Selected Publications

Lee J, Robinson ME, Ma N, Artadji D, Ahmed MA, Xiao G, Sadras T, Deb G, Winchester J, Cosgun KN, Geng H, Chan LN, Kume K, Miettinen TP, Zhang Y, Nix MA, Klemm L, Chen CW, Chen J, Khairnar V, Wiita AP, Thomas-Tikhonenko A, Farzan M, Jung JU, Weinstock DM, Manalis SR, Diamond MS, Vaidehi N & Müschen M. Nature 588:491-497 (2020).

B-cell activation and acute activation of oncogenes induce dramatic changes of cell membrane lipid composition, including >40-fold increases of PIP3-concentrations within seconds. To elucidate the mechanistic and structural requirements for PIP3 accumulation during normal B-cell activation and acute oncogenic transformation, we identified PIP3-interacting proteins by cell-surface proteomic analyses. In addition to proteins known to bind PIP3 with their pleckstrin homology (PH) domain, we identified the short 133-aminoacid protein IFITM3 (interferon-inducible transmembrane protein 3) with the greatest enrichment. Among known cell membrane lipids, PIP3 has the highest negative charge. Instead of a PH-domain, IFITM3 laterally sequestered PIP3 through electrostatic interactions with two basic lysine residues (K83 and K104) located at the membrane–solution interface.

IFITM3 was previously identified as endosomal protein that blocks viral infection. Analyses on gene expression data from patients with B-cell leukemia and lymphoma identified IFITM3 as a particularly strong predictor of poor clinical outcome. In normal resting B-cells, Ifitm3 was minimally expressed and mainly localized in endosomes. However, B-cell activation and oncogenic kinases induced phosphorylation at IFITM3-Y20, resulting in massive accumulation at the cell surface. Ifitm3ˉ/ˉ naïve B-cells developed at normal numbers; however, germinal center formation and production of antigen-specific antibodies were compromised. Oncogenes that induce development of leukemia failed to transform Ifitm3ˉ/ˉ B-cells. Conversely, the phospho-mimetic IFITM3-Y20E induced oncogenic PI3K-signaling and initiated transformation of pre-malignant B-cells. In Ifitm3ˉ/ˉ B-cells, lipid rafts were depleted of PIP3, resulting in defective expression of >60 lipid raft-associated surface receptors and impaired PI3K-signaling. We conclude that phosphorylation of IFITM3 upon B-cell activation induces a dynamic switch from antiviral effector functions in endosomes to PI3K-amplification at the cell-surface. IFITM3-dependent amplification of PI3K-signaling is critical to enable rapid expansion of activated B-cells. In addition, multiple oncogenes depend on IFITM3 to assemble PIP3-dependent signaling complexes and amplify PI3K-signaling for malignant transformation.

Chan LC, Murakami MA, Caeser R, Hurtz C, Kume K, Sadras T, Shojaee S, Hong C, Pölönen P, Nix MA, Ugale A, Chen Z, Lee J, Cosgun KN, Geng H, Chen CW, Chen J, Vogt A, Heinäniemi M, Lohi O, Wiita AP, Izraeli S, Weinstock DM & Müschen M. Nature 583: 845-851 (2020)

Cells often acquire hundreds of mutations before they develop characteristics of cancer. The concept of multi-step cancer progression suggests that the acquisition of mutations in addition to an existing set of mutations invariably accelerates tumor-progression. Studying how oncogenic drivers across multiple signaling pathways interact in 1,148 patient-derived B-cell leukemia samples, we found that individual mutations did not promote leukemogenesis unless they converged on one single oncogenic pathway. Oncogenic drivers that were not aligned with the central oncogenic driver and instead promote growth and survival in divergent directions were only found in 3% of leukemia samples (P=2.2E-16). Even in rare cases of co-occurrence in the same sample, single-cell mutation and phosphoprotein analyses revealed that non-aligned mutations were mutually exclusive and reflected two competing clones.

Genetic reactivation of divergent (suppressed) pathways engaged opposing biochemical and transcriptional programs, causing ‘friction’ with the principal oncogenic driver and subverted malignant transformation. Conversely, Cre-mediated deletion of divergent pathway components precipitated tumor-initiation and accelerated development of fatal, drug-resistant leukemia. Thus, the persistence of divergent signaling pathways represents a powerful barrier to overt malignant transformation.

