PATHFINDER

Before tumour cells metastasize, it seems that chemoattractants induce the accumulation of macrophages and haematopoietic progenitor cells in distant organs. Yoshiro Maru and colleagues have delineated a pathway that could be important in this process.

Previous work by the authors showed that distant primary tumours induce the expression of the chemoattractants S100A8 and S100A9 in the lungs of mice, thereby recruiting MAC1⁺ myeloid cells. However, they also noted that conditioned medium from pre-metastatic lungs (lung conditioned medium, LCM) stimulated with S100A8 induced migration of MAC1⁺ myeloid cells more effectively than S100A8 alone, prompting them to search for other factors that might be involved downstream of S100A8.

One gene strongly upregulated in pre-metastatic lungs, serum amyloid A 3 (SAA3), had a pattern of expression similar to that of S100A8 and S100A9. Specifically, Saa3 was strongly expressed in MAC1⁺ myeloid cells and endothelial cells from mice bearing tumours formed by Lewis lung carcinoma (LLC) cells. The authors then showed that S100A8 induced Saa3 mRNA and secretion of SAA3 protein specifically in cultured lungs, but not in livers, and that SAA3 secretion could be blocked by a neutralizing S100A8 antibody. Similar results were found in vivo.

What effect does secretion of SAA3 have? The results of a control experiment suggested that Toll-like receptor 4 (TLR4) might be a functional receptor for SAA3. Stimulation of TLR4-expressing Ba/F3 mouse haematopoietic cells with a SAA3–GST fusion protein induced activation of nuclear factor B (NFB) comparable to that of established TLR4 ligands. SAA3–GST also induced NFB activation through TLR4 in primary mouse macrophages, as shown by phosphorylation of inhibitor of NFB (IB). SAA3–GST induced chemotaxis of TLR4-expressing Ba/F3 cells and primary mouse macrophages from wild-type, but not Tlr4^-/-, mice. The authors also found that LCM from cultured lungs stimulated with SAA3–GST was a stronger chemoattractant than SAA3 alone. Analysis of the components of the SAA3-stimulated LCM indicated that there is a positive feedback loop in which SAA3–GST induces further secretion of SAA3 in wild-type, but not Tlr4^-/-, mice.

The authors have previously described a mouse model that separates pre-metastatic and metastatic stages. Mice are subcutaneously injected with LLC cells; the growth of the tumour at the primary site is the pre-metastatic phase. Subsequently more tumour cells are injected into the tail vein to create the metastatic phase. Primary tumour sizes were not significantly different in wild-type and Tlr4^-/- mice; however, there were fewer MAC1⁺ myeloid cells in the lungs of tumour-bearing Tlr4^-/- mice. A neutralizing antibody against SAA3 reduced the numbers of both MAC1⁺ myeloid cells and bone marrow-derived cells in the lungs of tumour-bearing wild-type mice. Furthermore, during the metastatic phase there was a reduction in tumour cell colonization in the lungs in wild-type, but not Tlr4^-/-, mice treated with the anti-SAA3 antibody.

These results indicate that the S100A8–SAA3–TLR4 pathway enhances metastasis to the lungs by promoting the accumulation of macrophages and haematopoietic progenitors and that blocking this pathway could help prevent metastasis.

Author: Sarah Seton-Rogers