Abstract
The invention relates to a method for determining the physicochemical properties of a nanoscale system (NSS) using analytical ultracentrifugation), comprising the steps of generating a multi-dimensional sedimentation analysis map associated with the NSS of interest; selecting sample-dependent parameters; determining sedimentation coefficient value/parameter in the sample; inserting sample sedimentation coefficient values onto the multi-dimensional sedimentation analysis map to obtain a NSS sample map value; and inferring from the NSS sample map value the physicochemical properties of the NSS sample. Furthermore, the invention relates to a system for performing the method, a computer program product and a computer readable storage medium.
Claims
1. A method for determining the physicochemical properties of a nanoscale system (NSS) sample using analytical ultracentrifugation (AUC), comprising the steps of a) generating a multi-dimensional sedimentation analysis map associated with the NSS of interest; b) selecting sample-dependent parameters; c) determining sedimentation coefficient value/parameter in the sample; d) inserting sample sedimentation coefficient values onto the multi-dimensional sedimentation analysis map to obtain a NSS sample map value; e) inferring from the NSS sample map value the physicochemical properties of the NSS sample.
2. The method according to claim 1, wherein the nanoscale system (NSS) is a nanoscale drug delivery system (NDDS).
3. The method according to claim 1, wherein the step of generating the multi-dimensional sedimentation analysis map comprises the steps of a) selecting parameters representative of the NSS of interest; and b) generating the multi-dimensional sedimentation analysis map utilizing AUC sedimentation coefficient values/parameter/concentration series of the NSS and any individual NSS component.
4. The method according to claim 1, wherein the parameter is selected from the group comprising density, absorption/emission, refraction, degradation kinetics.
5. The method according to claim 4, wherein the parameter degradation kinetics comprises physical parameter dependent degradation and/or chemical parameter dependent degradation and or biological parameter dependent degradation.
6. The method according to claim 3, wherein the step of generating the multi-dimensional sedimentation analysis map additionally comprises the step of determining a viscosity of a solvent used for suspending the NSS of interest.
7. The method according to claim 6, wherein at least three different sample-dependent parameters are selected in the method for determining the physicochemical property of an NSS sample using AUC, wherein at least three sample-dependent parameters comprise the parameters density, viscosity, and optical detection.
8. The method according to claim 1, wherein the physicochemical property is selected from the group consisting of NSS size, NSS molecular weight/molar mass, NSS concentration, NSS dispersity, NSS aggregate, NSS integrity, and NSS stability.
9. The method according to claim 8, wherein the physicochemical property NSS integrity comprises the integrity of the NSS in a solvent selected from water, a salt solution, or a biologically relevant fluid.
10. The method according to claim 1, wherein the NSS comprises at least two different components.
11. The method according claim 10, wherein at least one independent AUC measurement/sample has to be performed addressing the NSS and, preferably, one independent measurement/sample has to be performed addressing a solvent.
12. The method according to claim 11, wherein the number of measurements to be performed per sample is two plus at least one measurement/component changed relative to the NSS of interest.
13. The method according to claim 10, wherein the NSS comprises a solvent-derived physical and/or chemical modification.
14. The method according to claim 10, wherein the NSS is a complex polymeric system, wherein preferably the complex polymeric system comprises a biodegradable polymer and an encapsulated drug.
15. The method according to claim 14, wherein the complex polymeric system comprises a targeting dye moiety and/or a stabilizer, preferably surfactant.
16. The method according to claim 1 wherein step d) is implemented by a computer, comprising means adapted for receiving the data from step c), means for mapping the one or more sedimentation coefficient values onto the multidimensional sedimentation analysis map, preferably by multiple regression analysis.
17. The method according to claim 16, wherein step e) is implemented by the computer, comprising calculation means for inferring from the NSS sample map value the physicochemical properties of the NSS.
18. A system for performing the method according to claim 1, comprising multi-dimensional sedimentation analysis map generating means for generating a multi-dimensional analysis map associated with the NSS of interest; means for selecting sample-dependent AUC parameters; determination means for determining sedimentation coefficient value/parameter in the sample; means for inserting sample sedimentation coefficient values onto the sedimentation map to obtain an NSS sample map value; calculation means for inferring from the NSS sample map value the physicochemical properties of the NSS sample.
19. A computer program product comprising instructions which, when the program is executed by a computer, cause the computer to carry out the steps of the method of claim 1.
20. A computer readable storage medium comprising instructions which, when executed by a computer, cause the computer to carry out the steps of the method of claims 1-15.
Description
BRIEF DESCRIPTION OF THE FIGURES
[0036] FIG. 1 depicts a flowchart of the inventive method. Herein, a schematic protocol is shown for characterizing an unknown, complex NSS with respect to different parameters and physicochemical properties.
[0037] In FIG. 2 (a)-(c), schematic protocols are depicted, exemplifying the steps required (links) for characterization of unknown NSS and partially known NSS samples. FIG. 2 (a) depicts a schematic protocol for characterizing a NDDS sample containing a known NSS, which is, however, only partly identical to an established (“characterized”) NSS. FIG. 2 (b) depicts a schematic protocol for characterizing a new NSS sample corresponding to an established NSS but differing with respect to the solvent, whereas in FIG. 2 (c), a schematic protocol for a new NDDS sample is shown, wherein the drug and targeting moiety are identical to known components, whereas the matrix of the NDDS as well as additives and stabilizing components are unknown.
[0038] FIG. 3 depicts sedimentation velocity profiles of a NDDS sample after two weeks of storage under ambient conditions (3(a)) and after 10 weeks of storage at ambient conditions (3(b)), wherein sedimentation is tracked by absorbance detection at A=462 nm. In the upper panels of both, FIG. 3 (a) and FIG. 3 (b), the sedimentation velocity profile is depicted right after accelerating the AUC to 3,000 rpm (black line) and after 52 minutes (black dotted line) before accelerating to 10,000 rpm. The remaining sedimentation velocity profiles were recorded (gray squares) at 10,000 rpm. In the bottom panels of both, FIG. 3 (a) and FIG. 3 (b), the corresponding residuals plot from the Is-g*(s) model (solid lines) is shown.