We propose that a diverse spectrum of signaling input reflects interactions of normal cells with their environment, while convergence on one centralized pathway is characteristic of cancer. Tracking early stages of leukemia-initiation, we identified convergence on one principal oncogenic driver and inactivation of diverging pathways as an early critical step. Pharmacological reactivation of divergent signaling pathways to blunt transformation was achievable by kinase-agonists. Proof-of-concept studies in patient-derived leukemia cells revealed that pharmacological reactivation of suppressed divergent circuits can be leveraged as a previously unrecognized strategy to achieve long-term remission of leukemia and to overcome drug-resistance. Pharmacological pathway reactivation restores a diverse spectrum of signaling input just as in normal cells and creates a signaling environment that is not permissive to transformation.

Xiao G, Chan LN, Klemm L, Braas D, Chen Z, Geng H, Cosgun KN, Sadras T, Lee J, Salgia R, Ernst T, Hochhaus A, Jumaa H, Jiang X, Weinstock DM, Graeber TG & Müschen M. Cell 173: 644-656 (2018)

Upon encounter of antigen, B-cells mount a proliferative burst that exceeds the rate of cell division of any other cell type. Likewise, transformed B-cells in leukemia and lymphoma proliferate at higher rates than any other type of cancer. To understand how activated B-cells can fuel the energy demands of a massive surge in cell division, we performed mass spectrometry-based metabolite profiling and carbon fate analyses based on 13C isotopic labels.

In epithelial cells lacking burst capacity, glucose carbon flux through glycolysis and the pentose phosphate pathway (PPP) were equally strong. In contrast, glucose carbon flux in B-cells was diverted towards glycolysis with a 25-fold reduction of flux through the PPP. The shift from PPP to glycolysis was the result of transcriptional repression of the rate-limiting PPP enzyme G6PD by the B-cell transcription factors PAX5 and IKZF1.

B-cell-specific diversion of carbon flux from PPP to glycolysis enabled explosive proliferation, but at the expense of the PPP-product NADPH, a central effector of antioxidant protection. To avoid cytotoxic accumulation of ROS, B-cells have evolved mechanisms to terminate proliferative bursts and to restore balanced carbon flux between PPP and glycolysis. Mass spectrometry analyses revealed that this mechanism depends on the phosphatase PP2A and its ability to activate fructose-2,6-biphosphatase 2 (Pfkfb2) to restore antioxidant PPP-functions while paring down glycolysis.

Compared to their normal counterparts, transformed B-cells expressed very low levels of PP2A, resulting in permanent skewing of glycolysis-PPP carbon flux to enable permanent burst-like proliferation. These findings identified PP2A as a switch between glycolysis and PPP carbon flux and as a unique vulnerability in B-cell-tumors: Cre-mediated deletion of PP2A induced complete loss of PPP-activity, overwhelming ROS-accumulation and cell death. Strikingly, a newly developed PP2A-small molecule inhibitor (LB-100) leveraged this vulnerability and had strong preclinical activity in mice bearing patient-derived B-cell leukemia and lymphoma xenografts.

Chan LN, Chen Z, Braas D, Lee JW, Xiao G, Geng H, Cosgun KN, Hurtz C, Shojaee S, Cazzaniga V, Schjerven H, Hochhaus A, Kornblau SM, Konopleva M, Cazzaniga G, Milne TA, Koeffler HP, Ross TS, Sanchez-Garcia S, Borkhardt A, Yamamoto KR, Dickins RA, Graeber TG & Müschen M. Nature 542: 479-483 (2017)

Studying bioenergetic features of blood cell development, we observed that hematopoietic progenitors retained similar levels of energy abundance throughout differentiation. However, when common lymphoid progenitor cells expressed the B-cell transcription factors PAX5 and IKZF1 to commit to the B-cell lineage, we noticed a sharp drop in cytoplasmic volume and ATP levels of these cells. Metabolomic analyses revealed 500-fold increased AMP-ATP ratios suggesting that B-cell lineage commitment involves acute energy-depletion. Functioning as an energy-stress sensor, high AMP-ATP ratios induced LKB1-mediated phosphorylation of AMPK.

