CACaroline Alvares

Caroline Alvares

*This is an algorithmically generated author record based on publicly available information.

Current Organizations / Institutes

The organizations/institutes from the scholar's latest article

Previous Organizations / Institutes

The organizations / institutes from the scholar's other articles in the past ten years


Caroline Alvares has a total of 153 co-authors. Here are the top co-authors by number of shared publications:


Latest Publications

  • Open Access

    Rapid proteasomal degradation of mutant feline McDonough sarcoma-like tyrosine kinase-3 overcomes inhibitor resistance mechanisms of acute myeloid leukemia cells

    by Melisa Halilovic, Mohamed Abdelsalam, Andreas O. Mieland, Sarah Neuroth, Ramy Ashry, Marie E. S. Schwarting, Joanna Zabkiewicz, Michelle Lazenby, Caroline Alvares, Matthias Schmidt, et al.

    The hyperactive kinase FLT3-ITD causes acute myeloid leukemia (AML) and is a prognostic biomarker. We synthesized and characterized MA191 as a novel von Hippel-Lindau (VHL) E3 ubiquitin ligase-based proteolysis targeting chimera (PROTAC) for FLT3. MA191 induces apoptosis in cultured and primary AML patient cells harboring FLT3-ITD (IC 50 =9.6-11 nM; EC 50 =0.001-0.883 µM). Comparison with a VHL-binding-deficient stereoisomer of MA191 demonstrates that FLT3-ITD degradation is superior to kinase inhibition ( p < 0.03). Rapid depletion of FLT3-ITD by MA191 (DC 50 =10 nM) requires VHL, neddylation, proteasome activity, and the pro-apoptotic BH3-only protein BIM. MA191 reduces FLT3-ITD prior to apoptosis, uncovering apoptosis-independent roles of BIM for protein stability. Higher expression levels of FLT3 in leukemia cells than in healthy cells ( n  = 3675/ n  = 1249) are associated with worse AML patient survival ( p = 0.0099). MA191 does not harm healthy blood and bone marrow progenitor cells or myeloid cell differentiation. In Danio rerio larvae, MA191 halts AML cell proliferation without significant toxicity. Human stromal cell interactions and mutations in the tyrosine kinase domain of FLT3-ITD that confer resistance to FLT3 inhibitors do not abolish anti-leukemic effects of MA191. Accordingly, leukemia cells with acquired resistance to MA191 lack such mutations. We offer a framework for fast-acting next-generation FLT3 degraders that overcome resistance mechanisms to classical inhibitors.

  • Open Access

    Rapid proteasomal degradation of mutant feline McDonough sarcoma-like tyrosine kinase-3 overcomes tyrosine kinase inhibitor resistance of acute myeloid leukemia cells

    by Melisa Halilovic, Mohamed Abdelsalam, Andreas O. Mieland, Sarah Neuroth, Ramy Ashry, Joanna Zabkiewicz, Michelle Lazenby, Caroline Alvares, Matthias Schmidt, Walburgis Brenner, et al.