[0039] FIG. 4 depicts normalized UV-spectra of targeting dye and drug in DMSO solutions (solid lines) and diluted in water (dotted lines). The maximum for the dye shifts from λ=660 nm in DMSO down toward λ=635 in 95% water/5% DMSO, and the maximum for the drug shifts from λ=475 nm in DMSO down toward λ=462 nm in 95% water/5% DMSO.
[0040] FIG. 5 depicts the differential distributions of sedimentation coefficients Is-g*(s) of NDDS.sub.1 samples which were stored at ambient conditions from 0 to 12 weeks with 2 week intervals at a mass concentration of particles of c=6.31 mg ml.sup.−1. For optical detection, an interference optical detection system (RI) (5(a)) was used, and absorbance was tracked at λ=462 nm (5(b)), and λ=660 nm (5(c)). In experiments, the stored stock sample was diluted to a concentration of NDDS.sub.1 of c≈1.88 mg ml.sup.−1. Solid lines in (a)-(c) refer to the first measurement at 10,000 rpm, dotted lines refer to the measurements performed 10,000 rpm after shaking of the cells after the first experimental run with the aim to re-establish concentration equilibrium in each cell. (d) Plot of the intensities of RI and absorbance at λ=462 nm in particles (filled circles) and SN equal to free drug (empty circles) as well as λ=660 nm (filled triangles) for the NDDS.sub.1 sample prepared at the day of measurement against prepared NDDS.sub.1 solution concentration.
[0041] FIG. 6 depicts the differential distributions of sedimentation coefficients Is-g*(s) of NDDS.sub.1 sample in water, in culture medium (CM) and culture medium with added serum (CMS) at a similar NDDS.sub.1 sample concentration of c≈1.88 mg ml.sup.−1. The measurements were performed at 10,000 rpm.
[0042] FIG. 7 depicts a plot of UV signal intensity at λ=462 nm of the different components in solution over NDDS.sub.1 storage time including large undefined aggregates, drug in the persistent major nanoparticle (NP) population, and drug in supernatant (SN) solution (7 (a)) with collapse of the system after 10 weeks of storage (indicated by the dotted line). (b) Plot of absorbance at λ=462 nm corresponding to drug dissolved in pure water (gray squares, open: first measurement, filled: repetition) and in water with surfactant PVA (gray circles, open: first measurement, filled: repetition) used in NDDS.sub.1 formulation (circles). Shown also is the detector response for a completely degraded solution stored at c=1.26 mg ml.sup.−1 for 7 weeks at a temperature of T=37° C. and diluted for the shown concentration range (black filled circles), while the black filled squares correspond to the encapsulated drug in NP.
[0043] FIG. 8 depicts the spiking of drug to an already degraded solution of NDDS.sub.1 (filled circles) and the check of recovery at the respective wavelength used for detection at λ=462 nm (empty circles).
[0044] FIG. 9 depicts the differential distributions of sedimentation coefficients Is-g*(s) of NDDS.sub.2 (c=0.83 mg m.sup.−1) using the interference optical detection system (RI) as well as tracking absorbance at λ=462 nm and λ=660 nm after one night of storage at T=4° C. and T=37° C. (9(a), stored at c=4.29 mg ml.sup.−1). (b) shows the percentage values of drug located in nanoparticles (NP) and those in the supernatant (SN).
[0045] FIG. 10 depicts differential distributions of sedimentation coefficients Is-g*(s) of NDDS.sub.2 (stored at c=4.29 mg ml.sup.−1, measured at c=0.83 mg ml.sup.−1) using the interference optical detection system (RI) as well as tracking UV absorbance at λ=462 nm and λ=660 nm at different timescales after NDDS formulation and different storage conditions. For these experiments the AUC was spun at 10,000 rpm. (10 (a)) Storage at 4° C. for 1 and 15 days, (10(b)) Storage at 37° C. for 1 and 15 days. The bottom graphs show all detector intensities recorded during the experiments, particularly also the existence of PVA in the supernatant (SN), sedimenting at 42,000 rpm.
[0046] FIG. 11 depicts RI signal intensities of PVA utilized for NDDS formulation. For these experiments sedimentation velocity experiments were performed at a rotor speed of 42,000 rpm for 24 hours and at a temperature of T=20° C. The Is-g*(s) model has been utilized to confirm linearity of the RI detector by plotting the interference fringes against solution concentration.
[0047] FIG. 12 depicts normalized differential distributions of sedimentation coefficients Is-g*(s) of NDDS.sub.2 (stored at c=4.29 mg ml.sup.−1, measured at c≈0.83 mg ml.sup.−1) using the interference optical detection system (RI) as well as tracking UV absorbance at λ=462 nm and λ=660 nm at different timescales after NDDS formulation and stored at T=4° C. (12(a)). (b) Signal intensities after the different timescales of storage for all detectors, (c) varying NDDS.sub.2 populations observed while sedimentation takes places in human serum (40% water/60% serum, w/w) at a wavelength of λ=660 nm, and (d) area under the differential distribution of sedimentation coefficients Is-g*(s) at λ=660 nm against NDDS.sub.2 solution concentration.
[0048] FIG. 13 depicts differential distributions of intrinsic sedimentation coefficients Is-g*([s]) of NDDS.sub.2 (stored at c=4.29 mg ml.sup.−1, measured at c≈0.83 mg ml.sup.−1) in water (solid line) and human serum (dotted line) by taking into account the density and viscosity of solvent as well as the partial specific volume of the particles.
[0049] FIG. 14 depicts differential distributions of intrinsic sedimentation coefficients Is-g*([s]) of two different batches of NDDS: SDL_NP 4.1 (dashed line) and SDL_NP 4.2 (black line) measured in water. These particles differed with respect to the concentration of the targeting dye (in example DY635), which in SDL_NP 4.1 was 33.7 μg/mL and in SDL_NP 4.2 amounted to 2.6 μg/mL. Both batches were measured at c=1 mg/mL, respectively, and the measurements were performed at T=20° C., n=7500 rpm for 20 hours.
