阵，它记录了与每个细胞条形码相关的每个基因的UMIs数量。对于人类数据，测序reads与hg38人类转录组比对，然后使用BDRhapsody全转录组分析流水线定量转录本在每个细胞中的表达。除非另有说明，所有下游分析都是使用R版本3.5.2和软件包Seurat版本2.3.4进行的(参考文献56)。由于数据质量不同，低质量的单元格使用样本特定的截止值进行过滤(补充表1)。使用默认参数执行NormalizeData函数，以消除单元格间测序深度的差异。在图1所示的实验中，来自四个样本的细胞被汇集并分析在一起。经过严格的质量控制，我们获得19,582个高质量细胞，平均每个细胞有1,241个基因，结果在所有细胞中共检测到18,269个小鼠基因(扩展数据图1e)。图5中描述的实验,排除劣质细胞后,共有25897个细胞,包括4421个细胞来自BM(eBM_Gr1), 6232个细胞从PB (ePB_Gr1), 5989个细胞从脾(eSP_Gr1),4435个细胞从肝脏(eLV_Gr1)和4169个细胞腹腔(ePC_Gr1),用于分析(扩展数据图6 e、f)。

Batch correction. There was substantial variability between cells obtained from different samples, probably reflecting a combination of biological and technical differences. In this case, the batch had little effect on partitioning cell types and thus cell clustering into neutrophils, B cells, T cells, monocytes, dendritic cells, erythrocytes and progenitors. However, when clustering neutrophils alone, cells clustered first by sample rather than by biological clusters. Therefore, the ScaleData function was used to eliminate cell–cell variation in gene expression driven by batch and mitochondrial gene expression. Importantly, the batch-corrected data were only used for principal component analysis (PCA) and all steps relying on PCA (for example, clustering and UMAP visualization). All other analyses (for example, differential expression analysis) were based on the normalized data without batch correction.

批处理校正。从不同样品中获得的细胞之间有很大的差异，这可能反映了生物学和技术上的综合差异。在这种情况下，批次对分配细胞类型几乎没有影响，因此细胞聚集为中性粒细胞、B细胞、T细胞、单核细胞、树突状细胞、红细胞和祖细胞。然而，当中性粒细胞单独聚类时，细胞首先通过样本聚类，而不是通过生物聚类。因此，我们使用ScaleData函数来消除batch和线粒体基因表达驱动的基因表达的细胞-细胞差异。重要的是，批量校正后的数据仅用于主成分分析(PCA)和所有依赖于PCA的步骤(例如，聚类和UMAP可视化)。所有其他分析(例如，差分表达式分析)都是基于未批修正的标准化数据。

Dimension reduction. Dimension reduction was performed at three stages of the analysis: the selection of variable genes; PCA; and UMAP57. The FindVariableGenes function (y.cutoff = 1 for control total cells; y.cutoff = 1.2 for control neutrophils; y.cutoff = 0.7 for E. coli-challenged total cells) was applied to select highly variable genes covering most of the biological information contained in the whole transcriptome. Then, the variable genes were used for PCA implemented with the RunPCA function. Next, we selected principal components 1–20 (for total cells) or 1–15 (for neutrophils) as input to perform the RunUMAP function to obtain bidimensional coordinates for each cell.

降维。在分析的三个阶段进行了降维:变量基因的选择;主成分分析;和UMAP57。FindVariableGenes功能(y.cutoff =1为控制总细胞;y.cutoff = 1.2为中性粒细胞对照;对于大肠杆菌攻毒的总细胞，y.cutoff =0.7)被用于选择高度变异的基因，涵盖了整个转录组中包含的大部分生物学信息。然后，将变量基因用于RunPCA函数实现的PCA。接下来，我们选择主成分1-20(总细胞)或1-15(中性粒细胞)作为输入，执行RunUMAP函数，获得每个细胞的二维坐标。

