Article pubs.acs.org/ac
Simple Nanoparticle-Based Luminometric Method for Molecular Weight Determination of Polymeric Compounds Sari Pihlasalo,* Maria Virtamo, Nicolas Legrand, Pekka Han̈ ninen, and Harri Har̈ ma ̈ Laboratory of Biophysics and Medicity Research Laboratory, University of Turku, Tykistökatu 6A, 20520 Turku, Finland S Supporting Information *
ABSTRACT: A nanoparticle-based method utilizing timeresolved luminescence resonance energy transfer (TR-LRET) was developed for molecular weight determination. This mixand-measure nanoparticle method is based on the competitive adsorption between the analyte and the acceptor-labeled protein to donor Eu(III) nanoparticles. The size-dependent adsorption of molecules enables the molecular weight determination of differently sized polymeric compounds down to a concentration level of micrograms per liter. The molecular weight determination from 1 to 10 kDa for polyamino acids and from 0.3 to 70 kDa for polyethylene imines is demonstrated. The simple and cost-effective nanoparticle method as microtiter plate assay format shows great potential for the detection of the changes in molecular weight or for quantification of differently sized molecules in biochemical laboratories and in industrial polymeric processes.
M
range is broad.10,11 The macromolecules are measured with dynamic light scattering (DLS), whereas the static multiangle laser light scattering (MALLS) equipped with sensitive photodetectors enables the measurement of molecular weights below 1 × 103 Da.10,17 MALLS is a nondestructive and relatively simple, accurate, and rapid technique.15,18 However, as the scattered light signal is proportional to molecular weight and concentration, high sample concentrations (typically few grams per liter12) are required for small molecules.15 The experiments with available techniques discussed above require expensive equipment, are cumbersome, possess rather slow procedures,3,4 or have limitations for especially low or high molecular weight range.19 Most of the available methods are not suitable tools for routine analysis, and none of them provide an easy-to-use and cost-effective way for quickly controlling the quality during processes (for instance polymerization) or in the end of the reactions. Instead, the investments into expensive instrumentation and even toward experts to perform the analyses may be required. Thus, there is a great need to develop less costly, simpler, and faster, but still equally sensitive, methods. Previously, we have reported a protein quantification assay utilizing luminescence resonance energy transfer (LRET) and adsorption competition of analyte protein and acceptor-dye labeled protein onto donor nanoparticles.20 Here, we developed an application for this method to demonstrate the secondary type molecular weight measurement of different molecules. The methods utilize adsorption of analyte molecules onto the
olecular weight is one of the important parameters to characterize molecules, e.g., different polymers, at wide molecular weight range. Major efforts directed toward developing quick and accurate methods to determine the molecular weight have enabled the development of handful of tools, such as electrophoresis, NMR, mass spectrometry, vapor pressure osmometry, melting point depression, the measurement of viscosity, size exclusion chromatography, the determination of density, ultracentrifugation, and light scattering.1−12 Electrophoresis has been used to measure the molecular weight of mainly peptides and proteins. In gel electrophoresis, the separations of the molecules between 2 × 103 and 3 × 105 Da have been reported.13 However, a quite high quantity of the sample between 0.1 and 5 μg13 is required, and the method fails to measure the molecular weight of highly basic proteins and glycoproteins correctly.14 NMR is applicable for both small and large molecular weights as high as 3 × 106 Da.5,6 Laser desorption/ionization time-of-flight mass spectrometry (LDI TOFMS) is a sensitive and accurate method for small compounds, while the addition of matrix (matrix-assisted laser desorption/ionization time-of-flight mass spectrometry, MALDI TOFMS) enables the study of relatively large molecules.15 However, the upper weight limit is 100 kDa.16 With analytical ultracentrifugation and size-exclusion chromatography (SEC), a wide range of molecular weights can be studied.1,2 However, they have several drawbacks; for instance the experiments may take several hours or even days.3,4 SEC is the most commonly used, simple, and reproducible technique.1 However, the coverage of broad molecular weight range requires the production of gels with different pore sizes.1 Light scattering is probably the only potential primary type technique for routine analysis. The covered molecular weight © 2013 American Chemical Society
Received: June 16, 2013 Accepted: December 13, 2013 Published: December 13, 2013 1038
dx.doi.org/10.1021/ac401797m | Anal. Chem. 2014, 86, 1038−1044
Analytical Chemistry
Article
Figure 1. Principle for the determination of the molecular weight with the Eu(III) nanoparticle-based luminescence resonance energy transfer (LRET) method. (a) Small molecules have lower affinity to the Eu(III) nanoparticles compared to the affinity of larger molecules, and the Alexa labeled molecule adsorbs to the Eu(III) nanoparticles leading to the high LRET signal. (b) In the presence of large molecules, the LRET signal is decreased, as the large molecules cover efficiently the Eu(III) nanoparticles.
