Heidi Haikala
Professor
Research Interests
Make your next move.
Prepare your application
Review drafts before you send them.
Explore related searches
Positions (1)
Heidi Haikala
University of Helsinki
5 months ago
Postdoctoral Researcher opportunity in computational immuno-oncology at the Institute for Molecular Medicine Finland (FIMM) , HiLIFE, University of Helsinki . The research group led by Heidi Haikala is seeking a curiosity-driven computational immunologist , computational biologist , or cancer immunologist with strong data analytical skills. The project focuses on tumor-immune interactions , therapy resistance mechanisms in lung cancer , and the identification of novel therapeutic vulnerabilities and predictive biomarkers for immunotherapy responses . This postdoc sits at the interface of immunology , cancer biology , data science , and translational medicine . Application deadline: July 31 . Apply through the linked University of Helsinki job portal.
5 months ago
Articles (8)
Differential activation of T cell clones dependent on their prior cell state. A, The UMAP view of TCR distribution displaying clonotype frequency in CD8 + (i) and CD4 + (ii) T cell populations. B, Bar plot showing the relative abundance of clones among patients with RCC (RCC_1–RCC_8) in CD8 + (i) and CD4 + (ii) T cells. C, Average clonality of TCR repertoires at the cluster level, shown separately for CD8 + (i) and CD4 + (ii) T cells. D, Bar plot showing the relative abundance of clones among treatments (anti–PD-1, VEGFRi, anti–PD-1 + VEGFRi, and anti-CD3/CD28/CD2) in CD8 + (i) and CD4 + (ii) T cells. E, Scatter plot showing the expansion levels of T cell clones in control and different treatments. Clonal expansion or contraction is measured as the ratio of cells from a clonotype to the number of T cells. Each point represents a T cell clone from an individual patient. For each clone, Fisher exact test was performed to evaluate the statistical significance of expansion or contraction, and clones with FDR ≤0.05 were considered significantly altered and labeled in the figure. F, Gene set enrichment analysis (GSEA) of pathways upregulated or downregulated by the treatments in selected CD8 + and CD4 + clones. Heatmap shows the GSEA results at the clone level for the five largest clones of each patient (RCC_4–RCC_8). The most prevalent cell phenotype (cluster) in the control condition is indicated for each clone. Signatures with FDR ≤0.01 are highlighted within the heatmap tiles. G, A dot plot showing the proportion of cell at exhausted and activated cluster in control and each treatment condition for the largest clones, labeling those clones that are less responsive to anti-CD3/CD28/CD2. The clones are shown in the same order as in F . H, A volcano plot showing the results of Fisher exact test, identifying upregulated genes (FDR ≤0.05; log 2 FC ≥ 0.25; Wilcoxon rank-sum test) that are overrepresented in anti-CD3/CD28/CD2-responsive clones compared with less responsive clones. Categorization to “responsive” and “less responsive” is based on the analysis presented in G . FC, fold change.
Year:
2026
Characterization of the effects of a strong T cell activator, ImmunoCult (anti-CD3/CD28/CD2), on the RCC TME. A, Volcano plot illustrating changes in cytokine secretion in response to anti-CD3/CD28/CD2. Each point represents a cytokine, and statistically significantly upregulated cytokines are annotated and highlighted in dark red. B and C, CyTOF analysis for comparison of granzyme B, TNF (TNF-α), CCL4, or INF-γ and TNF expression between control and anti-CD3/CD28/CD2 in CD8 + ( B ) and CD4 + ( C ) T cells, respectively. Gated populations depict the percentage of positive cells for each marker. D, Heatmap showing changes in the percentage of marker-positive cells between control and anti-CD3/CD28/CD2 conditions in CD8 + and CD4 + T cells. E, UMAP of CD8 + T cells highlighting a newly formed “activated CD8 cell” cluster in response to anti-CD3/CD28/CD2 treatment. Bar plot to the right shows the proportion this newly formed cluster occupies in CD8 + T cells of each of the eight patients in anti-CD3/CD28/CD2 treatment. F, UMAP of CD4 + T cells highlighting a newly formed “ImmunoCult CD4 cells” cluster in response to anti-CD3/CD28/CD2 treatment. Bar plot to the right shows the proportion this newly formed cluster occupies in CD4 + T cells of each of the eight patients in anti-CD3/CD28/CD2 treatment. G and H, Volcano plot illustrating changes in gene expression in response to anti-CD3/CD28/CD2 in CD8 + T cells ( G ) or CD4 + T cells ( H ). Each point represents a gene, the significantly upregulated genes are colored dark red, whereas downregulated genes are dark blue. Only significantly regulated genes are labeled [adjusted P value 2 FC) > 0.5]. I, Dot plot of selected immunology-related hallmark pathways, comparing anti-CD3/CD28/CD2 and control conditions. Genes were identified using the FindMarkers method and included if the P value was P value, whereas color indicates the percentage of genes in our dataset that are upregulated (+) (in red) or downregulated (−) (in blue). J, Volcano plot illustrating changes in cluster size in response to anti-CD3/CD28/CD2. Each point represents a T cell cluster (CD8 + T cells and CD4 + T cells). Cell populations that are enriched in anti-CD3/CD28/CD2 and statistically significant (adjusted P value 2 FC > 0) are highlighted in dark red, and cell populations that are diminished in anti-CD3/CD28/CD2 and statistically significant (adjusted P value 2 FC K, Signature score distributions for terminally exhausted, progenitor exhausted, and IFN CD8 + T cell signatures comparing control and anti-CD3/CD28/CD2 cells. Statistical significance of differential signature enrichment ( P value) between subtypes was determined by two-sided Wilcoxon rank-sum test (****, P ≤ 0.0001).
