Abstract

   The interaction between SCAF4 and RNA polymerase II (POLR2A) is crucial
   for proper mRNA termination, with its dysregulation leading to
   truncated mRNAs and nonfunctional proteins, impairing cellular growth.
   Despite its potential relevance, the role of this interaction in
   triple‐negative breast cancer (TNBC) remains unexplored due to the lack
   of effective molecular tools. To address this, we employed SRiApt, an
   aptamer generated through the recently established Blocker‐SELEX
   pipeline. Its biological stability is improved by phosphorothioate
   modifications to form PT[f]‐SRiApt. Using this aptamer, the critical
   role of the SCAF4‐POLR2A interaction in driving TNBC tumor growth and
   immune regulation is uncovered. PT[f]‐SRiApt effectively inhibits tumor
   growth and induces cell cycle arrest in TNBC cells with elevated SCAF4
   and POLR2A expression. Additionally, PT[f]‐SRiApt promotes premature
   mRNA termination, boosting antigen presentation and promoting T‐cell
   infiltration. Analysis of patient samples further confirmed the
   negative correlation between SCAF4‐POLR2A interaction and the
   effectiveness of immunotherapy, highlighting the potential of
   PT[f]‐SRiApt in improving immune efficacy. Together, the work provides
   a powerful tool not only for dissecting previously “undruggable”
   protein‐protein interactions but also for enhancing tumor
   immunogenicity and reshaping the tumor microenvironment.

   Keywords: aptamers, cancers, personalized medicine, protein‐protein
   interactions, transcription factors
     __________________________________________________________________

   A phosphorothioate‐modified aptamer PT[f]‐SRiApt capable of disrupting
   “undruggable” transcription factor SCAF4‐POLR2A interaction with high
   inhibitory activity and biological stability is presented here. Through
   this inhibition, PT[f]‐SRiApt successfully arrested cell cycle
   progression, induced cell death, and elevated immune recruitment in
   SCAF4‐POLR2A‐high triple‐negative breast cancer. This novel inhibitor
   also demonstrated potential clinical efficacy in combination with
   immune checkpoint blockade in a wide range of cancers.

   graphic file with name ADVS-12-e00433-g005.jpg

1. Introduction

   Triple‐negative breast cancer (TNBC) is an aggressive subtype of breast
   cancer distinguished by the presence of unique super‐enhancers that
   regulate critical oncogenes involved in tumor progression.^[ [50]^1 ^]
   These enhancer regions are characterized by sustained active
   transcription and exhibit heightened sensitivity to disruptions in
   transcriptional activity.^[ [51]^2 , [52]^3 ^] While the targeting of
   kinases that influence the transcriptional process, such as eEF2K and
   various cyclin‐dependent kinases that modulate transcription factor
   complexes,^[ [53]^4 , [54]^5 , [55]^6 ^] has emerged as a prominent
   approach to TNBC treatment, there remains a lack of therapeutic
   strategies that directly target transcription factors (TFs) in this
   context. One important TF interaction driving transcriptional stability
   is between SCAF4 and POLR2A.^[ [56]^7 , [57]^8 , [58]^9 , [59]^10 ^]
   This interaction is critical for proper mRNA termination, ensuring the
   full‐length expression of functional proteins by suppressing early
   alternative polyadenylation sites, which otherwise produce truncated,
   non‐functional mRNAs.^[ [60]^7 ^] While this interaction's role in
   normal cellular function is well‐characterized, its contribution to
   cancer biology remains understudied, largely due to the lack of
   small‐molecule inhibitors – an issue attributed to the “undruggable”
   nature of the involved protein surfaces.^[ [61]^11 ^]

   Aptamers, short oligonucleotides that bind targets with high
   specificity and affinity, represent a powerful alternative approach for
   modulating protein‐protein interactions.^[ [62]^12 ^] Unlike small
   molecules, aptamers exhibit remarkable versatility in binding a broad
   range of targets, including proteins, lipids, and metabolites,
   positioning them as valuable chemical tools in regulating
   protein‐protein interactions.^[ [63]^13 , [64]^14 ^] Several
   aptamer‐based therapeutics targeting protein‐protein interactions have
   achieved significant regulatory milestones. Pegaptanib sodium (Macugen)
   and avacincaptad pegol (IZERVAY) have received the U.S. Food and Drug
   Administration (FDA) approval, underscoring their clinical relevance.^[
   [65]^15 , [66]^16 ^] Other candidates, such as lexaptepid pegol
   (NOX‐H94) and BB‐031, are currently in phase II clinical trials.^[
   [67]^17 , [68]^18 , [69]^19 ^] Additionally, the FDA has granted orphan
   drug designation to Apc001PE.^[ [70]^20 ^] Collectively, these advances
   highlight the clinical viability and growing potential of aptamers as
   therapeutic agents.

   To develop aptamers that can specifically block protein‐protein
   interactions, we recently reported the Blocker‐SELEX pipeline.^[
   [71]^21 ^] Among the inhibitory aptamers identified, SRiApt has shown
   practical promise by effectively inhibiting the interaction between
   SCAF4/8 and POLR2A. However, the poor biological stability of SRiApt
   has limited its cellular applicability and hindered a comprehensive
   evaluation of the therapeutic potential of targeting SCAF4‐POLR2A
   interactions.

   In this study, we first addressed this issue by improving the
   biological stability of SRiApt through phosphorothioate (PT)
   modifications.^[ [72]^22 , [73]^23 ^] Leveraging this PT‐modified
   SRiApt, we demonstrated that it selectively induces cell cycle arrest
   and impairs the growth of TNBC cell lines characterized by elevated
   SCAF4‐POLR2A expression. Disruption of SCAF4‐POLR2A interaction with
   PT[f]‐SRiApt promotes transcription early termination and thus
   activates the major histocompatibility complex (MHC) class II pathway.
   Moreover, in vivo assays demonstrated that PT‐modified SRiApt activates
   the antigen presentation pathway, leading to increased infiltration of
   T cells. Collectively, our findings reveal the function and therapeutic
   potential of blocking SCAF4‐POLR2A interaction in TNBC and underscore
   the utility of PT[f]‐SRiApt as a molecular tool for modulating tumor
   immunity in TNBC.