Unsupervised clustering and annotation. We performed the FindClusters function (resolution: 0.3, 0.6 and 0.2 for control total cells, neutrophils and E. coli-challenged total cells, respectively) to cluster cells using the Louvain algorithm based on the same principal components as for the RunUMAP function. Clusters G1–G5 were neutrophils at different maturation stages. G1 and G2 were early-stage neutrophils with a higher expression of Elane, Mpo, Fcnb and Camp (Fig. 1d,e). Neutrophils are terminally differentiated. The transition from a proliferative cell to terminal differentiation was accompanied by a dramatic change in the expression of the important cell cycle regulatory proteins, so we next performed a single-cell-resolution analysis of cell cycle activation during neutrophil differentiation based on the expression of G1/S- and G2/M-phase-specific genes58,59 (Fig. 1i). Cells in the G0 to G2 stages underwent active proliferation, while cell division stopped abruptly thereafter. CDC28 protein kinase regulatory subunit 2 (CKS2), Mki67 and Cdc20 were all strongly downregulated at the messenger RNA (mRNA) level.

无监督的聚类和注释。我们执行FindClusters函数(分辨率:0.3、0.6和0.2分别用于控制总细胞、中性粒细胞和大肠杆菌感染的总细胞)，使用基于与RunUMAP函数相同的主成分的鲁汶算法对细胞进行聚类。簇状G1-G5为不同成熟阶段的中性粒细胞。G1和G2为早期中性粒细胞，具有较高的Elane、Mpo、Fcnb和Camp表达(图1d,e)。中性粒细胞终末分化。从增殖细胞终端分化的转变是伴随着一个戏剧性的变化在重要的细胞周期调控蛋白的表达,所以我们接下来执行single-cell-resolution分析细胞周期激活中性粒细胞分化期间基于G1/ S的表达式和G2 / M-phase-specific genes58, 59(图1)。G0 -G2期细胞增殖活跃，细胞分裂停止。CDC28蛋白激酶调节亚基2 (CKS2)、Mki67和Cdc20均在信使RNA (mRNA)水平上强烈下调。

Identification of DEGs. We used the FindMarkers or FindAllMarkers function (test.use = ‘‘t’’ , logfc.threshold = log[1.5]) based on normalized data to identify DEGs. P value adjustment was performed using Bonferroni correction based on the total number of genes in the dataset. DEGs with adjusted P values > 0.05 were filtered out. Gene Ontology analysis was performed by using the R package clusterProfiler60. In the experiment described in Extended Data Fig. 7, we conducted differential gene expression analysis in each neutrophil subpopulation using the non-parametric Wilcoxon rank-sum test and identified DEGs with an average expression fold-change > 2.

度的识别。我们使用FindMarkers或FindAllMarkers函数(test。使用=“t”，logfc。阈值=log[1.5])基于归一化数据识别差异集。根据数据集中的基因总数，使用Bonferroni校正对P值进行调整。P值调整后>0.05的DEGs被过滤掉。基因本体分析使用R包clusterProfiler60进行。在扩展数据图7中描述的实验中，我们使用非参数Wilcoxon秩和检验在每个中性粒细胞亚群中进行差异基因表达分析，并鉴定出平均表达倍数变化>2的DEGs。

Developmental trajectory inference. Pseudo-time was generated with Monocle version 2 (ref. 18) to infer the potential lineage differentiation trajectory. The newCellDataSet function (lowerDetectionLimit = 0.5; expressionFamily = negbinomial.size) was used to build the object based on the above highly variable genes identified by Seurat version 2.3.4.

发育轨迹推理。用Monocle version2(参考文献18)产生伪时间来推断潜在的谱系分化轨迹。newCellDataSet函数(lowerDetectionLimit =0.5;使用expressionFamily = negbinomial.size)基于上述Seurat 2.3.4版本识别出的高变异基因构建对象。

Bulk RNA-seq analysis. The quality of sequencing reads was evaluated using FastQC and MultiQC. Adaptor sequences and low-quality score bases were trimmed using trimmomatic/0.36. The resulting reads were then mapped to the mouse reference sequence (GRCm38/mm10; Ensembl release 81) and counted using STAR2.5.2b alignment software. Gene differential expression analysis was performed using the R package EdgeR.

