Identifying viral particles and extracellular vesicles using nanopores coupled with optical detection

Abstract

The present invention relates to a method for identifying the types and / or subtypes of viral particles and / or extracellular vesicles using nanoporous membranes coupled with optical detection. The invention also relates to the use of this method, in particular for identifying / naming the type and / or subtypes of viral particles and / or extracellular vesicles present in a biological sample. More specifically, transporting viral particles and extracellular vesicles through nanopores allows the type of particle to be identified through its specific interaction with the nanopore and the modelling of this phenomenon, particularly using a bottleneck model.

Description

[0001] DESCRIPTION

[0002] Identification of viral particles and extracellular vesicles by nanopores coupled with optical detection

[0003] The present invention relates to a method for identifying the types and / or subtypes of viral particles and / or extracellular vesicles using nanoporous membranes coupled with optical detection.

[0004] It also concerns the use of this method, in particular to identify, that is to say in particular to name, the type and / or subtypes of viral particles and / or extracellular vesicles present in a biological sample. More precisely, the transport of viral particles and extracellular vesicles through nanopores makes it possible to identify the type of particle, thanks to its specific interaction with the nanopore and to the modeling of this phenomenon, for example by a traffic jam model.

[0005] Currently, the rapid and accurate identification of viral particles and extracellular vesicles remains a difficult objective to achieve at the industrial, clinical and academic levels. Commercially available tests do not fully meet the specifications required in several application areas: such as basic research (to study viral cycles and extracellular vesicle production mechanisms, etc.), or industrial bioengineering (quality assessment and production yield monitoring of viral or extracellular vesicle vectors used in vaccines or gene therapy), and even in clinical diagnostics (e.g., live monitoring of virus levels in patient samples).At the time of this invention, a compromise had to be made between rapid techniques that were not very precise and too specific, and therefore incapable of detecting several viral types (antigenic tests), and expensive and time-consuming techniques (qPCR, ELISA) that required complex intermediate steps. Although multiple improvements to these methods are proposed in the scientific literature (J. Eid, M. Mougel, and M. Socol, “Advances in Continuous Microfluidics-Based Technologies for the Study of HIV Infection,” Viruses, vol. 12, no. 9, pp. 1-16, 2020; J. Deng et al., “Rapid One-Step Detection of Viral Particles Using an Aptamer-Based Thermophoretic Assay,” J. Am. Chem. Sec., vol. 143, no. 19, pp. 7261-7266, May 2021), their inability to detect multiple viral types in the same sample remains problematic.

[0006] Nanopore-based systems offer single-molecule (or particle)-level and real-time detection capabilities. Such systems are also of increasing interest for the study of viral particle properties (L. Yang and T. Yamamoto, “Quantification of virus particles using nanopore-based resistive-pulse sensing techniques,” Front. Microbiol., vol. 7, pp. 1-7, 2016; A. Arima et al., “Selective detections of single-viruses using solid-state nanopores,” Sci. Rep., vol. 8, no. 1, pp. 1-7, 2018; L. Chazot-Franguiadakis et al., “Optical Quantification by Nanopores of Viruses, Extracellular Vesicles, and Nanoparticles,” Nano Lett., 2022; Z.D. Harms, L. Selzer, A. Zlotnick, and S.C. Jacobson, “Monitoring Assembly of Virus Capsids with Nanofluidic Devices,” ACS Nano, vol. 9, no. 9, pp. 9087-9096, 2015) and extracellular vesicles (L. Chazot-Franguiadakis et al., “Optical Quantification by Nanopores of Viruses, Extracellular Vesicles, and Nanoparticles,” Nano Lett., 2022; SLN Maas, J. De Vrij, and MLD Broekman, “Quantification and Size-profiling of Extracellular Vesicles Using Tunable Resistive Pulse Sensing,” J. Vis. Exp., vol. 92, pp. 1-7, Oct. 2014; R. Vogel et al., “A standardized method to determine the concentration of extracellular vesicles using tunable resistive pulse sensing,” J. Extracell. Vesicles, vol. 5, no. 1, pp. 1-13, 2016; S. Ryuzaki et al., “Rapid Discrimination of Extracellular Vesicles by Shape Distribution Analysis,” Anal. Chem., vol. 93, no. 18, p. 7037-7044, 2021). Indeed, the small diameter and high surface-to-volume ratio of nanopores allow for fine detection of viral characteristics and extracellular vesicles (A. Arima, M. Tsutsui, T. Washio, Y. Baba, and T. Kawai, “Solid-State Nanopore Platform Integrated with Machine Learning for Digital Diagnosis of Virus Infection,” Anal. Chem., vol. 93, no. 1 , pp. 215-227, Jan. 2021 ).

[0007] To probe the viral particle or extracellular vesicle as a whole, most nanopore-based methods rely on particle detection by electrical forcing and detection. These techniques, known as Resistive Pulse Sensing (RPS), are based on analyzing resistance changes using nanoscale constriction. SPR is a versatile technique that can be used to detect nanoparticles (M. Platt, GR Willmott, and GU Lee, “Resistive Pulse Sensing of Analyte-Induced Multicomponent Rod Aggregation Using Tunable Pores,” Small, vol. 8, no. 15, pp. 2436-2444, Aug. 2012) or biological objects such as DNA, bacteria, proteins, viruses, and extracellular vesicles (S. Howorka and Z. Siwy, “Nanopore analytics: Sensing of single molecules,” Chem. Soc. Rev., vol. 38, no. 8, pp. 2360-2384, Jul. 2009; JA Oukhaled, L. Bacri, M. Pastoriza-Gallego, JM Betton, and J.Pelta, “Sensing proteins through nanopores: Fundamental to applications,” ACS Chem. Biol., vol. 7, no. 12, pp. 1935-1949, Dec. 2012; DH Stoloff and M. Wanunu, “Recent trends in nanopores for biotechnology,” Curr. Opin. BiotechnoL, vol. 24, no. 4, pp. 699-704, Aug. 2013; GR Willmott and BG Smith, “Modelling of Resistive Pulse Sensing: Flexible Methods for Submicron Particles,” ANZIAM J., vol. 55, no. 3, pp. 197-213, 2014). Specifically, SPR techniques use a system consisting of two chambers filled with an electrolyte solution and connected by a nanopore. In practice, a voltage is applied between the two sides of the pore, and the variations in ionic current are measured. When an object passes through the pore, the ions in the electrolyte solution are momentarily evacuated from the pore, which results in a drop in the ionic current, in the form of a pulse. The measured signal contains a multitude of information representing the shape(EC Yusko et al., “Real-time shape approximation and fingerprinting of single proteins using a nanopore,” Nat. Nanotechnol., vol. 12, no. 4, pp. 360-367, Dec. 2016), le volume ( H. Yasaki et al., “Substantial Expansion of Detectable Size Range in Ionic Current Sensing through Pores by Using a Microfluidic Bridge Circuit,” J. Am. Chem. Sec., vol. 139, no. 40, pp. 14137-14142, Oct. 2017; A. Arima, M. Tsutsui, and M. Taniguchi, “Volume discrimination of nanoparticles via electrical trapping using nanopores,” J. Nanobiotechnology, vol. 17, no. 1 , pp. 1-6, Mar. 2019; R. Vogel et al., “Quantitative sizing of nano / microparticles with a tunable elastomeric pore sensor,” Anal. Chem., vol. 83, no. 9, pp. 3499-3506, 2011 ), la masse (M. Tsutsui, K. Yokota, A. Arima, Y. He, and T. Kawai, “Solid-State Nanopore Time-of-Flight Mass Spectrometer,” ACS Sensors, vol. 4, no. 11 , pp. 2974-2979, Nov. 2019) et la charge de surface des particules transportées (A. Sikora, A. G. Shard, and C.Minelli, “Size and ^-Potential Measurement of Silica Nanoparticles in Serum Using Tunable Resistive Pulse Sensing,” Langmuir, vol. 32, no. 9, pp. 2216-2224, Mar. 2016; A. Arima, M. Tsutsui, and M. Taniguchi, “Discrimination of equi-sized nanoparticles by surface charge state using low-aspect-ratio pore sensors,” Appl. Phys. Lett., vol. 104, no. 16, pp. 1-4, Apr. 2014; N. Arjmandi, W. Van Roy, L. Lagae, and G. Borghs, “Measuring the electric charge and zeta potential of nanometer-sized objects using pyramidal-shaped nanopores,” Anal. Chem., vol. 84, no. 20, pp. 8490-8496, Oct. 2012). La validité du signal de détection est généralement évaluée par comparaison avec plusieurs modèles physiques (I. W. Leong, M. Tsutsui, S. Murayama, Y. He, and M. Taniguchi, “Electroosmosis-Driven Nanofluidic Diodes,” J. Phys. Chem. B, vol. 124, no. 32, pp. 7086- 7092, Aug. 2020).