2. Results and Discussion

2.1. PT Modification Improves the Stability and Performance of SRiApt

   SRiApt, developed through our Blocker‐SELEX platform, effectively
   disrupts SCAF4/8‐POLR2A interaction in vitro.^[ [74]^21 ^] However, its
   application is limited by its susceptibility to metabolic degradation.
   Given the crucial role of chemical modifications in optimizing the
   drug‐like properties of nucleic acid‐based therapeutics, we sought to
   improve the stability and efficacy of SRiApt through PT modification, a
   well‐established approach to improve both stability and affinity of
   nucleic acid drugs.^[ [75]^22 , [76]^23 ^]

   In this study, we introduced three levels of PT modification to SRiApt:
   one‐third, one‐half, and full PT substitution (Figure [77] 1A). To
   evaluate the biological stability of these modified aptamers, we
   incubated FAM (Fluorescein Amidite)‐labeled SRiApt variants in DMEM
   with 20% FBS at 37 °C. Electrophoretic analysis and quantification
   revealed that the fully modified PT[f]‐SRiApt exhibited the highest
   stability with only minimal degradation observed, while partial
   modifications (PT[1/3] and PT[1/2]) resulted in more significant
   degradation (Figure [78]1B,C). Next, we evaluated the binding
   affinities of the SRiApt variants to SCAF4, using fluorescence
   polarization (FP) assays. The binding affinities (K[D]) were determined
   to be 0.469±0.010 µm for unmodified SRiApt, 0.223±0.030 µm for
   PT[1/3]‐SRiApt, 0.121±0.054 µm for PT[1/2]‐SRiApt, and 0.073±0.003 µm
   for PT[f]‐SRiApt, with the fully modified aptamer showing the strongest
   binding affinity (Figure [79]1D). Additionally, PT‐modified variants
   exhibited comparable IC[50] values in competitive FP assays for
   inhibiting SCAF4‐POLR2A interaction against unmodified SRiApt:
   0.463±0.202 µm for unmodified SRiApt, 0.377±0.026 µm for
   PT[1/3]‐SRiApt, 0.386±0.034 µm for PT[1/2]‐SRiApt, and 0.379±0.030 µm
   for PT[f]‐SRiApt (Figure [80]1E). Noticeably, PT modification resulted
   in a steeper IC50 binding curve, implying increased binding affinity or
   the stabilization of binding‐competent conformations.^[ [81]^23 ,
   [82]^24 , [83]^25 ^]

Figure 1.

   Figure 1
   [84]Open in a new tab

   Characterization of the biological stability, binding affinities, and
   IC[50] values of PT‐modified SRiApt variants. A) Schematic of SRiApt
   variants with different degrees of PT modification. B‐C) Evaluation of
   the metabolic stability of PT‐modified SRiApt variants in serum,
   showing their susceptibility to degradation. D) Binding affinities of
   PT‐modified SRiApt variants to SCAF4 protein, as determined using
   fluorescence polarization (FP) assays. Data represent the mean ± SD
   (n = 3). E) IC[50] values of PT‐modified SRiApt variants in inhibiting
   SCAF4‐POLR2A interaction, as measured by competitive FP assays. Data
   represent the mean ± SD (n = 3). F) Confocal microscopy images
   depicting the endocytosis of PT[f]‐SRiApt, PT[f]‐CS, and SRiApt in 4T1
   cells. Cells were treated with 1 µm of corresponding FAM‐labelled
   oligonucleotide for 24 h at 37 °C, followed by co‐staining with a
   lysosome tracker and DAPI after formaldehyde fixation. G)
   Quantification of fluorescence intensities from FAM‐PT[f]‐SRiApt (top),
   FAM‐PT[f]‐CS (middle), and FAM‐SRiApt (bottom), lysosome tracker, and
   DAPI in grayscale values following the red line in F).

   These findings demonstrate that PT[f]‐SRiApt exhibits the highest
   binding affinity for SCAF4 with comparable IC[50] values across
   variants. Given these improved characteristics, PT[f]‐SRiApt was
   selected for further cellular intake studies using three cell lines
   (4T1, AGS, MCF1A). When FAM‐labeled PT[f]‐SRiApt (1 µm) was introduced
   to the culture medium, it was efficiently internalized by all cells
   tested within 24 h (Figure [85]1F; Figure [86]S1A, Supporting
   Information). To determine the effects of PT modification on the
   intracellular localization of SRiApt, we performed colocalization
   studies using a lysosome tracker and DAPI nuclear stain. The results
   revealed that FAM‐labeled PT[f]‐SRiApt was evenly distributed
   throughout the cytoplasm with no detectable accumulation in lysosomes,
   indicating effective cellular delivery without lysosomal degradation
   (Figure [87]1G; Figure [88]S1B, Supporting Information). In contrast,
   the unmodified aptamer showed no detectable cellular uptake,
   highlighting the importance of chemical modification for membrane
   permeability. Interestingly, the modified control sequence also
   exhibited cellular uptake comparable to PT[f]‐SRiApt, suggesting that
   the observed internalization was determined by PT modifications
   (Figure [89]1G; Figure [90]S1B, Supporting Information). This
   phenomenon was coherently observed across all three cell lines tested,
   indicating PT modification's consistent and broad enhancement of
   cellular uptake efficiency, in alignment with previously reported
   findings in the literature.^[ [91]^26 ^] In summary, PT modification
   significantly augments SRiApt, improving its biological stability,
   binding affinity, and cellular uptake, reinforcing its potential as a
   modulating agent for inhibiting SCAF4/8‐POLR2A interactions.

2.2. PT[f]‐SRiApt Selectively Impairs the Growth of TNBC Cancer Cells with
Elevated SCAF4 and POLR2A Expression Levels

   The interaction between SCAF4 and POLR2A is integral to maintaining
   transcriptional fidelity, yet its role in cancer remains unexplored.^[
   [92]^8 ^] Through Gene Expression Profiling Interactive Analysis
   (GEPIA)^[ [93]^27 ^] of breast cancer datasets, we uncovered a strong
   positive correlation between SCAF4 and POLR2A gene expression across
   multiple breast cancer subtypes (Figure [94] 2A). Survival analysis
   further revealed that co‐expression of these genes is significantly
   associated with poorer clinical outcomes, including reduced
   relapse‐free survival (RFS) and diminished distant metastasis‐free
   (DMFS) survival (p = 0.036), as analyzed by Kaplan‐Meier Plotter,^[
   [95]^28 ^] highlighting the clinical relevance of this interaction in
   breast cancer progression (Figure [96]2B; Figure [97]S2, Supporting
   Information). Further analysis indicated that TNBC samples exhibited
   significantly higher SCAF4 expression compared to other breast cancer
   subtypes (Figure [98]2C). This suggests that SCAF4 may play a pivotal
   role in transcription regulation and stability in TNBC, making the
   SCAF4‐POLR2A interaction a compelling target for therapeutic
   intervention in this particularly aggressive cancer subtype.

Figure 2.