大部分RNA-seq分析。使用FastQC和MultiQC评估测序reads的质量。使用trimmomatic/0.36修剪适配器序列和低质量评分基础。然后将读取结果映射到小鼠参考序列(GRCm38/mm10;使用STAR2.5.2b校准软件进行计数。使用Rpackage EdgeR进行基因差异表达分析。

Scoring of biological processes. Individual cells were scored for their expression of gene signatures representing certain biological functions. For all signatures except neutrophil aging, functional scores were defined as the average normalized expression of corresponding genes. Aging score was defined as the weighted average of Z scores of age-related genes, where the Z scores were calculated by scaling the normalized expression of a gene across all cells. Gene weights were set to either 1 or −1 to reflect positive or negative relationships. The neutrophil maturation signature was derived by identifying the top 50 DEGs (as listed in Supplementary Table 4) with the highest fold-changes and adjusted P values < 0.05 between the mature cluster (G4) and immature clusters (G0–G3). Granule signatures were from ref. 20. Other functional signatures were derived from the Gene Ontology database61, with the full gene list provided in Supplementary Table 4. For instance, to access the phagocytosis function at the transcript level, we determined a phagocytosis score by calculating the average expression of genes in the Gene Ontology term ‘phagocytosis, engulfment’ (GO: 0006911). The apoptosis score was measured by the upregulation of the integrated proapoptotic pathway (Fig. 4b). To further dissect apoptotic heterogeneity in G5 populations independent of transcriptome-based sub-clustering, we fit a two-component Gaussian mixture model to the apoptotic score of all G5 cells using the R package mixtools version 1.1.0 (ref. 62). We then chose the distribution with the higher mean as the apoptotic group and assigned each cell to one of the two groups based on its posterior (Fig. 4c). Age-related genes were summarized from the previous literature (Fig. 2i). Aging is a main mechanism that accounts for neutrophil heterogeneity7,63: aged neutrophils are smaller with fewer granules and granular multi-lobed nuclei and produce more neutrophil extracellular traps (NETs). Related to function, aged neutrophils express less of the adhesion molecule l-selectin (CD62L; encoded by Sell) and more CD11b (αM; encoded by Itgam), lymphocyte function-associated antigen-1 (CD11a/β2), CD49d (integrin α4; encoded by Itga4), TLR4, ICAM-1, CD44 and CD11c (encoded by Itgax). Additionally, aged neutrophils express more surface CXCR4 and less CXCR2, which regulates their release from and return to BM. CXCR4 may also play a role in clearing aged, senescent neutrophils, particularly at BM sites. Anti-CXCR4 antibodies or CXCR4 antagonists impede neutrophil homing to BM64,65. Finally, aged neutrophils exhibit increased expression of CD24 (a glycosylphosphatidylinositol-linked glycoprotein that induces apoptosis when crosslinked) and reduced expression of CD47 (the ‘don’t eat me’ signal that inhibits efferocytosis—a process leading to clearance of dead neutrophils). ROS-mediated pathogen killing is a major host defense mechanism. In neutrophils, ROSs are mainly produced by the phagocytic NADPH oxidase (aka the NOX2 complex). During cell activation, cytosolic components of the NADPH oxidase NCF2 (p67phox), Rac1 and/or Rac2, NCF4 (p40-phox) and NCF1 (p47phox) are recruited to the membrane to form a complex with membrane proteins CYBA (p22-phox) and CYBB (gp91 or cytochrome-b 558 subunit beta). We evaluated the NADPH oxidase score based on the expression of the seven NADPH oxidase-related genes (Supplementary Fig. 6d).

