Sulfisoxazole does not inhibit the secretion of small extracellular vesicles

Recently, the study by Im et al. focused on blocking the release of extracellular vesicles (EVs) by cancer cells, as a strategy to block metastasis, by deploying a drug repurposing screen. Upon screening the library of FDA approved drugs in breast cancer cells in vitro, the authors reported the ability of the antibiotic Sulfisoxazole (SFX) in inhibiting EV biogenesis and secretion. SFX was also effective in reducing breast primary tumor burden and blocking metastasis in immunocompromised and immunocompetent mouse models. As we seek a compound to block EV biogenesis and secretion in our current in vivo studies, we intended to use SFX and hence performed in vitro characterization as the first step. However, treatment of two cancer cells with SFX did not reduce the amount of EVs as reported by the authors.

Over the last two decades, extracellular vesicles (EVs) have been implicated in intercellular communication and utilised as drug delivery vehicles and as reservoirs of disease biomarkers 1-5 . As they continue to garner interest, several seminal studies have established that EVs regulate various pathophysiological processes in favour of cancer progression, including remodelling the tumour microenvironment, immune evasion, coagulation, vascular leakiness, establishing the pre-metastatic niche, tropism for metastasis and transfer of chemoresistance [6][7][8][9][10][11] . Hence, there is growing interest in blocking the release of EVs and limiting their systemic circulation as a novel therapeutic avenue to treat cancer 12 . As anti-metastatic therapies are scarce, it is speculated that FDA approved drugs that target EVs could possibly fill the void.
Recently, the study by Im et al. focused on blocking the release of EVs by cancer cells, as a strategy to block metastasis, by deploying a drug repurposing screen 13 . The rationale was to screen the existing FDA approved library to identify drugs which can inhibit EV biogenesis or secretion with the obvious advantage of known mode of action, efficacy and toxicity profile and hence has the potential of immediate clinical utility. Upon screening the library of FDA approved drugs in metastatic breast cancer cells in vitro, the authors reported the ability of the antibiotic Sulfisoxazole (SFX) in inhibiting EV biogenesis and secretion. The authors also reported that SFX was effective in reducing breast primary tumor burden and blocking metastasis in immunocompromised and immunocompetent mouse models. SFX was proposed to target Endothelin receptor A (ETA) previously in 1994 14 and hence the authors validated that SFX targets ETA which can positively regulate EV biogenies and secretion. The findings in this study thus present SFX as a potential novel EV-targeted therapeutic alternative. As a group interested in EVs, the outcomes proposed in the study were encouraging and attractive as FDA approved drugs targeting EV release are limited. Recently, Datta et al. also performed a repurposing screen to identify drugs that modulate the release of EVs in prostate cancer cells 12 . However, the identified drugs are yet to be tested in vivo.
As we seek a compound to block EV biogenesis and secretion in our current in vivo studies, we intended to use SFX and hence performed in vitro characterization as the first step. However, treatment of 4T1 breast cancer cells with SFX did not reduce the amount of EVs as reported by the authors. We acknowledge the fact that our EV isolation protocol 15 was different from the study 13 ( Fig. 1 -Mathivanan laboratory protocol) and hence could have attributed to the varied results. In order to rule out the possibility of variations in the method of EV isolation or cell-type dependency, three researchers exactly followed the protocol employed by the authors (Fig. 1 -protocol   1 and 2) to isolate EVs from two different cell types (4T1 and MDA-MB231) that were used by the authors. All these assays were performed with 3 technical replicates (same batch of cells seeded in three different plates) for every biological replicate.
Consistent with our previous observations, treatment of the cancer cells with varying concentrations of SFX (50, 100 and 200 µM) did not impede the release of EVs while the positive control ceramide inhibitor GW4869, at low concentration (5 µM), inhibited EV secretion. Upon EV isolation, we quantified the protein amount, particle number and performed Western blotting for EV enriched proteins (TSG101, Alix), all normalised to equal cell number (Fig. 2). Contrary to the authors claim of SFX treatment led to a 3-fold decline in EV particle number, we observed a significant increase in particle number upon SFX (200 µM) treatment.
Overall, we report that SFX does not reduce the release of EVs (3 independent researchers) and emphasise caution in using SFX (Merk 31739) as a drug to block EV release. However, we do acknowledge the fact that our findings do not challenge the authors main conclusion of SFX mediated reduction of primary tumour burden and metastasis though our results suggest that the phenotype observed may not be cancer