*4.4. Particle Concentration*

SRFM, NTA and flow virometry were compared for the quantification of nanoparticles in HEK 293 and Sf9 samples (Figures 4–6 and Table 2). Compared to SRFM and NTA, flow virometry yielded a lower concentration of nanoparticles being around 30-fold lower for VLPs and >60-fold lower for EVs (Table 2). Similar differences in quantification by flow virometry have also been reported by van der Pol and co-authors in the analysis of urinary vesicles [5]. Differences in nanoparticle quantification between techniques could be related to the swarm effect, which is based on the fact that more than one particle passes through the detector simultaneously. However, the analysis of the same sample dilutions by flow virometry and NTA resulted in a linear correlation between both methods (VLPs, R<sup>2</sup> > 0.99; total particles R2 > 0.91), indicating that the swarm effect could not be the main reason behind these results (Supplementary materials S3). On the other hand, it is also possible that the detection of small particles fell in the range of the background signal of the cytometer, thus contributing to reduce the final titers. Even though the average particle size of EVs and VLPs was above the minimum detection particle diameter of flow virometry, additional refinement of the Mie correlation could still be required to adequately explain these data.

Analysis of the different nanoparticle populations in each platform revealed that at least 50% of the total particles produced did not correspond to VLPs (Table 2). The possibility to discriminate but also simultaneously quantify VLPs and the rest of specimens by NTA and flow virometry represents a remarkable advantage for nanoparticle-based bioprocesses. However, further refinement of light scattered-based methodologies is required since a remarkable background signal is detected in medium solutions devoid of particles (Supplementary materials S1).

#### **5. Conclusions**

The selection of an adequate analytical method is essential in VLP characterization processes. Among the six methodologies studied in this work, cryo-TEM was shown as the best method to

resolve nanoparticle structures while maintaining its native conformation. This technique allowed for a detailed characterization of the different EV and BV subpopulations co-produced with VLPs in HEK293 and Sf9 productions. Alternatively, the high-throughput analysis of VLPs and their differentiation from other contaminant particles was achieved by flow virometry, SRFM and NTA. Among them, flow virometry showed to be the fastest method for PSD analysis while also allowing to simultaneously quantify different nanoparticle subpopulations. Nonetheless, further improvements of these methods are required since different quantification results are observed between techniques.

**Supplementary Materials:** The following are available online at http://www.mdpi.com/1999-4915/12/2/223/s1, Figure S1: Negative control analyses: cell culture and conditioned media controls, Figure S2: Mie correlation for EV analysis with FCMPASS software, Figure S3: Nanoparticle correlation between flow virometry and NTA.

**Author Contributions:** Conceptualization, I.G.-D. and E.P.-M.; methodology, I.G.-D. and E.P.-M.; formal analysis, I.G.-D. and E.P.-M.; investigation, I.G.-D. and E.P.-M.; writing—original draft preparation, I.G.-D. and E.P.-M.; review, editing and supervision, L.C. and F.G. All authors have read and agreed to the published version of the manuscript.

**Funding:** Irene González-Domínguez (FPU16/02555) and Eduard Puente-Massaguer (FPU15/03577) are recipients of an FPU grant from Ministerio de Educación, Cultura y Deporte of Spain. The research group is recognized as 2017 SGR 898 by Generalitat de Catalunya.

**Acknowledgments:** The help of Manuela Costa (Servei de Cultius Cel·lulars, Producció d'Anticossos i Citometria, UAB) and Ángel Calvache and Jorge Fomaro (Beckman Coulter) for facilitating access to the Cytoflex LX equipment and the Megamix-Plus beads is very appreciated. The authors would like to acknowledge the help of Marti de Cabo and Emma Rossinyol (Servei de Microscòpia, UAB, Barcelona, Spain) for the assistance with EM methodologies, Mònica Roldan for the support with SRFM (Servei d'Anatomia Patològica, Hospital Sant Joan de Déu, Barcelona, Spain) and Joshua Welsh (NIH, Bethesda, MD, USA) for the help with FCMPASS Software.

**Conflicts of Interest:** The authors declare no conflict of interest. The funders had no role in the design of the study; in the collection, analyses, or interpretation of data; in the writing of the manuscript, or in the decision to publish the results.