Methods. Labeling of γ-Globulin and Polyethyleneimine with Alexa Fluor 680. Alexa Fluor 680 carboxylic acid, succinimidyl ester, was conjugated to γ-globulin (γG-Alexa) and PEI with molecular weight 70 kDa (PEI70-Alexa) as recommended by the manufacturer. Determination of Molecular Weight. The determinations of molecular weight using nanoparticle application were performed in microtiter wells. To determine the molecular weight of PAAs, 70 μL of the sample in 5 mM glycine buffer pH 3.0 and 10 μL of the Eu(III) nanoparticles in water were mixed. γG-Alexa was added in 10 μL of water. The final concentrations of Eu(III) nanoparticles and γG-Alexa were 0.36 and 53 pM, respectively. The determination of PEI molecular weights was enabled in optimized assay conditions for two different molecular weight ranges. The volumes and the order of the additions were similar to the assay for PAAs; however, the buffer, components, and their concentrations varied. The determination of low molecular weights (LM) was performed in 5 mM glycine buffer pH 4.0 with 0.36 pM Eu(III) nanoparticles and 68 pM γG-Alexa, and the determination of high molecular weights (HM) was performed in 5 mM glycine buffer pH 3.0, 0.15 M NaCl with 0.12 pM Eu(III) nanoparticles and 0.98 nM PEI70-Alexa. An additional third condition for the determination of medium molecular weights (MM) may be used to differentiate molecular weights from 1.2 to lower than 10 kDa more precisely (see the Supporting Information). This determination was performed in 5 mM glycine buffer pH 2.0 with 0.36 pM Eu(III) nanoparticles and 44 pM γG-Alexa. Luminescence emission intensities for TR-LRET were measured using two or three replicates with the Victor2 multilabel counter (Wallac, Perkin-Elmer Life and Analytical Sciences, Turku, Finland) using 340 nm excitation and 730 nm emission wavelengths. Optimization of Assay Parameters. The molecular weights of PAAs were determined using assay parameters optimized earlier by us21 for quantification of proteins. The determination of PEI molecular weights for the complete molecular weight range 0.3−70 kDa was achieved by utilizing two assay conditions, which were optimized separately for buffer, pH, salt concentration, the molecule (γG or PEI70) labeled with Alexa, and the concentrations of Eu(III) nanoparticles and γGAlexa or PEI70-Alexa. The assay parameters may have to be optimized separately for each unknown polymer type.
Eu(III) doped nanoparticles and competitive assay principle to assess the change in the adsorption, when the size of the adsorbing molecule changes. The detection of change in molecular weight is based on the differences in the adsorption of molecules (Figure 1) and the shift or change of the calibration curves. When the calibration curves overlap or the molecular weight is known, the measurement of the concentration is enabled. The ability to measure the molecular weight or mass concentration was shown with two different molecules available in several molecular weights: polyamino acids (PAA) and polyethyleneimines (PEI). The measurements are simple to perform in a homogeneous assay format in microtiter wells.
■
EXPERIMENTAL SECTION Materials. γ-Globulins from bovine blood (158 kDa), albumin from chicken egg white (44 kDa), catalase from bovine liver (250 kDa), thyroglobulin from bovine thyroid (670 kDa), lysozyme from chicken egg white (14.6 kDa), and aprotinin from bovine lung (6.512 kDa) were purchased from SigmaAldrich Co. (St. Louis, MO). Insulin from bovine pancreas (5.733 kDa), insulin chain A oxidized ammonium salt from bovine pancreas (2.532 kDa), and insulin chain B oxidized from bovine pancreas (3.496 kDa) were kind gifts from Dr. John E. Fox (Alta Bioscience, Birmingham, U.K., manufacturer SigmaAldrich Co.). Angiotensin I human (1.296 kDa) was a kind gift from Dr. Garry Corthals (Turku Centre for Biotechnology, Turku, Finland, manufacturer Sigma-Aldrich Co.). [D-Penicillamine2,5]-enkephalin (DPDPE, 0.646 kDa) was purchased from Tocris Bioscience (Avormouth, U.K.) and epidermal growth factor human recombinant (EGF, 6.222 kDa) from ProSpec-Tany TechnoGene Ltd. (Rehovot, Israel). Branched PEIs (0.3, 0.6, 1.2, 1.8, 10, and 70 kDa) were obtained as kind gifts from Nippon Shokubai Co., Ltd. (Osaka, Japan). Carboxylate modified Eu(III) polystyrene nanoparticles 92 nm in diameter were purchased from Seradyn Inc. (Indianapolis, IN) and Alexa Fluor 680 carboxylic acid, succinimidyl ester, from Molecular Probes (Eugene, OR). NAP-5 gel filtration columns were ordered from GE Healthcare (Uppsala, Sweden). All the reagents used to prepare buffer solutions were obtained from Sigma-Aldrich (St. Louis, MO). High purity Milli-Q water was used to prepare all aqueous solutions. 1039
dx.doi.org/10.1021/ac401797m | Anal. Chem. 2014, 86, 1038−1044
Analytical Chemistry
■
Article
RESULTS AND DISCUSSION