   Figure 2
   [99]Open in a new tab

   Biological effects of PT[f]‐SRiApt on arresting breast cancer
   progression. A) Spearman correlation analysis of SCAF4 and POLR2A
   expression levels in breast cancer samples, as analyzed using GEPIA. B)
   Kaplan‐Meier survival analysis showing reduced relapse‐free survival
   (RFS) in breast cancer patients with high co‐expression of SCAF4 and
   POLR2A. C) Comparison of SCAF4 expression in TNBC versus other subtypes
   from The Cancer Genome Atlas (TCGA) dataset; **** indicates p < 0.0001,
   determined using two‐tailed Wilcoxon test. D) Expression profiles of
   SCAF4 and POLR2A across seven human TNBC cell lines. E) Growth curves
   of HCC1806 cells treated with PT[f]‐SRiApt or PT[f]‐CS for five days
   (n = 3). Data are presented as mean ± SD; **** indicates p < 0.0001,
   determined using a two‐tailed Student's t‐test. F) Anti‐tumorigenic
   effects of PT[f]‐SRiApt on HCC1806 cells, as determined by crystal
   violet staining assays (n = 3). p‐value determined by two‐tailed
   Student's t‐test. G) Volcano plot displaying the log[2] fold change of
   significantly upregulated and downregulated genes in 4T1 cells treated
   with PT[f]‐SRiApt. Genes with significant changes were shown in red
   (p < 0.05, Log[2] (fold change) > 1). H) GSEA analysis of pathways
   regulated by PT[f]‐SRiApt, showing significant activation of the RNA
   Polymerase II Transcription Termination pathway (p = 0.007) and
   significant downregulation of the Mammary Gland Epithelial Cell
   Proliferation (p = 0.012) and G1/S Cell Cycle pathways (p < 0.001). The
   p‐value was determined using a permutation test. I) Flow cytometric
   analysis of cell cycle distribution in 4T1 cells treated with 1 µm
   PT[f]‐SRiApt or PT[f]‐CS. Data are presented as mean ± SD with p‐values
   calculated using a two‐tailed Student's t‐test. n.s. Indicates not
   significant. J) GSEA of the REACTOME_MHC_Class_II_antigen_presentation
   pathway, significantly upregulated following PT[f]‐SRiApt treatment
   (p = 0.0045). The p‐value was determined using a permutation test.

   To evaluate the antiproliferative effects of PT[f]‐SRiApt, we conducted
   a proliferation screening assay across seven human‐derived TNBC cancer
   cell lines. PT[f]‐SRiApt exhibited potent inhibitory activity,
   particularly in the HCC1806 cell line, while PT[f]‐CS, a fully
   PT‐modified control sequence, had no significant effect (Figure
   [100]S3, Supporting Information). Notably, HCC1806 cells, which were
   the most sensitive to PT[f]‐SRiApt (IC[50] = 4.38 ± 0.04 µm), exhibited
   elevated co‐expression of SCAF4 and POLR2A (Figure [101]2D,E). In
   contrast, cell lines with discordant expressions of these genes showed
   reduced sensitivity to treatment (Figure [102]S3, Supporting
   Information). This trend was similarly observed in gastric cancer cell
   lines AGS and HGC27, further supporting the specificity of
   PT[f]‐SRiApt's anti‐proliferative activity (Figure [103]S4, Supporting
   Information). These data suggest that PT[f]‐SRiApt impairs cell
   proliferation by specifically targeting SCAF4‐POLR2A interaction, with
   therapeutic activity depending on the high expression of these two
   genes. In addition to its anti‐proliferative effects, PT[f]‐SRiApt
   markedly inhibited long‐term tumorigenic potential in HCC1806 cells
   compared to PT[f]‐CS control, as demonstrated by colony formation
   assays (Figure [104]2F). This dual impact on proliferation and
   tumorigenesis strengthens the antitumor activity of PT[f]‐SRiApt for
   TNBC with high SCAF4‐POLR2A expression.

   To explore the underlying molecular mechanisms, we conducted bulk RNA
   sequencing on PT[f]‐SRiApt‐treated HCC1806 cells. Compared to cells
   treated with the control sequence, the transcriptomic analysis revealed
   significant changes in gene expression, with 272 up‐regulated and 182
   downregulated genes up‐on treatment (p < 0.05, Log[2] (fold change) >
   1) (Figure [105]2G). Gene set enrichment analysis (GSEA) identified
   activation of the RNA Polymerase II Transcription Termination pathway
   (p = 0.007), suggesting the on‐target effect of PT[f]‐SRiApt promotes
   premature transcription termination by disrupting the SCAF4‐POLR2A
   interaction (Figure [106]2H). Concurrently, the Mammary Gland
   Epithelial Cell Proliferation pathway was significantly downregulated
   (p = 0.012), correlating with the observed inhibition of cell
   proliferation (Figure [107]2H). Further pathway analysis highlighted
   the downregulation of G1/S cell cycle transition genes, implicating
   PT[f]‐SRiApt in inducing cell cycle arrest at the G0/G1 phase
   (Figure [108]2H). Flow cytometry corroborated this finding, showing a
   marked increase in G0/G1 phase cell population and a corresponding
   decrease in S phase cell population (Figure [109]2I; Figure [110]S5,
   Supporting Information), further supporting the antiproliferative
   activity of PT[f]‐SRiApt.

   Interestingly, PT[f]‐SRiApt treatment also led to the enrichment of the
   MHC class II antigen presentation pathway (Figure [111]2J). This
   suggests that polypeptides derived from prematurely terminated
   transcripts, resulting from the disruption of SCAF4‐POLR2A interaction,
   may undergo rapid proteasomal degradation, thereby promoting neoantigen
   generation and subsequent MHC‐mediated presentation.^[ [112]^29 ^] The
   improved antigen presentation may represent an additional mechanism
   contributing to the antitumor efficacy of PT[f]‐SRiApt, suggesting that
   it could promote immune‐mediated tumor suppression beyond its direct
   inhibitory effects on proliferation.

2.3. PT[f]‐SRiApt Suppresses Tumor Growth In Vivo

   To evaluate the therapeutic potential of PT[f]‐SRiApt in vivo, we first
   investigated its cytotoxic effects on murine‐derived TNBC cell lines.
   PT[f]‐SRiApt exhibited pronounced cytotoxicity toward 4T1 cells with an
   IC[50] of 0.336 ± 0.070 µm, while EMT6 cells displayed limited response
   (Figure [113] 3A). This differential sensitivity aligns with elevated
   expression levels of SCAF4 and POLR2A in 4T1 cells, supporting our
   earlier observations that high SCAF4‐POLR2A co‐expression enhances
   responsiveness to PT[f]‐SRiApt (Figure [114]3B). PT[f]‐SRiApt
   significantly inhibited tumorigenesis in 4T1 cells, as demonstrated by
   a colony formation assay, further supporting its potential for in vivo
   applications (Figure [115]3C).

Figure 3.

   Figure 3
   [116]Open in a new tab

   Pharmacokinetic characterization of PT[f]‐SRiApt in murine 4T1 breast
   cancer model. A) Growth curves of 4T1 and EMT6 cells treated with
   PT[f]‐SRiApt over 5 days. B) Relative expression levels of SCAF4 and
   POLR2A in 4T1 and EMT6 cell lines. The p‐value was determined using a
   two‐way ANOVA test. C) Anti‐tumorigenesis effects of PT[f]‐SRiApt or
   PT[f]‐CS on 4T1 cells, as assessed via crystal violet staining assay
   (n = 3). The p‐value was determined by a two‐tailed Student's t‐test.
   D) Schematic illustration of the fluorescent measurement schedule of
   mice injected with PT[f]‐SRiApt‐Cy7 (n = 5) and Cy7 (n = 5). E)
   Epi‐fluorescence images showing the in vivo distribution of free Cy7
   dye versus PT[f]‐SRiApt‐Cy7 in mice over a 96‐h period post‐injection.
   F) Quantification of fluorescence intensity from panel E over a 96‐h
   period. The line plot depicts the fluorescence signal of each mouse for
   free Cy7 and PT[f]‐SRiApt‐Cy7. The p‐value was determined by a
   two‐tailed Student's t‐test. ****, p < 0.0001.

