Blog

Public perceptions about banks have improved since the launch of the Consumer Duty two years ago, which set higher standards of consumer protection, a survey indicates.

The Consumer Duty was launched in July 2023, requiring financial firms to put customers at the heart of what they do, including when designing products and communicating with customers.

Back in March 2023, YouGov asked people how they felt banks and financial services companies were performing. Just over two years later, it has carried out another survey, indicating improvements in public opinion.

In the most recent survey, YouGov recorded 11 percentage-point increases in the proportion of people who believe banks provide information that is easy to understand (rising from 44% in 2023 to 55% in 2025); that banks protect customers from potential harm (from 40% in 2023 to 51%); and that they provide value for money (up from 36% to 47%).

There was also a three-point rise in the proportion of people who believe that banks provide good customer service (from 60% to 63%).

The proportion of people who believe banks are doing a poor job of communicating risk to consumers has dropped seven percentage points from 47% to 40%.

Under the duty, financial firms should provide helpful and accessible customer support, so it is as easy to sort out a problem, switch or cancel a product, as it was to buy it in the first place.

They should also provide timely and clear information, helping people to make good financial decisions. Important information should not be buried in lengthy small print.

Providers should also offer products and services that are right for the customer and products and services should also provide fair value.

Firms also need to consider whether someone is in a vulnerable situation, for example due to poor health or financial troubles.

The latest research was carried out in June 2025, among more than 2,100 people across Britain.

The 2023 survey involved 2,000 people across Britain.

Cancer cell cultures and assessment of drug sensitivity

The human TNBC cell line MDA-MB-468 (RRID:CVCL_0419) was obtained from the American Type Culture Collection (ATCC) (Manassas, VA, USA). A cell variant resistant to paclitaxel, named MDA-MB-468-PR, was established in-house by exposing MDA-MB-468 cells to paclitaxel (3–4.5 nM) for 2 months. The doxorubicin-resistant variants of TNBC cell line MDA-MB-231 and estrogen receptor positive cell line MCF7 (described previously [29]) were provided by Dr. Synnøve Yndestad (Haukeland University Hospital, Bergen, Norway) (see Supplementary materials and methods for additional information). The MDA-MB-468 cells were cultured in high glucose DMEM containing GlutaMAX (Gibco, Life Technologies, Carlsbad, CA, USA) and supplemented with 10% heat-inactivated fetal bovine serum (FBS) (Sigma-Aldrich, St. Louis, MO, USA). MDA-MB-468-PR cells were periodically exposed to 3 nM paclitaxel to assure preservation of the resistant phenotype. All cell cultures were maintained at 37 °C in a humidified atmosphere containing 5% CO2 and were mycoplasma negative. Both cell variants have been authenticated in the past 3 years, using Short Tandem Repeat (STR) analysis performed by Genetica Labcorp (Burlington, NC, USA). The authenticity has been confirmed by comparing with the ATCC cell database.

To assess the sensitivity to paclitaxel, 1.3 × 10⁴ cancer cells/cm² were seeded in white clear bottom plates, before the drug was added at the desired concentration the following day. Cell proliferation was evaluated by measuring cell confluence over time with Incucyte S3 (Sartorius, Göttingen, Germany). On day 4, cell viability was assessed by measuring metabolic activity with CellTiter-Glo (Promega, Madison, WI, USA). The half maximal inhibitory concentration (IC50) of paclitaxel was estimated using GraphPad Prism 10.4.1 (GraphPad, La Jolla, CA, USA).

Cancer cell secretome collection and analysis

Cancer cells at 4 × 10³ cells/cm² density were seeded in RPMI 1640 medium containing GlutaMAX (Gibco) and supplemented with 10% FBS and 2 mM HEPES (Gibco) (further referred to as complete RPMI medium). After 3 days, half of the culture medium was replaced with fresh complete RPMI medium, and on day 6, the conditioned medium (CM) was collected. To remove cellular debris, CM was centrifuged at 300 g for 3 min, followed by filtration through a membrane with pore size 0.22 μm (Avantor, VWR, Radnor, PA, USA). Cell-free CM were aliquoted and stored at −80 °C until further use. CM was analyzed for the cytokine levels using a Bio‐Plex Pro™ Human Cytokine 27‐plex Assay (Bio‐Rad, Hercules, CA, USA; #M50‐0KCAF0Y); a Luminex 200™ instrument (Luminex, Austin, TX, USA); and the Bio‐plex manager 6.0 software (Bio‐Rad), following the manufacturer’s instructions. Information on analysis utilizing the custom-developed 73-plex inflammatory protein assay is provided in Supplementary materials and methods.