生物过程的评分。单个细胞的得分是根据它们代表某种生物学功能的基因签名的表达。对于除中性粒细胞老化外的所有特征，功能得分被定义为相应基因的平均归一化表达。衰老得分被定义为年龄相关基因Z得分的加权平均值，其中Z得分是通过将一个基因在所有细胞中的归一化表达比例来计算的。基因权重设为1或−1，以反映正向或负向关系。通过鉴定成熟集群(G4)和未成熟集群(G0-G3)中fold-changes最高且P值调整<0.05的前50位DEGs(见补充表4)，得出中性粒细胞成熟特征。颗粒特征来自参考文献20。其他功能签名来自基因本体论数据库61，完整基因列表见补充表4。例如，为了在转录水平上了解吞噬功能，我们通过计算基因本体术语“吞噬、吞噬”中的基因平均表达来确定吞噬评分(GO:0006911)。通过上调整合促凋亡通路来测量凋亡评分(图4b)。为了进一步分析独立于基于转录组的亚聚类的G5细胞凋亡异质性，我们使用R包mixtoolsversion1.1.0对所有G5细胞的凋亡评分拟合了一个双组分高斯混合模型(参考文献62)。然后我们选择均值较高的分布作为凋亡组，根据其后验值将每个细胞分为两组(图4c)。从以前的文献中总结了年龄相关基因(图2i)。衰老是导致中性粒细胞异质性的主要机制7,63:衰老的中性粒细胞体积较小，颗粒和多叶粒状核较少，产生更多的中性粒细胞胞外陷阱(net)。与功能相关，衰老中性粒细胞粘附分子l-选择素(CD62L;和更多的CD11b(αM;Itgam编码)、淋巴细胞功能相关抗原-1(CD11a/β2)、CD49d(整合素α4);Itga4编码)、TLR4、ICAM-1、CD44和CD11c(Itgax编码)。此外，衰老的中性粒细胞表达更多的表面CXCR4和更少的CXCR2，这调节了它们从BM的释放和返回。CXCR4也可能在清除衰老的中性粒细胞中发挥作用，特别是在BM部位。抗CXCR4抗体或CXCR4拮抗剂可阻止中性粒细胞向BM64,65归巢。最后，衰老的中性粒细胞表现出CD24(一种糖基磷脂酰肌醇连接的糖蛋白，交联时诱导细胞凋亡)表达增加，CD47(“不要吃我”信号，抑制efferocytodisease-一个导致清除死亡中性粒细胞的过程)表达减少。ros介导的病原体杀灭是一种主要的寄主防御机制。在中性粒细胞中，ROSs主要由吞噬细胞NADPH氧化酶(又名NOX2复合物)产生。在细胞激活过程中，NADPH氧化酶NCF2(p67phox)、Rac1和/或Rac2、NCF4 (p40-phox)和NCF1 (p47phox)的胞质成分被招募到细胞膜上，与膜蛋白CYBA(p22-phox)和CYBB (gp91或cytochrom -b558亚基β)形成复合物。我们根据7个NADPH氧化酶相关基因的表达来评估NADPH氧化酶的评分(补充图6d)。

Comparison of scRNA-seq-defined populations with morphology-defined neutrophil subpopulations. To benchmark single-cell transcriptomic neutrophil classification against existing morphological classification schemes, we deconvoluted bulk RNA-seq profiles based on the expression of scRNA-seq-identified group-specific signatures. This approach was similar to other existing deconvolution methods such as CIBERSORT66, but we used a linear regression model with the constraint of non-negative coefficients (that is, the non-negative least-squares problem) instead of the linear support vector regression in CIBERSORT. Although we manually chose 20 genes with the highest fold-changes as signatures for each single-cell group, we noted that the deconvolution in our case was robust to the choice of signatures. The regression model was built using the R package nnls (version 1.4)67. Bulk profiles were quantile normalized. At different morphology-defined neutrophil differentiation stages, neutrophils produce different granules containing distinct enzymes and antimicrobial compounds. Thus, we also examined the expression of various granule genes in differentiating neutrophils. Genes related to primary (azurophilic) granules such as Mpo started to be expressed in some G0 cells, peaked in G1 cells and then rapidly decreased in G2 cells (Fig. 2a,b). Myeloperoxidase (MPO)-negative granules can be divided into granules containing LTF but no gelatinase (MMP9), granules that contain both and granules that contain gelatinase but no LTF68. We found sequential production of these granules in maturing neutrophils, with LTF-containing granules emerging in G2 cells, LTF and gelatinase-containing granules emerging in G3 cells, and gelatinase-containing granules (LTF low) emerging in G4 cells (Fig. 2a–c). Of the proteins that localize exclusively to secretory vesicles such as FPR1 (encoded by Frp1) and V AMP2 (encoded by Vamp2), their cognate mRNA profiles peaked in G4 cells in BM and continued to be expressed in PB neutrophils.

