Bcl-2 Gene

We performed an immunohistochemistry (IHC) using tissues obtained before treatment. Estrogen receptor (ER), progesterone receptor (PR), human epidermal growth factor receptor 2 (HER2), p53, bcl-2, and Ki-67 expressions were evaluated. IHC was performed as previously described [14 (link),22 (link)]. ER and PR positivity was defined as ≥10% positive tumor cells with nuclear staining. HER2 positivity was defined as either HER2 gene amplification by fluorescent in situ hybridization or scored as 3+ by IHC [23 (link)]. In case of HER2 2(+), fluorescent in situ hybridization was performed to determine HER2 positivity. TNBC was defined as ER(-), PR(-), and HER2(-), regardless of the expression of EGFR and basal cytokeratins. Only cytoplasmic staining was scored as positive for bcl-2, regardless of the intensity of the stained cells. Cells stained for Ki-67 and p53 were counted and expressed as a percentage. The percentage was determined by the number of Ki-67 positive cells among the total number of counted tumor cells. High expression of Ki-67 was defined as ≥10%, because 10% as cutoff provided the best prognosis-prediction results in our institute [14 (link)]. Specimens with no residual invasive carcinoma in the both breast and lymph nodes were classified as pathologic complete response (pCR). Residual ductal carcinoma in situ was also included in the pCR category [24 (link)]. Otherwise the specimens which did not achieve pCR category were classified as residual disease

Starting from BCL files obtained from Illumina sequencing, we ran cellranger mkfastq to extract sequence reads in FASTQ format, followed by cellranger count to generate gene-count matrices from the FASTQ files. Since our data are from single nuclei, we built and aligned reads to genome references with pre-mRNA annotations, which account for both exons and introns. Pre-mRNA annotations improve the number of detected genes significantly compared to a reference with only exon annotations^{15 (link)}. For human and mouse data, we used the GRCh38 and mm10 genome references, respectively. To compare samples of interest (e.g., different loading concentrations), we pooled their gene-count matrices together, and filtered out low-quality nuclei identified based on any one of the following criteria: (1) a total number of expressed genes <200; (2) a total number of expressed genes > = 6000; or (3) a percentage of RNA UMIs from mitochondrial genes > = 10%. We then normalized and transformed the filtered count matrix to natural log space as follows: (1) selected genes that were expressed in at least 0.05% of all remaining nuclei; (2) normalized the count vector of each nucleus such that the total sum of normalized counts from selected genes is equal to 100,000 (transcripts per 100 K, TP100K); (3) transformed the normalized matrix into the natural log space by replacing each normalized count c with $log\left(c+1\right)$ (log(TP100K+1)). We performed dimensionality reduction, clustering and visualization on the log-transformed matrix using a standard procedure^{16 (link),17 (link)}. Specifically, we selected highly variable genes^{18 (link)} with a z-score cutoff at 0.5, performed PCA on the standardized sub-matrix consisting of only highly variable genes and selected the top 50 principal components (PCs)^{19 (link)}, clustered the data based on the 50 selected PCs using the Louvain community detection algorithm^{20 (link)} with a resolution at 1.3. We identified cluster-specific gene expression by differential expression analyses between nuclei within the cluster and outside of the cluster^{16 (link)} using Welch’s t test and Fisher’s exact test; controlled false discovery rates (FDR) at 5% using the Benjamini–Hochberg procedure²¹ , and annotated putative cell types based on legacy signatures of human and mouse brain cells. We visualized the reduced dimensionality data using tSNE²² with a perplexity at 30. Note that in experiments 1 and 4 (Supplementary Data 1), we identified one cluster that did not express any known cell-type markers and had the lowest median number of RNA UMIs among all clusters. We removed it from further analysis, and repeated the above analysis workflow, except the low-quality nucleus filtration step.

Primer sequences specific to Bid, Bak, Bcl-xL, Bax, Bcl-2 and Bad genes were used to ensure targeted amplification. The expression levels of these genes were normalized against a housekeeping gene (GAPDH). The oligonucleotides of the primers used in the present study were supplied by Thermo Fisher Scientific, Inc., Waltham, MA, USA, and their sequence is presented in detail in Table 1.
The analysis was conducted using the 2^−ΔΔCt method, which allowed for the determination of fold changes in gene expression relative to untreated control cells.

As focal points, BAX, BAK, BCL-2, IRF-3, IRF-7, and IFN-β genes, along with an internal reference gene (β-actin), were selected. To acquire the gene sequences, we utilized the NCBI database; subsequently, Allele ID software was employed to formulate the primer sets. The oligonucleotide sequences are displayed in Table 1.Table 1

Primer sequences for BAX, BAK, BCL-2, IRF3, IRF7, IFN-β, and β-actin genes.

Genes	Orientation primer	Sequence (5′ → 3′)
BAX	Forward	F: AGGGTGGCTGGGAAGGC
	Reverse	R: TGAGCGAGGCGGTGAGG
BAK	Forward	F: GCCTACTGACCCAGAGATGG
	Reverse	R: CTCATAGGCGTTGTCTGCTG
BCL-2	Forward	F: ATGTGTGTGGAGACCGTCAA
	Reverse	R: GCCGTACAGTTCCACAAAGG
IRF-3	Forward	F: CGGAAAGAAGTGTTGCGGTTAG
	Reverse	R: TTTGCCATTGGTGTCAGGAGAG
IRF-7	Forward	F: TGCAAGGTGTACTGGGAG
	Reverse	R: TCAAGCTTCTGCTCCAGCTCCATAAG
IFN-β	Forward	F: CAACTTGCTTGGATTCCTACAAAG
	Reverse	R: TATTCAAGCCTCCCATTCAATTG
β-actin	Forward	F: CCTGGCACCCAGCACAAT
	Reverse	R: GCCGATCCACACGGAGATCT