[0050] FIG. 15 depicts differential distributions of sedimentation coefficients of SDL_NP4.2 at c=0.498 mg/mL (gray dashed line), c=0.746 mg/mL (gray line), c=1.068 mg/mL (black dashed line), c=1.531 mg/mL (black line) in 55 w % human serum, with absorbance detection at λ=369 nm. The prominent component at a sedimentation coefficient of about 27 S, marked with a bold gray dashed line, refers to 55 w % human serum whose differential sedimentation coefficient distribution is illustrated by the black dotted line.
[0051] FIG. 16A depicts differential distributions of sedimentation coefficients of SDL_NP4.2 at c=0.498 mg/mL (black line), c=0.746 mg/mL (black dashed line), c=1.068 mg/mL (gray line), c=1.531 mg/mL (gray dashed line) in 55 w % human serum, with absorbance detection at λ=635 nm (GC-33). The prominent component of the distribution at a sedimentation coefficient of about 27 S refers to (55 w %) human serum, as shown in FIG. 14, marked with a bold gray dashed line. For comparison, c=0.994 mg/mL of SDL_NP4.2 in pure water is shown as black dotted line. In FIG. 16B, the corresponding linear correlation of optical densities measured at λ=635 nm plotted against the different concentrations (in mg/mL) is shown.
DETAILED DESCRIPTION
[0052] FIG. 1 shows a schematic protocol for characterizing an unknown, complex NSS with respect to different parameters and physicochemical properties. This schematic protocol makes reference to the initial step of the method of the invention, i.e. generating the multi-dimensional sedimentation analysis map. Thus, term “unknown NSS” refers to a NSS comprising a matrix, such as PLGA, in addition to known components at known concentrations, wherein the physicochemical properties of the complete system and the respective components are yet uncharacterized utilizing AUC methods. In the example, the NSS of interest is a complex structure comprising the following components: matrix (which is not studied individually), a drug, a targeting moiety, a stabilizer (surfactant) and optionally one or more additives. Additionally, the NSS is suspended/dissolved in a solvent. Concentration series of the complete NSS and of each component of the NSS are subjected to AUC under different NSS conditions (complex parameter: degradation kinetics, which depends, e.g. on temperature, pH, etc.) using different parameters of measurement, such as optical detection (depicted here: absorption, refractive index) and rotor velocity (rpm). As explained above, density is a complex parameter (like the parameter “degradation kinetics”), which is inversely related to the partial specific volume of a NSS and may be, e.g., derived from AUC sedimentation equilibrium techniques by subjecting the NSS to AUC in a solvent gradient until equilibrium with the solvent is attained. Alternatively, the density of a particle may be determined utilizing the sedimentation coefficients of the particle in two solvents of (known) different density and viscosity, such as e.g. mixtures of D.sub.2O/H.sub.2O, and calculating the partial specific volume respectively, by its inverse, the density. The viscosity of the solvent and of the suspended NSS may be determined from measurements; alternatively, it may be derived from reference tables or the like. For an uncharacterized NSS (“unknown system”), the complete NSS and its main components have to be subjected to AUC analysis in order to generate a multi-dimensional sedimentation analysis map. For complete characterization of the exemplary NSS, all “links”, which are depicted in FIG. 1 as solid, dashed and dotted lines (19 links), have to be resolved, wherein an experimental resolution by AUC is necessarily required with respect to parameters such as degradation kinetics, density, and optical detection. From the dilution series of AUC measurements of the complete NSS and its components (e.g. sedimentation coefficient values/parameters/concentration series of NSS) a multi-dimensional map is generated. From the measurement results, the physicochemical properties of the NSS may be derived, namely properties relating to size, mass and concentration, to components, to integrity and stability as well as to aggregation of NSS. Once derived, the NSS of interest is a characterized NSS (“established and validated”). For resolving the links between the NSS and its components on one side and the parameters on the other side, only a limited number of AUC measurements are required, i.e. it is possible to resolve multiple links in one AUC experiment (“run”). For example, degradation kinetics may be assessed by measuring samples simultaneously under different optical detection conditions (absorbance, refraction, emission).
[0053] In FIG. 2 (a)-(c) schematic protocols are depicted exemplifying the steps required for characterization of unknown NSS and partially known NSS samples. In case of a fully unknown NSS, the schematic protocol of FIG. 1 has to be performed.
[0054] FIG. 2 (a) depicts a schematic protocol for characterizing a NSS sample, in particular a NDDS sample, wherein “NDDS” refers to a nanoscale drug delivery system, which is a NSS additionally comprising a drug. The depicted NDDS sample is, however, only partly identical to an established NDDS. In particular, the NDDS comprises the same main constituents as described in the above example (see FIG. 2 (a)) but differs with respect to the drug and the targeting moiety. For determining the physicochemical properties of this NDDS sample, i.e. for characterizing or qualifying the particular NDDS, only a limited number of links has to be resolved (7 links, shown as solid and dotted lines in bold). As depicted, the links between the new NDDS and the components drug and targeting moiety on the one side and the parameters density, viscosity, and optical detection (depicted here: absorption and refraction) have to be resolved, resulting in a very significant reduction of the required AUC measurements. By mapping the sedimentation coefficient values/parameter/concentration series of the complete NDDS and its possible components (e.g. drug, targeting moiety) onto the multi-dimensional sedimentation analysis map, the physicochemical properties of the NDDS contained in the sample may be determined with high precision and in a timely manner.
[0055] In FIG. 2 (b), the same schematic protocol for characterizing a NSS (no drug load) sample based on gold-nanorods is shown. The sample containing the gold-nanorods differs from an established gold-nanorod NSS with respect to the suspension medium. For example, a patient derived NSS sample may be characterized, requiring the characterization of patient-specific medium (blood or a dilution thereof) in order to derive physicochemical properties from the multi-dimensional sedimentation analysis map, which—for reasons of general applicability —does not comprise patient specific data. For characterizing the sample, only 6 links connecting the NSS and its components to the selectable parameters have to be resolved, wherein said links relate to the matrix (gold-nanorods) and the solvent. Hence, in order to characterize the NSS sample of interest, the number of individual AUC measurements required for determining the physicochemical properties of the NSS sample is limited significantly.