A simple mix-and-measure nanoparticle-based method utilizing time-resolved luminescence resonance energy transfer TRLRET20 was developed for the molecular weight determination of polymeric compounds. The method was demonstrated with polyamino acids (PAAs) and polyethyleneimines (PEIs) available with different molecular weights. Adsorption of acceptor labeled molecule to Eu(III) doped polystyrene nanoparticles was followed, and a change in luminescence resonance energy signal was related to the displacement of the labeled molecule (γG- or PEI70-Alexa) with the analyte molecule. The sensitized emission was monitored at 730 nm using excitation wavelength of 340 nm and time-resolved luminescence detection. The measurement of molecular weight is based on the size-dependent adsorption of different molecules to the nanoparticles: a small molecule occupies the nanoparticle surface less efficiently than a large molecule (Figure 1). To study the concept, we measured the TR-LRET signal as a function of analyte concentration for chemically similar molecules and investigated the changes in signal as a function of the molecular weight in the developed nanoparticle method. The surface chemistry of the Eu(III) polystyrene nanoparticles was dominated by ionizable carboxylic groups giving a negative surface charge [pKa(−COOH) 4.7, pKa(−SO3H) 1.9, and pKa(−SO4H) −3]. The low assay pH provides protonated carboxyl groups and, thus, reduces negative charge on the nanoparticle surface. However, the nanoparticles carry a negative surface charge due to the sulfite and sulfate groups remaining from the polymerization process of polystyrene. In the assay, γ-globulin or PEI (molecular weight 70 kDa) labeled with Alexa Fluor 680 (γG-Alexa or PEI70-Alexa) and the analyte were adsorbed to the nanoparticles at pH 2−4. These assay pH values are below the isoelectric points of γG (pI 6.4−8.822) and PEI70 [pKa(−NH3+) 4.5, pKa(−NH2+) 6.7, and pKa(−NH+) 11.623]. Thus, the Eu(III) nanoparticles and the labeled or analyte molecules have opposite total charges, and the electrostatic interaction is attractive. The TR-LRET signal as a function of analyte concentration was measured for 11 PAAs and 6 PEIs. We chose to measure PEIs because they are commercially available in a relatively wide molecular weight range and the structural changes within the size series of the polymers are small. In contrast, the selected PAAs vary, as the amino acid sequence differs between PAAs having different pI values, molecular weights, and tertiary structures. Therefore, it is conceivable that PEIs lead to less biased information regarding the size-related adsorption to nanoparticles as these polymers are structurally equal within the polymer series. In Figures 2 and 3, the normalized TR-LRET signals as a function of analyte concentration for PAAs and PEIs are shown. The average coefficient of variation for the replicates in the assays was 5% and 6% for PAAs and PEIs, respectively. The calibration curves for PAAs were measured in 5 mM glycine at pH 3.0 using γG-Alexa as a labeled protein (Figure 2). The calibration curves for proteins with molecular weight higher than 15 kDa overlapped. Therefore, the mass concentration of large proteins can be measured relatively accurately independent of molecular weight at very low concentration. In fact, the method for measuring total protein concentration is approximately 100-fold more sensitive than the most sensitive commercial methods.24,25 The calibration curves
Figure 2. Calibration curves for differently sized polyamino acids. The data was fitted to the logistic function. Abbreviations: DPDPE = [Dpenicillamine2,5]-enkephalin and EGF = epidermal growth factor human recombinant.
Figure 3. Normalized energy transfer signal as a function of polyethyleneimine concentration for differently sized polyethyleneimines measured at two assay conditions optimized for the determination of (a) low and (b) high molecular weights. The vertical dashed lines display the concentrations for the optimal separation of different molecular weights.
for proteins and peptides having molecular weight below 10 kDa shifted clearly to large mass concentrations. The sizedependent gradual shift of the calibration curves suggests that the IC50 could be related to the molecular weight with the information on mass concentration. The assay pH provides opposite total charges for Eu(III) nanoparticles and different PAAs, leads to similar interactions on the surface, and suggests that our observations are strongly dependent on the molecular weight rather than differences in these electrostatic interactions. In Figure 4, a standard curve based on the IC50 values is shown. 1040
dx.doi.org/10.1021/ac401797m | Anal. Chem. 2014, 86, 1038−1044
Analytical Chemistry
Article
with PEI70-Alexa due to the similar adsorption properties for the labeled and analyte molecules. At the HM condition with PEI70-Alexa as a labeled molecule, the TR-LRET signal for the largest PEIs as a function of PEI concentration did not overlap at PEI concentrations higher than 40 μg/L allowing the determination of PEI molecular weight larger than 2 kDa (Figure 3b). The small PEIs (