Year:
2026
Characterization of CD4 + T cells in control. A, UMAP of 2,136 CD4 + T cells in untreated control condition captured across eight patient samples (RCC_1–RCC_8), colored by characterized cell types. B, Bar plot showing distribution of characterized CD4 + cell types across all eight patients, ordered by the proportion of regulatory CD4 + cells in each patient. C, Dot plot showing average expression of selected genes across identified cell clusters. Each dot represents a specific gene–cluster combination, in which the size of the dot indicates the percentage of cells within the cluster expressing the gene, and the color intensity reflects the average expression level of the gene in those cells. D, CyTOF analysis of CD4 + T cells in control condition from nine patients (RCC_9–RCC_17). Marker expression patterns are visualized using t-distributed stochastic neighbor embedding (t-SNE) plots. E, Signature score for naïve, activation/effector, exhaustion, TCR signaling, cytotoxicity, cytokines, IFN response, Treg, and co-stimulation CD4 + T cell signatures plotted on CD4 + T cell UMAP. F, UMAP depicting the pseudotime inference of CD4 + T cells, showing three distinct paths. Heatmaps on the right show the top 20 genes that change their expression over the course of path progression.
Year:
2026
Characterization of CD8 + T cells in control. A, UMAP of 5,177 CD8 + T cells in untreated control condition, captured across eight patient samples (RCC_1–RCC_8), colored by characterized cell subtypes. B, Bar plot showing distribution of characterized CD8 + cell subtypes across all eight patients, ordered by the proportion of exhausted CD8 + cells in each patient. C, Dot plot showing average expression of selected genes across identified cell clusters. Each dot represents a specific gene–cluster combination, in which the size of the dot indicates the percentage of cells within the cluster expressing the gene and the color intensity reflects the average expression level of the gene in those cells. D, CyTOF analysis of CD8 + T cells in control condition from nine patients (RCC_9–RCC_17). Marker expression patterns are visualized using t-distributed stochastic neighbor embedding (t-SNE) plots. E, Signature score for naïve, activation/effector, cytotoxicity, exhaustion, progenitor exhaustion, and terminal exhaustion CD8 + T cell signatures plotted on CD8 + T cell UMAP. F, UMAP depicting the pseudotime inference of CD8 + T cells, showing three distinct paths. Heatmaps on the right show the top 20 genes that change their expression over the course of path progression.
Year:
2026
Cell–cell communication in the RCC TME. A, Heatmap showing cell–cell communication network for selected signaling pathways in the untreated, control condition. Normalized pathway network centrality scores (importance) are visualized for galectin, NECTIN-TIGIT, CD39, and CD86–CTLA4 signaling pathways, showing signal for cell types as “sender” and “receiver.” More intense color indicates higher relative “sender” or “receiver” signal. B, Heatmap showing the average expression of selected ligands and receptors in the control for specified immune cell types. C, Ridge plot showing the pseudobulk analysis comparing each treatment with the control condition. The change in the ligand–receptor expression is expressed as mean Wald statistic, in which positive values denote increase and negative values decrease in ligand–receptor expression. D, The results of the pseudobulk analysis comparing each treatment with the control condition were visualized as a dot plot. In this plot, each row represents a ligand–receptor pair, and each column corresponds to a cell type and treatment combination. Ligand and receptor results were visualized in separate panels. Dot color indicates the log 2 FC between treatment and control, whereas dot size reflects −log 10 FDR. Dots with FDR ≤0.25 were highlighted with a circle, in which circle color denotes the direction of change: red for higher expression in treatment and blue for lower expression. FC, fold change.