Isolation of monocytes

Monocytes were isolated from buffy coats obtained from healthy, human anonymous donors. The buffy coats were provided by The Blood Bank of Oslo University Hospital (OUS), and the project was approved by the Norwegian Regional Committee for Medical and Health Research Ethics (REK 95986). Human peripheral blood mononuclear cells (PBMCs) were isolated using Lymphoprep density gradient medium (STEMCELL, Serumwerk, Bernburg, Germany). Monocytes were isolated from PBMCs by positive selection using a CD14⁺ isolation kit (Miltenyi, Bergisch Gladbach, Germany). All procedures were performed according to the manufacturer’s instructions. CD14⁺ cell purity was above 95%, as assessed by flow cytometry. Monocytes were cryo-preserved in FBS supplemented with 10% DMSO for further use. Monocytes from 7 donors were used in this study.

Monocyte recruitment

The recruitment of monocytes was determined using transwell chambers with 3 μm pore membrane in 12-well plates (Costar, Corning, Corning, NY, USA). Approximately, 2 × 10⁵ monocytes were seeded in the top chamber in serum-free RPMI medium, while CM supplemented with 20% complete RPMI medium (hereafter CM) was added to the bottom chamber. Complete RPMI medium alone or supplemented with 250 ng/mL recombinant human CCL2 (ImmunoTools, Friesoythe, Germany) was included as negative and positive control, respectively. After 24 h, monocyte abundance in the bottom chamber was assessed by measuring metabolic activity with CellTiter-Glo.

Differentiation of monocytes and culturing of monocyte-derived cells

Monocytes at 1.5 × 10⁵ cells/cm² density were cultured for 6 days, in plates pre-coated with 50 μg/mL collagen type I (Corning, Discovery Labware, Bedford, MA, USA). To evaluate the influence of CM, the monocytes were cultivated in CM, which was exchanged every other day. On day 6, the monocyte-derived cells were analyzed as specified in sections “Phagocytosis assay”, “Morphology assessment”, “Monocyte-derived cell cytokine analysis”, “Response to classical activation stimuli”, “Immunofluorescent (IF) staining” and “RNA extraction and RT-PCR”. Where indicated, monocytes were differentiated to MΦs or DCs by standard protocols [30, 31]. In brief, monocytes were cultured for 6 days in complete RPMI medium supplemented with recombinant human cytokines; 50 ng/mL M-CSF to generate MΦs and 70 ng/mL GM-CSF and 50 ng/mL IL-4 (all from ImmunoTools) to generate DCs. For co-culture experiments, 4 × 10³ cancer cells/cm² and 2 × 10⁴ monocytes/cm² were co-cultured for 6 days, where half of the medium was exchanged after 3 days. The collected cells were analyzed by flow cytometry as described in section “Flow cytometry”.

Phagocytosis assay

Monocyte-derived cells were incubated with PHrodo red zymosan bioparticles (Life Technologies, #P35364) for 1 h. Simultaneously, the cultures were stained with green calcein dye (Life Technologies) for live cell identification. The cellular uptake of bioparticles (red signal) was followed over time by Incucyte. Integrated red fluorescence intensity and green live cell counts per well were measured over time using the Basic Analyzer module in the Incucyte S3. The average red signal intensity per live cell was calculated for each time point, and the area under the curve (AUC) was calculated using GraphPad.

Morphology assessment

Morphology of the monocyte-derived cells was analyzed by measuring eccentricity using the cell-by-cell module in Incucyte S3 or the image J software (version Java 1.8.0_322 [64-bit]). Eccentricity value > 0.9 was set as a threshold for the identification of high eccentricity cells.

Monocyte-derived cell cytokine analysis

Monocyte-derived cells cultured in CM, were washed with PBS and further cultured in complete RPMI medium for 24 h. The collected culture medium was processed and analyzed by Bio-Plex Pro human cytokine 27-plex assay, as specified in section “Cancer cell secretome collection and analysis”.

Response to classical activation stimuli

To assess the responsiveness to classical activation stimuli, monocyte-derived cells were stimulated with 20 ng/mL recombinant human IFNγ (ImmunoTools) and 50 ng/mL LPS (Sigma-Aldrich; #L3024). After 24 h, cells were collected and the expression levels of cytokines and chemokines were analyzed by Real-time PCR (RT-PCR), as specified in section “RNA extraction and RT-PCR”. Non stimulated cells were used as controls.