   To evaluate the in vivo behavior of PT[f]‐SRiApt, we first assessed its
   pharmacokinetics in a 4T1 tumor‐bearing BALB/c mouse model. Once tumors
   were palpable, mice received intratumoral injections of either free Cy7
   dye or PT[f]‐SRiApt conjugated with Cy7, followed by longitudinal
   epi‐fluorescence imaging over 96 h (Figure [117]3D). Compared to free
   Cy7, which exhibited a rapid clearance with a half‐life of ≈2 h,
   PT[f]‐SRiApt‐Cy7 remained confined to the tumor site with no detectable
   systemic distribution and demonstrated significantly prolonged
   intratumoral retention, with a half‐life exceeding 72 h and visible
   clearance by 96 h (Figure [118]3E,F). This clearance profile suggests
   effective metabolism without long‐term accumulation, mitigating the
   risk of chronic toxicity. Overall, these results indicate that
   PT[f]‐SRiApt achieves localized and sustained tumor exposure following
   intratumoral administration, with a favorable safety profile.

   Upon establishing the pharmacokinetic properties of PT[f]‐SRiApt, we
   first investigated its efficacy in vivo employing a 4T1 tumor‐bearing
   model in immunodeficient nude mice. Once the tumors became palpable,
   mice were administered daily with intratumoral injections of 1 µg kg^−1
   of PT[f]‐SRiApt or PT[f]‐CS (Figure [119]S6A, Supporting Information).
   Tumor volumes were measured daily and calculated using the standard
   formula. PT[f]‐SRiApt significantly inhibited tumor growth compared to
   PT[f]‐CS control (Figure [120]S6B, Supporting Information). Body weight
   monitoring revealed no detectable systemic toxicity at this dosage
   (Figure [121]S6C, Supporting Information). We next tested PT[f]‐SRiApt
   in an immunocompetent BALB/c mouse model to further explore its
   therapeutic potential. Mice were grouped into three cohorts, including
   PT[f]‐SRiApt, PT[f]‐CS, and PBS, and treated with 1 µg kg^−1 of
   PT[f]‐SRiApt, PT[f]‐CS, or PBS administered daily via intratumoral
   injection once the tumors became palpable (Figure [122] 4A).
   Remarkably, PT[f]‐SRiApt significantly suppressed tumor growth compared
   to PT[f]‐CS and PBS with no observable body weight loss or organ
   toxicity through H&E histological staining of heart, kidney, liver,
   lung, and spleen (Figure [123]4B,C, Figures [124]S7 and [125]S8,
   Supporting Information). Furthermore, no alterations to complete blood
   count, liver function, and kidney function test results were visible
   after PT[f]‐SRiApt treatment, suggesting a lack of toxicity
   (Figure [126]4D–F). Importantly, the absence of significant immune
   response in the PT[f]‐CS group suggests that the antitumor activity of
   PT[f]‐SRiApt is specifically mediated through targeted disruption of
   SCAF4‐POLR2A interaction, rather than through nonspecific immune
   activation by exogenous DNA.

Figure 4.

   Figure 4
   [127]Open in a new tab

   Antitumor efficacy of PT[f]‐SRiApt in murine 4T1 breast cancer model.
   A) Schematic illustration of the treatment schedule with PT[f]‐SRiApt,
   PT[f]‐CS, and PBS in the in vivo mouse experiments. B) Averaged tumor
   volume curves of mice treated with PBS (black, n = 7), PT[f]‐CS (blue,
   n = 7), and PT[f]‐SRiApt (red, n = 8). Data represent mean ± SD of
   tumor volumes calculated from the diameter of the tumor mass.
   Differences at day 11 were considered statistically significant if the
   p‐value < 0.05, determined by one‐way ANOVA with Tukey's multiple
   comparisons test. **, p < 0.01; ns, not significant. C) H&E imaging
   with zoomed regions (black box) of the main organs (heart, liver,
   kidney, lung, and spleen) treated with PT[f]‐SriApt (n = 4), PT[f]‐CS
   (n = 4), or PBS (n = 4) in mice. D) Complete blood count of 4T1 bearing
   BALB/c mice after PT[f]‐SRiApt (n = 4), PT[f]‐CS (n = 4), or PBS (n =
   4) treatment (WBC: white blood cell, RBC: red blood cell, HGB:
   hemoglobin, PLT: platelet). E) Liver function test of 4T1 bearing
   BALB/c mice after PT[f]‐SRiApt (n = 4), PT[f]‐CS (n = 4), or PBS
   (n = 4) treatment (AST: aspartate transaminase, ALT: alanine
   transaminase). F) Kidney function test of 4T1 bearing BALB/c mice after
   PT[f]‐SRiApt (n = 4), PT[f]‐CS (n = 4), or PBS (n = 4) treatment (BUN:
   blood urea nitrogen, UA: uric acid). Significance was determined using
   a one‐way ANOVA test. ns, not significant.

2.4. Single‐cell Characterization of PT[f]‐SRiApt's Antitumor Effect

   To elucidate the underlying mechanism of PT[f]‐SRiApt's antitumor
   activity, we performed single‐cell RNA sequencing on tumors harvested
   from PT[f]‐SRiApt treated and control mice after 10 days of treatment.
   Post‐quality control, we analyzed 17482 high‐quality cells and
   categorized them into six major cell types based on canonical markers
   (Figure [128] 5A; Figure [129]S9, Supporting Information). PT[f]‐SRiApt
   treatment significantly reduced tumor cell proliferation and altered
   the cellular composition of tumors, as evidenced by a decreased
   proportion of tumor cells and an increased proportion of macrophages,
   cancer‐associated fibroblasts (CAFs), T cells, and neutrophils (Figure
   [130]S10, Supporting Information).

Figure 5.

   Figure 5
   [131]Open in a new tab

   Antitumor efficacy of PT[f]‐SRiApt in murine 4T1 breast cancer model.
   A) UMAP plot showing cell populations identified within tumors from
   both PT[f]‐SRiApt and PT[f]‐CS groups. Cell types are labeled on the
   right. B) Proportional changes of tumor cells in resistant and
   sensitive subgroups upon PT[f]‐SRiApt treatment. C) UMAP plot of
   subclustering of 4T1 cell‐derived tumors to resistant and sensitive
   groups. D) Expression levels of POLR2A and SCAF4 in resistant and
   sensitive subgroups. p‐values were determined by a two‐tailed Wilcoxon
   test. E) Enrichment analysis of GO terms (BP) in 4T1 cells treated with
   PT[f]‐SRiApt for 5 days. F) GSEA analysis of the RNA Pol II
   transcription regulatory pathway regulated by PT[f]‐SRiApt treatment
   with p = 1.73e‐21. G) Density plot of gene counts per cell for each
   experimental group across sensitive and resistant clusters. Vertical
   dashed lines indicate median gene counts within each group. The
   p‐values were determined using the Wilcoxon rank sum test. ****,
   p < 0.0001.

