1. In vivo validation of drug vulnerability based on NBM features of ONB by establishment of PDX
Immunodeficient NOD SCID gamma (NSG) and Balb/c(Nu/Nu) mice (male, 18g-20 g, 6-8 weeks at the time of receipt, Vital River Laboratory Animal Technology Co., Ltd. Zhejiang, China) were used for PDX model development (Extended Data Fig.3c-3f). The procedures of animal handling and tissue harvesting were approved by the Experimental Animal Department of Fudan University (Permit Number: 202104036S). All animal studies were conducted in accordance with the Animal Research Reporting of In Vivo (ARRIVE) guidelines. NSG and Balb/c-nu mice were housed in individually ventilated polysulfone cages in a temperature- (20°C-26°C, humidity:40%-70%) and light-controlled room (12h light 12h dark cycle, lights on at 7:00 a.m.) Experiment protocol will end when the maximal size of established tumors was around 2.0 cm in any direction. The maximal tumor size was not exceeded in this study.
In order to establish tumor xenografts of ONB-311, tumor tissue collected during endoscopic surgery was added to a 1.5 ml tube with cell freezing medium (Invitrogen, Carlsbad, CA, USA) and minced into approximately 1-mm3 fragments and then inoculated subcutaneously into NSG mice. Subsequent passages were made in a similar fashion in NSG mice. Xenografted tumor for the generation of second-passage in NSG mice were harvested and evenly minced into 2-mm3 fragments and implanted subcutaneously into the right flank of each Balb/c-nu mouse for further in vivo efficacy test of drugs and radiotherapy.
Drug administration started at 15 days after implantation. Mice bearing established tumors (149–168 mm3) were randomly allocated into four groups according to the size of the tumors, with the average volume of each group basically the same. The four groups were set as Radiotherapy (N=5), Radiotherapy + Endostatin (N=5), Radiotherapy + Doxorubicin (N=4) and a vehicle control group (N=5). Mice in Radiotherapy, Radiotherapy + Endostatin and Radiotherapy + Doxorubicin group received 4Gy Cs137 γ-ray radiation at the first day before drug administration (D1). Then, mice in Radiotherapy + Endostatin group were treated with endostatin (Endostar, Simcere-Medgenn Bio-pharmaceutical Co.,Ltd. Shandong, China) for 14 days (50 mg/kg, iv, qd, D1-D14). Mice in Radiotherapy + Doxorubicin group were treated with doxorubicin (ChemeGen, Shanghai, China) at the first day of drug administration (8mg/kg, iv, qd, D1).
Throughout the study, the tumor diameter was measured with calipers twice a week. The tumor volume was calculated using the following formula: tumor volume (TV, mm3) = (a × b2)/2, where a is the longest diameter of the tumor and b is the shortest diameter of the tumor; relative tumor volume (RTV, mm3) = Vt/Vo, where Vo is the tumor volume measured at the time of cage separation (D1), and Vt is the tumor volume at each measurement. The relative tumor proliferation rate T/C (%) and the relative tumor proliferation inhibition rate were both adopted for the evaluation of anti-tumor activity, where T/C (%) = TRTV/CRTV◊100%, and the relative tumor proliferation inhibition rate (%) = (1-T/C)◊100%. Average tumor volumes were calculated for each group, and tumor growth curves were generated as a function of time, where T test was conducted. Animals were sacrificed at 20 days after first drug administration (34 days after tumor implantation). Tumors from each group were collected at the end of the experiment for further analysis.
2. Construction of Pseudo-time Trajectory
The R package Monocle2 (version 2.14.0) was adopted to perform pseudo-time analysis in order to reveal putative trajectories of differentiation and cell-state transition. The developmental trajectory of CD8+ T cells, CD4+ T cells, macrophages and DCs, were constructed respectively. Practically, the data of the indicated clusters calculated in Seurat was fed directly into Monocle2. The function DDRTree was adopted for dimensionality reduction, after which the cells were placed on pseudo-time trajectories by calling the function orderCells. Then the immune cell differentiation trajectories were visualized by the plot_cell_trajectory function with the default parameters of Monocle2. The pseudo-heatmaps reflecting the expression of marker genes varying along the developmental trajectory were generated by the plot_genes_branched_heatmap function. DEGs over the pseudo-time trajectory between cells in different differentiation fate branches were generated by applying the differentialGeneTest function in Monocle2.