[0008] The applications of SPR to study the properties of viral particles and extracellular vesicles are numerous. In particular, SPR has been used to measure: size (knowing the pore characteristics) (L. Yang and T. Yamamoto, “Quantification of virus particles using nanopore-based resistive-pulse sensing techniques,” Front. Microbiol., vol. 7, pp. 1-7, 2016; R. Vogel et al., “Quantitative sizing of nano / microparticles with a tunable elastomeric pore sensor,” Anal. Chem., vol. 83, no. 9, pp. 3499-3506, 2011; Z.D. Harms, DG Haywood, AR Kneller, L. Selzer, A. Zlotnick, and SC Jacobson, “Singleparticle electrophoresis in nanochannels,” Anal. Chem., vol. 87, no. 1, pp. 699-705, 2015; W. Anderson, D. Kozak, VA Coleman, Â. K. Jamting, and M. Trau, “A comparative study of submicron particle sizing platforms: Accuracy, precision and resolution analysis of polydisperse particle size distributions,” J. Colloid Interface Sci., vol. 405, p.322-330, Sep. 2013), zeta potential (N. Arjmandi, W. Van Roy, L. Lagae, and G. Borghs, “Measuring the electric charge and zeta potential of nanometer-sized objects using pyramidal-shaped nanopores,” Anal. Chem., vol. 84, no. 20, pp. 8490-8496, Oct. 2012) and electrophoretic mobility (ZD Harms, DG Haywood, AR Kneller, L. Selzer, A. Zlotnick, and SC Jacobson, “Single-particle electrophoresis in nanochannels,” Anal. Chem., vol. 87, no. 1, pp. 699-705, 2015) as well as particle concentration (L. Yang and T. Yamamoto, “Quantification of virus particles using nanopore-based resistive-pulse sensing techniques,” Front. Microbiol., vol.7, pp. 1-7, 2016; RW DeBlois and RK Wesley, “Sizes and concentrations of several type C oncornaviruses and bacteriophage T2 by the resistive-pulse technique.,” J. Virol., vol. 23, no. 2, pp. 227-233, 1977). SPR also allows for qualitative probing of viral particle mechanics (A. Darvish et al., “Mechanical characterization of HIV-1 with a solid-state nanopore sensor,” Electrophoresis, vol. 40, no. 5, pp. 776-783, Mar. 2019; B.I. Karawdeniya et al., “Adeno-associated virus characterization for cargo discrimination through nanopore responsiveness,” Nanoscale, vol. 12, no. 46, pp. 23721-23731, Dec. 2020) and the dynamics of viral capsid assembly (Z.D. Harms, L. Selzer, A. Zlotnick, and S.C. Jacobson, “Monitoring Assembly of Virus Capsids with Nanofluidic Devices,” ACS Nano, vol. 9, no. 9, pp. 9087-9096, 2015). Finally, the identification of viral types, or even viral subtypes, by SPR represents a significant challenge. The ability to distinguish particles that have similar characteristics requires detecting unique features in the electrical signal for each particle type. One possible strategy, which was employed by Alrima et al., consists of coupling the detection system with computational data analysis techniques (e.g. Deep Learning) (A. Arima et al., “Selective detections of single-viruses using solid-state nanopores,” Sci. Rep., vol. 8, no. 1 , pp. 1-7, 2018; A. Arima et al., “Digital Pathology Platform for Respiratory Tract Infection Diagnosis via Multiplex Single-Particle Detections,” ACS Sensors, vol. 5, no. 11 , pp. 3398-3403, 2020). Using this type of analysis, Arima et al. were able to build a classification model, which is a function of the characteristics of the electrical signal for a given virus type (A. Arima et al., “Selective detections of single-viruses using solid-state nanopores,” Sci. Rep., vol. 8, no. 1 , pp. 1-7, 2018; A. Arima et al., “Digital Pathology Platform for Respiratory Tract Infection Diagnosis via Multiplex Single-Particle Detections,” ACS Sensors, vol. 5, no. 11 , pp. 3398- 3403, 2020).However, overall the SPR technique has major drawbacks. First, in terms of detection specificity since any object passing through the pore will be detected (including irrelevant particles), as there is no specific labeling of an object type. Therefore, the measurement does not provide direct information on the general nature of the object. Second, the SPR technique relies on the use of a single nanopore (no parallelization of the system), which is a constraint both for the concentration sensitivity of the measurement (lower concentration limit of 10. 7particles / mL (RW DeBlois and RK Wesley, “Sizes and concentrations of several type C oncornaviruses and bacteriophage T2 by the resistive-pulse technique.,” J. Virol., vol. 23, no. 2, pp. 227-233, 1977)) but also for the lifetime of the experiment, the pore can become blocked very quickly and thus prevent the measurement (A. Arima et al., “Selective detections of single-viruses using solid-state nanopores,” Sci. Rep., vol. 8, no. 1, pp. 1-7, 2018). This obstruction is irreversible, so the nanopore must be replaced. Finally, the pore size must be close to that of the particle to be detected, otherwise there will be no signal (L. Yang and T. Yamamoto, “Quantification of virus particles using nanopore-based resistive-pulse sensing techniques,” Front. Microbiol., vol. 7, pp. 1-7, 2016). This suggests the need to adapt the nanopore to the particle diameter.Nanopore size adjustment can also be achieved using Tunable RPS systems, consisting of elastomeric and tunable nanopores whose pore diameter can be adjusted in situ. However, such systems are more complex and the size detection limit is high (around 80 nm, L. Yang and T. Yamamoto, “Quantification of virus particles using nanopore-based resistive-pulse sensing techniques,” Front. Microbiol., vol. 7, pp. 1-7, 2016; R. Vogel et al., “A. standardized method to determine the concentration of extracellular vesicles using tunable resistive pulse sensing,” J. Extracell. Vesicles, vol. 5, no. 1 , pp. 1-13, 2016; R. Vogel et al., “Quantitative sizing of nano / microparticles with a tunable elastomeric pore sensor,” Anal. Chem., vol. 83, no. 9, pp. 3499-3506, 2011 ), while some viral particles or extracellular vesicles can reach 30 nm in diameter or less.

[0009] In the context of the present invention, the inventors have developed a method using a nanoporous membrane that can be functionalized, combined with an optical detection system, to identify viral particles and extracellular vesicles labeled by fluorescence. The strength and originality of this method lies in its ability to detect and identify a wide range of viral particles (enveloped or not, DNA or RNA, and of all sizes) and extracellular vesicles directly from a biological sample, with high sensitivity in concentration. Viral particles and extracellular vesicles are detected in their entirety, i.e. the entire particle is detected, and not just the genetic material as in PCR (Polymerase Chain Reaction) tests, or a specific protein as in ELISA (Enzyme-Linked ImmunoSorbent Assay) tests.The method described in the present invention does not require intermediate purification or amplification steps. Compared to existing methods, the method of the present invention makes it possible to identify viral particles and extracellular vesicles without the need for the test to be designed for a particular type / subtype of viral particles or extracellular vesicles.

[0010] Furthermore, unlike the RPS technique, the method described in the present invention has a modulatable detection specificity. Indeed, in the RPS method, the particles are not labeled, which is a major drawback when studying complex samples, since any object present in solution (cellular debris, DNA, proteins, etc.) will be detected. This greatly disrupts the measurement with numerous spurious signals. On the other hand, the method according to the present invention involves fluorescent labeling of the particles to be detected. This labeling is carried out before the start of the measurement, and its detection specificity can be modulated, since it is possible to label a specific viral particle or extracellular vesicle (using an antibody-fluorophore combination) or a group of particles.This marking represents a first level of detection, making it possible to distinguish groups of particles from the rest of the biological objects which are always part of a complex sample.

[0011] Furthermore, the RPS technique relies on the use of a single nanopore, whereas the nanoporous membrane described in the present invention has the considerable advantage of having a large number of pores in parallel (approximately 10,000 times higher than the RPS). Having a large number of nanopores in parallel significantly increases the measurement statistics, resulting in a much higher concentration sensitivity (lower concentration limit of 10 3 -10 5particles / mL). In addition, high parallelization also avoids the irreversible phenomenon of nanopore clogging, which frequently occurs with the SPR technique and prevents measurement. Finally, the method does not require the nanopore to be close in size to the particle diameter. Therefore, for different types of viral particles or extracellular vesicles, the same nanopore diameter can be used.

[0012] With the method according to the present invention, the identification of viral particles and extracellular vesicles is simpler, faster (minutes versus hours for qPCR for example) and precise.

[0013] The inventors of the present invention have developed a method using nanoporous membranes to increase the sensitivity of virus detection. By measuring the frequency of passage through the nanoporous membrane (the number of fluorescent particles passing through the nanopores per unit of time) as a function of a control parameter (pressure, concentration of viral particles or extracellular vesicles, etc.), it is possible to extract the dynamic and / or physicochemical parameters of interaction of the viral particle or extracellular vesicle with the nanopore and / or of the interaction rate between the viruses and / or extracellular vesicles themselves or between the viruses and / or extracellular vesicles and the surface of the nanopore.These parameters being unique for a given viral type / subtype (or extracellular vesicle), their measurement allows the creation of a specific signature and a database allowing the identification of the types and / or subtypes of viral particles and / or extracellular vesicles.

[0014] Thus, the present invention relates to a method for identifying the type and / or subtypes of viral particles and / or extracellular vesicles comprising: a. labeling the viral particles and / or extracellular vesicles present in a biological sample using a fluorescent marker; b. passing said sample comprising the labeled viral particles and / or extracellular vesicles through a nanoporous membrane; c. measuring the transport of said viral particles and / or extracellular vesicles through said nanoporous membranes by optical detection as a function of at least one control parameter; and d.the extraction of physicochemical and / or dynamic parameters and / or interaction rates between the viruses and / or extracellular vesicles themselves or between the viruses and / or extracellular vesicles and the surface of the nanopore, characteristics of the transport of said viral particles and / or extracellular vesicles from a physical model linking the transport to the control parameter, for example a traffic jam model.

[0015] Advantageously, these parameters extracted in step d. allow the identification of the type and / or subtype of viral particles and / or extracellular vesicles, in particular by linking them to a database.

[0016] The present invention also relates to the use of the method according to the invention for identifying the viral type and / or subtype present in a patient, the quality of the production of viral vectors or the identification of zoonosis in biological samples.

[0017] The present invention also relates to the use of the method according to the invention for identifying the type and / or subtype of extracellular vesicles in a patient or the quality of the production of vesicular vectors. Detailed description

[0018] As mentioned above, the present invention relates to a method for identifying viral particles and / or extracellular vesicles comprising: a. labeling the viral particles and / or extracellular vesicles present in a biological sample using a fluorescent marker; b. passing said sample comprising the labeled viral particles and / or extracellular vesicles through a nanoporous membrane; c. measuring the transport of said viral particles and / or extracellular vesicles through said nanoporous membranes by optical detection as a function of at least one control parameter; and d. extracting physicochemical and / or dynamic parameters and / or interaction rates, characteristic of the transport of said viral particles and / or extracellular vesicles from a physical model linking the transport to the control parameter, for example a traffic jam model.

[0019] Advantageously, these parameters can be linked to a database to enable identification of the type and / or subtype of viral particles and / or extracellular vesicles.

[0020] Identification refers to the process of associating detected viral particles and / or extracellular vesicles with their corresponding type(s) and / or subtype(s). Identification does not only refer to the detection of viral particles and / or extracellular vesicles present in the sample, but also refers to specifically naming the type(s) and / or subtype(s) of virus and / or extracellular vesicles present in a sample. For example, this may involve identifying whether the sample contains influenza virus (InfV) or hepatitis B virus (HBV), which is referred to as identifying the viral type. Another example is identifying whether the sample contains influenza A(H1N1) or A(H3N2), which is referred to as identifying the viral subtype.