根据现有的形态学分类方案，我们基于scrna-seq识别的群体特异性签名的表达来反卷积整体RNA-seq谱。这种方法类似于其他现有的反卷积方法，如CIBERSORT66，但我们使用了一个带有非负系数约束的线性回归模型(即非负最小二乘问题)，而不是CIBERSORT中的线性支持向量回归。尽管我们在每个单细胞组中手动选择了20个变化幅度最大的基因作为签名，但我们注意到，在我们的案例中，反卷积对签名的选择是强有力的。回归模型是使用R包nnls(version1.4)建立的67。批量配置文件被分位数规范化。在不同形态定义的中性粒细胞分化阶段，中性粒细胞产生不同的颗粒，含有不同的酶和抗菌化合物。因此，我们也检测了不同颗粒基因在分化中性粒细胞中的表达。Mpo等原代(嗜蓝)颗粒相关基因在部分G0细胞中开始表达，在G1细胞达到高峰，然后在G2细胞中迅速下降(图2a,b)。髓过氧化物酶(MPO)阴性颗粒可分为含LTF不含明胶酶(MMP9)颗粒、含两者的颗粒和含明胶酶不含LTF68颗粒。我们发现这些颗粒在成熟的中性粒细胞中相继产生，G2细胞中出现含LTF颗粒，G3细胞中出现LTF和含明胶酶颗粒，G4细胞中出现含明胶酶颗粒(LTFlow)(图2a-c)。在特异性定位于分泌囊泡的蛋白中，如FPR1(由Frp1编码)和VAMP2(由Vamp2编码)，它们的同源mRNA谱在BM的G4细胞中达到峰值，并继续在PB中性粒细胞中表达。

SCENIC analysis. SCENIC is a computational tool that infers regulatory modules or regulons by analyzing the co-expression of transcription factors and their putative target genes characterized by enrichment of corresponding transcription factor-binding motifs in their regulatory regions40. Regulatory network analysis was performed on all control and E. coli-challenged samples using the Python package pySCENIC (version 0.9.11)40 with default parameters. We scaled the network inference step by first inferring regulons on a 6,000-cell subset, then calculated AUCell scores for all 32,888 cells included in this analysis. Specifically, we randomly sampled 300 cells from each neutrophil population in each condition and 1,200 non-neutrophil cells as the training set for network inference. Output co-expression modules were trimmed with cisTarget databases (mm10_refseq-r80_500bp_up_ and_100bp_down_tss, mm9-tss-centered-10kb-7species and mm9-500bp-upstream- 10species). The identified 413 regulons were then scored to determine their activities in each cell. k-means clustering was performed on the first 20 principle components of the regulon activity matrix with the cluster number k = 7.

风景优美的分析。SCENIC是一种计算工具，通过分析转录因子及其预期靶基因的共表达来推断调控模块或调控子，其特征是在其调控区域富集相应的转录因子结合基序40。使用带有默认参数的Python包pySCENIC(version0.9.11)40对所有控制样本和大肠杆菌感染样本进行了调节网络分析。我们通过首先推断6000个细胞子集上的规则来扩展网络推理步骤，然后计算包含在该分析中的所有32,888个细胞的AUCell分数。具体来说，我们从每个中性粒细胞群中随机抽取300个细胞，在每个条件下抽取1200个非中性粒细胞作为网络推理的训练集。输出的共表达模块使用cisTarget数据库(mm10_refseq-r80_500bp_up_和_100bp_down_tss、mm9-tss-centered10kb-7species和mm9-500bp-upstream10species)进行修剪。然后对识别出的413个调控子进行评分，以确定它们在每个细胞中的活动。对调控子活性矩阵的前20个主成分进行k-means聚类，聚类数为k= 7。

Differential activity analysis of SCENIC regulons. To assess the effect of biological conditions on regulon activity, we applied a GLM as reported in ref. 41. We compared the AUCell score of each regulon with different baseline clusters corresponding to different biological questions such as neutrophil cluster transition and infection response. The GLM results were further filtered by P values and visualized using the R package ComplexHeatmap69.