Year:
2026
The effects of anti–PD-1 (pembrolizumab), VEGFRi (axitinib), and their combination on immune cell functionality in the RCC TME. A, Volcano plot illustrating changes in cytokine secretion in response to anti–PD-1, VEGFRi, or anti–PD-1 + VEGFRi combination. Each point represents a cytokine, and upregulated and statistically significant cytokines are highlighted in dark red, whereas downregulated and statistically significant cytokines are colored in dark blue [−log 10 (P) > −log 10 (0.05) or P > 0.05]. B, Heatmap comparing change in immune effector markers expression between control, anti–PD-1, VEGFRi, or anti–PD-1 + VEGFRi combination in CD8 + and CD4 + T cells. C, Volcano plot illustrating changes in gene expression in response to anti–PD-1, VEGFRi, or anti–PD-1 + VEGFRi combination in CD8 + T cells (i) or CD4 + T cells (ii). Each point represents a gene, the upregulated genes are colored dark red, whereas downregulated genes are dark blue. Only significant genes are labeled [adjusted P value 2 FC) > 0.5, and in the case of anti–PD-1 adjusted P value 2 FC) > 0.3]. D, Dot plot of selected immunology-related hallmark pathways, comparing anti–PD-1, VEGFRi, or anti–PD-1 + VEGFRi to control condition. Genes were identified using the FindMarkers method and included if the P value was P value, whereas color indicates the percentage of genes in our dataset that are upregulated (+) or downregulated (−). E, Volcano plot illustrating changes in relative cluster size in response to anti–PD-1, VEGFRi, or anti–PD-1 + VEGFRi combination treatments. Each point represents a T cell cluster (CD8 + T cells and CD4 + T cells). Cell populations that are enriched in treatment and statistically significant (adjusted P value 2 FC > 0) are highlighted in dark red, and cell populations that are diminished in treatments and statistically significant (adjusted P value 2 FC F, Signature score distributions for terminally exhausted, progenitor exhausted and IFN CD8 + T cell signatures comparing control with anti–PD-1, VEGFRi, anti–PD-1 + VEGFRi, and anti-CD3/CD28/CD2. Statistical significance of differential signature enrichment ( P value) between subtypes was determined by two-sided Wilcoxon rank-sum test (ns, P > 0.05; ****, P ≤ 0.0001).
Year:
2026
Immune cells in the tumor microenvironment (TME) are attractive therapeutic targets; however, their responses to immunotherapy and targeted therapy remain incompletely understood. We developed a patient-derived ex vivo system to profile baseline immune characteristics and model treatment-induced activation at the single-cell level. The model was utilized to study T cell responses to PD-1 blockade and VEGFR inhibition (VEGFRi) in patients with renal cell carcinoma (RCC). The baseline RCC TME was highly infiltrated by T cells, characterized by diverse cytotoxic, memory, exhausted, or regulatory phenotypes. T cells were activated by direct CD3/CD28/CD2 stimulation, upregulating the IFN-γ, TNF, and IL2 signaling pathways. However, activation capacity varied noticeably depending on the baseline phenotype. PD-1 blockade induced modest T cell activation, whereas VEGFRi downregulated several immune markers, including signaling pathways and immune activation-related cytokines. Our study reveals the suppressive features of the RCC TME and the challenge of fully overcoming it with PD-1 blockade and VEGFRi. Significance: We developed a patient-derived ex vivo model to study immune cell therapy responses within the TME of RCC. Immune activation toward PD-1 blockade and VEGFRi was attenuated and depended on the immune cell state.
Year:
2026
Residual disease remains a major obstacle for achieving durable responses in patients treated with oncogene-targeted therapy. Drug-tolerant persister (DTP) cells emerging under treatment and persisting in residual tumors are considered to be the root of acquired resistance, yet their contribution to immune evasion in on-treatment tumors is poorly defined. Here, we show in the context of EGFR-mutant lung cancer that DTP cells actively contribute to the formation of an immunosuppressive tumor microenvironment during EGFR tyrosine kinase inhibitor (TKI) therapy. In syngeneic mouse models and in patients, EGFR TKI therapy leads to an accumulation of immunosuppressive macrophages, which is strictly treatment-dependent and fully reversible upon treatment cessation or progressive disease, respectively. Quiescent DTP cells directly drive the recruitment and immunosuppressive reprogramming of monocytes and macrophages through a YAP-driven secretome, and the DTP-reprogrammed monocytes suppress T cell proliferation and effector functions in vitro . Co-targeting YAP with a TEAD inhibitor ORM-47286 rewires the DTP secretome and inhibits macrophage reprogramming in vitro , and prevents immunosuppressive macrophage accumulation and improves the efficacy of EGFR TKI therapy in immunocompetent mouse models. Our findings highlight the previously unappreciated role of DTP cells in modulating the tumor microenvironment in on-treatment tumors, and position the treatment-induced YAP/TEAD activity in DTP cells as an important driver of adaptive immunosuppression during EGFR-targeted therapy.
Year:
2026
Collaborators (9)
Anna Kreutzman
-
Karita Dolores Peltonen
University of Helsinki
Paloma Cejas
Head of Innovation/Director
Dana-Farber Cancer Institute
Tuomas Mirtti
University of Helsinki
Milos Gojkovic
University of Helsinki
Maija Hollmén
Principal Investigator
University of Turku
Lynn Heasley
Professor and Chair
University of Colorado Anschutz Medical Campus
Satu Mustjoki
University of Helsinki
Steffen Boettcher
Assistant professor
University Hospital of Zurich
Make your next move.
Prepare your application
Review drafts before you send them.