**Abbreviations**

The following abbreviations are used in this manuscript:


#### **Appendix A. Magnetic Particles Aggregation Study**

**Figure A1.** (**A**) Kinetic of NS1 labeling for 4 types of microbeads. (**B**) 20 magnification photographs of NS1 cells incubated 2 h with: (i) Dynabeads 2.8 μm (ii) Micromod 1.5 μm (iii) Dynabeads 1 μm (iv) Micromod 0.5 μm (v) Adembeads 200 nm.

mAbs-functionalized 200 nm carboxyl-Adembeads (Ademtech -R ), 500 nm streptavidin nanomag -R -D (Micromod), 1 μm DynabeadsTM MyOneTM Streptavidin T1 (InvitrogenTM), 1.5 μm streptavidin sicastar -R -M (Micromod) and 2.8 μm DynabeadsTM M-280 Tosylactivated (InvitrogenTM) suspensions were analyzed in terms of aggregate distribution (data not shown). Then, cell immunocapture kinetic studies were realized with an optical microscope. Data from 200 nm beads were hardly exploited but it was clear that beads aggregated considerably while no excessive labeling was visible at cell surface (Figure A1B(v)). The other 4 types of MPs gave quantified results presented on Figure A1 as plots of average number of beads per cell against time. The 500 nm-diameter Micromod beads labelled poorly the cells after 2 h. 1.5 μm diameter Micromod beads labelled the cells correctly but aggregates were far too numerous for those beads to be of any interest. The best options were the two kinds of DynabeadsTM. The estimation of number and size of aggregates for these beads showed that 1 μm beads seemed to form labeled cells more distinguishable from beads aggregates than the 2.8 ones. However, to confirm that, there were attempts to decrease the number of aggregates for 2.8 μm beads as they reached the maximum cell labeling quicker. They were sonicated for 30 s to 1 minute for 1 to 3 times before use with no conclusive results. Then, they were vortexed and sonicated alternatively 3 times with cycles of 2 and 1 min respectively. Finally they were added to several detergent alone or in combinations at 0.1% each (SDS (sodium dodecyl sulfate), tween 20, triton X-100, CHAPS (3-[(3-cholamidopropyl)dimethylammonio]-1-propanesulfonate)) either in PBS or in deionized water before being submitted to the 3 cycles of sonication-vortices described above. While most of these attempts remained unsuccessful, the suspension of beads in deionized water containing all detergents and the suspension in deionized water containing SDS and tween 20 showed significant improvements: the percentage of single beads improved from 66% in the suspension in PBS without detergent to 90% and 84% respectively. Once this observation was done, the cell survival was evaluated in the presence of different detergent concentrations ranging from 0.01% to 1%. All cells were lysed after 2 h contact with all these detergent concentrations. DynabeadsTM of 1 μm were thus the optimum and detergent use was abandoned. Sonication and vortex did not improve their aggregation.