   To further investigate the heterogeneity of tumor cells, we divided
   tumor cells into seven distinct subclusters based on their expression
   profiles (Figure [132]S11A, Supporting Information). Comparison of
   these subclusters before and after PT[f]‐SRiApt treatment revealed a
   compositional change in tumor cells post‐treatment: clusters 0, 3, and
   6 increased, while clusters 1, 2, 4, and 5 decreased (Figure [133]S11B,
   Supporting Information). This differential response suggests varying
   sensitivity to PT[f]‐SRiApt among the clusters. To explore the basis of
   this differential sensitivity, we grouped the proportionally increased
   clusters (0, 3, and 6) into a resistant group and the proportionally
   decreased clusters (1, 2, 4, and 5) into a sensitive group
   (Figure [134]5B). UMAP visualization demonstrated spatial similarities
   within clusters of the same subgroup and distinct separation between
   the subgroups suggesting different expression profiles across sensitive
   and resistant groups (Figure [135]5C). Expression analysis showed that
   the sensitive group exhibited significantly higher POLR2A and SCAF4
   expression levels compared to the resistant group, providing evidence
   of PT[f]‐SRiApt's specificity at the single‐cell level
   (Figure [136]5D). Pathway enrichment analysis revealed that the
   resistant group was enriched in transcriptional regulatory pathways,
   particularly in RNA polymerase II‐related transcription regulatory
   pathway (Figure [137]5E,F; Figure [138]S11C, Supporting Information).
   Moreover, the total transcription level was downregulated in the
   resistant group as evidenced by the lower average number of transcripts
   detected per cell compared to the sensitive group (Figure [139]5G).
   Higher transcriptional activity in the sensitive group suggested
   “transcriptionally addicted” cancer cells are more vulnerable to
   transcriptional dysregulation, a finding that is further supported by
   existing literature.^[ [140]^30 ^] These findings shed light on the
   mechanistic basis of PT[f]‐SRiApt's selective anti‐proliferative
   effects on tumor cells in a SCAF4‐POLR2A interaction‐dependent manner.

2.5. PT[f]‐SRiApt Promotes T‐Cell Infiltration and Modulates Tumor
Microenvironment in the 4T1 Mouse Model

   To determine the phenotype of mice tumor cells post PT[f]‐SRiApt
   treatment at the single‐cell level, we performed GSEA analysis
   comparing PT[f]‐SRiApt treated with PBS‐treated control tumor cells.
   Consistent with bulk RNA sequencing results, antigen binding is one of
   the most enriched pathways in PT[f]‐SRiApt treated tumor cells
   (Figure [141] 6A). Additionally, enrichment in the MHC complex,
   activation of alpha‐beta T cell, and mononuclear cell migration suggest
   PT[f]‐SRiApt treatment induces antigen binding and presentation in
   tumor cells, which in turn recruits and activates leukocytes. This
   observation was further supported by notable alterations in the
   cellular composition of 4T1 tumors treated with PT[f]‐SRiApt,
   particularly a pronounced increase in T cell infiltration
   (Figure [142]6B, Figure [143]S10, Supporting Information). Flow
   cytometry analysis was conducted to experimentally validate this
   finding, revealing a substantial enrichment of T cells in
   PT[f]‐SRiApt‐treated tumors compared to PT[f]‐CS and PBS controls
   (Figure [144]6C,D, Figure [145]S12, Supporting Information).
   Additionally, a comparison of PT[f]‐SRiApt's therapeutic efficacy in
   immunocompetent versus immunodeficient mice revealed significantly
   better tumor suppression in immunocompetent mice (p = 0.0109), further
   underscoring the role of immune activation in PT[f]‐SRiApt's antitumor
   activity (Figure [146]S13, Supporting Information).

Figure 6.

   Figure 6
   [147]Open in a new tab

   PT[f]‐SRiApt improves T cell populations in breast cancer. A) GSEA
   enrichment of GO terms in PT[f]‐SRiApt tumor cells versus control PBS
   tumor cells. B) Proportion of T cells in tumors treated with
   PT[f]‐SRiApt or PBS. C) Representative flow cytometric analysis of T
   cells in tumors treated by PT[f]‐SRiApt, PT[f]‐CS, or PBS. D)
   Populations of T cells in total tumor cells. Statistical analysis from
   four biological replicates. ***, p < 0.001; ns, not significant, as
   determined by one‐way ANOVA with Tukey's multiple comparisons test. E)
   GSEA enrichment of GO terms in PT[f]‐SRiApt T cells versus control PBS
   T cells. F) GSEA plot showing the enrichment of cytokine production and
   T cell activation pathways in T cells of PT[f]‐SRiApt treated mice. The
   p‐value was determined using a permutation test. G) Tumor growth curves
   of 4T1 tumor‐bearing mice treated with PBS (n = 4), SRiApt (n = 5),
   anti‐PD‐1 (n = 4), or a combination of SRiApt and anti‐PD‐1 (n = 5).
   Data represent mean ± SD of tumor volumes. *, p < 0.05; **, p < 0.01;
   and ***, p < 0.001, determined using unpaired t‐test at day 21.

   To investigate the mechanism by which PT[f]‐SRiApt enhances T cell
   infiltration into the tumor microenvironment, we first examined whether
   this effect could be attributed to direct modulation of T cell
   activity. In vitro assays using CCK8 for proliferation and flow
   cytometric analysis for activation marker CD69 revealed that
   PT[f]‐SRiApt alone does not directly influence T cell proliferation or
   activation (Figure [148]S14A,B, Supporting Information). Furthermore,
   co‐culture experiments comparing tumor cells treated with PT[f]‐SRiApt
   alone versus those co‐cultured with both PT[f]‐SRiApt and T cells
   showed no significant difference in tumor cell killing, indicating that
   PT[f]‐SRiApt does not directly induce T cell‐mediated cytotoxicity
   under these conditions (Figure [149]S14C, Supporting Information). In
   contrast, single‐cell RNA sequencing of tumors from
   PT[f]‐SRiApt–treated mice revealed transcriptional signatures
   consistent with increased T cell activation within the tumor
   microenvironment (Figure [150]6E,F). These results align with the bulk
   and single‐cell transcriptomic analyses, which revealed upregulation of
   MHC and antigen presentation pathways in tumor cells following
   PT[f]‐SRiApt treatment (Figures [151]2J and [152]6A), implicating
   enhanced antigen presentation as a mechanism underlying the observed
   immunomodulatory effects.

   Given that enhanced immune infiltration and activation are often
   associated with improved responsiveness to immune checkpoint blockade
   (ICB),^[ [153]^31 , [154]^32 ^] we next evaluated whether PT[f]‐SRiApt
   could potentiate immunotherapeutic efficacy. In the 4T1 syngeneic tumor
   model, treatment with either a reduced dose of PT[f]‐SRiApt
   (0.5 mg kg^−1) or anti‐PD‐1 antibody (200 µg) alone had no significant
   effect on tumor growth. However, combination therapy with a reduced
   dose of PT[f]‐SRiApt and anti‐PD‐1 led to a significant reduction in
   tumor volume (Figure [155]6G; Figure [156]S15, Supporting Information).
   These results suggest that PT[f]‐SRiApt synergizes with ICB therapy,
   likely by reshaping the tumor immune landscape to favor increased T
   cell activation and infiltration.