[0056] Correspondingly, the schematic protocol in FIG. 2 (c) exemplifies a liposome based NDDS sample (unknown matrix), carrying a known targeting moiety and harboring a known encapsulated drug (=NDDS), but which differs with respect to the stabilizing component (surfactant) and also with regard to additive components. For characterizing the sample, only 8 links have to be resolved connecting the experiment parameters with the new system and those components, which differ from the established and validated NDDS (upon which the multi-dimensional sedimentation analysis map is based). Yet again, in order to characterize the sample containing the NDDS of interest, the number of individual AUC measurements required for determining the physicochemical properties of the NDDS sample is limited significantly.
[0057] Hence, the method of the present invention allows for a precise and rapid characterization of NSS samples once the underlying NSS system and its components have been established and validated. The system also may be used to study conveniently modifications of existing established and validated NSS systems, wherein modifications only relate to components of NSS. In that way, the method of the invention may also be used in the design and testing of new NSS species.
EXAMPLES
[0058] Instrumentation
[0059] Sedimentation velocity experiments were performed with a Proteomelab XL-I analytical ultracentrifuge (Beckman Coulter Instruments, Brea, Calif.), using double-sector epon centerpieces with a 12 mm optical path length. The cells were placed in an An-50 Ti eight-hole rotor. Typically, sedimentation velocity profile scans were performed right after accelerating the AUC to 3,000 rpm and after 52 minutes. Subsequently, a rotor speed of 10,000 rpm was used for NSS sedimentation velocity experiments. In instances, a rotor speed of 42,000 rpm was used to resolve further details of remaining supernatant (SN), particularly PVA, that was as well investigated via sedimentation velocity experiments at 42,000 rpm. In fact, the sedimentation protocol can simply be adapted to the individual NSS by choosing a speed between 1,000 to 60,000 rpm or more. The cells were filled with ca. 440 μl sample solution in water and with ca. 420 Il H.sub.2O or D.sub.2O/H.sub.2O as the reference. The experiments were conducted for 10 h at a rotor speed of 10,000 rpm after which the cells were taken out of the centrifuge rotor, shaken and subjected to the same procedure. After this second round of centrifugation, in some experiments the centrifuge was accelerated to 42,000 rpm. All sedimentation experiments were performed at a temperature of T=20° C. Sedimentation profile scans were recorded with interference optics (refractive index (RI)) and absorbance detection systems (at λ=462 nm and λ=660 nm) at 12 min time intervals.
[0060] The density of solvent, solvent mixtures, culture medium with and without human serum as well as human serum used for sedimentation velocity experiments were determined with a DMA4100 density meter (Anton Paar, Graz, Austria) at T=20° C.
[0061] Viscosities of solvent, solvent mixtures, culture medium without (CM) and with human serum (CMS), and human serum (HS) in accordance with its concentration with particle dilutions were determined with an Automated Micro Viscometer (AMVn, Anton Paar, Graz, Austria) at T=20° C. via a capillary/ball combination at a tilting angle of the capillary of 50°.
[0062] Data Analysis
[0063] Suitable scans were selected for data evaluation with SEDFIT using the Is-g*(s) model, i.e. by least squares boundary modelling with a Tikhonov-Phillips Regularization procedure and by assuming non-diffusing species. This model results in an apparent differential distribution of sedimentation coefficients, s. Respective integration of the distributions was used for estimating signal (weight) averages of sedimentation coefficients, s, and concentrations, c, via the integral of the differential sedimentation coefficient distributions.
[0064] The partial specific volume of NSS or NDDS, ν.sub.NSS or ν.sub.NDDS, and, respectively, their density Q.sub.NDDS or Q.sub.NSS, were determined as described by Mächtle et al. (W. Mächtle, L. Börger, “Analytical Ultracentrifugation of Polymers and Nanoparticles”, Springer, Berlin, 2006). For this purpose, densities and viscosities of water and respective mixtures of D.sub.2O/H.sub.2O in which sedimentation velocity experiments of NDDS were performed, were measured. Equating the intrinsic sedimentation coefficient [s] in the two solvents and resolving to the partial specific volume, ν.sub.NDDS, results in the following equation:
[00001]
where s.sub.1 is the sedimentation coefficient of NDDS diluted in water, η.sub.1 the viscosity of water, and ρ.sub.1 the density of water. Dilutions with D.sub.2O were aimed at achieving a similar solution concentration of the NDDS when compared to dilutions with water.
[0065] NDDS
[0066] Cell culture media, Dulbecco's MEM (DMEM, containing 1.0 g L.sup.−1 D-glucose) were obtained from Biochrom (Berlin, Germany). Human serum was purchased from Sigma Aldrich (obtained from human male AB plasma, USA origin, sterile-filtered).
[0067] Nanoscale drug delivery systems (NDDS) were provided by SmartDyeLivery GmbH (Jena, Germany). The NDDS consists of a poly (lactid-co-glycolid) (PLGA) copolymer functionalized with a targeting dye unit and with an encapsulated drug. The NDDS carrier system in solution, as provided, further contains the surfactant poly (vinyl alcohol) (PVA) used for the NDDS formulation process. NDDS and utilized surfactant PVA used for NDDS formulation were received via a materials transfer agreement with the following characteristics:
[0068] NDDS.sub.1 (Example 1 & Example 2) was composed of the biodegradable polymer poly (lactid-co-glycolid) (PLGA) partially functionalized by a targeting-dye moiety, carrying an encapsulated drug, and are as well stabilized in solution by surfactant partly incorporated in the NDDSs due to their preparation process. Total NDDS.sub.1 concentration 6.31 mg ml.sup.−1, total drug content 167 μg ml.sup.−1, total PVA concentration 1.90 mg ml.sup.−1, total volume 30 ml.
[0069] NDDS.sub.2 (Example 3 & Example 4) was composed of the biodegradable polymer poly (lactid-co-glycolid) (PLGA), partially functionalized by a targeting-dye moiety, carrying an encapsulated drug, and are as well stabilized in solution by surfactant partly incorporated in the NDDSs due to their preparation process. Total NDDS.sub.2 concentration 4.29 mg ml.sup.−1, total drug content 162 μg ml.sup.−1, total PVA concentration 1.28 mg ml.sup.−1, total volume 30 ml.