[0021] “Viral particles” means biological entities related to the class of viruses. This includes viruses of all types (enveloped, non-enveloped, DNA viruses, RNA viruses, ...) and also includes all derivatives, i.e. viruses obtained by means other than natural infection. This includes laboratory-produced viruses (in complete or partial form, such as virus-like particles or viral protein assemblies), as well as viral vectors and the bioproduction of viruses for vaccination processes. Viruses vary in size (typically about 20 to 400 nm in diameter), geometry, and surface and mechanical properties. In one embodiment according to the invention, the viral particles are selected from enveloped or non-enveloped viral particles, which may contain DNA or RNA.

[0022] "Viral type" means the species of virus as established in the official classification (International Committee on Taxonomy of Viruses). For example, influenza A virus (Influenza A) corresponds to a viral type.

[0023] "Viral subtype" means the serotype of the virus as established in the official classification (International Committee on Taxonomy of Viruses). For example, influenza A(H1N1) or A(H3N2) virus both correspond to a viral subtype of influenza A virus (Influenza A).

[0024] Extracellular vesicles are biological particles consisting of a lipid bilayer that are naturally released by almost all cell types. Extracellular vesicles are a complex and heterogeneous class of biological nanoparticles present in many biological fluids, including, for example, urine and blood. Extracellular vesicles are highly dispersed in size, typically ranging from 30 to 300 nm in diameter. The size, morphological, mechanical, and biochemical properties of extracellular vesicles are highly variable. Extracellular vesicles and their synthesis pathways are also used by tumors and infectious diseases to facilitate their propagation and evade the immune system.The present invention includes extracellular vesicles naturally secreted by cells, those produced in the laboratory, as well as vectors of bioactive macromolecules.

[0025] In one embodiment, the extracellular vesicles are selected from vesicles produced by healthy or cancerous cells, or from vesicles derived from non-cellular bioproduction.

[0026] Extracellular vesicle type refers to a population of extracellular vesicles from a single cell lineage / type. Extracellular vesicles are generally classified into types based on their cellular origin and the presence of viral infection or tumors in the cellular environment.

[0027] Extracellular vesicle subtype refers to the different populations of extracellular vesicles that can be found within the same extracellular vesicle type (i.e., extracellular vesicles from the same cell line). The subtype is defined based on size, morphological, mechanical, and biochemical properties.

[0028] A "biological sample" means a sample from a biological environment. A "biological environment" means the specific conditions and components of the medium in which biological particles such as viral particles or extracellular vesicles can be stored. Typically, this is a complex medium including a buffer, a cell culture medium, a cell supernatant, a real medium (water sample, agri-food sample), biological samples of living organisms (including patient samples).

[0029] The biological sample can be, for example, a sample taken from a patient, but also a sample from a bioproduction process (in a laboratory or in industrial production), or a sample taken from the environment and in particular from river water, or agri-food samples.

[0030] The term "fluorescent marker" has the usual meaning for a person skilled in the art. In particular, the following examples may be cited: membrane fluorescent markers (of the lipid bilayer), markers of genetic material, markers linked to a protein, markers binding to a specific protein, markers linked to a primary or secondary antibody, this list not being exhaustive. Examples are given in Table 5 of the experimental part. Examples include:

[0031] - YOYO-1 (Molecular Probes) which is capable of penetrating the viral capsid and marking the viral genetic material,

[0032] - SYTO-13 (Molecular Probes) which is capable of penetrating through the viral envelope and marking viral genetic material,

[0033] - DiO (Vybrant Cell Labelling, Molecular Probes), which is a lipophilic carbocyanine that emits weak fluorescence in water but is highly fluorescent and photostable when incorporated into membranes or lipid environments,

[0034] - GFP, fused to the C-terminal position of the structural protein Gag.

[0035] -Alexa Fluor 488® linked to the gp64-Baculovirus antibody (AcV1, Biotechne), which corresponds to a fluorophore linked to a primary antibody targeting the gp64 envelope protein of the baculovirus.

[0036] Labeling can be carried out by any method known to those skilled in the art, for example by incubation for a variable duration between the sample and the chosen fluorophore, or incorporation of the fluorophore at the time of bioproduction. Fluorescent labeling allows for detection with adjustable specificity.

[0037] By "detection specificity" is meant the ability of a measurement to accurately detect a particular particle group / class, while excluding or minimizing interference from other types of particles or unwanted noise. A specific measurement is therefore capable of providing accurate and reliable results for the specific particle group / class under study (in the present invention, these are extracellular vesicles and / or viral particles), without being significantly influenced by other particles or sources of error, such as, for example, the presence of cellular debris, bacteria, contamination.

[0038] Within the scope of the present invention, the detection specificity can be modulated.

[0039] By "modulatable detection specificity level" is meant that the detection specificity level of the measurement can be modulated. In the present invention, the detection specificity is ensured by several factors, including the presence of the nanoporous membrane which can be functionalized and the use of fluorescent labels. The use of a functionalized nanoporous membrane with a pore size of the order of a few tens to hundreds of nanometers (for example nanopores of 50 to 400 nm in diameter) acts as a filter and prevents the passage of undesirable elements through the pores (for example bacteria or large cellular debris).In addition, fluorescent labeling allows the labeling of a specific group of particles (viral particles and / or extracellular vesicles), with a precision that can be adjusted since it is also possible to use a very specific labeling such as an antibody-fluorophore combination to label a specific type or subtype of viral particle and / or extracellular vesicle.

[0040] The term "nanoporous membrane" means solid organic or inorganic membranes with pores on the nanoscale.

[0041] In the context of the present invention, the membranes comprise several pores in parallel, for example between 10 6 and 10 9 pores / cm 2 The membranes have pores with diameters ranging from 10 to 600 nm.

[0042] In one embodiment, the nanoporous membrane is functionalized.

[0043] By "functionalized nanoporous membrane" we mean solid organic or inorganic membranes whose nanometric-sized pores are grafted with synthetic polymers.

[0044] By "electrografted polymers" is meant polymers which contain at least one electroactive function allowing the grafting of the polymer onto a surface. More particularly, the electrografted polymers are chosen from "poly(2-alkyl-2-oxazoline) polymers" well known in the art.

[0045] In particular, the term “alkyl” means an alkyl chain comprising between 1 and 4 carbons. Poly(2-methyl-2-oxazoline), poly(2-ethyl-2-oxazoline), poly(2-isopropyl-2-oxazoline) and poly(2-n-propyl-2-oxazoline) may be cited as examples. In particular, the polymers according to the invention comprise between 2 and 1000 monomers, more particularly between 10 and 800.

[0046] In one embodiment, said polymers are linked to electroactive probes, in particular of the diazonium type. This allows electrochemical grafting of these polymers onto the nanoporous surface.

[0047] The process for electrografting polymers onto a functionalized nanoporous membrane is as described in patent application WO2022195059A1.

[0048] Nanoporous membranes can also be functionalized by non-covalent grafting, in particular with Pluronic © or poly(vinylpyrrolidone), methylcellulose, albumin, or fetal bovine serum.

[0049] Detection refers to the process of determining the presence or absence of viral particles or extracellular vesicles in a sample. Detection differs from identifying the type / subtype of viral particles and / or extracellular vesicles. In particular, detection occurs in real time.

[0050] By "real-time detection" we mean that the transport of particles (viral particles or extracellular vesicles) through nanopores is observed in real time at the single-particle level.

[0051] In the context of the present invention, the method is capable of detecting a virus and / or an extracellular vesicle in a biological sample, but also a mixture of viruses and / or extracellular vesicles present in the same sample.

[0052] As mentioned above, the measurement of the transport of said viral particles and / or extracellular vesicles through said nanoporous membranes is carried out by optical detection as a function of at least one control parameter.

[0053] In particular, Zero Mode Waveguide microscopy (intensification of the electromagnetic field in the vicinity of a nanopore covered with a metal film) can be used. This extremely sensitive optical system allows the real-time detection of the transport of single fluorescent particles at the level of a single pore.

[0054] By "control parameter" we mean a parameter controllable by the experimenter and in particular the concentration of viral particles and / or extracellular vesicles, pressure, temperature, salinity and pH.

[0055] By "measured parameter" is meant a parameter observed and / or quantified during the experiment, and in particular the frequency of passage of the fluorescent particles, i.e. the number of particles passing through the pores per unit of time, the intensity of the transport events, the duration of the transport events or the trajectory of the transport events. In the context of the present invention, the terms "transport", "translocation", "passage", and "event" are used interchangeably for the transport / passage / translocation of the viral particles / extracellular vesicles through the nanoporous membrane. Advantageously, in step b), the passage of the sample through the nanoporous membrane can be done by hydrodynamic entrainment.Unlike electrical drive (used in the case of SPR presented in the introduction), hydrodynamic drive allows the application of weak and controlled forces on the transported particles. For example, forces of the order of magnitude of thermal agitation can be used and controlled. Moreover, the modeling of hydrodynamic forces is straightforward and simple. It does not involve coupling with other physical quantities / phenomena as in the case of electrical drive (e.g. electro-osmosis, dielectrophoresis).

[0056] By "transport event intensity" we mean the spatio-temporal profile of the light intensity of events. In the case of an event, i.e. the passage of a fluorescently labeled viral particle or extracellular vesicle, the light intensity is extracted from the raw signal.

[0057] The term "trajectory of transport events" refers to the evolution of the spatial coordinates of events as a function of the intensity of transport events over time. Once an event has been identified, the coordinates of its center, for example, can be recorded over time.

[0058] This results in the extraction of physicochemical and / or dynamic parameters, and / or interaction rates between the viruses and / or extracellular vesicles themselves or between the viruses and / or extracellular vesicles and the surface of the nanopore, characteristic of the transport of said viral particles and / or extracellular vesicles, from a physical model, for example of traffic jam, linking the measured transport parameter to the control parameter.

[0059] By "extraction of physicochemical and / or dynamic parameters and / or interaction rates" is meant the fact of quantifying the physicochemical and / or dynamic parameters and / or interaction rates using a physical model giving the transport as a function of one or more control parameters, in particular by adjustment to this physical model. The physicochemical and / or dynamic parameters and / or interaction rates correspond in particular to at least one parameter selected from the transport rate, the affinity, and the association or dissociation rates of the viral particles and / or extracellular vesicles with each other or with the surface of the nanopore.

[0060] By "interaction rate" is meant the rates of association and / or dissociation of the viral particles and / or extracellular vesicles with each other or with the surface of the nanopore. Preferably, step d. is the extraction of physicochemical parameters and / or rates of interaction between the viruses and / or extracellular vesicles themselves or between the viruses and / or extracellular vesicles and the surface of the nanopore from the modeling of the measurement of the transport of said viral particles and / or extracellular vesicles by a physical model, preferably a traffic jam model, linking the transport to the control parameter. These parameters are linked to a database to allow the identification of the type and / or subtype of viral particles and / or extracellular vesicles.