    Background: Feline McDonough sarcoma (FMS)-related receptor tyrosine kinase 3 with activating internal tandem duplications (FLT3-ITD) causes acute myeloid leukemia (AML). Targeted protein degraders for FLT3 have evolved as drugs against leukemia. Methods: We synthesized and characterized MA191 as novel von Hippel-Lindau (VHL)-based proteolysis targeting chimera (PROTAC) for FLT3. We analyzed protein expression, protein degradation mechanisms, and posttranslational modifications by immunoblot. Selective proteasome modulation, an inactive stereoisomer of MA191, and siRNA confirmed the event-driven degradation of FLT3-ITD. Hematopoietic cell survival and differentiation were determined by flow cytometry using apoptosis and cell surface markers. As models, we used cultured and primary human AML cells, FLT3 inhibitor-resistant AML cells, mature blood cells, and hematopoietic stem cells. We scrutinized the databases DepMap, GEPIA2, Hemap, and HPA to assess FLT3 expression and patient survival. Experiments with Danio rerio larvae verified in vivo anti-leukemic activity of MA191. ANOVA and Bonferroni correction were used for statistics. Results: MA191 is a rapid nanomolar apoptosis inducer in AML cells harboring FLT3-ITD (IC 50 =10.16-11.6 nM; EC 50 =0.015-0.883 µM). A stereoisomer of MA191 that cannot recruit VHL demonstrates that elimination of FLT3-ITD is superior to its inhibition. MA191 abrogates FLT3 inhibitor resistance from rebound activation of mitogen-activated kinases. Rapid depletion of FLT3-ITD by MA191 (DC 50 =10 nM) requires VHL, neddylation, and the pro-apoptotic BH3-only protein BIM. Reduction of FLT3-ITD by MA191 precedes apoptosis. This reveals an apoptosis-independent function of BIM on protein stability. Leukemia cells express more FLT3 than healthy cells (n=3675/n=1249) and FLT3 expression is associated with worse AML patient survival (p=0.0099). MA191 does not harm blood cells and bone marrow progenitor cells and does not disturb myeloid blood cell differentiation. In Danio rerio , MA191 halts AML cell proliferation without significant toxicity. Anti-leukemic effects of MA191 are not susceptible to anti-apoptotic effects of human stromal cells and mutations in the tyrosine kinase of FLT3-ITD that confer resistance to selective FLT3 inhibitors. Conclusions: these insights and the disclosure of the structure of MA191 provide a framework for an improved design of PROTACs that target mutant FLT3 and are not vulnerable to extrinsic and intrinsic resistance mechanisms. Degradation kinetics appear as determinant of such resistance breakers. Abstract Figure:


  • Open Access

    Survival Prediction in Acute Myeloid Leukemia at Distinct Treatment Time Points: A Performance Comparison of Random Survival Forest and Elastic-Net Regularized Cox Regression. (Preprint)

    by Oisin Padraig Brady, Carolina Fuentes Toro, Sean James Johnson, Peter Giles, Caroline Alvares, Joanna Zabkiewicz

    Risk group stratification based on the prediction of survival of patients with acute myeloid leukemia (AML) is complex. Despite common risk group categorization guidelines, the overall prognosis remains poor. Machine learning techniques have been shown to provide more accurate risk group stratification than conventional approaches using trial data. However, many time-to-event (TTE) models do not use training sets constrained to specific time windows, instead using aggregations of trial data. This study aimed to evaluate the performance of (1) random survival forest (RSF) and (2) Cox proportional hazard regression with elastic net regularization (CoxNet) for survival prediction of patients with AML within a censoring window trained with available data recorded at discrete time points during the United Kingdom National Cancer Research Institute Acute Myeloid Leukaemia 17 randomized controlled trial (AML17). For each stage in the AML17 trial, separate models were trained for each exhaustive k-choice combination of available AML17 data subsets. Data combinations for each model were further constrained according to the respective trial stage to avoid data leakage. Preliminary Pearson correlation methods were used to remove directly correlating features with the TTE prediction (time-to-death/5-y censoring point). Repeated k-fold stratified cross-validation was used on each dataset ablation to find candidate models. Permutation importance and elastic net regularization were used to monitor stability across validation folds and reduce the feature set of the highest performing stage RSF and Cox proportional hazard regression models, respectively. Finally, selected ablated models were re-evaluated using the nested, k-fold, stratified sampling cross-validation method with bootstrapping. Concordance index ranked the best models for data constricted up to the end of induction (RSF=0.68, CoxNet=0.67), stages 1 (RSF=0.69, CoxNet=0.68), 2 (RSF=0.68, CoxNet=0.66), and 3 (RSF=0.69, CoxNet=0.63) of the trial. This study details the high prediction accuracy for time-to-survival-event predictions when training sets of CoxNet and RSF models, which are sequentially constricted to data measured up to the end of respective AML17 trial stages. The performance of these sequential TTE models is intended to justify their use as part of a wider digital twin system simulating multiple TTE outcomes for patients with AML.