景区规则的差异活动分析。为了评估生物条件对调控子活性的影响，我们应用了参考文献41中报道的GLM。我们将每个调控子的AUCell评分与不同的基线簇进行比较，这些基线簇对应不同的生物学问题，如中性粒细胞簇过渡和感染反应。GLM结果通过P值进一步过滤，并使用R包ComplexHeatmap69进行可视化。

RNA velocity analysis. Cell RNA velocity analysis was performed using the Velocyto program34. This approach uses the relative proportion of unspliced and spliced mRNA abundance as an indicator of the future cell state34. The calculated RNA velocity is a vector that predicts individual cell transition, with the direction and speed of each transition assessed based on the amplitude and direction of individual cell velocity arrows on the UMAP plot. Accordingly, the hierarchical relationship between two cell populations can be inferred by the directional flow in the RNA velocity vector field. Annotation of spliced and unspliced reads was first performed using velocyto.py command-line tools. Then, downstream analysis was performed using the velocyto.R pipeline. We retained the genes expressed in at least one cell population. In total, 4,815 genes were used for the analysis. RNA velocities of each cell were estimated using the gene.relative.velocity.estimates function. Finally, the velocity field was projected onto the existing UMAP space.

RNA速度分析。使用Velocyto程序进行细胞RNA速度分析34。这种方法使用未剪接和剪接的mRNA丰度的相对比例作为未来细胞状态的指示器34。计算出的RNA速度是预测单个细胞转变的矢量，每个转变的方向和速度是根据UMAP图上单个细胞速度箭头的振幅和方向评估的。因此，两个细胞群之间的层次关系可以通过RNA速度矢量场的方向流动来推断。拼接和未拼接读取的注释首先使用velocyto.py命令行工具执行。然后，使用velocyto进行下游分析。R管道。我们保留了在至少一个细胞群中表达的基因。总共有4815个基因被用于分析。利用基因。相对。速度。估计函数估计每个细胞的RNA速度。最后，将速度场投影到现有的UMAP空间上。

Cell label transfer. Total cells were partitioned into distinct cell types annotated by the expression of known marker genes. Neutrophils in their steady state were partitioned into eight clusters based on gene expression profiles annotated according to their development order. E. coli-challenged neutrophils were annotated using a well-accepted method42. Briefly, we first identified pairwise correspondences (also known as anchors) between single cells across datasets (before and after E. coli challenge) to quantify the batch effect. Each cell in the E. coli-challenged dataset was then annotated based on the transcriptomic similarity between this cell and cells in the reference dataset. Specifically, cells would receive corresponding labels with the highest similarity scores, whereas cells with the highest similarity score lower than 0.5 were defined as unassigned. The unassigned cells counted for <10% of the total cell population, distributed randomly on the UMAP plot, and thus were excluded from further investigation. In this way, each neutrophil from the new stimulated dataset was assigned a cluster name, and neutrophils sharing similar transcriptomic profiles were placed into the same cluster. Hence, each cell in the bacterial infection state was assigned to one of the nine cluster labels. This transfer procedure was implemented using the FindTransferAnchors (dims = 1:15) and TransferData (dims = 1:15) functions in Seurat version 3.0.2 (ref. 42) with the combination of top 100 DEGs of each cluster.

电池标签转移。通过已知标记基因的表达，将总细胞分成不同的细胞类型。根据中性粒细胞发育顺序注释的基因表达谱，将稳定状态的中性粒细胞划分为8个簇。大肠杆菌感染的中性粒细胞用一种公认的方法注释42。简单地说，我们首先确定了跨数据集(大肠杆菌挑战之前和之后)单个细胞之间的成对对应(也称为锚定)，以量化批量效应。然后根据大肠杆菌挑战数据集中的每个细胞与参考数据集中的细胞之间的转录组相似性进行注释。具体来说，相似度评分最高的细胞将得到相应的标签，而相似度评分最高低于0.5的细胞被定义为未分配。未分配细胞数小于细胞总数的10%，随机分布在UMAP图上，因此被排除在进一步的调查之外。通过这种方式，每个来自新的受刺激数据集的中性粒细胞被分配一个簇名，和具有相似转录组谱的中性粒细胞被放置到相同的簇中。因此，每个处于细菌感染状态的细胞都被分配到9个簇标签中的一个。使用Seuratversion 3.0.2 (ref. 42)中的FindTransferAnchors (dims = 1:15)和TransferData (dims =1:15)函数，结合每个集群的前100个DEGs，实现了这个转移过程。