[0061] By "traffic jam" or "jamming", or even "plug", we mean the presence of a state of congestion due to the presence of viruses and / or extracellular vesicles inside a nanopore which induces a slowdown in the transport of viruses and / or extracellular vesicles crossing the nanopore. From an experimental point of view, this traffic jam phenomenon is notably reflected by the reduction in the frequency of passage with the increase in the concentration of viruses and / or extracellular vesicles.

[0062] By "transport measurement" is meant the measurement of the frequency of passage of viruses and / or extracellular vesicles through the nanopores and / or another parameter measured as a function of a control parameter. In particular, these measurements correspond to the relationships between the translocation frequency and the applied pressure, to the relationships between translocation frequency and the concentration of particles or vesicles, as well as to measurements of the fluorescence intensity as a function of time.

[0063] The "physical model" or "identification model" or "transport measurement modeling" corresponds to the transport model used to extract specific physicochemical and / or dynamic parameters through the measurement of the transport of viral particles or extracellular vesicles through the nanoporous membrane. In particular, a model describing a particle plug phenomenon in the nanopore linking the frequency of particle passage to their physicochemical and / or dynamic interaction parameters can be used.

[0064] The "physical traffic jam model" corresponds to the physical transport model describing the relationship between the frequency of passage of viruses and / or extracellular vesicles and the control parameters. This model depends on a finite set of physicochemical parameters and / or interaction rates between the viruses and / or extracellular vesicles themselves or between the viruses and / or extracellular vesicles and the nanopore surface. Each type / subtype of virus or extracellular vesicle can be associated with a given set of these physicochemical parameters and / or interaction rates. A given set of these physicochemical parameters and / or interaction rates therefore corresponds to the specific signature of a type / subtype of virus or extracellular vesicle.

[0065] By "extraction of physicochemical parameters and / or interaction rates between the viruses themselves as well as between the virus and the surface of the nanopore" we mean the quantification of said parameters using an adjustment of the experimental data ("transport measurement") by the physical traffic jam model.

[0066] The transport measurement may correspond to the curves produced during the transport of viral particles or extracellular vesicles across the nanoporous membranes. In particular, these curves correspond to the translocation frequency curves as a function of the pressure applied to the system curves, to the translocation frequency curves as a function of the particle or vesicle concentration, as well as to the event intensity curves as a function of time. In particular, the physicochemical parameters may be extracted from the "transport measurement" using the adjustment with the "identification model", and in particular the "physical traffic jam model". The transport parameters constitute the identity card of the particle (viral particle or extracellular vesicle), since they make it possible to trace back to the associated type or subtype of the particle using the method described in the present invention.

[0067] In particular, the use of the identification model allows extracting these physicochemical and / or dynamic parameters, and / or interaction rates between the viruses themselves, as well as between viruses and the nanopore surface. These parameters are linked to the definition of a unique domain for a viral particle or a type or subtype of extracellular particle in the identification database.

[0068] Advantageously, the method of the invention comprises a step e. of identifying said viral particles and / or extracellular vesicles present in the sample, by comparing the parameters extracted in step d., to a database comprising the specific signature of each viral type or subtype and / or each type or subtype of extracellular vesicle.

[0069] A specific signature may be associated with a “reference sample”, which corresponds to a biological sample of viral particles or extracellular vesicles of a specific type or subtype and used to construct the database described in the context of the present invention.

[0070] Indeed, in particular, the physicochemical parameters and / or interaction rates between the viruses and / or extracellular vesicles themselves or between the viruses and / or extracellular vesicles and the surface of the nanopore of said viral particles and / or extracellular vesicles are compared to a database comprising the specific signature of each viral type or subtype and / or each type or subtype of extracellular vesicle in order to identify said viral particles and / or extracellular vesicles present in the sample.

[0071] More particularly, the specific signature of each viral type and / or each type of extracellular vesicle allowing the identification of the labeled viral particles and / or extracellular vesicles is obtained by carrying out steps a) to d) with samples comprising known viral particles and / or extracellular vesicles.

[0072] By "database" is meant a structured collection of different viral particles or extracellular vesicles, where physicochemical parameters and / or interaction rates between the viruses and / or extracellular vesicles themselves or between the viruses and / or extracellular vesicles and the nanopore surface and / or specific dynamic parameters have been measured for each type / subtype of particle or vesicle, allowing to construct a distinct domain for each specific type or subtype of particle or vesicle in the overall database. The elements composing this database correspond to reference samples of a specific type or subtype of viral particles or extracellular vesicles for which specific transport parameters have been measured according to the method described in the present invention.

[0073] As mentioned above, the present invention also relates to the use of the method according to the invention for identifying the viral type and / or subtype present in a patient, the quality of viral vector production or the identification of zoonosis in biological samples.

[0074] The present invention also relates to the use of the method according to the invention for identifying the type and / or subtype of extracellular vesicles in a patient or the quality of the production of vesicular vectors.

[0075] Thus, the present method finds application in the medical field, in particular in the clinical or diagnostic field, in the field of bioproduction, in particular in the manufacture of vaccines and gene therapies, in the field of health monitoring of human, plant or animal pathologies, as well as in the agri-food field to identify the presence of viral particles.

[0076] The invention is illustrated by the following figures and examples.

[0077] Figures

[0078] [Fig 1]: Zero Mode Waveguide setup for translocation of viral particles or extracellular vesicles through functionalized nanoporous membranes. The cis chamber is connected to a pressure controller and contains the fluorescently labeled viral particles or extracellular vesicles. Upon application of the pressure difference between the two chambers, the particles are transported through the nanopore into the trans chamber. A laser is directed to the trans side and the particles are illuminated as soon as they cross the evanescent field region at the end of the pore. The Zero Mode Waveguide effect consists of an amplification of the fluorescent signal in the vicinity of the gold layer. Once they leave the nanopore, the fluorescently labeled particles become blurred (they go out of focus) and are bleached.

[0079] [Fig 2]: A. Evolution of translocation frequency as a function of pressure for HIV particles (Gag-GFP, IRIM) at different concentrations, f is the translocation frequency divided by a prefactor, k. P is the applied pressure divided by a critical pressure Pc.

[0080] B. Evolution of translocation frequency as a function of viral concentration for different examples of viral types. The pressure is fixed at 8 mbar. G. Evolution of translocation frequency as a function of pressure for extracellular vesicles (HeLa cells (no virus production), IRIM) at different concentrations. For A., ​​B. and C. The solid lines are examples of fits by an identification model (here the traffic jam model, or plug model, or jamming model). Nanoporous membrane with a pore diameter of 200 nm. The experimental errors are the standard error of the mean, and the number of repetitions is greater than 8 for each experimental series.

[0081] [Fig 3]: A. Evolution of translocation frequency as a function of pressure for HIV particles (Gag-GFP, IRIM, 70.10 5particles / mL) for different experimental conditions: nanoporous membrane with a conventional buffer (with 5 mM NaCl), functionalized nanoporous membrane (grafted with poly(2-methyl-2-oxazoline) of polymerization degree 387) with a conventional buffer (with 5 mM NaCl) and nanoporous membrane with a buffer containing more salt (with 150 mM NaCl). f is the translocation frequency divided by a prefactor, k. P is the applied pressure divided by a critical pressure Pc.

[0082] B. Evolution of translocation frequency as a function of pressure for extracellular vesicles (HeLa cells (no virus production), IRIM, 30.10 5particles / mL) for different experimental conditions: nanoporous membrane with a conventional buffer (with 5 mM NaCl) or a buffer containing more salt (with 150 mM NaCl). For A., ​​B. The solid lines are examples of fits by an identification model (here the bottling model, i.e., a cap). Nanoporous membrane with a pore diameter of 200 nm. The experimental errors are the standard error of the mean, and the number of repetitions is greater than 8 for each experimental series.

[0083] [Fig 4]: Voronoi diagram representing several domains, each associated with a type of viral particle. A color is associated with each domain, the borders of which are shown in black. Some points correspond to the location of each viral domain, i.e., the average physicochemical parameters that are associated with a specific viral type. A particular point corresponds to an “unknown sample” correctly assigned to the influenza domain (Influenza, InfV).

[0084] Definitions

[0085] “Unknown sample or unknown particle” means a “biological sample of viral particle or extracellular vesicle” where the type or subtype of viral particle and / or extracellular vesicle present in the sample is not known.

[0086] By "identification specificity" we mean the probability of obtaining a negative test if the particle sought is not present (in other words, it is the probability of obtaining a negative test in non-diseased people). In particular, it can be given by the following formula: Specificity = (True Negative) / (True Negative + False Positive) as will be detailed in the experimental part.

[0087] By "concentration sensitivity" we mean the detection limit of the system, i.e. the smallest concentration (of viral particles or extracellular vesicles) that can be detected and worked with.

[0088] By "identification sensitivity" we mean the probability that the test correctly assigns the unknown particle to the correct area (in other words, it is the probability that the test will be positive if the disease is present). In particular, it can be given by the following formula: Sensitivity = (True Positive) / (True Positive + False Negative), as will be detailed in the experimental section.

[0089] By “accuracy” we mean the probability corresponding to the number of unknown samples that were correctly assigned to the correct type / subtype of viral particles or extracellular vesicles among all the unknown samples that were assigned to that type / subtype of viral particles or type / subtype of extracellular vesicles (in other words, it corresponds to the number of samples that were correctly classified as positive among all the positives.) In particular, it can be given by the following formula: Accuracy=(True Positive) / (True Positive+False Positive), as will be detailed in the experimental part. By “F1 score” we mean the harmonic mean of the identification accuracy and sensitivity, which is a measure of the performance of the model’s classification ability.In particular, it can be given by the following formula: F1 -score = 2 (Precision x Sensitivity) / (Precision + Sensitivity), as will be detailed in the experimental part.

[0090] Examples

[0091] Example of preparation of nanoporous membrane according to the invention

[0092] The nanoporous membranes used are track-etched polycarbonate membranes from Whatman Nucleopore (GE Healthcare). Polycarbonate was chosen because of its low interaction with proteins. These commercially available nanoporous membranes consist of a 6-10 μm thick polycarbonate layer and have cylindrical nanopores with diameters and densities ranging from 50-400 nm and 1-6 10, respectively. 8 pores / cm 2 .

[0093] Membrane specifications are summarized in Table 1.