2.6. SCAF4‐POLR2A Interaction Predicts Poor Tumor Immunity in TNBC Patients

   To explore the clinical relevance of targeting the SCAF4–POLR2A
   interaction, we next sought to determine whether its expression is
   associated with immune activity in human tumors. Leveraging publicly
   available TNBC spatial transcriptomic (ST) data, we conducted a
   comprehensive analysis of 94 publicly available samples to explore the
   spatial correlation between SCAF4‐POLR2A interaction and tumor
   immunity.^[ [157]^33 ^] After imputing dropout values, we calculated
   the Pearson correlation between SCAF4 and POLR2A expression across
   spatial transcriptomic spots. The majority of samples (83 out of 94)
   showed a statistically significant positive correlation, with 38
   displaying strong spatial co‐expression (Pearson R > 0.5), indicating
   that SCAF4 and POLR2A are frequently co‐expressed in the tumor
   microenvironment (Figure [158] 7A).

Figure 7.

   Figure 7
   [159]Open in a new tab

   SCAF4‐POLR2A interaction promotes tumor proliferation and inhibits
   immune infiltration in TNBC patients. A) Per bin Pearson correlation
   coefficient and significance between SCAF4 and POLR2A expression in
   TNBC ST samples (n = 94). B) Per bin Pearson correlation coefficient
   and significance between SCAF4‐POLR2A score and MKI67 expression in
   TNBC ST samples (n = 94). C‐E) Pearson correlation between (C)
   MKI67/SCAF4‐POLR2A, (D) PTPRC/SCAF4‐POLR2A and (E) CD3E/SCAF4‐POLR2A
   Pearson coefficient with SCAF4/POLR2A Pearson coefficient (n = 94). F)
   Comparison between ESTIMATE immune score in SCAF4‐POLR2A high and low
   interaction ST samples (n = 94). G) Comparison between MKI67 expression
   in SCAF4‐POLR2A high and low interaction ST samples (n = 94). H)
   Spatial visualization of SCAF4 and POLR2A co‐expression. I) Spatial
   visualization of CD3E and PTPRC co‐expression. J) Comparison between
   ESTIMATE immune score in SCAF4‐POLR2A high and low interaction
   proteomic samples (n = 55). K) Comparison between MKI67 expression in
   SCAF4‐POLR2A high and low interaction proteomic samples (n = 55).
   Panels F, G, J, K: p‐values determined using two‐tailed Wilcoxon test.
   L) pCR rate of pembrolizumab treatment in patients grouped by
   SCAF4‐POLR2A expression in the I‐SPY2 clinical trial. G) Association of
   SCAF4‐POLR2A expression with T cells and NK cells deconvolution across
   TCGA cancer types. *, p < 0.05; **, p < 0.01; and ***, p < 0.001,
   determined using asymptotic confidence intervals based on Fisher
   transformation.

   To determine whether this spatial interaction reflects changes in tumor
   proliferation and immunity, we first calculated SCAF4‐POLR2A
   co‐expression by averaging the normalized expression of both genes at
   each spot. We then assessed the spatial correlation between the
   SCAF4‐POLR2A co‐expression and the proliferation marker MKI67, which
   revealed that most samples displayed a strong positive correlation,
   suggesting that regions with high SCAF4–POLR2A co‐expression are
   associated with increased tumor proliferation (Figure [160]7B). Further
   analysis revealed that samples with high SCAF4‐POLR2A co‐localization
   also exhibit a strong spatial correlation between MKI67 expression and
   SCAF4‐POLR2A co‐expression (Figure [161]7C). In contrast, these same
   samples show a low spatial correlation between immune markers (PTPRC
   and CD3E) and SCAF4‐POLR2A co‐expression (Figure [162]7D,E). Together,
   these results suggest that strong SCAF4‐POLR2A co‐expression is
   localized with tumor progression and linked to spatial exclusion of
   immune infiltration.

   To further illustrate this trend, we stratified samples into high‐ and
   low‐interaction groups based on pseudobulk expression profiles. Tumors
   with above‐average expression of both SCAF4 and POLR2A were defined as
   high‐interaction, while those with below‐average expression of both
   genes were classified as low‐interaction. High‐interaction tumors
   showed more negative spatial correlations between immune markers and
   the interaction score, further supporting the association between
   SCAF4–POLR2A activity and immune exclusion (Figure [163]7F,G).
   Representative spatial plots further illustrate this dissociation
   (e.g., TNBC6 CN3_E1; Figure [164]7H,I).

   Complementing these findings at the protein level, analysis on a public
   TNBC proteomic dataset by Anurag et al.^[ [165]^34 ^] showed that high
   SCAF4–POLR2A interaction was associated with significantly lower immune
   scores based on ESTIMATE deconvolution, supporting the negative
   relationship between this interaction and immune infiltration
   (Figure [166]7J). Additionally, while MKI67 expression trended higher
   in the high‐interaction group, this difference was not statistically
   significant (p = 0.16; Figure [167]7K), but remains consistent with the
   transcriptomic data. These results suggest that in TNBC clinical
   samples, SCAF4–POLR2A co‐expression is linked to tumor proliferation
   and immune exclusion, reinforcing the potential of targeting this axis
   to improve immunotherapeutic outcomes.

   The pembrolizumab (Keytruda, anti‐PD‐1 antibody) treatment arm of the
   I‐SPY2 clinical trial was then analyzed to examine the effect of
   SCAF4‐POLR2A interaction in immunotherapy.^[ [168]^35 ^] Patients with
   high expression levels of both genes exhibited poor response rates to
   pembrolizumab, whereas patients with moderate expression levels showed
   intermediate responsiveness, and those with low expression experienced
   the most favorable outcomes. This suggests that SCAF4 and POLR2A
   expressions could serve as predictive biomarkers for immunotherapy
   response, positioning PT[f]‐SRiApt as a potential booster of
   immunotherapy efficacy by disrupting SCAF4‐POLR2A interaction
   (Figure [169]7L).

   SCAF4‐POLR2A expression also correlates to immune activation in
   pan‐cancer. Using Cancer Genome Atlas (TCGA) data,^[ [170]^36 ^] we
   observed a broad negative correlation between SCAF4‐POLR2A expression
   and activated T cell levels, as well as a positive correlation with
   resting T cell levels. This phenomenon was also observed in NK cell
   populations as elevated SCAF4‐POLR2A expression was negatively
   correlated with activated natural killer (NK) cells and positively
   correlated with inhibitory NK cell populations (Figure [171]7M). These
   findings further underscore the broad clinical potential of targeting
   the SCAF4‐POLR2A axis with PT[f]‐SRiApt, which could reprogram the
   tumor microenvironment and potentiate the efficacy of immune checkpoint
   inhibitors and other immunotherapeutic strategies.