Immunofluorescent (IF) staining

Monocytes were cultured in 8-well chamber slides (Nunc, Thermo Fisher Scientific, Waltham, MA, USA) in the presence of CM. The cells were fixed with 4% paraformaldehyde (Chemi-Teknik, Oslo, Norway) for 10 min at room temperature (RT) and subsequently blocked with 10% horse serum (HS) in IF-buffer (PBS with 0.1% BSA, 0.2% Triton X-100, and 0.05% Tween-20) for 1 h at RT. Staining with primary antibodies in IF-buffer supplemented with 1% HS, was performed overnight at 4 °C, followed by secondary antibodies and DAPI staining (1:2400, Sigma-Aldrich) for 1 h at RT. Antibodies are listed in Table S1. IF imaging was performed by using an Olympus IX81 microscope, controlled by CellSens Standard software (version 4.1) (Olympus, Tokyo, JP). Images were analyzed with image J software (version Java 1.8.0_322 [64-bit]). Additional information on data processing is provided in Supplementary materials and methods.

RNA extraction and RT-PCR

Total RNA was isolated using TRIzol reagent (Ambion, Life Technologies) and purified using sodium acetate solution. RNA was reverse transcribed using cDNA synthesis kit (Quanta Biosciences, Gaithersburg, MD, USA), according to manufacturer’s instructions. RT-PCR reactions were performed in duplicate using 25 ng cDNA per reaction. For RT-PCR, either TaqMan gene expression assays (Applied Biosystems, Thermo Fisher Scientific) or FAM-labeled probes with Universal ProbeLibrary designed primers were used (Roche Applied Science, Penzberg, Germany). For the later, each primer mix contained 200 nM FAM-labeled probe, 300 nM of each primer and 1 × Perfecta qPCR supermix (Quanta Biosciences). For TaqMan gene expression assays, each reaction mix contained 1 × TaqMan universal PCR mastermix (Applied Biosystems) and 1 × TaqMan gene expression assay. Primers and probes are listed in Table S2. The PCR was performed using Bio-Rad CFX Connect Real Time PCR machine (Bio-Rad). Data were analyzed using Bio-Rad CFX Manager software (version 3.0). Relative gene expression was calculated by the ∆ΔCt method. GAPDH, HRPT1 and/or YARS were used as housekeeping genes.

RNA sequencing

RNA sequencing (RNAseq) was performed as described in [32]. In short, sequencing libraries were prepared according to manufacturer’s protocol, using KAPA RNA HyperPrep Kit (Roche, Basel, Switzerland) followed by mRNA capture with Twist Human Comprehensive Exome (Twist Bioscience, South San Francisco, CA, USA). Samples were sequenced at the Genomics Core Facility, Institute for Cancer Research, OUS, on the Illumina NextSeq500 or Illumina NovaSeq6000 with Illumina NextSeq500/550 reagent kit v2.5 (Illumina, San Diego, CA, USA) or Illumina NovaSeq6000 reagent kit v1.5, respectively, generating pair-end reads at 75 or 101 bp read length. Sequencing data was aligned as previously described, with the UCSC KnownGene database (downloaded March 7th 2022) as transcript reference. All downstream analyses were performed as previously described [32].

T cell activation

Two types of T cell receptor (TCR)-engineered T cells, either Radium-1 TCR T cells (recognizing a TGFβRII frameshift mutation) [33] or DMF5 TCR T cells (recognizing the native MART-1 antigen) [34] were used. The TCR-engineered T cells were generated as previously described (REK 2019–121) [35]. Briefly, PBMCs from healthy donors were transduced with a retrovirus encoding the respective TCR. Spinoculation of PBMC was performed with 1 Volume of retroviral supernatant in plates pre-coated with 20 μg/mL retronectin (Takara Bio, Shiga, Japan). The transduced T cells were further expanded by culturing in the presence of Dynabeads CD3/CD28 and 100U/mL IL-2 (Proleukin, Clinigen Healthcare, Schiphol, The Netherlands). Expression of Radium-1 TCR was assessed by staining with Vβ3-FITC antibody. The TGFβRII frameshift peptide p621_127–145 (hereafter p621) [35] and the MART-1_26–35 peptide (hereafter MART-1) [34] were manufactured by ProImmune Ltd (Oxford, UK) and were specifically recognized by Radium-1 and DMF5 TCR, respectively, in the context of HLA-A2. Antigen presenting cells (APCs), the Epstein-Barr virus transformed lymphoblastoid cell line (EBV-LCL), was generated by immortalization of B cells from HLA-A2 donors [36].

