SULF1 suppresses Wnt3A-driven growth of bone metastatic prostate cancer in perlecan-modified 3D cancer-stroma-macrophage triculture models

Bone marrow stroma influences metastatic prostate cancer (PCa) progression, latency, and recurrence. At sites of PCa bone metastasis, cancer-associated fibroblasts and tumor-associated macrophages interact to establish a perlecan-rich desmoplastic stroma. As a heparan sulfate proteoglycan, perlecan (HSPG2) stores and stabilizes growth factors, including heparin-binding Wnt3A, a positive regulator of PCa cell growth. Because PCa cells alone do not induce CAF production of perlecan in the desmoplastic stroma, we sought to discover the sources of perlecan and its growth factor-releasing modifiers SULF1, SULF2, and heparanase in PCa cells and xenografts, bone marrow fibroblasts, and macrophages. SULF1, produced primarily by bone marrow fibroblasts, was the main glycosaminoglycanase present, a finding validated with primary tissue specimens of PCa metastases with desmoplastic bone stroma. Expression of both HSPG2 and SULF1 was concentrated in αSMA-rich stroma near PCa tumor nests, where infiltrating pro-tumor TAMs also were present. To decipher SULF1’s role in the reactive bone stroma, we created a bone marrow biomimetic hydrogel incorporating perlecan, PCa cells, macrophages, and fibroblastic bone marrow stromal cells. Finding that M2-like macrophages increased levels of SULF1 and HSPG2 produced by fibroblasts, we examined SULF1 function in Wnt3A-mediated PCa tumoroid growth in tricultures. Comparing control or SULF1 knockout fibroblastic cells, we showed that SULF1 reduces Wnt3A-driven growth, cellularity, and cluster number of PCa cells in our 3D model. We conclude that SULF1 can suppress Wnt3A-driven growth signals in the desmoplastic stroma of PCa bone metastases, and SULF1 loss favors PCa progression, even in the presence of pro-tumorigenic TAMs.

ALK+ Anaplastic Large Cell Lymphoma (ALCL)-Derived Exosomes Carry ALK Signalling Proteins and Interact With Tumor Microenvironment

The oncogenic pathways activated by the NPM-ALK chimeric kinase of ALK+ anaplastic large cell lymphoma (ALCL) are well characterised, however, the potential interactions of ALK signalling with the microenvironment are not yet known. Here we report that ALK+ ALCL-derived exosomes contain critical components of ALK signalling as well as CD30 and that exosome uptake by lymphoid cells led to increased proliferation and expression of critical antiapoptotic proteins by the recipient cells. The bone marrow fibroblasts highly uptake ALK+ ALCL - derived exosomes and acquire cancer-associated fibroblast (CAF) phenotype. Moreover, exosome-mediated activation of stromal cells altered the cytokine profile of the microenvironment. These interactions may contribute to tumor aggressiveness and possibly resistance to treatment.

Bone Marrow Fibroblasts Derived from Vitally Frozen Mononuclear Cells As a Source for Germ Line DNA in Acute Myeloid Leukemia (AML)

The introduction of next generation sequencing (NGS) techniques has revolutionized the genetic characterization of malignant diseases. Still, optimal determination of somatic mutations in patients requires not only tumor DNA but also germ line DNA and finding a reliable source for germ line DNA is then crucial. In AML, germ line sources such as skin biopsies and buccal swabs/washes can be hampered by contamination of leukemic cells as well as low amounts of DNA and the use of T-cells is limited by the presence of evolutionary early somatic mutations that are also present in the malignant clone. As AML samples commonly are preserved and biobanked as vitally frozen mononuclear cells, we aimed to find a reliable method using these samples taken at the time of AML diagnosis as a source of germ line DNA. For this purpose, we choose to focus the non-hematopoietic population present in the diagnostic AML bone morrow sample and hypothesized that fibroblasts would fit the purpose. Bone marrow fibroblasts (BMF) were expanded in vitro by thawing vitally frozen mononuclear cells and subsequently grow adherent cells with the aim was to (i) get rid of any leukemic cells (ii) obtain up to 5 µg of germ line genomic DNA. BMFs were cultured from 6 AML patients harboring a monosomy 7 karyotype as well as fibroblast from normal donors. Cells were cultured in MyeloCult™ for up to 6 weeks. After 48 hours, unattached cells washed away while BMF attached to the bottom of the flask and were monitored by light microscopy for their ability to form colonies and retain a flattened shape. At 2, 4 and 6 weeks after initiation of the culture, cells were counted and characterized by a differentiation assay, FISH for monosomy 7, and FACS for surface markers. Doubling time for BMFs was on average 2 weeks although proliferation was age dependent with faster growth in younger patients. After 6 weeks of expansion culture, the number of cells had reached the intended 5.106 BMFs and ~5 μg of DNA for most patients. FISH results showed BMF to lack monosomy 7 in contrast to AML to cells. Surface markers were CD73+, CD105+, CD90+, CD45-, CD34- and CD33- and thus showed fibroblast phenotype though also similar with that of mesenchymal stem cells (Table 1). For the differentiation assay, BMFs were cultured with accurate types of medium for 10-14 days allowing cells to differentiate into either bone, fat or cartilage cells. The differentiation assay showed that BMFs were able to differentiate to bone and fat but not cartilage cells. In conclusion, our findings show the feasibility of deriving large amounts of non-hematopoietic cells defined as fibroblasts as a source of germ line DNA from vitally frozen diagnostic AML samples. We consider this as an attractive source of germ line DNA in analyses such as whole genome exome sequencing of AML patients. Table 1. Cell surface markers expressed on cultures bone marrow fibroblasts and AML cells in two AML patient. CD7 CD11b CD13 CD14 CD19 CD33 CD34 CD38 CD45 CD56 CD73 CD90 CD105 CD117 HLA-DR P1a BMF + - - + + + P1bAML blast + - + + + + + + + - - - + + P2c BMF + - - + + + P2dAML Blast - - + - - + + + + - - - - + + P1: patient1 (a BMF; b blast cells), P2: patient2 (c BMF; d blast cells). Disclosures No relevant conflicts of interest to declare.