Example 1: Conceptual Suitability of AUC
[0070] FIG. 3 shows initial scans and the evolution of sedimentation profiles during sedimentation velocity experiments monitored at λ=462 nm of NDDS.sub.1, representative of the encapsulated drug (FIG. 4). Experiments were performed at 10,000 rpm for sixteen hours. NDDS.sub.1 was stored for two weeks at room temperature (FIG. 3(a)) and 10 weeks of storage under the same conditions (FIG. 3(b)), and diluted to a concentration of approximately c≈1.26 mg ml.sup.−1. The experimental profiles (gray squares) were analyzed with the Is-g*(s) approach, i.e. by least squares boundary modelling with a Tikhonov-Phillips Regularization procedure and by assuming non-diffusing species. This model results in an apparent differential distribution of sedimentation coefficients, s, representative of the population of sedimenting species (vide infra). In both cases, the numerical solution to sedimentation velocity profiles shown as solid lines (top) represent the experiment very well, as seen by the respective residual plots (bottom), showing fluctuations of a few percent only. As well clear from FIG. 3 is remaining absorbance (A) after the population of NDDS.sub.1 has migrated toward the cell bottom, larger for NDDS.sub.1 stored at room temperature for 10 weeks as compared to only two weeks under same storage conditions. As well, aggregates can be seen. These aggregates are absent for two weeks under identical storage conditions of the NDDS.sub.1 (see solid black and dotted lines in FIG. 3).
[0071] This just described set of experiments was performed via multi-detection, i.e. RI and absorbance at two wavelengths, after NDDS.sub.1 was freshly received after preparation and purification (practically no storage) and after two weeks storage at room temperature at a concentration of c=6.31 mg ml.sup.−1. The fresh and stored samples were appropriately diluted in a concentration range of c=0.63-1.88 mg ml.sup.−1 and (i) universal RI detection, (ii) absorbance detection at a wavelength of λ=462 nm, representative of the encapsulated drug (FIG. 3, FIG. 4), and (iii) at a wavelength of λ=660 nm, representative of the targeting dye functionality (FIG. 4) were carried out. At day 15, one NDDS.sub.1 sample, prepared at a concentration of c=1.26 mg ml.sup.−1, was stored at T=37° C.
[0072] AUC experiments have the advantage that the mass balance in the closed sector-shaped cell volume is preserved. FIG. 5 shows the evolution of the differential distribution of sedimentation coefficients, Is-g*(s), for the particles at c=1.88 mg ml.sup.−1 starting from the freshly prepared NDDS, sample in two weeks storage steps. RI detection (FIG. 5 (a)) as well as absorbance detection at λ=462 nm, representative of encapsulated drug (FIG. 5 (b)), and at λ=660 nm, representative of targeting dye functionality (FIG. 5 (c)) were performed. Solid lines represent experiments from the first sedimentation cycle at 10,000 rpm, after which the cells were taken out the centrifuge rotor and shaken to achieve re-suspension of sedimented material. Afterwards, another sedimentation velocity experiment was performed under exactly same conditions followed by the respective sedimentation analysis (dotted lines). All solid and dotted lines are very similar, showing repeatability of the obtained results with a solution at specific storage times. In fact, all profiles from RI and absorbance detection show a similar appearance in terms of average distribution of sedimentation coefficients, Is-g*(s), as well as their trend toward storage time. The experiments were repeatable showing also that sedimentation at 10,000 rpm did not significantly affect particle integrity and estimated signal intensities of NDDS.sub.1 (see solid and dotted lines). It also demonstrated that for two days at room temperature (the timescale over which the experiments were performed), changes to solution properties under conserved mass balance were barely traceable. Further clear, drug and targeting dye were located within the population of sedimenting particles since the distributions from absorbance detection (FIGS. 5 (b) and (c)) were virtually identical to that derived from the universal RI detection (FIG. 5 (a)). Another interesting observation was the apparent increase in average sedimentation coefficients after two weeks of NDDS.sub.1 storage at room temperature. Further storage primarily decreased the apparent area under the curve of the differential distribution of sedimentation coefficients, Is-g*(s), in all detection modes (vide infra). FIG. 5 (d) clearly shows that all detectors correspond linearly against NDDS.sub.1 concentration including the observation of the population of NDDS.sub.1 by RI, observation of encapsulated drug in NDDS.sub.1 at λ=462 nm, and targeting dye located in NDDS.sub.1 at λ=660 nm. The existence of a small amount of free drug persisting in the supernatant (SN) at λ=462 nm was also verified. These results clearly support the inventive idea of quantitative assessment of individual concentrations of NSSs, encapsulated drug, and targeting dye functionality of NSSs in a very repetitive manner when dispersed in water.
[0073] Typically, NSSs in solution are described by the size of hydrodynamic equivalent spheres, i.e. their hydrodynamic diameter, d.sub.h,NSS. Based on the solid sphere concept, d.sub.h,NSS is accessible via the relation: d.sub.h,NSS=3√{square root over (2)}√{square root over ([s]ν.sub.NSS)} from sedimentation velocity experiments, where ν.sub.NDDS is the partial specific volume of NSSs and [s]=s.sub.NSSη.sub.0/(1−ν.sub.NSS![custom-character]()
.sub.0) the intrinsic sedimentation coefficient defined with the solvent viscosity, η.sub.0, and solvent density, ![custom-character]()
.sub.0. ν.sub.NSS, in first approximation, is inversely proportional to NSS density, Q.sub.NSS, and desired for NSS size estimations. It was estimated by sedimentation velocity experiments of the particle NSS populations diluted in water only but as well D.sub.2O whose density and viscosity was determined. The partial specific volume of NDDS.sub.1 was ν.sub.NDDS=0.76 cm.sup.3g.sup.−1 and hence their density Q.sub.NDDS=1.32 g cm.sup.3, a value close to the known bulk density of PLGA.