[0094] [Table 1]

[0095] Healthcare) (“GE Polycarbonate Track Etch ( PCTE ) Membrane”, Ge Water Process Technol., pp. 1-2, 2010).

[0096] To visualize the translocation of viral particles and extracellular vesicles through the nanopores, a 50 nm thick gold layer was deposited on top of the membrane using a clean room evaporator. The evaporation process was carried out at a pressure below 10' 6Pa and gold was deposited at a rate of 0.1 nm / s. The deposition rate and thickness were controlled by a quartz balance. The gold-coated membranes were characterized using both scanning electron microscopy (SEM) and atomic force microscopy (AFM) (P.J. Kolbeck et al., “Thermally Switchable Nanogate Based on Polymer Phase Transition,” Nano Lett., 2023). The diameters measured by SEM are grouped in Table 2 (L. Chazot-Franguiadakis et al., “Optical Quantification by Nanopores of Viruses, Extracellular Vesicles, and Nanoparticles,” Nano Lett., 2022; P.J. Koibeck et al., “Thermally Switchable Nanogate Based on Polymer Phase Transition,” Nano Lett., 2023; T. Auger et al., “Zero-mode waveguide detection of flow-driven DNA translocation through nanopores,” Phys. Rev. Lett., vol. 113, no. 2, pp. 1-5, 2014).

[0097] [Table 2]

[0098] Table 2: Pore diameters determined by SEM measurements with an average over more than 200 pores.

[0099] To increase the sensitivity of the measurements, it is advantageous to work with functionalized nanoporous membranes. For example, different types of polymers have been used: electrochemically grafted poly(2-alkyl-2-oxazolines) of two types (poly(methyl-2-oxazolines) and poly(2-n-propyl-2-oxazolines). The synthesis of these polymers (with different degrees of polymerization varying between 2 and 1000, more particularly between 10 and 800) and their grafting are described in patent application WO2022195059, which details their preparation in detail. Briefly, the polymer grafting is carried out above the gold layer of the nanoporous membrane by electrografting (connection between the polymers and the gold surface via an azo coupling).

[0100] Example of a nanoporous membrane coupled to an optical system according to the invention Experimental setup

[0101] The experimental setup consists of an optical detection and pressure control system to induce the transport of viral particles and extracellular vesicles across a nanoporous membrane and has been previously described (L. Chazot-Franguiadakis et al., “Optical Quantification by Nanopores of Viruses, Extracellular Vesicles, and Nanoparticles,” Nano Lett., 2022; T. Auger et al., “Zero-mode waveguide detection of flow-driven DNA translocation through nanopores,” Phys. Rev. Lett., vol. 113, no. 2, pp. 1-5, 2014; T. Auger et al., “Zero-mode waveguide detection of DNA translocation through FIB-organized arrays of engineered nanopores,” Microelectron. Eng., vol. 187-188, pp. 90-94, 2018) (Figure 1). More precisely, the device consists of two chambers separated by a nanoporous membrane.The optical system consists of a conventional epifluorescence microscope (Axiovert 200, ZEISS), on which a water immersion objective (ZEISS C-Apochromat, Korr M27) ​​is mounted. The objective is equipped with chromatic correction, and has a numerical aperture of 1.2 and a magnification of 63x. The microscope also has a filter box (combining several emission / excitation filter pairs and a dichroic mirror). The laser (Cobolt blues 50, Cobolt AB) emits at a wavelength of 473 nm and has a power of 50 mW, making it suitable for fluorescence measurements on biomolecules. The laser follows an optical path, and is notably broadened by a telescope, so that the beam is parallel to the objective exit. Finally, the microscope is connected to an EMCDD camera (Electron Multiplying Charged Coupled Device (Andor, iXon+)). Typically, in experiments, a frame rate of 33 Hz is used (with 512x512 pixels).In addition, the camera is cooled to -60°C, which is essential to limit thermal noise. The entire setup, including the laser, light path, fluorescence microscope, dual-chamber system, and camera, is located on a stable, air-cushioned optical table.

[0102] The heart of the experimental setup consists of two chambers connected by the functionalized nanoporous membrane, as shown in Figure 1:

[0103] -The lower chamber (trans chamber) consists of a Teflon ring on which a glass slide (0.17 mm thick) is glued. The glue used is always a biocompatible silicone glue (Silcomet JS 533 Red, Loctite).

[0104] -The upper chamber (cis chamber) consists of a cap at the end of which a hole is drilled and the functionalized nanoporous membrane glued. The upper part of the cap serves as a reservoir in which the viral particle or extracellular vesicle solution is deposited. The upper chamber is also connected to a pressure controller (MFCS-FLex, Fluigent). The pressure controller includes four channels, three of which range from 0 mbar to 1 bar with an accuracy of 1 mbar and a more precise channel from 0 mbar to 69 mbar with an accuracy of 0.1 mbar.

[0105] The trans and cis chamber assembly is placed above the objective. This assembly can be moved in all three spatial directions using a turntable, with micrometric precision in both X and Y directions.

[0106] For a typical experiment, fluorescently labeled viral particles or extracellular vesicles are placed in the cis chamber, and by applying a pressure difference between the two chambers, they are transported through the functionalized nanoporous membrane into the trans chamber. The nanoporous membrane is illuminated by a laser beam on the trans side, and the particles are observed exiting the pores. The detection is based on the Zero-Mode Waveguide effect which consists of an amplification of the fluorescence signal in the vicinity of the gold layer ( L. Chazot-Franguiadakis et al., “Optical Quantification by Nanopores of Viruses, Extracellular Vesicles, and Nanoparticles,” Nano Lett., 2022 ; , P.J. Koibeck eta / ., “Thermally Switchable Nanogate Based on Polymer Phase Transition,” Nano Lett., 2023; T. Auger et al., “Zeromode waveguide detection of flow-driven DNA translocation through nanopores,” Phys. Rev. Lett., vol. 113, no. 2, pp. 1-5, 2014; T.Auger, “Translocation of biopolymers through natural or artificial pores,” Paris Cité, 2016; B. Molcrette et al., “Experimental study of a nanoscale translocation ratchet,” Proc. Natl. Acad. Sel., vol. 119, no. 30, pp. 1-10, Jul. 2022; N. Klughammer and C. Dekker, “Palladium zero-mode waveguides for optical singlemolecule detection with nanopores,” Nanotechnology, vol. 32, no. 18, pp. 1-14, Feb. 2021). This effect makes it possible to work at the single-particle scale. Therefore, the presence of the gold layer is essential to promote the Zero Mode Waveguide effect, but the gold also acts as a mirror, blocking the view of fluorescent particles upstream of the nanopore, so that the only visible particles are those exiting the pores (Figure 1). The fluorescence of the particles decreases due to bleaching and defocusing as the particles move away from the membrane. Finally, the field of view on the nanoporous membrane is 125x125pm.2 (measured using a Thorlabs R1 L3S2P micrometer ruler). For a pore density between 1 .10 8 and 6.10 8 pores / cm 2 (depending on the pore diameter, see Table 1), the number of pores is estimated to be approximately between 15,600 (±2,500) and 93,700 (±15,000) pores in parallel across the field of view. This high number of pores provides good statistics and thus high concentration sensitivity for measuring the average translocation frequency of viral particles and extracellular vesicles. This is a huge advantage of the method described here over that of the SPR technique.

[0107] Example of preparation of samples of viral particles and extracellular vesicles according to the invention

[0108] Samples of viral particles and extracellular vesicles were obtained courtesy of various virology research laboratories. Briefly, viral particles were obtained by cellular expression of viral proteins, plasmid transfection processes, or viral cell culture. Extracellular vesicles were harvested from the culture medium of HeLa, NIH3T3, or 293THEK cells producing or not producing viral particles. Examples of viral particle types and subtypes tested are summarized in Table 3, while examples of extracellular vesicles tested are summarized in Table 4. [Table 3]

[0109] Table 3: Examples of viral types and subtypes used in this method.

[0110] [Table 4]

[0111] Table 4: Example of types of extracellular vesicles (EVs) used in this method.

[0112] Overall, these samples are complex biological samples collected from an unpurified biological medium (e.g., a cell supernatant) without intermediate purification steps. Furthermore, tests were also carried out on water collected from the Rhône River, for example (the collected samples were filtered through 450 nm pores before carrying out the experiment).

[0113] Experimental conditions

[0114] Typically, experiments are performed in a buffered medium (at a fixed pH). The buffer used is a conventional buffer, Tris-EDTA buffer composed of 10 mM tris(hydroxymethyl)aminomethane potassium (Tris-KCl), and 1 mM ethylenediaminetetraacetic acid (EDTA). Once mixed, the pH is adjusted to 7.5 using a 0.1 M hydrochloric acid solution (Sigma Aldrich). It is then filtered through a Whatman nylon membrane (pore diameter 0.2 μm), to avoid the presence of impurities. The buffer is stored at 4°C. The buffer used can also be PBS (1X, Mg-free, Ca-free, Sigma Aldrich).

[0115] Fluorescence labeling

[0116] To visualize viral particles or extracellular vesicles emerging from the ZMW-functionalized nanoporous membrane, the particles must be fluorescently labeled. Different strategies are possible, depending on the target of the fluorescent label. For example, four types of labels have been used. They target respectively: the genetic material (DNA, RNA) inside a viral capsid or through a lipid envelope, the lipid envelope itself, or are inherent to particle production, as shown in Table 5. The different labels are chosen to have excitation and emission wavelengths compatible with the experimental setup. In addition, the fluorescent labels used are passive labels that do not interact with nanopores and do not alter the translocation of viral particles and extracellular vesicles. [Table 5]

[0117] Table 5: Overview of the different fluorescent labels available for viral particles and extracellular vesicles.

[0118] Y0Y0-1 marking

[0119] YOYO-1 (Molecular Probes) is able to penetrate the viral capsid and label viral genetic material. It has been used to label non-enveloped viruses (e.g., AAV, H BV, and AdV) (L. Chazot-Franguiadakis et al., “Optical Quantification by Nanopores of Viruses, Extracellular Vesicles, and Nanoparticles,” Nano Lett., 2022). The goal is to obtain the strongest possible fluorescence emission by saturating the DNA molecules with YOYO-1. For example, the same protocol was used for all types of viral particles and consists of adding 0.7 pL of YOYO-1 (1 mM) directly to 1 pL of viral solution. The solution is left to incubate for 10 min at room temperature, after which it can be rediluted according to the concentration required for the experiment.