  • Open Access

    MHC-related protein 1–restricted recognition of cancer via a semi-invariant TCR-α chain

    by Garry Dolton, Hannah Thomas, Li Rong Tan, Cristina Rius Rafael, Stephanie Doetsch, Giulia-Andreea Ionescu, Lucia F. Cardo, Michael D. Crowther, Enas Behiry, Théo Morin, et al.

    The T cell antigen presentation platform MR1 consists of 6 allomorphs in humans that differ by no more than 5 amino acids. The principal function of this highly conserved molecule involves presenting microbial metabolites to the abundant mucosal-associated invariant T (MAIT) cell subset. Recent developments suggest that the role of MR1 extends to presenting antigens from cancer cells, a function dependent on the K43 residue in the MR1 antigen binding cleft. Here, we successfully cultured cancer-activated, MR1-restricted T cells from multiple donors and confirmed that they recognized a wide range of cancer types expressing the most common MR1*01 and/or MR1*02 allomorphs (over 95% of the population), while remaining inert to healthy cells including healthy B cells and monocytes. Curiously, in all but one donor these T cells were found to incorporate a conserved TCR-α chain motif, CAXYGGSQGNLIF (where X represents 3–5 amino acids), because of pairing between 10 different TRAV genes and the TRAJ42 gene segment. This semi-invariance in the TCR-α chain is reminiscent of MAIT cells and suggests recognition of a conserved antigen bound to K43.


  • Open Access

    Selective degradation of mutant FMS-like tyrosine kinase-3 requires BIM-dependent depletion of heat shock proteins

    by Melisa Halilovic, Mohamed Abdelsalam, Joanna Zabkiewicz, Michelle Lazenby, Caroline Alvares, Matthias Schmidt, Walburgis Brenner, Sara Najafi, Ina Oehme, Christoph Hieber, et al.

    Internal tandem duplications in the FMS-like tyrosine kinase-3 (FLT3-ITD) are common mutations in acute myeloid leukemia (AML). Proteolysis-targeting chimeras (PROTACs) that induce proteasomal degradation of mutated FLT3 emerge as innovative pharmacological approach. Molecular mechanisms that control targeted proteolysis beyond the ubiquitin-proteasome-system are undefined and PROTACs are the only known type of FLT3 degraders. We report that the von-Hippel-Lindau ubiquitin-ligase based FLT3 PROTAC MA49 (melotinib-49) and the FLT3 hydrophobic tagging molecule MA50 (halotinib-50) reduce endoplasmic reticulum-associated, oncogenic FLT3-ITD but spare FLT3. Nanomolar doses of MA49 and MA50 induce apoptosis of human leukemic cell lines and primary AML blasts with FLT3-ITD (p < 0.05-0.0001), but not of primary hematopoietic stem cells and differentiated immune cells, FLT3 wild-type cells, retinal cells, and c-KIT-dependent cells. In vivo activity of MA49 against FLT3-ITD-positive leukemia cells is verified in a Danio rerio model. The degrader-induced loss of FLT3-ITD involves the pro-apoptotic BH3-only protein BIM and a previously unidentified degrader-induced depletion of protein-folding chaperones. The expression levels of HSP90 and HSP110 correlate with reduced AML patient survival (p < 0.1) and HSP90, HSP110, and BIM are linked to the expression of FLT3 in primary AML cells (p < 0.01). HSP90 suppresses degrader-induced FLT3-ITD elimination and thereby establishes a mechanistically defined feed-back circuit.