[0074] FIG. 6 shows Is-g*(s) distribution from sedimentation velocity experiments performed in water, and in culture medium (CM) without and with human serum (10 wt %) (CMS) after initial NDDS.sub.1 preparation. Using absorbance detection at λ=462 nm demonstrated feasibility to as well study these solution conditions for the NDDS.sub.1 population (FIG. 6). The differential distributions of sedimentation coefficients, Is-g*(s), shifted to slightly larger values at a reduced overall absorbance intensity, more so for CMS as opposed to that of CM. As well, as opposed to CM alone, even at this relatively low speed of 10,000 rpm used for NDDS.sub.1 sedimentation, a distinct population of smaller sedimentation coefficients below 10 S, originating from serum proteins, was seen. In contrast to dilutions in water, shaking solutions after the sedimentation experiment in CM and more so in CMS impose increased difficulties of complete solution reconstitution. Likely, protein components in CM such as insulin and in CMS with human serum albumin support consolidation of NDDSs centrifuged against the cell bottom. Knowledge of NDDS.sub.1 properties and the assumption that the partial specific volume of the NDDSs, ν.sub.NDDS, does not change in different fluids also shows that particles in CM and CMS are slightly larger on average, an observation later re-affirmed for another batch of NDDSs further dealt with experiments performed in human serum (HS). Summarizing, it has been demonstrated that the multi-detection concept is highly useful to observe the sedimenting population of NDDSs with desirable linearity of respective detectors (RI and A) against concentration, allowing for quantitative considerations (FIG. 5 (d)). As well, it was indicated that sedimentation velocity experiments and the respective behaviour of NDDS populations under more realistic conditions such as CM and CMS media are feasible.
Example 2: Quantitative Access to Particle Integrity and Degradation
[0075] The inventors further focused on an encapsulated drug for treatment purposes. FIG. 7 shows the evolution of signal intensities at λ=462 nm over the timescale of three months including repetitive measurements after each sedimentation velocity experiment (indicated by the small symbols). For this purpose, integrated distributions (Is-g*(s)) of FIG. 5(b) were utilized, representative of the drug in NDDS.sub.1, as well as the remaining SN (e.g. FIGS. 3 and 5 (d)). Several features in FIG. 7 are evident. Very initially, the amount of free drug observed in the SN was low when compared to the NDDS.sub.1. Upon storage, the amount of drug in the NDDS.sub.1 substantially decreased with a concomitant increase of free drug in the SN. For example, after two weeks of storage of NDDS.sub.1 at ambient temperature, 20% of drug has entered the particle surrounding liquid concomitant by 20% reduced signal intensity from the particle entity. Also observed was the absence of aggregates for the first two weeks of NDDS.sub.1 storage (see also FIG. 3 (a)) but at apparently increased particle sizes as seen in the Is-g*(s) distribution shifting to larger values of sedimentation coefficients (FIG. 5 (a)-(c)). Allowing for further uncontrolled NDDS.sub.1 degradation at room temperature, resulted in the observation of aggregates after four weeks storage at ambient. These are appearing very pronouncedly at 10 weeks of storage (see also FIG. 3 (b)). Summing all signals associated to the drug (aggregates, NDDSs, SN) showed quantitative recovery in the complex degraded solution up to 10 weeks of storage, i.e. the overall correctness of the mass balance without apparent loss of the most crucial system component, i.e. the drug. After 10 weeks of storage, solid material at the bottom of the NDDS suspension was observed, which could not be re-suspended even by vigorous vortexing and shaking. At this timescale of NDDS.sub.1 storage, degradation made a significant amount of NDDS.sub.1 lose identity, precipitating out from solution by forming insoluble rather undefined material. It was noted that NDDSs formed by nanoprecipitation were located in a metastable region that, due to changes in the formulation in solution, particularly by degradation, may approach their energetic minimum by precipitation of polymer from solution forming water-insoluble aggregated polymeric components.
[0076] In addressing validation of the apparent quantitative situation implied by FIG. 7 (a), and the perception that the drug shows poor water solubility, additional control experiments were performed. The goal was to dissolve the drug in water with and without added surfactant poly(vinyl alcohol) (PVA) that was used for NDDS.sub.1 formulation. In both cases signal intensities at λ=462 nm in the AUC cells responded poorly to concentration (FIG. 7 (b)). This was accompanied by the visual observation of solid drug material precipitating at the bottom of the flasks before filling the AUC cells. These experiments also indicated that the apparently high signal intensities of free drug in SN at λ=462 nm in FIG. 3 (a) and FIG. 7 (b) originated from an enhanced solubility of the drug. The only origin being possible for this scenario was the existence of degradation products of PLGA, a polymer that appears excellently suited to encapsulate the hydrophobic drug via nanoprecipitation. While drug in NDDSs corresponds linearly to concentration it was observed as well for a completely degraded sample stored at a concentration of c=1.26 mg ml.sup.−1 at a temperature of T=37° C. for 7 weeks, as mentioned later. The response to concentration was very similar (drug in NDDS vs. drug in completely degraded solution, FIG. 7 (b)) indicating similar absorptivity of the drug in both scenarios at λ=462 nm. Spiking drug dissolved in dimethyl sulfoxide (DMSO) to the solution of known concentration of the drug (FIG. 8), while keeping the ratio of DMSO to water constant in each spiked sample, showed quantitative recovery of additional amount of drug. The recorded signal intensities practically merged in the established signal intensity-concentration relationship (FIG. 8).
[0077] FIG. 7 (a), therefore, is a quantitative representation of existent drug (in aggregates, encapsulated in NDDS.sub.1 or in SN), i.e., the opportunity to quantitatively access drug load in (i) potential aggregates, in (ii) NDDS.sub.1, (iii) free drug in surrounding liquids (i.e. SN), as well as (iv) mechanistic insight in particle degradation has been demonstrated. This situation as well could be correlated to a first slight increase in NDDS.sub.1 size followed by further reduction of average sizes at similar densities, i.e. their degradation. All these insights are provided by the single instrumental basis of multi-detection AUC.