[0120] SYTO-13 Labeling SYTO-13 (Molecular Probes) is able to penetrate the viral envelope and label viral genetic material. It has been used to label enveloped viral particles (e.g., HIV, HSV, MLV, MeV, SC2, etc.). For example, the same protocol has been used for all types of viral particles and consists of adding 0.5 μL of SYTO-13 (5 mM) directly to 1 μL of viral solution. The solution is left to incubate for 10 min at room temperature, after which it can be rediluted to the concentration required for the experiment.

[0121] DiO marking

[0122] In cases where genetic material is inaccessible or absent, it is possible to fluorescently label the lipid envelope of particles. This is done using DiO (Vybrant Cell Labeling, Molecular Probes), which is a lipophilic carbocyanine that emits weak fluorescence in water but is highly fluorescent and photostable when incorporated into membranes or lipid environments. Labeling of extracellular vesicles and enveloped viral particles (e.g., HIV, MLV, etc.) has already been achieved with this type of fluorophore (L. Chazot-Franguiadakis et al., “Optical Quantification by Nanopores of Viruses, Extracellular Vesicles, and Nanoparticles,” Nano Lett., 2022).For example, the protocol used to label enveloped viral particles or extracellular vesicles corresponds to a classic lipid envelope labeling protocol and consists of adding 0.5 μL of DiO (0.1 mM) to a solution containing 1 μL of vesicles or virus. The entire solution is incubated for 15 min at 37°C (in a dry bath). This solution can then be rediluted according to the concentration required for the experiment.

[0123] GFP marking

[0124] The fluorophore can be incorporated directly at the time of production, as in the case of HIV and MLV, where fluorescent labeling is achieved using GFP, fused to the C-terminal position of the structural protein Gag (J. Eid, “Etude du relachage du VIH-1 en temps réel à l’échelle de la cellule unique par la Viro-fluidique,” ​​Montpellier, 2022; S. Nydegger, M. Foti, A. Derdowski, P. Spearman, and M. Thali, “HIV-1 egress is gated through late endosomal membranes,” Traffic, vol. 4, no. 12, pp. 902-910, Dec. 2003).

[0125] Fluorescent antibody (primary or secondary) labeling

[0126] A specific antibody (primary or secondary) fluorescently labeled typically with Alexa Fluor 488 ® or FITC can also be used. This type of labeling is interesting to increase the specificity of the measurement and also allows the identification of a viral type / subtype or extracellular vesicles. The incubation time varies depending on the fluorescent antibody / antigen pair considered. For example, to fluorescently label baculovirus samples, the gp64-baculovirus antibody (AcV1, Biotechne) coupled with Alexa Fluor 488 ® was incubated with the baculovirus sample for 15 min at 20°C, with a ratio of 1:5000 (virus:antibody).

[0127] Example of use of a nanoporous system coupled with optical detection according to the invention

[0128] For a typical experiment, fluorescently labeled viral particles or extracellular vesicles are placed in the cis chamber, and by applying a pressure difference between the two chambers, they are transported across the membrane into the trans chamber. Specifically, after placing a drop of water on the objective, the trans chamber (Teflon ring + glass surface) is placed on the stage in contact with the objective and filled with 500 pL of Tris-EDTA buffer. The internal cavity of the cap (cis chamber) is filled with the solution containing the biological sample of interest. The minimum capacity of the chamber is approximately 1 pL while the maximum capacity of the chamber is approximately 1000 pL. The exact volumes and concentrations used vary depending on the experiment. The cap (cis chamber) is screwed onto the support connected to the pressure controller and rests on the trans chamber.The membrane (trans side) is illuminated by a laser beam, and the particles are observed at the exit of the pores. Detection is based on the Zero-Mode Waveguide effect which consists of an amplification of the fluorescence signal in the vicinity of the gold layer.

[0129] Data acquisition

[0130] Data acquisition and parameter selection are performed using the Andor Solis software, which is connected to the camera. As mentioned earlier, the camera has a built-in cooling system, allowing all measurements to be taken at -60°C, thus reducing thermal noise. In addition, it is possible to set the gain value corresponding to the system's signal amplification ratio, which is set to 300.

[0131] An event corresponds to the complete translocation of a viral particle or an extracellular vesicle through a nanopore. In the experiments, the aim is to evaluate the evolution of the translocation frequency as a function of the pressure difference applied to the system, or the concentration of objects in the system. The experiments correspond to single particle measurements, based on the detection of isolated events and their quantification during the acquisition period. Image sequences are acquired with the camera, and the resolution, the number of images and the time interval between each image are defined. Typically, the mode used is a real-time display mode, with an image resolution of 512 x 512 pixels by 125 x 125 pm 2, and an acquisition frequency of 33 Hz. The sequences are typically 250 images, corresponding to a 7.6 s movie. Four sequences are executed successively, in order to increase the statistics. Finally, the temporal resolution can be improved by increasing the binning and reducing the observed area. By adapting these parameters (256x256 pixels on 62.5x62.5 pm 2 with binning set to 2), it is possible to increase the acquisition frequency of the detection system up to 112 Hz. In addition, it is also possible to monitor the evolution of the intensity of each individual event as a function of time.

[0132] Translocation frequency as a function of pressure

[0133] Once the biological sample is loaded into the cis chamber, a series of films covering a range of pressures is produced, from the minimum (typically 20 mbar for a pore diameter of 50 nm and 0.1 mbar for a pore diameter of 400 nm) to the maximum (typically 80 mbar for a pore diameter of 50 nm and 2 mbar for a pore diameter of 400 nm). The values ​​of the minimum and maximum pressures, as well as the intervals between each measurement, vary according to the pore diameter considered. In addition, it is typically recommended to wait 3 min before acquiring the first point, and generally 1 min between each new point to allow the pressure to stabilize. For each point, at a given pressure, 4 successive films are produced. For the same cap (cis chamber), two sequences on two different areas are also generally acquired to increase the statistics.

[0134] Translocation frequency as a function of the concentration of viral particles or extracellular vesicles

[0135] For this type of measurement, the pressure value is fixed. Initially, the concentration in the plug (cis chamber) is minimal, then gradually increased. More precisely, for each concentration, 4 films are made on a first area, then a change of area on the membrane allows 4 more films to be made. Once a concentration has been measured, a volume of the solution of viral particles and / or extracellular vesicles can be added to the cis chamber to obtain a new concentration. After homogenization of the contents of the plug, it is allowed to equilibrate for 3 min, before starting a new acquisition. This procedure is repeated as many times as there are different concentrations.

[0136] Event intensity as a function of time

[0137] For this type of measurement, the pressure and particle concentration are fixed. Each individual event intensity in the videos can be analyzed over time. Typically, the normalized event intensity is expressed as I nor m = -^ with l(t) nax the intensity of the event in time and l ma x the maximum intensity of the event. It is then possible to fit the curve obtained by a model and extract parameters on the intensity signal. For example, an adjustment by a decreasing exponential can be used.

[0138] Image analysis

[0139] Once the measurements are made, the data is analyzed by counting the number of events film by film. For example, the simplest method for detecting and counting events in an image sequence is manual detection by eye. To facilitate event detection, several operations are performed on the sequence to be analyzed using Image J software:

[0140] - From the initial sequence, the Z-project function averages all the images over time, to obtain the average intensity per pixel. The result is an image containing an average noise value for each pixel, taking into account the static elements.

[0141] - The resulting image is then subtracted from the initial image sequence, using the Image Calculator function of Image J. This subtraction eliminates fixed parasitic elements (membrane defects, stuck objects) during acquisition, and reduces noise.

[0142] Additionally, it is also possible to use a program dedicated to automatic event detection. An automated algorithm to track and quantify translocation events was used. The program is written with Matlab and partly uses functions already available in Matlab. Briefly, this consists of using several filters to process the image (Gaussian filter, Wiener filter, ...) and then tracking the events.

[0143] Finally, from the number of events obtained during image analysis, the corresponding translocation frequency is calculated according to the following formula: f = — Nxfacq - picture x<jX With N, the number of events, f acq , the acquisition frequency (f acq = 33 Hz), or the porous density (o = 1-6.10 8 pores / cm 2 , see Table 1), S the field of observation (S = 125 x 125 pm 2 ).

[0144] Normalization allows working with the translocation frequency rather than the number of events, which allows comparing the results obtained between different pore sizes. Indeed, membranes with different pore diameters do not have the same pore density (see Table 1). Example of data analysis and results for the identification of viral particles and extracellular vesicles according to the invention

[0145] The present invention relates to the identification of viral particles and extracellular vesicles based on their specific interaction with the nanopore during transport. By measuring the translocation frequency through the nanoporous membrane as a function of a control parameter (pressure, particle concentration), as well as the intensity of the event as a function of time, it is possible to extract physicochemical parameters and / or interaction rates between the viruses and / or extracellular vesicles themselves or between the viruses and / or extracellular vesicles and the surface of the nanopore specific to a type / subtype of viral particle or extracellular vesicle, thanks to the traffic jam model. This set of parameters being specific to a given type of particle, its measurement allows the constitution of a specific signature and therefore of an identification database.

[0146] Construction of the identification database

[0147] The identification database is constituted by the collection of experimental transport measurements from reference samples. These data are then analyzed to extract, using a physical model, physicochemical parameters (and / or interaction rates between the viruses and / or extracellular vesicles themselves or between the viruses and / or extracellular vesicles and the surface of the nanopore) specific to each type or subtype of viral and extracellular vesicles. Typically, once the parameters have been measured for all the particles composing the initial database described in the present invention (see Tables 3 and 4), specific domains are associated with each viral type / subtype (or type / subtype of extracellular vesicles) within a space common to all viral types (or types of extracellular vesicles).This representation constitutes the database necessary for the identification of unknown particles, as will be described later.

[0148] Collection of baseline measurements

[0149] For a reference sample corresponding to each viral type / subtype (or each extracellular vesicle type / subtype), several experimental measurements are carried out and will be detailed below. For all these experiments, nanoporous membranes or functionalized nanoporous membranes are used.

[0150] -Translocation frequency versus pressure curves at a fixed particle concentration. At least four curves are performed, each at a fixed concentration of viral particles (or extracellular vesicles), and the pressure range varies depending on the pore diameter (see Table 6). For example, for 200 nm pores, the pressure can vary from 0 to 8 mbar in steps of 0.2 mbar, then 0.5 mbar and 1 mbar. For each curve, each pressure point is measured four times in a row. Each curve is also performed at least twice (on two different observation areas), for a total number of repetitions greater than 8 for each point.