We also applied Slingshot (Version 1.5.1), a cell lineage inference algorithm, to uncover the differentiation trajectories and bifurcations of olfactory neural and sustentacular lineages. Slingshot generated principal curves and cell developmental distances for each lineage, which was then mapped to UMAP projection for visualization. The ordering provided by Slingshot, analogous to pseudotime, is referred to herein as developmental order. Based on the inferred developmental distance, heatmaps displaying the average scaled expression profiles of NBM (Neural/Basal/Mesenchymal) ONB subtypes and cell cycle-related features were produced, with cells ordered according to their developmental positions within the olfactory neuronal and sustentacular cell lineages.
The scRNA profiles of normal adult olfactory epithelial cells (OM2) and one Basal subtype ONB tumor dataset (ONB-311) were integrated using the Seurat function merge(). To place ONB and OM cells in a pseudotime trajectory, the integration was re-processed using R package Monocle2. Expression files were normalized using the estimateSizeFactors() function, with genes expressed in fewer than 3 cells excluded. Using the estimateDispersions() function to estimate negative binomial over-dispersion for each gene, genes with mean expression value > 0.1 and variance greater than the empirical dispersion (the best fit mean-dispersion trend-line) were selected as ordering genes for pseudotime analysis.
3. Hematoxylin and Eosin (H&E) Staining, Immunohistochemistry (IHC) and immunofluorescence (IF)
Immunohistochemical (IHC) procedures as well as Hematoxylin and Eosin (H&E) staining were performed on 3-5 µm sections of formalin-fixed, paraffin-embedded human tumor or normal tissues as described previously. The following antibodies were used for IHC: anti-Ki67 (1:1500, Abcam, #ab15580), anti-CTGF/CCN2 (1:200, Thermofisher, #PA5-109248), anti-Chromogranin (1:50, Gene Tech (Shanghai), #GT211429), anti-Tubulin β3/TUBB3 (1:300, BioLegend, #MMS-435P) and anti-CD34 (1: 2000, Abcam, #ab81289). The Dako REAL EnVision System (K5007, DAKO, Glostrup, Denmark) was used as the biotinylated second antibody and color substrate solution. Microvessels were detected by CD34, and microvessel density (MVD) was calculated according to methodology described in our previous studies82. Expression levels of CCN2 and CHG were evaluated semi-quantitatively with histochemistry scores (H-scores), based on the percentage of staining positive cells and staining intensity: H-score = ΣPi(i+1). Here, Pi represented the percentage of the number of positive cells in the total number of cells in a selected field, and i stands for staining intensity evaluated by two qualified pathologists respectively. Proliferation index (Ki-67% index) referred to the percentage of Ki-67 positive cells in the total number of cells in a selected field.
Immunofluorescence (IF) specific for markers including CD8 (1:800, Abcam, #ab23T710), CD4 (1:400, Abcam, #ab133616), CD3 (1:200, Abcam, #ab16669), FOXP3 (1:100, Abcam, #ab215206), LAG3 (1:100, Abcam, #ab209236), CD68 (1:00, Abcam, #ab955), VEGFA (1:100, Abcam, #ab52917) and S100A4 (1:1000, Abcam, #ab197896) was performed. Multiple process including dewaxing, rehydration, block of endogenous peroxidase activity and non-specific antigens, retrieval of heating induced epitope were performed on 3-5 µm paraffin-embedded slides. After incubating with primary antibodies and horseradish peroxidase-conjugated secondary antibody, the slides were incubated with tyramide signal amplification (TSA) Fluorochromes (Runnerbio, Shanghai, China) for 10-30 min at room temperature. After the second or third round of antibody incubation and PBS washing, the slides were mounted using Antifade Mounting Medium with DAPI (Beyotime, P0131, Jiangsu, China).