Example 3: Study of Storage Conditions
[0078] With the above-described experiments on NDDS.sub.1 for the quantitative utility of AUC for the described inventive purposes, another batch of NSSs (NDDS.sub.2) was utilized to gain different insights. After preparation, this sample was stored at different temperatures of T=4° C. and at T=37° C. and at a concentration of c=4.29 mg mL.sup.−1. Knowledge of the detector responses was shown as well to allow a quantitative estimation of concentration of the drug located in NDDSs and surrounding liquid at λ=462 nm (FIG. 7). A wavelength of λ=660 nm appeared representative of the targeting dye (FIGS. 4 & 5), while universal RI detection was broadly applicable for the observation of sedimenting NDDS populations supporting all observed changes.
[0079] FIG. 9 (a) clearly demonstrates that aliquots of NDDS.sub.2 stored one night at T=4° C. in the fridge and at T=37° C. are already different in their solution appearance, i.e. the population of NDDS.sub.2 stored at T=37° C. showed larger average sedimentation coefficients than that stored at T=4° C. This was in agreement with the situation that, upon storage and potential degradation, average values of the sedimentation coefficients and apparent hydrodynamic diameters, d.sub.h,NDDS, increases as shown for NDDS.sub.1. Estimations of the percentage of the drug located in NDDS.sub.2 and SN are shown in FIG. 9 (b). The results as well re-affirm with this batch of NDDS.sub.2 the accompanied release of drug from the NDDSs and its existence in the surrounding liquid after storage overnight, more so at T=37° C. as opposed to T=4° C.
[0080] FIG. 10 makes use of all detector responses and shows that even after 15 days of NDDS.sub.2 storage at T=4° C. individual signals from the RI and UV-detector at λ=462 nm and λ=660 nm appear invariant, including that of the free drug in SN (bottom of FIG. 10 (a)). Such conditions are therefore suitable for long-term storage of NSSs without affecting their integrity and drug load. After 16 h of sedimentation experiments, observing the population of NDDS.sub.2, the AUC was accelerated to 42,000 rpm allowing for further detailing of existent SN that should contain the PVA as smaller colloidal species used for NDDS formulation. In this case as well, similar RI intensities, presumably stemming from the PVA, were apparent at T=4° C. Control experiments of varying concentrations of PVA used for formulation of NDDS.sub.2 and dissolved in water clearly supported this consideration (FIG. 11). Established signal intensity-concentration relationships allowed for gauging PVA polymer concentration in SN, vice versa also the amount of PVA associated to the NDDSs.
[0081] Storage at T=37° C. for 15 days (FIG. 10 (b)) indeed led to severe degradation. This degradation was clearly identifiable by a significant reduction in the RI signal intensity, absorbance at λ=462 nm in the NDDS.sub.2 population and its pronounced observation in the SN (FIG. 10 (b), bottom), a result in excellent agreement to that of NDDS.sub.1 particles (FIG. 5). Coupled to a significant reduction of RI intensity and that of absorbance at λ=462 nm is as well the reduction of signal intensity associated to the dye measured in NDDS.sub.2 at λ=660 nm, a situation again comparable to that of NDDS.sub.1 (FIG. 5C). While the maximum of absorbance of the drug was independent of its location (in NDDSs, aggregates, and SN, see FIG. 7b), the apparent loss of dye signal in NDDS.sub.2 was observed without dominant appearance in the SN. While the dye has its maximum absorbance located at λ=660 nm in DMSO, a solvatochromic shift downward in wavelength to a maximum absorbance located at λ=635 nm in water was observed (FIG. 4). This supported the apparent loss of dye signal upon NDDS degradation.
[0082] Additionally, spinning the centrifuge at 42,000 rpm after NDDS.sub.2 sedimentation, clearly showed an increased concentration of PVA present in the SN (FIG. 10b, bottom), a situation in concert with the assumption that it becomes partly incorporated in the particle during formulation and acting as a surfactant. It also allowed maintaining overall NDDS integrity by preventing aggregation with a stealth-like effect when NDDSs were exposed to protein-containing media (vide infra). In due course of NDDS.sub.2 degradation, additional PVA entered the free solution state.
Example 4: Experiments in Human Serum
[0083] FIG. 12 (a) highlights that the normalized distribution of sedimentation coefficients Is-g*(s) for NDDS.sub.2 stored at T=4° C. for a timescale of 15 days is almost undistinguishable in all detection modes (RI, absorbance at λ=462 nm and λ=660 nm), as well supported by the very similar absolute signal intensities (FIG. 12 (b)).
[0084] After 9 days, NDDS.sub.2 sample stored at T=4° C. was diluted with human serum such that the fraction of aqueous phase and serum (40/60, w/w) remained identical but varying NDDS.sub.2 concentration. Unsurprisingly, an overall increase in the optical density was observed. Notwithstanding, at a wavelength of λ=660 nm, a wavelength that allowed tracing of the NDDS population virtually identical to that of a wavelength of λ=462 nm and RI detection (FIG. 5 for NDDS.sub.1 and FIG. 10 for NDDS.sub.2), was used to observe the NDDS.sub.2 population. Interestingly, increased NDDS.sub.2 concentrations correspond to higher signal intensities at λ=660 nm as clearly seen in FIG. 12 (c). At the same time, a rather undefined distribution of small sedimentation coefficients was apparent. These could be associated to sedimentation of rather large serum components even at relatively low speeds of 10,000 rpm. Integration of the Is-g*(s) distributions resulted in a concentration dependence that was again highly linear (FIG. 12 (d)). All NDDS.sub.2 populations in serum sediment slightly slower than the NDDS.sub.2 population in water. This is well-supported by the increased viscosity of human serum when compared to water. With the respective density and viscosity of the liquid media at hand, the inventors were able to establish an intrinsic sedimentation scale, [s], where [s]=sη.sub.0/(1−ν![custom-character]()
.sub.0) (see FIG. 13). ν is the partial specific volume of the NDDSs in cm.sup.3g.sup.−1 (vide supra).