[0151] [Table 6]

[0152] Table 6: Pressure range typically used in pressure-dependent translocation frequency experiments.

[0153] -Translocation frequency curve as a function of particle concentration at a fixed pressure.

[0154] For this type of measurement, the pressure value is set at a significant pressure related to the diameter of the nanopores (e.g. 8 mbar for 200 nm pores or 2 mbar for 400 nm pores). Initially, the concentration in the plug (cis chamber) is minimal, then gradually increased. The concentration range used varied from 10 3 particles / mL at 10 9 particles / mL depending on the type of particles. For each concentration, 4 films are made on a first area, then a change of area on the membrane allows 4 more films to be made. The total number of repetitions is therefore greater than 8 for each point.

[0155] Examples of translocation frequency versus pressure and translocation frequency versus particle concentration curves are given for viral particles and extracellular vesicles in Figure 2.

[0156] Other experimental conditions

[0157] Several experiments were performed to test changes in experimental conditions (including the type of medium or membrane). In particular, experiments were performed by changing the reference buffer medium (Tris-EDTA) used (initially at 5 mM NaCl) by directly adding salt, resulting in a final extra-saline solution at 150 mM NaCl. In this case, as before, the transport of viral particles or extracellular vesicles across nanoporous membranes can be followed. Examples of translocation frequency curves as a function of pressure are shown in Figure 3.

[0158] Furthermore, functionalized nanoporous membranes have also been used, including membranes functionalized with poly(2-methyl-2-oxazoline). Nanoporous membranes are used in the same way whether they are functionalized or not and under the same experimental conditions. For example, poly(2-methyl-2-oxazolines) with a degree of polymerization of 387 were synthesized and then grafted onto a nanoporous membrane with a pore diameter of 200 nm. More details on the synthesis, characterization and grafting of these polymers are described in patent application WO2022195059. An example of transport of viral particles through this functionalized nanoporous membrane is also given in Figure 3.

[0159] Data analysis: parameter extraction

[0160] Once the experimental data of the reference samples are generated, the idea is to extract a set of physicochemical parameters and / or interaction rates between the viruses and / or extracellular vesicles themselves or between the viruses and / or extracellular vesicles and the nanopore surface specific to each viral type / sub-type (or extracellular vesicle type / sub-type).

[0161] For example, to model the experimentally obtained curves (frequency / pressure and frequency / concentration), a physical model has been developed. It is called the "traffic jam model". It describes the accumulation of particles inside the nanopore observed experimentally. It is based on the following experimental observations:

[0162] - Existence of a critical pressure below which no viral particles or extracellular vesicles exit the pore. This critical pressure (Pc) is assumed to be correlated with the adhesion of the particle inside the pore (minimum pressure required to prevent the attachment of particles to the pore surface).

[0163] - At high pressures, we observed the presence of a plateau for the frequency (for the frequency / pressure curves), which is more pronounced as the concentrations of viral particles or extracellular vesicles are high.

[0164] -The increase in the concentration of viral particles or extracellular vesicles is accompanied by a decrease in the frequency of translocation of viral particles or extracellular vesicles and a greater saturation, which constituted a distinctive sign of a traffic jam phenomenon (also called a plug).

[0165] Overall, these observations are consistent with the presence of a traffic jam phenomenon (also called a plug) linked to the strong confinement of viral particles or extracellular vesicles under flow. The proposed physical model describes the phenomenon as aggregation under flow. Therefore, the total translocation time of a virus through a nanopore can be decomposed into two durations, (time between the entry of two consecutive viral particles or extracellular vesicles into the pore) and T2 (time between the exit of two consecutive viral particles or extracellular vesicles from the pore).

[0166] -For the part entering the pore, it was assumed that a viral particle or an extracellular vesicle can only enter the nanopore if another viral particle or an extracellular vesicle that already occupied the pore has left its place there. This depends on the probability that the pore is occupied by another viral particle or an extracellular vesicle (probability Pparticie). A second assumption is that the time to bring a viral particle or an extracellular vesicle to the entrance of the pores by advection is negligible. Then, assuming that the equilibrium between free viral particles or extracellular vesicles in solution and those associated with the traffic jam (i.e., the plug) is given by the Langmuir equilibrium, the duration can be expressed by: with k o cl ° 3, the characteristic dissociation rate at which the viral particle or extracellular vesicle detaches from the bottled part (i.e., the cap), eï K^ l ° 3 , the dissociation constant between the viral particle or extracellular vesicle and the bottled part (i.e., of the cap).

[0167] -For the part exiting the pore, the first hypothesis is that in stationary regime, inside the nanopore, the concentration of viral particles or extracellular vesicles, C n (x, t), can be described as a competition between particle adhesion to the pore surface and advection by the flow. Based on the conservation equation of the viral particle or extracellular vesicle: =

[0168] -k 3 ° re C n (x, t) = o; with v, the flow speed and k 3 ° re , the characteristic rate to which a viral particle or extracellular vesicle adheres to the central channel of the nanopore, it is possible to access the concentration of viral particles or extracellular vesicles in the nanopore: with C, the concentration of viral particles or extracellular vesicles in the cis chamber (at x=0).

[0169] Then, using the Hagen-Poisseuille law in a cylindrical geometry (like that of the pore), it is possible to introduce the experimental parameter of the pressure difference between the two chambers P = where R and L are respectively the radius and length of the pore, and q, the viscosity of water.

[0170] Therefore, the time between the exit of two consecutive viral particles or extracellular vesicles (T2) can be expressed as:

[0171] 8pk P ° re L 2 by introducing a prefactor k = k 3 ° re CLR2 and a critical pressure P c = °” — .

[0172] Finally, the translocation frequency of a viral particle or an extracellular vesicle through a nanopore can be written:

[0173] All experimental data (frequency / pressure and frequency / concentration curves) for different viral particles (or extracellular vesicles) were successfully fitted by this traffic jam (i.e., plug) model, as can be seen by the solid lines shown in the examples in Figure 2 and Figure 3. Specifically, to fit the data with the traffic jam (i.e., plug) model, Matlab was used. Analysis scripts were developed and built-in tools such as Curve Fitting Tool were also used.

[0174] Typically, three physicochemical and / or interaction rate parameters between the viruses and / or extracellular vesicles themselves or between the viruses and / or extracellular vesicles and the nanopore surface specific to each type / subtype of viral particle or extracellular vesicle are extracted (P c , related to the interaction of viral particles or extracellular vesicles with the pore surface and k r O c : f 1 ?f , K^ U l ° 3 related to the interaction of viral particles or extracellular vesicles with each other in the bottled part (i.e., the cap). The procedure is as follows:

[0175] -The frequency / pressure curves are fitted at their first points (before the onset of saturation) by the following part of the model: f = k — exp (— ). This allows

[0176] Pc Pc to extract the critical pressure parameter, P c. This process is performed for all frequency / pressure curves and the P c final is obtained by averaging over all P c found by adjustment.

[0177] -The frequency / concentration curves are fitted by the entire traffic jam (i.e., plug) model (considering the term -1 exp() as a constant since the pressure is fixed). The assembly allows access to both parameters

[0178] Examples of physicochemical parameters and / or interaction rates between viruses and / or extracellular vesicles themselves or between viruses and / or extracellular vesicles and the extracted nanopore surface for different reference samples of viral particle and extracellular vesicle types are given in Table 7. Similar experiments were conducted on subtypes.

[0179] [Table 7]

[0180] Table 7: Example of transport (or physicochemical) parameters extracted from the traffic jam model (i.e., plug) for different viral types and extracellular vesicle types. For viral types, the results were averaged over the different viral subtypes composing a single viral type (which were also averaged over the different experiments). Error bars were calculated using error propagation and the jackknifing process. This step is essential for the construction of the database since the physicochemical parameters and / or the interaction rates between the viruses and / or extracellular vesicles themselves or between the viruses and / or extracellular vesicles and the nanopore surface collected are specific to each type / subtype of viral particle or extracellular vesicle and therefore allow the construction of a unique type / subtype signature.

[0181] Additionally, it is also possible to fit the data with other models such as the following Gaussian model:

[0182] The 4 parameters collected by the fit (a, b, c and d) then constitute the transport parameters associated with each viral type / subtype (or extracellular vesicle type / subtype).

[0183] Database representation

[0184] From the parameters extracted above, a representation of the database is generated in nD plot. For example, we chose to create a 3D plot with the transport (or physicochemical) parameters extracted from the traffic jam model (i.e., traffic jam). Each axis corresponds to a parameter (P c , k o c f l f , K^ l ° 3). The objective is then to establish a specific domain for each viral type / subtype (or extracellular vesicle type / subtype). For this purpose, a Voronoi algorithm is generated. This algorithm was adapted from an existing Matlab algorithm. The principle of the Voronoi diagram is the division of a cube into domains (adjacent regions) from a set of points called "sites". Each domain contains a site and forms the set of points in the plane closer to this site than to any other. The boundaries between domains are constructed to be perfectly equidistant from the nearest sites.

[0185] A cube (comprising the entire database of viral particles and / or extracellular vesicles) is thus obtained and divided into several domains. Each domain is associated with a subtype or type of viral particle (or extracellular vesicle). For greater readability, the axes have been normalized by the maximum and minimum values ​​of the physicochemical parameters and / or interaction rates between the viruses and / or extracellular vesicles themselves or between the viruses and / or extracellular vesicles and the nanopore surface. An example of a Voronoi diagram is shown for the different viral types in Figure 4. The values ​​of the physicochemical parameters correspond to the average parameters that were measured for the different subtypes associated with the same viral type. The red dots correspond to the site of each viral domain, i.e. the average parameters that are associated with a specific viral type.A Voronoi diagram containing all virus subtypes was also produced.

[0186] Overall, this representation constitutes the identification database necessary for the identification of unknown particles, as will be described later.

[0187] Additionally, other types of representation than the Voronoi diagram can be used, such as the k-nearest neighbors algorithm or the support vector machine (SVM) as a supervised learning model.