3. Conclusion

   TNBC is the most aggressive and diverse breast cancer subtype,
   accounting for 15%–20% of cases. Characterized by the absence of ER,
   PR, and HER2 receptors, TNBC is unresponsive to common HER2‐targeted
   and hormonal therapies.^[ [172]^37 , [173]^38 , [174]^39 ^] This makes
   the development of new therapeutic targets and effective treatments for
   TNBC an urgent priority. In this study, we utilized inhibitory aptamers
   to target the unique transcriptional landscape in TNBC and established
   them as versatile chemical tools for illustrating the potential of
   protein‐protein interaction targets for therapeutic intervention.

   We presented PT[f]‐SRiApt as a first‐in‐class aptamer‐based chemical
   tool that selectively disrupts the “undruggable” protein‐protein
   interaction between SCAF4 and POLR2A, uncovering a previously
   unrecognized chemical intervention target in TNBC treatment. Extensive
   in vitro and in vivo analyses demonstrated that PT[f]‐SRiApt
   preferentially suppresses the proliferation of “transcriptionally
   addicted” TNBC cancer cells with heightened SCAF4 and POLR2A
   expression. Mechanistically, this suppression involved the promotion of
   RNA polymerase II transcription termination, activation of the MHC
   class II pathway, cell cycle arrest, and elevated immune infiltration,
   all of which synergistically contribute to a marked inhibition of
   breast cancer progression in murine models.

   Moreover, leveraging PT[f]‐SRiApt, we further revealed the critical
   role of the SCAF4‐POLR2A interaction in intrinsic immune response and
   tumor microenvironment. Blocking this interaction with PT[f]‐SRiApt
   activates the MHC pathways, promotes leukocyte infiltration, and
   enhances their antitumor immunity. Additionally, co‐expression analysis
   of SCAF4 and POLR2A further supports the predictive role of their
   interaction in immunotherapy. These findings uncover a novel function
   of the SCAF4‐POLR2A interaction in TNBC and highlight the potential of
   PT[f]‐SRiApt in tumor immunotherapy.

   Interestingly, while PT[f]‐SRiApt could bind to both SCAF4 and SCAF8
   and inhibit their bindings to POLR2A due to the structural similarity
   of their CID domains, we observed that the expression level of
   SCAF8‐POLR2A did not correlate with SRiApt sensitivity. This supports
   the distinct functional roles of SCAF4 and SCAF8 in transcription
   regulation.^[ [175]^7 ^] Consequently, PT[f]‐SRiApt not only holds
   promise as a therapeutic agent but also serves as a valuable tool for
   distinguishing the functional specificities of SCAF4 and SCAF8, further
   highlighting the broader potential of aptamers in elucidating
   protein‐protein interactions and their biological roles.

4. Experimental Section

Fluorescence Polarization

   Experimental procedures were carried out using a Perkin Elmer EnVision
   2104 Multilabel Reader (BioTek, USA, v1.14) equipped with 485 nm
   excitation and 535 nm emission filters for the FITC. FP measurements
   were performed using a Bioland 96‐well plate (product #PB06‐96S). Both
   parallel and perpendicular fluorescence intensity (Fǁ and F⊥) relative
   to linearly polarized excitation light were determined to calculate the
   FP signal. The affinity experiments were conducted in three biological
   replicates, each containing three technical replicates. The average
   affinity value was determined using the GraphPad Prism 7 program
   (GraphPad Software, Inc., USA) through curve fitting. The non‐linear
   fit model and the one‐site specific binding to fluorescent population
   ratio are used for these calculations at the respective aptamer
   concentrations. The FP assay buffer was constituted of 30 mM MES buffer
   pH 6.5, 25 mm NaCl, 2 mm β‐ME, 1 mm CHAPS, and 0.002 mg mL^−1 BSA.

Competition Assays of SRiApt

   Competition assays of PT‐modified SRiApt for SCAF4‐POLR2A interaction
   were performed using Fluorescence Polarization (FP) assays. In brief,
   FP signals resulting from the interaction were detected by combining
   FITC‐labeled peptides with SCAF4 (40 nm) in the FP competition buffer
   (30 mm MES pH 6.0, 25 mm NaCl, 2 mm β‐ME, 1 mm CHAPS, and
   0.002 mg mL^−1 BSA). Each well of a 96‐well plate was loaded with
   100 µL of the assay solution. SRiApt or PT‐modified SRiApt were
   serially diluted to varying concentrations and added to each well for
   the competition assay. Experiments were performed in three biological
   replicates, each containing three technical replicates, and the average
   inhibition constant values were determined by performing curve fitting
   and data analysis using GraphPad Prism 7 (GraphPad Software, Inc., USA)
   with the Log(inhibitor) versus normalized response model.

Cell Lines and Mice

   TNBC cell lines were obtained from ATCC and cultured according to
   ATCC's recommendations with Dulbecco's modified Eagle's medium (DMEM;
   Gibco) or RPMI 1640 (YESEN, 41402ES76) with 10% FBS (Sunrise,
   SR100180.03) and 1% penicillin/streptomycin (Biosharp, BL505A). Cells
   were cultured at 37 °C in an incubator containing 5% CO[2] and were
   passaged no more than 25 times. Six‐ to eight‐week‐old female Balb/c
   mice (GemPharmatech Co., Ltd., Nanjing, China) or Balb/c nude mice
   (Qizhen, Hangzhou, China) were housed in a temperature‐controlled,
   pathogen‐free room. All animal procedures were performed in accordance
   with the approved protocols and guidelines of the Hangzhou Institute of
   Medicine Animal Experiment Ethics Committee (Hangzhou, China). The
   Animal Experiment Ethics number: AP2024‐08‐0169, AP2024‐10‐0309,
   AP2025‐04‐0495.

Cell Cycle Assay

   4T1 Cells were plated in 12‐well plates and treated with PT[f]‐SRiApt
   aptamer or CS aptamer (0.5 µm) as a control for 24 h. A Cell Cycle and
   Apoptosis Analysis Kit (Beyotime, C1052) was used for the following
   analysis. Cells were fixed in 1 mL 75% ethanol for 1 h at 4 °C and
   stained with Propidium (PI) solution for 30 min in a 5% CO[2] incubator
   at 37 °C. Cells were analyzed using a Thermo Attune NxT flow cytometer.
   The number of live cells in each stage was determined by FlowJo
   software (version 10.8.1).

Cell Colony Assay

   4T1 cells or HCC1806 cells were plated in 12‐well plates and treated
   with PT[f]‐SRiApt aptamer or CS aptamer (5 µm) as a control for 5 days.
   After treatment, cells were washed with PBS and fixed in 4% multigrade
   formaldehyde (Biosharp, BL539A) for 10 min at room temperature. Cells
   were then washed with ddH[2]O twice and stained with Crystal Violet
   Staining Solution (Beyotime, C0121) for 10 min. After staining, cells
   were washed with ddH[2]O. Cells were observed under the scan and
   analyzed by ImageJ (version 1.8.0).