[0085] In fact, the differential distribution of intrinsic sedimentation coefficients, Is-g*([s]) for NDDS.sub.2 shifts to slightly larger values, however, with some influence on distribution width. The calculated hydrodynamic diameter d.sub.h,NDDS=3√{square root over (2)}√{square root over ([s]ν.sub.NDDS)} based on the weight average sedimentation coefficient of the Is-g*(s) distribution is calculated as d.sub.h=17 nm in water and d.sub.h=18 nm in human serum, respectively. These results showed that though particles did not lose their integrity, their apparent hydrodynamic effective size only slightly increased. Maintenance of NDDS.sub.2 integrity in HS appeared to originate from the “stealth-like” effect of PVA used in NSS formulation. These inventive approaches and experimental results demonstrate the opportunity and present invention of an in situ and quantitative assessment of medical NSSs under conditions most close to intended real-life conditions, i.e. in human fluids.
[0086] Further studies in human serum were conducted with two different batches of NDDS: SDL_NP 4.1 and SDL_NP 4.2, which were firstly characterized in water and then characterized in situ in human blood serum. Samples with concentrations of nanoparticles between 0.5 to 1.5 mg/mL, containing an overall 55 w % human serum, were prepared and measured at T=20° C. and n=7500 rpm in the AUC with the refractive index and absorbance detection at λ=369 nm, i.e. identical to experiments in water. Firstly, differential distributions of intrinsic sedimentation coefficients Is-g*([s]) of two different batches of NDDS: SDL_NP 4.1 and SDL_NP 4.2 were measured/determined in water. These particles differ with respect to the concentration of the targeting dye DY635, which in SDL_NP 4.1 was 33.7 μg/mL and in SDL_NP 4.2 amounted to 2.6 μg/mL.
[0087] Further characteristics are summarized in Table 1:
TABLE-US-00001 TABLE 1 characteristics of SDL_NP 4.1 and SDL_NP 4.2 NDDS. Z-average [DY635] [NP] [Drug] PDI DLS in NaCl SDL_NP4.1 33.7 μg/mL 5.982 mg/mL 394 μg/mL 0.079 82.5 nm SDL_NP4.2 2.6 μg/mL 5.855 mg/mL 348 μg/mL 0.138 99.4 nm
[0088] Both batches were measured at c=1 mg/mL, respectively, and measurements were performed at T=20° C., n=7500 rpm for 20 hours (FIG. 14). Clearly, the differential distributions of sedimentation coefficients of the sample with higher concentration of DY635 according to the refractive index detector was broadly distributed compared to samples with lower concentration of DY635. The situation qualitatively confirmed the DLS results regarding the increased dispersity of both samples, reflected by the increased PDI (Table 1). A PDI of 0.079 according to DLS for SDL_NP4.1 was reflected in the AUC results by a relatively narrow distribution, whereas SDL_NP4.2 with a PDI of 0.138 (according to DLS) indicated a broader distribution. Additionally, the size difference of 16.9 nm based on Z-Average DLS results (Table 1) between both nanoparticles was reflected by different maxima of differential sedimentation coefficient distributions in AUC results. Distribution of SDL_NP4.2 therefore was shifted to higher average sedimentation coefficients compared to SDL_NP4.1.
[0089] Differential distributions of sedimentation coefficients of SDL_NP4.2 at different concentrations, namely c.sub.1=0.498 mg/mL, c.sub.2=0.746 mg/mL, c.sub.3=1.068 mg/mL, c.sub.4=1.531 mg/mL, measured in 55 w % human serum with absorbance detection at λ=369 nm, are depicted in FIG. 15. The differential distributions of sedimentation coefficients obtained for each concentration of SDL_NP4.2 revealed a distribution centred at about 27 S, probably representing human serum components. Measurements of pure 55 w % human serum yielded confirmation by exhibiting also the peak in the distributions at 27 S. Furthermore, a broader distribution between 50 to 300 S appeared for each concentration. Also, the different concentrations of samples were only slightly observable in the distributions by specified peak maxima obtained from sedimentation velocity recorded at λ=396 nm. Due to the relatively high background optical density, a new wavelength had to be chosen, allowing for sufficient signal intensity (with respect to the background) in the investigation of nanoparticles of the NSS. As there is a specific absorbance maximum at λ=635 nm for the targeting dye (DY635), it was established that distributions determined from refractive index and absorbance measurements closely overlap in the DY635-carrying nanoparticles, indicating a similar differential distribution of sedimentation coefficients. Differential distributions of sedimentation coefficients of SDL_NP4.2 at the four different concentrations listed above, namely c.sub.1=0.498 mg/mL, c.sub.2=0.746 mg/mL, c.sub.3=1.068 mg/mL, c.sub.4=1.531 mg/mL, were measured in 55 w % human serum with absorbance detection at λ=635 nm. Results are depicted in FIG. 16A, showing that different nanoparticle concentrations are distinguishable using absorbance detection. The absorbance intensities were higher with increasing concentrations. By plotting the concentration of nanoparticles (as seen from increased detector responses) for each respective sample against the corresponding measured optical density at λ=635 nm, the linearity of signal intensities and their dependence on concentrations could be determined (FIG. 16B). Comparable with FIG. 15, a peak at about 27 S was observed due to components of human serum, while the distribution covering sedimentation coefficients between 30 to 400 S correspond to the nanoparticles analyzed. By using absorbance detection at λ=635 nm, nanoparticles in 55 w % human serum were resolved satisfactorily. Different nanoparticle concentrations were detected in a biofluid, reflecting conditions for applications of these drug delivery systems (i.e. NSS). The result of the measurement further revealed the integrity and comparability of nanoparticles analyzed in human serum-containing media, which is essential for applications in medicine/pharmacy. In particular, the comparison of differential sedimentation coefficient distributions of samples in 55 w % human serum vs. water established the similarity in appearance, such that stability conditions of nanoparticle suspensions close to application reality may be tested in the future. Moreover, the demonstration of identical properties of nanoparticles SDL_NP4.1 and SDL_NP4.2, regarding their apparent differential sedimentation coefficient distributions, also suggested that nanoparticles with a lower dye concentration, e.g. of the mentioned DY635, would exhibit the same integrity in human serum.
