Secretomes from metastatic breast cancer cells, enriched for a prognostically unfavorable LCN2 axis, induce anti-inflammatory MSC actions and a tumor-supportive premetastatic lung.

Cancer metastasis is responsible for the clear majority of cancer-related deaths. Survival and expansion of cancer cells at secondary sites requires that these premetastatic microenvironments be primed by primary tumor cells and their secreted factors. Efforts to date have been limited by immune-deficient in vivo models and/or the need for finely-tuned analysis time points that reduce contributions from early-disseminating cancer cells. In this regard, we developed a tumor cell-free syngeneic breast cancer model for characterizing tumor cell secretome-mediated reprogramming of premetastatic tissues. We demonstrate that secretomes from metastatic breast cancer cells differentially regulate the lung and brain, promoting a tumor-supportive lung microenvironment with both elevated CD73 expression and decreased TNFα expression. Using in vitro models of CD73-positive mesenchymal stem cells (MSCs) and macrophages/monocytes, we tested whether MSCs can mediate anti-inflammatory effects of metastatic breast cancer cells. Notably, conditioned media from metastatic Py230 cells reprogrammed the secretomes of MSCs toward an anti-inflammatory state. Mining transcriptome data from Py8119 and Py230 cells revealed a lipocalin 2 (LCN2) axis that is selectively expressed in the metastatic Py230 cells, predicts poor breast cancer patient survival and is elevated in circulating serum of mice chronically treated with conditioned media from Py230 cells. Taken together, these results establish the utility of an immune-competent tumor cell-free model for characterizing the mechanisms of breast cancer cell priming of the premetastatic niche, demonstrate that MSCs can mediate the anti-inflammatory effects of metastatic breast cancer cells and substantiate LCN2 as a promising therapeutic target for blocking breast cancer progression.

Lectin binding and effects in culture on human cancer and non-cancer cell lines: examination of issues of interest in drug design strategies.

By using a non-cancer and a cancer cell line originally from the same tissue (colon), coupled with testing lectins for cell binding and for their effects on these cell lines in culture, this study describes a simple multi-parameter approach that has revealed some interesting results that could be useful in drug development strategies. Two human cell lines, CCL-220/Colo320DM (human colon cancer cells, tumorigenic in nude mice) and CRL-1459/CCD-18Co (non-malignant human colon cells) were tested for their ability to bind to agarose microbeads derivatized with two lectins, peanut agglutinin (Arachis hypogaea agglutinin, PNA) and Dolichos biflorus agglutinin (DBA), and the effects of these lectins were assessed in culture using the MTT assay. Both cell lines bound to DBA-derivatized microbeads, and binding was inhibited by N-acetyl-D-galactosamine, but not by L-fucose. Neither cell line bound to PNA-derivatized microbeads. Despite the lack of lectin binding using the rapid microbead method, PNA was mitogenic in culture at some time points and its mitogenic effect displayed a reverse-dose response. This was also seen with effects of DBA on cells in culture. While this is a simple study, the results were statistically highly significant and suggest that: (1) agents may not need to bind strongly to cells to exert biological effects, (2) cell line pairs derived from diseased and non-diseased tissue can provide useful comparative data on potential drug effects and (3) very low concentrations of potential drugs might be initially tested experimentally because reverse-dose responses should be considered.

Analysis of unconventional approaches for the rapid detection of surface lectin binding ligands on human cell lines.

For over a decade our laboratory has developed and used a novel histochemical assay using derivatized agarose beads to examine the surface properties of various cell types. Most recently, we have used this assay to examine lectin binding ligands on two human cell types, CCL-220, a colon cancer cell line, and CRL-1459, a non-cancer colon cell line. We found that CCL-220 cells bound specific lectins better than CRL-1459, and this information was used to test for possible differential toxicity of these lectins in culture, as a possible approach in the design of more specific anti-cancer drugs. Although we have examined the validity of the bead-binding assay in sea urchin cell systems, we have not previously validated this technique for mammalian cells. Here the binding results of the bead assay are compared with conventional fluorescence assays, using lectins from three species (Triticum vulgaris, Phaseolus vulgaris, and Lens culinaris) on the two colon cell lines. These lectins were chosen because they seemed to interact with the two cell lines differently. Binding results obtained using both assays were compared for frozen, thawed and fixed; cultured and fixed; and live cells. Both qualitative and quantitative fluorescence results generally correlated with those using the bead assay. Similar results were also obtained with all of the three different cell preparation protocols. The fluorescence assay was able to detect lower lectin binding ligand levels than the bead assay, while the bead assay, because it can so rapidly detect cells with large numbers of lectin binding ligands, is ideal for initial screening studies that seek to identify cells that are rich in surface binders for specific molecules. The direct use of frozen, thawed and fixed cells allows rapid mass screening for surface molecules, without the requirement for costly and time consuming cell culture.

Direct targeting of cancer cells: a multiparameter approach.

Lectins have been widely used in cell surface studies and in the development of potential anticancer drugs. Many past studies that have examined lectin toxicity have only evaluated the effects on cancer cells, not their non-cancer counterparts. In addition, few past studies have evaluated the relationship between lectin-cell binding and lectin toxicity on both cell types. Here we examine these parameters in one study: lectin-cell binding and lectin toxicity with both cancer cells and their normal counterparts. We found that the human colon cancer cell line CCL-220/Colo320DM bound to agarose beads derivatized with Phaseolus vulgaris agglutinin (PHA-L) and wheat germ agglutinin (WGA), while the non-cancer human colon cell line CRL-1459/CCD-18Co did not. When these lectins were tested for their effects on cell viability in culture, both cell lines were affected by the lectins but at 6, 48 and 72 h incubation times, PHA-L was most toxic to the cancer cell line in a concentration dependent manner. At 48 h incubation, WGA was more toxic to the cancer cell line. The results suggest that it may be possible to develop lectin protocols that selectively target cancer cells for death. In any case, examination of both malignant cells and their non-malignant counterparts, analysis of their binding characteristics to immobilized lectins, and examination of the toxicity of free lectins in culture, provides a multiparameter model for obtaining more comprehensive information than from more limited approaches.

Early changes of LIFR and gp130 in sciatic nerve and muscle of diabetic mice.

Peripheral neuropathy is a common complication of diabetes mediated by alterations of growth factors. Members of the neuropoietic cytokine family, which include IL-6, LIF, and CNTF among others, have been shown to be important regulators of peripheral nerves and the muscles that they innervate. To investigate their potential role in diabetic nerve and muscle, we studied the expression of the shared receptor subunits, LIFR and gp130 in a mouse model of streptozotocin (STZ)-induced diabetes. The results of Western blotting and densitometric analysis showed that both LIFR and gp130 protein expression were increased in diabetic sciatic nerve compared to control mice at early time points following STZ injection. In diabetic gastrocnemius muscle, LIFR and gp130 were increased from 3 days to 24 weeks following STZ injection. In contrast, both LIFR and gp130 protein expression were decreased in diabetic soleus muscle at 3-days post-injection. Our results suggest that hyperglycemia results in changes to nerve and muscle soon after the onset of diabetes and that cytokines may play a role in this process.