Known as essential transcriptional determinants of B-cell identity, PAX5 and IKZF1 also prevent the development of autoimmune disease and B-cell tumors through unknown mechanisms. For instance, 85% of B-cell leukemia cases carry deletions of these transcription factors, the significance of which was not known.

By combining ChIP- and RNA-seq with a genetic dCas9-VP64-based screen of transcriptional PAX5- and IKZF1-targets, we discovered a novel B-lymphoid transcriptional program (termed ‘metabolic gatekeeper’) that restricts nutrient uptake and energy metabolism during early B-cell development. Through transcriptional repression of central mediators of glucose- and glutamine transport (GLUT1, INSR, SLC1A5), PAX5 and IKZF1 limit energy supply of developing B-cells.

Inducible deletion of PAX5 and IKZF1 relieved transcriptional repression and increased glucose-uptake and ATP levels by >25-fold. In the presence of functional metabolic gatekeepers, inducible activation of the oncogenic BCR-ABL1 kinase was not sufficient to induce overt B-cell leukemia. However, concurrent relief of transcriptional repression of GLUT1, INSR and SLC1A5 enabled the development of full-blown B-cell leukemia. Likewise, lipophilic dimethyl-conjugates of glycolysis-metabolites that readily permeate the cell membrane bypassed transcriptional repression of glucose-transport to fuel malignant transformation of BCR-ABL1-transgenic B-cells. Since B-cells frequently give rise to autoimmune disease and leukemia, we propose that B-lymphoid transcription factors function as ‘metabolic gatekeeper’ to eliminate pathogenic B-cells through limitation of energy supply.

Chen Z, Shojaee S, Buchner M, Geng H, Lee JW, Klemm L, Titz B, Graeber TG, Park E, Tan YX, Satterthwaite A, Paietta E, Hunger SP, Willman CL, Melnick A, Loh ML, Jung JU, Coligan JE, Bolland S, Mak TW, Limnander A, Jumaa H, Reth M, Weiss A, Lowell CA & Müschen M. Nature. 521:357-361 (2015).

Cell survival and proliferation depends on signals that are transmitted from the environment through surface receptors. Upon recognition of antigen, the B-cell receptor (BCR) transmits these signals in B-lymphocytes. Unlike other cell types, B-lymphocytes are predicated on a ‘Goldilocks’ principle of signaling strength. If B-lymphocytes fail to express a functional BCR, signaling output is too weak, resulting in ‘death by neglect’. If, on the other hand, the BCR binds to ubiquitous self-antigen, BCR-signals are overwhelmingly strong, resulting in negative selection of autoreactive B-cells. Only BCR-signals of intermediate strength are ‘just right’ to promote survival and proliferation.

In many B-cell malignancies, the transforming oncogene mimics a constitutively active BCR. For this reason, we asked whether B-cell-tumors are still bound by the Goldilocks principle and whether oncogenic signaling strength can be targeted for therapeutic benefit. According to this concept, targeted oncogene-inhibition induces the functional equivalent of ‘death by neglect’ (no functional BCR). Conversely, pharmacological hyperactivation of BCR-related tyrosine kinases could induce cell death by triggering mechanisms of negative selection for removal of autoreactive B-cells.

Despite oncogenic transformation, basic mechanisms of negative selection were still functional and B-cell tumors remained fully sensitive kinase-hyperactivation downstream of the BCR. Unlike other types of cancer, B-cell tumors were uniquely susceptible to clonal deletion induced by hyperactive signaling from an autoreactive BCR. Our experiments studying patient-derived B-cell tumors in an in vivo transplant setting showed that targeted activation of negative selection is achievable by short transient hyperactivation of BCR-signaling, for instance by targeted blockade of inhibitory phosphatases.

Targeted therapy of cancer typically focuses on agents that suppress oncogenic signaling below minimum thresholds for survival and proliferation. Leveraging the susceptibility of B-cells to negative selection, we propose that targeted hyperactivation of oncogenic signaling above maximum thresholds represents a novel concept to target drug-resistant B-cell tumors.