  • Open Access

    Inhibitors of the tyrosine kinases FMS-like tyrosine kinase-3 and WEE1 induce apoptosis and DNA damage synergistically in acute myeloid leukemia cells

    by Christoph Hieber, Al-Hassan M. Mustafa, Sarah Neuroth, Sven Henninger, Hans-Peter Wollscheid, Joanna Zabkiewicz, Michelle Lazenby, Caroline Alvares, Siavosh Mahboobi, Falk Butter, et al.


  • Open Access

    CNL and aCML should be considered as a single entity based on molecular profiles and outcomes

    by Gonzalo Carreño-Tarragona, Alberto Álvarez-Larrán, Claire Harrison, José Carlos Martínez-Ávila, Juan Carlos Hernández-Boluda, Francisca Ferrer-Marín, Deepti H. Radia, Elvira Mora, Sebastian Francis, Teresa González-Martínez, et al.

    Chronic neutrophilic leukemia (CNL) and atypical chronic myeloid leukemia (aCML) are rare myeloid disorders that are challenging with regard to diagnosis and clinical management. To study the similarities and differences between these disorders, we undertook a multicenter international study of one of the largest case series (CNL, n = 24; aCML, n = 37 cases, respectively), focusing on the clinical and mutational profiles (n = 53 with molecular data) of these diseases. We found no differences in clinical presentations or outcomes of both entities. As previously described, both CNL and aCML share a complex mutational profile with mutations in genes involved in epigenetic regulation, splicing, and signaling pathways. Apart from CSF3R, only EZH2 and TET2 were differentially mutated between them. The molecular profiles support the notion of CNL and aCML being a continuum of the same disease that may fit best within the myelodysplastic/myeloproliferative neoplasms. We identified 4 high-risk mutated genes, specifically CEBPA (β = 2.26, hazard ratio [HR] = 9.54, P = .003), EZH2 (β = 1.12, HR = 3.062, P = .009), NRAS (β = 1.29, HR = 3.63, P = .048), and U2AF1 (β = 1.75, HR = 5.74, P = .013) using multivariate analysis. Our findings underscore the relevance of molecular-risk classification in CNL/aCML as well as the importance of CSF3R mutations in these diseases.


  • Open Access

    Understanding a high-risk acute myeloid leukemia by analyzing the interactome of its major driver mutation

    by Claudia Chiriches, Nathalie Nicolaisen, Maria Wieske, Heba Elhaddad, Ecmel Mehmetbeyoglu, Caroline Alvares, Dörte Becher, Paul Hole, Oliver Gerhard Ottmann, Martin Ruthardt

    The WHO classifies t(6;9)-positive acute myeloid leukemia (AML) as a subgroup of high-risk AML because of its clinical and biological peculiarities, such as young age and therapy resistance. t(6;9) encodes the DEK/NUP214 fusion oncoprotein that targets only a small subpopulation of bone marrow progenitors for leukemic transformation. This distinguishes DEK/NUP214 from other fusion oncoproteins, such as PML/RARα, RUNX1/ETO, or MLL/AF9, which have a broad target population they block differentiation and increase stem cell capacity. A common theme among most leukemogenic fusion proteins is their aberrant localization compared to their wild-type counterparts. Although the actual consequences are widely unknown, it seems to contribute to leukemogenesis most likely by a sequester of interaction partners. Thus, we applied a global approach to studying the consequences of the aberrant localization of t(6;9)-DEK/NUP214 for its interactome. This study aimed to disclose the role of localization of DEK/NUP214 and the related sequester of proteins interacting with DEK/NUP214 for the determination of the biology of t(6;9)-AML. Here we show the complexity of the biological consequences of the expression of DEK/NUP214 by an in-depth bioinformatic analysis of the interactome of DEK/NUP214 and its biologically dead mutants. DEK/NUP214’s interactome points to an essential role for aberrant RNA-regulation and aberrant regulation of apoptosis and leukocyte activation as a significant determinant of the phenotype of t(6;9)-AML. Taken together, we provide evidence that the interactome contributes to the aberrant biology of an oncoprotein, providing opportunities for developing novel targeted therapy approaches.