Correlation of scRNA-seq-defined neutrophil populations with previously reported neutrophil subpopulations. Previous studies revealed a variety of distinct BM neutrophil subpopulations arising during differentiation and maturation. Using scRNA-seq coupled with a new analytical tool, iterative clustering and guide-gene selection and clonogenic assays, Olsson et al.25 analyzed discrete genomic states and the transitional intermediates that span myelopoiesis. They performed scRNA-seq on stem/multipotent progenitor cells, CMP cells, GMP cells and LK CD34+ cells (lin−c-Kit+CD34+) that included granulocytic precursors. We calculated the fraction of each scRNA-seq-defined cluster in the four samples. Each cell in these samples was annotated using the cell label transfer method described above. The cluster identity of each cell was inferred based on the transcriptomic similarity between this cell and the reference clusters (G0–G5) defined in the current study. This same method was used to analyze the C1 and C2 neutrophil clusters reported by Zhu et al.27. Recently, a proliferative unipotent neutrophil precursor that suppresses T cell activation and promotes tumor growth was identified in the mouse BM that generates neutrophils after intra-BM adoptive transfer. scRNA-seq analysis of BM Lin−c-kit+Ly6A/E-Ly6G+/low cells revealed two populations: an early-stage c-kit+Gfi1lowCebpahiLy6G−/low progenitor with stem cell morphology (C1) and a late-stage c-kit+Gfi1hiCebpalowLy6G+ precursor with morphological features similar to transient neutrophil precursors (C2)27. Further analysis showed that cluster C1 is the early-stage committed unipotent neutrophil progenitor. Interestingly, the late-stage progenitors were mostly similar to the preNeu population identified by Evrard et al.27. In the current study, the raw data related to C1 and C2 cells were retrieved and reanalyzed. We annotated each cell as described above, and performed PCA and t-distributed stochastic neighbor embedding and clustering using the same arguments as in the original publication27 (principal components 1–12; resolution parameter set at 0.03). Using mass cytometry (CyTOF) and cell cycle–based analysis, Evrard et al.26 identified three neutrophil subsets within BM: committed c-Kitlow/int proliferative neutrophil precursors expressing primary and secondary granule proteins (preNeu); CXCR2low non-proliferating immature neutrophils highly expressing secondary granule proteins (immature neutrophils); and CXCR2high mature neutrophils highly expressing gelatinase granule proteins (mature neutrophils). To reveal the correlation of these neutrophil subtypes with scRNA-seq-defined neutrophil populations, we applied the same regression-based deconvolution approach as was used for comparing scRNA-seq-defined populations with morphology-defined neutrophil subpopulations (see above). We deconvoluted bulk RNA-seq profiles of BM GMPs, preNeu cells, immature neutrophils and mature neutrophils, as well as PB neutrophils based on their expression of scRNA-seq-identified group-specific signatures. The 20 highest DEGs of each single-cell group (G0–G5) were selected as signatures for deconvolution. Finally, Giladi et al.13 also defined two BM neutrophil subpopulations. c-Kit+ stage I neutrophils express a set of genes used to approximate the neutrophil differentiation axis, while stage II neutrophils display a mature neutrophil signature defined by genes upregulated in the most terminally differentiated neutrophils. The initial increase in the differentiation of stage I neutrophils is independent of PU.1, but further neutrophil maturation and activation of stage II genes is completely blocked in PU.1 knockout13. To reveal the correlation of stage I and stage II neutrophils with scRNA-seq-defined neutrophil populations, individual cells in each scRNA-seq-defined neutrophil cluster were scored for their expression of stage I and stage II gene signatures13. The stage I and stage II scores were defined as the average normalized expression of corresponding genes.

