br In literature the metabolic
In literature, the metabolic coupling between fibroblasts and tumor U0126 is associated with the fibroblast activation process measured as α-SMA expression . The role of α-SMA levels in NSCLC is con-troversial, with studies indicating that it is an unfavorable prognostic factor  and studies in which α-SMA did not showed any prognostic information .
We characterized the ability to activate fibroblasts from the two NSCLC cell lines used in the co-culture experiments (A549 and H1299). Our results show that the A549 cell line and the H1299 cell line present diﬀerent ability to activate fibroblasts when co-cultured, measured as the induction of α-SMA expression in the fibroblasts (Fig. 3A). After 4 days of co-culture, A549 cells induce an increase in α-SMA expression (at both, mRNA and protein levels) in the two diﬀerent fibroblast cell lines. On the other hand, H1299 cells are only capable of activate the CF1, and to a much lesser extent than A549 cells.
This diﬀerential ability of the tumor cells correlates with the ex-pression of epithelial markers, such as EpCAM (Fig. 3B) and E-Cadherin (data not shown), as described before . The H1299 cells with more mesenchymal characteristics were less eﬀective at activating α-SMA expression than A549 cells with more epithelial characteristics.
Interestingly, since A549 is the cell line that induces more α-SMA in the fibroblasts but it is not able to reproduce the metabolic change, these results question the assumption in which the reverse Warburg eﬀect is associated with the fibroblast activation, suggesting that these phenomena are two diﬀerent entities that do not necessarily occur to-gether.
On the other hand, these in vitro results suggest that α-SMA level may not be a good marker for the metabolic reprogramming eﬀect in-duced by fibroblasts in NSCLC tumors. To test this hypothesis, we re-analyzed the patient's data from our cohort (data not shown) and from TCGA. The results show no correlation between α-SMA expression and PGC-1α or GAPDH/MT-CO1 (Fig. 3C).
3.4. Fibroblast basal OXPHOS function modifies tumor cells metabolic reprogramming
It seems clear that there is a metabolic change in both cell types, whereas the H1299 tumor cells increase its OXPHOS function, the fi-broblasts reduce it when co-cultured. We asked ourselves if this process occurred in a similar way if we co-cultivated the tumor cell line H1299 with a fibroblast with an aﬀected OXPHOS function.
Coincidentally, when we characterized the mitochondrial function of the fibroblasts present in the laboratory, we observed deep diﬀer-ences in their galactose to glucose growth rate ratios as well as in their MIMP for one fibroblast cell line (Aﬀected Fibroblast, AF) (Fig. 4A). This result would indicate that these fibroblasts present an OXPHOS defect of unknown origin. Interestingly, this OXPHOS defect is somehow compensated by an increased glycolytic function as evi-denced by their increased LDHA mRNA levels. Moreover, in order to have an isogenic model with a decreased OXPHOS function, we treated the CF1 with EtBr (ethidium bromide) . The treatment with EtBr greatly reduced the MIMP, and specifically stopped the growth of cells in galactose, without aﬀecting their growth in glucose (Fig. 4A).
We studied the levels of MCT4 (responsible for the lactate export from the cytoplasm to the microenvironment) for both, monoculture and co-culture conditions (Fig. 4B). The basal MCT4 levels of CF1 were not detectable by immunofluorescence. On the contrary, the CF1 treated with EtBr or the Aﬀected Fibroblast showed a notable MCT4 expression reinforcing their higher glycolytic activity. Remarkably, after the co-culture with tumor cells, an increase of MCT4 levels was induced on CF1, similarly to the LDHA increase observed (Fig. 2B) and therefore supporting a glycolytic shift in the fibroblasts. On the other hand, the CF1 treated with EtBr or the AF maintain their already high MCT4 levels that did not significantly change. Similar results were obtained analyzing MCT4 mRNA levels. CF1 showed lower MCT4 mRNA levels than the AF. Furthermore, MCT4 mRNA levels were in-duced in CF1 after tumor co-culture but no diﬀerences were observed after tumor co-culture in Aﬀected Fibroblast MCT4 mRNA levels.
Additionally, after 4 days of co-culture with the diﬀerent fibroblasts, the results showed a greater MIMP increase for the H1299 cell line when it is co-cultivated with fibroblasts with a diminished OXPHOS
Fig. 3. Metabolic reprogramming is not dependent on α-SMA expression. (A) Left panel, α-SMA confocal images of two diﬀerent fibroblasts cell lines co-cultured with A549 or H1299 tumor cells. Right panel, fibroblast α-SMA mRNA levels induced by tumor cell co-culture. Data are means from at least three diﬀerent experiments. Error bars indicate standard deviation. Student's t-test p-value was considered to be statistically significant when was < 0.05 (*=p ≤0.05). Cells in monoculture were used as controls. Note that the co-culture with A549 cells induce a higher α-SMA expression than H1299 co-culture (B) Epithelial marker EpCAM by flow cytometry. A549 cells are positive for EpCAM and E-Cadherin (data not shown) markers, on the other hand fibroblasts and H1299 cells are negative. (C) α-SMA expression do not correlate with PGC-1α or GAPDH/COX5A in patients.