  • Open Access

    Characterization of milk production systems in the county of Alegrete, RS, Brazil, based on productive indexes

    by Caroline Alvares Silva, Leonardo Ereno Tadielo, Neliton Flores Kasper, Othon Dalla Colletta Altermann, Taiani Ourique Gayer, Rodrigo Holz Krolow, Ricardo Pedroso Oaigen, Deise Dalazen Castagnara

    This study aimed to characterize dairy production systems in Alegrete, RS, Brazil, based on productive indices, management practices, and technification. The present study was conducted on 43 farms distributed in 22 localities of the county. The collection of data on milk production systems was carried out through visits to the properties, using a semi-structured guide questionnaire. The data obtained with the questionnaires were tabulated in Excel and with the aid of the IBM SPSS Statistics 20.0 software, through multivariate statistics, data were submitted to main component analysis (MCA) and hierarchical clusters analysis (HCA), allowing the division of 43 production units into homogeneous groups. The studied variables were summarized through the MCA in two main components (1 and 2), which clarified 71.53% of the explained variance. The alpha-Cronbach values observed for the two main components totaled 0.977, a result that confirms the reliability of the questionnaire used and reveals the high correlation between the answers obtained. From the hierarchical classification analysis, the dataset of the 43-farm studied was reduced to six groups (G1, G2, G3, G4, G5, and G6). The quadrants obtained from the insertion of the axes of the main components 1 and 2 allowed the interpretation of the groups of systems, according to the characteristics related to milk production. G2 presented the highest number of farms of the six systems formed, representing 41.86% of the establishments studied. These are characterized by being a more productive farm, an average 881-1 L day, with greater technological adoption of production and greater area destined to milk production, corresponding to the average of 78 hectares. The productive aspects that define the characteristics of milk production systems in the county were related to the structure of the herd, pasture area, daily production, disposal criteria, and milking management. The main differences found in the different groups are related to the productive indexes, suggesting that the technical assistance and rural extension actions in the dairy production systems in the county of Alegrete should be directed according to the individual need of each group formed.

  • Open Access

    Combination of a mitogen‐activated protein kinase inhibitor with the tyrosine kinase inhibitor pacritinib combats cell adhesion‐based residual disease and prevents re‐expansion of FLT3‐ITD acute myeloid leukaemia

    by Joanna Zabkiewicz, Michelle Lazenby, Gareth Edwards, Ceri A. Bygrave, Nader Omidvar, Lihui Zhuang, Steve Knapper, Carol Guy, Robert K. Hills, Alan K. Burnett, et al.

    Minimal residual disease (MRD) in acute myeloid leukaemia (AML) poses a major challenge due to drug insensitivity and high risk of relapse. Intensification of chemotherapy and stem cell transplantation are often pivoted on MRD status. Relapse rates are high even with the integration of first‐generation FMS‐like tyrosine kinase 3 (FLT3) inhibitors in pre‐ and post‐transplant regimes and as maintenance in FLT3‐mutated AML. Pre‐clinical progress is hampered by the lack of suitable modelling of residual disease and post‐therapy relapse. In the present study, we investigated the nature of pro‐survival signalling in primary residual tyrosine kinase inhibitor (TKI)‐treated AML cells adherent to stroma and further determined their drug sensitivity in order to inform rational future therapy combinations. Using a primary human leukaemia‐human stroma model to mimic the cell–cell interactions occurring in patients, the ability of several TKIs in clinical use, to abrogate stroma‐driven leukaemic signalling was determined, and a synergistic combination with a mitogen‐activated protein kinase (MEK) inhibitor identified for potential therapeutic application in the MRD setting. The findings reveal a common mechanism of stroma‐mediated resistance that may be independent of mutational status but can be targeted through rational drug design, to eradicate MRD and reduce treatment‐related toxicity.