scrna-seq定义的中性粒细胞群与先前报道的中性粒细胞亚群的相关性。以往的研究表明，在骨髓细胞分化和成熟过程中出现了多种不同的中性粒细胞亚群。Olsson等人利用scRNA-seq结合新的分析工具、迭代聚类、引导基因选择和克隆分析，分析了跨越骨髓生成的离散基因组状态和过渡中间产物。他们对包括粒细胞前体的干细胞/多能祖细胞、CMP细胞、GMP细胞和LKCD34+细胞(lin−c-Kit+CD34+)进行了scRNA-seq。我们计算了四个样本中每个scrna-seq定义的簇的比例。这些样本中的每个单元都使用上面描述的单元标签转移方法进行了注释。每个细胞的聚类特性是基于该细胞与本研究中定义的参考聚类(G0-G5)之间的转录组相似性推断的。同样的方法也用于分析Zhu等报道的C1和C2中性粒细胞簇27。最近，在小鼠骨髓内过继转移后产生中性粒细胞的小鼠骨髓中发现了一种可抑制T细胞激活并促进肿瘤生长的增生性单效中性粒细胞前体。BMLin−c-kit+Ly6A/E-Ly6G+/low细胞的scRNA-seq分析显示了两个群体:具有干细胞形态的早期c-kit+Gfi1lowCebpahiLy6G−/low祖细胞(C1)和晚期c-kit+Gfi1hiCebpalowLy6G+前体，其形态特征类似于短暂中性粒细胞前体(C2)27。进一步分析表明，簇C1是早期承诺单效中性粒细胞祖细胞。有趣的是，晚期祖细胞大多与Evrard等人鉴定的preNeu群体相似27。在本研究中，我们检索了与C1和C2细胞相关的原始数据并重新分析。我们如上所述对每个细胞进行注释，并使用与原始出版物相同的参数进行PCA和t分布随机邻居嵌入和聚类(主成分1-12;分辨率参数设置为0.03)。Evrard等26使用细胞计数法(CyTOF)和基于细胞周期的分析，在骨髓中发现了三种中性粒细胞亚群:表达初级和次级颗粒蛋白(preNeu)的c-Kitlow/int增殖中性粒细胞前体;CXCR2low非增殖未成熟中性粒细胞高表达二级颗粒蛋白(未成熟中性粒细胞);cxcr2高成熟中性粒细胞高度表达明胶酶颗粒蛋白(成熟中性粒细胞)。为了揭示这些中性粒细胞亚型与scrna-seq定义的中性粒细胞群体之间的相关性，我们应用了与比较scrna-seq定义的中性粒细胞群体与形态学定义的中性粒细胞亚群相同的基于回归的反卷积方法(见上文)。基于scrna -seq鉴定的群特异性特征，我们对BMgmp、preNeu细胞、未成熟中性粒细胞和成熟中性粒细胞以及PB中性粒细胞的整体RNA-seq谱进行了反卷。从每个单细胞组(G0-G5)中选出20个最高的DEGs作为反褶积特征。最后，Giladi等13也定义了两个BM中性粒细胞亚群。c-Kit+I期中性粒细胞表达一组用于接近中性粒细胞分化轴的基因，而II期中性粒细胞显示成熟的中性粒细胞信号，这是由最终末分化中性粒细胞中上调的基因所定义的。I期中性粒细胞分化的初始增加独立于pu1，但在pu1敲除后，进一步的中性粒细胞成熟和II期基因的激活被完全阻断。为了揭示I期和II期中性粒细胞与scrna-seq定义的中性粒细胞群的相关性，我们对每个scrna-seq定义的中性粒细胞群中的单个细胞进行了I期和II期基因标记表达评分13。I期和II期评分定义为相应基因的平均归一化表达。

Statistical analyses. For most experiments, comparisons were made using a two-tailed, unpaired Student’s t-test. The values shown in each figure represent means ± s.d. P < 0.05 was considered statistically significant. All experiments were repeated at least three times. All statistical analyses and graphics were made using GraphPad Prism (GraphPad) and R (The R Project for Statistical Computing).

统计分析。

在大多数实验中，比较使用的是双尾、未配对的t检验。各图所示为平均值±标准差，P <0.05为差异有统计学意义。所有实验都至少重复三次。所有统计分析和图形都是使用GraphPad Prism (GraphPad)和R(统计计算的R项目)。