T cells were cultured in X-Vivo 15 media (Lonza, Walkersville, MD, USA) supplemented with 5% human serum (TCS Biosciences Ltd, Buckingham, UK) and 100U/mL IL-2, while the EBV-LCL was cultured in complete RPMI medium supplemented with 50 μg/mL gentamicin (Gibco), both at 1 × 10⁶ cells/ml density. For peptide-specific T cell activation experiments, the EBV-LCL was pre-loaded with 10 μM peptide (p621 or MART-1), while T cells were pre-incubated with CM overnight. Approximately 2 × 10⁵ peptide-loaded or not EBV-LCL cells were added to 1 × 10⁵ respective T cells for further co-culturing in CM in 96-well round bottom plates. For detection of intracellular markers (IFNγ and TNFα), the co-culturing was performed for 6 h in the presence of protein transport inhibitors, GolgiStop (BD Biosciences, San Diego, CA, USA) and GolgiPlug (BD Biosciences) at 1:1000 dilution at 37 °C. For detection of cell surface markers (CD137, CD25 and CD69), the co-cultures were incubated overnight at 37 °C. For assessment of T cell proliferation, the T cells were pre-stained with CellTrace Blue (Life Technologies), according to manufacturer’s recommendations, before co-culturing for 6 days with the peptide-loaded EBV-LCL. Further analysis was performed by flow cytometry as described in section “Flow cytometry”.

Flow cytometry

Cells were harvested and stained with either Zombie NIR (1:1000, BioLegend, San Diego, CA, USA), Aqua (1:700, Life Technologies) or Near-IR (1:1000, Life Technologies) live/dead dye for 20 min at 4 °C. Subsequently, cells were stained with fluorescently labeled antibodies for 30 min at 4 °C. Before intracellular staining, cells were fixed and permeabilized using eBioscience™ FOXP3/Transcription Factor staining buffer set (Invitrogen, Life Technologies), according to the manufacturer’s instructions. Antibodies are listed in Table S1. Samples were analyzed on a FACSymphony A5 flow cytometer and data were processed by the FlowJo 10.9.0 software. For assessment of T cell proliferation, the proliferation modeling function in FlowJo was used.

Analyses in clinical cohort

In the NeoAva phase II clinical trial (ClinicalTrials.gov identifier: NCT00773695), patients in the NAC arm received FEC100 every 3 weeks for four cycles followed by taxane for 12 weeks. Tumor biopsies and serum samples were collected pre- and post-NAC. The biopsies were analyzed for gene expression as described in Silwal-Pandit et al. [37], and all transcriptomics data were retrieved from this study. Serum cytokine profiling was performed using a Bio-Rad 27-plex cytokine panel and presented in Jabeen et al. [38]. Only the data from patients with no pCR were used in the current study. Patient-matched pre- and post-NAC transcriptomics data were available for 51 patients out of which 26 had available cytokine data. To infer MΦ and T cell scores, the leukocyte gene signature matrix (LM22) was applied on the transcriptomics data with CIBERSORT [39]. For comparison, another TIME deconvolution method, xCell, [40], was applied on the same dataset (additional information is provided in Supplementary materials and methods). For heatmap generation, cytokine levels were z-score normalized across all samples, before Δ was calculated by subtracting concentrations in samples pre-NAC from samples post-NAC.

Statistical analyses

Data processing, analyses and figure generation were done in R (v4.3.1) programming language with RStudio (2023.09.0 + 463) or in GraphPad Prism (version 10.4.1). Figures in R were generated with the R package ggplot2 (v3.4.4) and ComplexHeatmap (v2.16.0). Statistics was performed using GraphPad Prism or rstatix (v0.7.2). Statistical significance was calculated by unpaired or paired two-tailed t-test between two groups, while one-way ANOVA or repeated measurement (RM) one-way ANOVA was used when comparing multiple groups. For multiple testing, Tukey’s multiple comparisons test was performed. For all tests, a p < 0.05 was considered statistically significant, and labeled as * p < 0.05; ** p < 0.01; *** p < 0.001; **** p < 0.0001 in figures. Correlations were calculated with the Spearman method. The data presented in the heatmaps are mean centered, unless indicated otherwise. Bars indicate mean ± SEM from biological replicates, where n is stated in the figure legends. The mean and SD for individual biological replicate were calculated based on multiple technical replicates.