Cell Proliferation Assay

   Different cells were plated at different cell numbers per well in
   96‐well plates with PT[f]‐SRiApt aptamer or CS aptamer for 3–5 days of
   treatment (4T1: 5000 cells per well; EMT6: 1500 cells per well;
   HCC1806: 10000 cells per well; MDA‐MB‐231: 8000 cells per well;
   MDA‐MB‐468: 15000 cells per well; CAL‐120: 10000 cells pre well;
   Hs‐578: 5000 cells per well; SUM159PT: 5000 cells per well). Imaged and
   analyzed with the Incucyte SX5 live cell imaging device (Sartorius,
   Germany).

Tumor Treatment

   1 × 10^6 viable 4T1 breast cancer cells (in 50 µL of PBS) were
   subcutaneously injected into the backs of Balb/c mice. When tumors grew
   to measurable sizes, they were randomly grouped according to
   experimental needs. The mice were given PBS, PT[f]‐SRiApt aptamer, or
   CS aptamer (1 mg kg^−1 in 50 µL of PBS) every day. The tumor volume was
   measured in a blinded manner using slide calipers and was calculated
   using the following formula: tumor volume (mm^3) = length × (width)^2
   /2.

Intratumor Infiltrating T‐Cell Analysis

   Tumors were collected from the mice and minced into fine pieces in a
   digestion buffer containing 2% FBS and collagenase IV (2 mg mL^−1,
   Gibco, 17104019). The samples were incubated in the digestion buffer at
   37 °C for 1 h with a shaker, filtered through a 100‐µm filter,
   hemolyzed in hemolysis buffer (Solarbio, R1010), and washed twice with
   PBS. The collected cells were stained with the following
   fluorescent‐labeled antibodies: CD45 (Clone: 30‐F11, Cat: # 563891, BD
   Horizon), CD3e (Clone:145‐2C11, Cat: # 562600, BD Horizon). All flow
   cytometry was performed on Agilent NovoCyte Quanteon(Agilent
   Technologies, Inc., USA), and the analyses were performed using
   NovoExpress software(version 1.6.2).

Anti‐PD‐1 Antibody and PT[f]‐SRiApt Aptamer Combination Treatment

   7.5 × 10^5 viable 4T1 breast cancer cells (in 50 µL of PBS) were
   subcutaneously injected into the backs of Balb/c mice. Tumor‐bearing
   mice were intratumorally treated with PBS or PT[f]‐SRiApt aptamer
   (0.5 mg kg^−1 in 50 µL of PBS) every day, and intraperitoneally treated
   with InVivoMAb anti‐mouse PD‐1 (CD279) (200 µg in 50 µL of PBS, Cat: #
   BE0273, Clone: 29F.1A12, BioXcell) every 2 days. The tumor volume was
   measured every day.

Pharmacokinetic (PK) / Pharmacodynamic (PD) Assay

   1 × 10^6 viable 4T1 breast cancer cells (in 50 µL of PBS) were
   orthotopically into the fourth mammary fat pad of BALB/c mice. When the
   tumor volume in mice reached ≈200 mm^3, Cy7 solution or
   PT[f]‐SRiApt‐Cy7 (3 nmol per mouse) was administered via intratumoral
   injection. The mice were photographed at 0, 2, 4, 6, 8, 10, 12, 24, 48,
   72, and 96 h via the In vivo imaging system IVIS Spectrum CT (Caliper
   Life Sciences, Inc.) utilized to monitor the time‐course
   biodistribution of Cy7 fluorescence in accordance with the
   manufacturer's protocols. Living image Software (4.8.2 version) was
   used to measure the average radiance of the region of interest (ROI).
   The half‐life (t1/2) of the Cy7 solution or PT[f]‐SRiApt‐Cy7 was
   determined by fitting the data to a plateau followed by a one‐phase
   decay model via GraphPad Prism 8 (GraphPad Software, Inc., USA).

Complete Blood Count and Biochemical Analysis

   Blood was sampled from the hearts of treated mice on day 11. Whole
   blood samples pre‐treated with heparin were measured by TEK8500
   automatic hematology animal blood analyzer (TECOM, Jiangxi, China),
   while serum levels ALT, AST, UA, CR, TBII, ALB, and BUN were
   immediately measured using an automatic biochemical analyzer 7180
   (HITACHI, Japan)

Differentially Expressed Genes Identification and Gene Set Enrichment
Analysis

   Differentially expressed genes for HCC1806 bulk mRNA sequencing were
   determined using R package DESeq2 (v1.40.2). Differentially expressed
   genes for single‐cell data were determined using the FindMarkers
   function in package Seurat (v4.4.0) with thresholds set to negative
   infinity. Significantly differentially expressed genes were selected by
   setting the threshold to p < 0.05 and the absolute value of Log[2](FC)
   > 1.

   Functional analysis was performed using R package fgsea (v1.26.0).
   Curated Molecular Signatures Database (MSigDB) Gene Ontology (GO),
   Reactome, and Chemical and Genetic Perturbations of human gene sets
   were assessed for HCC1806 bulk sequencing data. GO mouse gene sets were
   assessed for 4T1 mouse model single‐cell sequencing. Default unadjusted
   p values were used and displayed in the figures.

Gene Scoring and Immune Deconvolution

   SCAF4/POLR2A expression score was calculated by taking the mean of
   scaled expression of both gene expressions. Each gene expression was
   scaled to a distribution with a mean of zero and a standard deviation
   of one. Immune deconvolution was performed with the Cibersort option
   with default parameters using the R package IOBR (v0.99.8). Pearson
   correlation was calculated using cor.test function from the R stats
   package (v4.3.3), p values were calculated using asymptotic confidence
   intervals based on Fisher's z transformation.

Single‐Cell Sequencing Data Processing and Analysis

   Seurat standard workflow was implemented using the R package Seurat
   (v4.4.0). Cells were filtered using the following parameters:
   nFeatures_RNA > 200, nCount_RNA < 30 000, percent.mt < 10. Single‐cell
   data was clustered using 0.5 resolution, six‐cell subtypes were
   identified in total using markers: epithelial cell (Epcam, Krt18,
   Krt8), endothelial cell (Pecam1, Egfl7, Flt1), macrophage (Itgam,
   Csflr, Cd68, C1qa), T cell (Cd3e, Cd3g), neutrophil (Csf3r, Cd33,
   S100a8), and cancer‐associated fibroblasts (Col1a1, Col12a1, Pdpn).

Statistical Analysis

   Single‐cell RNA sequencing data was normalized and scaled using
   standard Seurat v4.4.0 workflow. All results are expressed as the mean
   ± standard deviation (SD). ANOVA with Tukey's multiple comparisons test
   was performed to calculate p‐values when comparing continuous tumor
   growth. Two‐tailed Student's t‐test was performed to calculate p‐values
   when comparing groups of biological replicates. Two‐tailed Wilcoxon
   test from R package ggpubr (0.6.0) was performed to calculate p‐values
   when comparing gene expression across groups. The standard permutation
   test from R package fgsea (v1.26.0) was used to calculate p‐values in
   GSEA. All statistical analyses were performed in R and GraphPad Prism.
   In all cases, significance was defined by p < 0.05.

Conflict of Interest

   The authors declare no conflict of interest.

Supporting information

   Supporting Information
   [176]ADVS-12-e00433-s001.docx^ (33.7MB, docx)

Acknowledgements