[0188] Testing the identification process

[0189] 1- “Unknown sample”

[0190] Once the identification database is established, the next step is to create a protocol for identifying an “unknown sample,” i.e., a sample potentially containing viral particles and / or extracellular vesicles of unknown type and / or subtype. The first step of this protocol is to measure the physicochemical parameters (and / or interaction rates between the viruses and / or extracellular vesicles themselves or between the viruses and / or extracellular vesicles and the nanopore surface) associated with the sample in the same way as when creating the identification database (see section Collection of reference measurements). However, in the case of an unknown sample, only one frequency / pressure curve and one frequency / concentration curve are performed in order to limit the experimental and processing time.The next step is to extract the physicochemical and / or interaction rate parameters between the viruses and / or extracellular vesicles themselves or between the viruses and / or extracellular vesicles and the nanopore surface (by fitting with the traffic jam model) and then to represent these new parameters in the identification database (e.g. in the Voronoi diagram). An example of an "unknown viral particle sample" is represented by a star in Figure 4. A similar measurement can be performed for extracellular vesicles.

[0191] To determine in which domain the "unknown sample" is located, and how close it is to the various neighboring domains (which is likely to give an indication of the measurement error), the distances between the "unknown sample" and all the sites in the domain are measured. The smallest distance is associated with the domain in which the sample is located and is therefore identified as the type / subtype associated with the domain.

[0192] 2- Measurement of the robustness of the identification Thanks to the tests carried out with the “unknown samples”, it is possible to provide criteria for evaluating the robustness of the method described in the present invention. For example, four “samples of unknown viral particles” were tested for each viral subtype. The result of the identification is given in Table 8.

[0193] [Table 8]

[0194] Table 8: Confusion matrix for “unknown viral samples” to evaluate the identification process. The data are generally distributed according to a classical process (3 / 4 for the construction of the identification database and 1 / 4 for testing). The input (“reference truth” or “ground truth”) corresponds to the actual subtype of the viral particle while the output (“experiment”) corresponds to the result of the identification process developed in this invention. The results are given in percentages. Four “unknown viral particle samples” were tested for each viral subtype.

[0195] From these results, it is then possible to calculate the identification specificity, identification sensitivity, and precision of the method (see Table 9). Identification specificity is the probability of obtaining a negative test if the viral particle sought is not present (in other words, it is the probability of obtaining a negative test in people who are not ill). Identification sensitivity is the probability that the test correctly assigns the "unknown" viral particle to the correct domain (in other words, it is the probability that the test is positive if the disease is present). Precision is the probability corresponding to the number of samples that have been correctly assigned to the correct subtype among all the samples that have been assigned to this subtype. [Table 9]

[0196] Table 9: Confusion matrix of possible outcomes when measuring the intrinsic validity of a test.

[0197] Based on Table 9, the identification specificity, identification sensitivity and precision of a method can be expressed as follows:

[0198] TN TP TP

[0199] Specificity = - ; Sensitivity = - ; Accuracy = -

[0200] 1 TN + FP TP + FN TP + FP

[0201] For the above-mentioned example (unknown sample test for each viral subtype) and with the method described in the present invention, the identification specificity was found to be 98%, the identification sensitivity to be 95% and the accuracy to be 98%.

[0202] In addition, the harmonic mean of identification accuracy and sensitivity gives a score called F1-score which is a measure of the performance of the model's classification ability. The F1-score is expressed as follows:

[0203] Precision x Sensitivity Fl — score = 2 -

[0204] Precision + Sensitivity

[0205] For the above-mentioned example (unknown sample test for each viral subtype) and with the method described in the present invention, the F1-score reaches 96%.

[0206] 3- Sample comprising a mixture of several types of viral particles and / or extracellular vesicles

[0207] Complex samples corresponding to mixtures of different types / subtypes of viral particles and / or extracellular vesicles present at different concentrations were tested.

[0208] For example, a sample containing viral particles (HIV Gag-GFP) and extracellular vesicles (EVs from HeLa cells used for HIV Gag-GFP production, fluorescently labeled using DiO) was prepared according to the protocol described above. The concentrations of viral particles (1.10 11 particles / mL) and extracellular vesicles (2.10 10particles / mL) were first quantified. Then, the measurement of the passage frequency as a function of pressure and viral particle concentration made it possible to extract the transport parameters (derived from the adjustment by the traffic jam model (i.e., plug)) specific to the viral particle. Finally, the viral type was identified using the database. The identification result corresponds well to the majority viral type (HIV Gag-GFP) and shows the ability to identify a viral type in a mixture including extracellular vesicles.

[0209] Furthermore, a sample containing two virus types, AAV-8 (fluorescently labeled using YOYO-1) and human InfV ​​A(H1N1) (fluorescently labeled using DiO), with a concentration ratio of 1:100 was also tested. The method developed above (transport of viral particles through nanopores / measurement of transport parameters / use of the identification database) was used to identify the predominant viral type. The identification result corresponds well to the predominant viral type human InfV ​​A(H1N1) and shows the ability to identify a viral type in a mixture including other viral types.

[0210] It should be noted that the concentration ratios discussed above are relevant in the case of viral particles and / or extracellular vesicles in patient samples. Indeed, in the case of co-infection, there is a significant concentration difference between the different virus types (usually a factor greater than 1:1000). This is also valid in the case of bioproduction of viral particles and / or extracellular vesicles, where the presence of unwanted particles is low compared to the amount of target particles produced.

[0211] 4-Patient samples

[0212] Patient samples corresponding to respiratory swabs were tested. Specifically, a patient respiratory swab stored in universal transport medium (Becton Dickinson), corresponding to a negative matrix (absence of viral or bacterial infection) was used to set up the sample pretreatment process. Indeed, highly purified samples are not required, as the nanoporous membrane itself functions as a filter, selectively allowing the passage of particles smaller than the pore size (200 or 400 nm in diameter). The method also requires very low concentrations, so the sample can be diluted at least 10 times, which avoids some of the problems associated with overly concentrated media.Furthermore, our method requires fluorescent labeling, which allows the exclusive observation of virus and / or extracellular vesicle translocation with an adjustable degree of specificity (standard labeling of the viral lipid bilayer or viral genetic material but also specific labeling with antibodies). Nevertheless, the transition to complex samples such as patient samples (e.g., respiratory swab) represents a technological advance and additional pretreatment may be necessary in this case. The objective of the pretreatment steps is therefore to make the sample suitable for analysis by removing various components that would completely clog the nanopores, such as bacteria, cells, and cellular debris.

[0213] We performed a simple pre-filtration step through a 0.45 pm filter (Agilent), and possibly an additional centrifugation step (affinity column, or simple centrifugation). We then added fluorescent beads (260 nm diameter polystyrene beads functionalized with carboxyl groups, Spherotech Inc) to the filtered test sample (negative matrix of a patient respiratory swab) and we sought to measure the transport through the pores of these beads in such a medium. For the same bead concentration, we measured the passage frequency as a function of the applied pressure under three conditions: beads in PBS buffer; beads in universal transport medium; and beads in the patient respiratory swab (negative matrix). Conclusively, the curves obtained under the three conditions overlap.This highlights the effectiveness of the pre-treatment protocol, allowing manipulations to be carried out under the same conditions as previously while avoiding the complexity of the environment.

Claims

CLAIMS 1. A method for identifying types and / or subtypes of viral particles and / or extracellular vesicles comprising: a. labeling viral particles and / or extracellular vesicles present in a biological sample using a fluorescent marker; b. passing said sample comprising the labeled viral particles and / or extracellular vesicles through a nanoporous membrane; c. measuring the transport of said viral particles and / or extracellular vesicles through said nanoporous membranes by optical detection as a function of at least one control parameter; and d.the extraction of physicochemical and / or dynamic parameters and / or interaction rates between the viral particles and / or extracellular vesicles themselves or between the viral particles and / or extracellular vesicles and the surface of the nanopore, characteristics of the transport of said viral particles and / or extracellular vesicles by a physical model linking the transport to the control parameter.

2. Method according to claim 1, wherein the parameters extracted in step d. allow the identification of the type and / or subtype of viral particles and / or extracellular vesicles.

3. Method according to claim 1 or 2, wherein step d. is the extraction of physicochemical and / or dynamic parameters and / or interaction rates between the viral particles and / or extracellular vesicles themselves or between the viral particles and / or extracellular vesicles and the surface of the nanopore from the modeling of the measurement of the transport of said viral particles and / or extracellular vesicles by a physical traffic jam model linking the transport to the control parameter.

4. Method according to any one of the preceding claims, in which the physicochemical parameters and / or interaction rate between the viruses and / or extracellular vesicles themselves or between the viruses and / or extracellular vesicles and the surface of the nanopore of said viral particles and / or extracellular vesicles are compared to a database comprising the specific signature of each type or subtype. viral and / or each type or subtype of extracellular vesicle in order to identify said viral particles and / or extracellular vesicles present in the sample.

5. Method according to any one of the preceding claims, said method comprising a step e. of identifying said viral particles and / or extracellular vesicles present in the sample, by comparing said physicochemical parameters and / or said interaction rates extracted in step d., to a database comprising the specific signature of each viral type or subtype and / or each type or subtype of extracellular vesicle.

6. Method according to any one of the preceding claims, wherein said at least one control parameter is chosen from the concentration of viral particles and / or extracellular vesicles, pressure, duration of transport events, intensity of transport events, trajectory of transport events, temperature, salinity and pH.

7. Method according to any one of the preceding claims, wherein the specific signature of each viral type and / or each type of extracellular vesicle allowing the identification of the labeled viral particles and / or extracellular vesicles is obtained by carrying out steps a) to d) with samples comprising known viral particles and / or extracellular vesicles.

8. Method according to any one of the preceding claims, in which the nanoporous membrane is functionalized, in particular by synthetic polymers, for example by poly-2-alkyl-2-oxazolines or by non-covalent grafting, in particular with Pluronic ©, poly(vinylpyrrolidone), methylcellulose, albumin, or fetal bovine serum.

9. Method according to claim 8, in which the polymers are linked to electroactive probes, in particular of the diazonium type.

10. Method according to any one of the preceding claims, wherein in step b), the passage of said sample through a nanoporous membrane is done by hydrodynamic entrainment.

11. Method according to any one of the preceding claims, in which the viral particles are chosen from enveloped or non-enveloped viral particles, DNA or RNA.

12. Method according to any one of claims 1 to 10, in which the extracellular vesicles are chosen from vesicles produced by healthy or cancerous cells, or from vesicles resulting from non-cellular bioproduction.

13. Use of the method according to any one of claims 1 to 9 for identifying the viral type and / or subtype present in a patient, the quality of the production of viral vectors or the identification of human, plant or animal pathologies in biological samples in the context of health monitoring, or for the identification of viral particles in agri-food samples.

14. Use of the method according to any one of claims 1 to 10 and 12, for identifying the type and / or subtype of extracellular vesicles in a patient or the quality of the production of vesicular vectors.

Images