Introduction 1 2 3 4 2 5 6 7 8 10 11 12 13 14 17 18 19 20 19 21 22 Escherichia coli 2+ 2+ 23 24 2+ 2+ 25 In this study, steady-state and time-resolved fluorescence spectroscopy experiments with Nile red are used for the sensitive detection of large β-galactosidase aggregates, present in solution at concentrations in the low nanomolar range. Denaturation and aggregation of native protein is induced by incubating aqueous solutions below and above the experimentally determined unfolding temperature of 57.4°C. The presence of aggregates in heat-treated solutions is confirmed by light scattering. Time-resolved fluorescence anisotropy and fluorescence correlation spectroscopy (FCS) are used for analyzing the size of the aggregates detected by Nile red. The results obtained with Nile red are compared with results from SEC analysis. Materials and methods Materials Escherichia coli 2 Methods Preparation of β-galactosidase solution 2 26 For incubation at elevated temperatures, 1 ml aliquots of β-galactosidase solution containing 0.10 μM protein were filled into 1.5 ml polypropylene Eppendorf tubes. They were heated for a defined period of time without shaking using an Eppendorf Thermomixer (Eppendorf, Hamburg, Germany), and then cooled at room temperature. Steady-state fluorescence measurements Instrument and measurement settings Steady-state fluorescence measurements were performed with a Fluorolog FL3-21 spectrofluorometer (Jobin Yvon – Horiba, Edison, NJ), equipped with a short-arc xenon lamp. The slit openings of the excitation and emission monochromators were set to a bandwidth of 3 nm. The integration time was 0.05 s, and the signals were corrected for lamp intensity fluctuations by a simultaneously recorded reference signal. Each sample was measured 3–5 times, and the average spectrum was calculated. The sample temperature was controlled by a water bath with a temperature sensor connected to the sample holder. Samples were measured in quartz cuvettes (Hellma GmbH, Muellheim, Germany). Intrinsic tryptophan fluorescence Measurements of intrinsic tryptophan fluorescence of β-galactosidase were performed by exciting 0.10 μM protein solutions at 298 nm and scanning emission between 310 and 450 nm. To establish a protein denaturation curve, the sample temperature was increased from 25 to 70°C in one degree steps. At every temperature, the sample was equilibrated for 5 min and then measured. Nile red fluorescence 21 Time-resolved fluorescence measurements Instrument and measurement settings 27 4 λ Measurements were performed at 25°C, and consisted of repeated 10 s sequences of measuring parallel and perpendicularly polarized fluorescence emission until a maximum peak content of at least 50,000 counts in the data files was reached. Samples of β-galactosidase solution without Nile red were measured for background correction. To minimize background luminescence, all solutions, including buffer, were prepared with fluorescence spectroscopy grade water (Fluka, Buchs, Switzerland). For the performance of a deconvolution procedure in data analysis, the dynamic instrumental response of the experimental setup was recorded using the fast and single-exponential fluorescence decay of the reference compound erythrosine B in water. 28 I t τ i 29 30 1 \documentclass[12pt]{minimal}
\usepackage{amsmath}
\usepackage{wasysym} 
\usepackage{amsfonts} 
\usepackage{amssymb} 
\usepackage{amsbsy}
\usepackage{mathrsfs}
\usepackage{upgreek}
\setlength{\oddsidemargin}{-69pt}
\begin{document}$$I(t) = \sum\limits_{i = 1}^3 {\alpha _i \cdot \exp ( - t/\tau _i )}$$\end{document} α i I t I II t I ⊥ t 2 \documentclass[12pt]{minimal}
\usepackage{amsmath}
\usepackage{wasysym} 
\usepackage{amsfonts} 
\usepackage{amssymb} 
\usepackage{amsbsy}
\usepackage{mathrsfs}
\usepackage{upgreek}
\setlength{\oddsidemargin}{-69pt}
\begin{document}$$I(t) = I_{\rm II} (t) + 2 \cdot G \cdot I_ \bot (t)$$\end{document} G g r t I II t I ⊥ t 3 \documentclass[12pt]{minimal}
\usepackage{amsmath}
\usepackage{wasysym} 
\usepackage{amsfonts} 
\usepackage{amssymb} 
\usepackage{amsbsy}
\usepackage{mathrsfs}
\usepackage{upgreek}
\setlength{\oddsidemargin}{-69pt}
\begin{document}$$r(t) = \frac{{I_{\rm II} (t) - G \cdot I_ \bot (t)}}{{I_{\rm II} (t) + 2G \cdot I_ \bot (t)}}$$\end{document} Fluorescence correlation spectroscopy measurements Instrument and measurement settings Fluorescence correlation spectroscopy measurements were performed with a system composed of a krypton-argon laser and an MRC1024 confocal laser-scanning microscope (Biorad, Hercules, CA), a TE300D inverted microscope (Nikon, Tokyo, Japan) with a water immersion objective lens (Plan Apo 60×, NA 1.2, collar rim correction, Nikon), and ALV-5000/E avalanche photodiode detectors (ALV GmbH, Langen, Germany). The 568 nm line of the krypton-argon laser was used for excitation of Nile red samples. Fluorescence emission above 585 nm was detected. Samples were prepared by mixing heated 0.10 μM β-galactosidase solution, non-heated 0.10 μM β-galactosidase solution or phosphate buffer with diluted Nile red stock solution in polypropylene Eppendorf tubes. Heated β-galactosidase solution was mixed with Nile red in the same tubes that were used for heating. The final concentration of Nile red in the samples was 18.75 nM. One hundred μl sample was transferred into a well of a 96 glass bottom well plate (Bio-one, Greiner, Frickenhausen, Germany). For measurements, the focal volume was positioned 100 μm above the bottom of the well. Raw data, i.e. fluctuations of fluorescence intensity with time, were collected during 30 s. The laser power was minimized to prevent photobleaching during this period. Every sample was measured ten times, and ten experimental autocorrelation functions were subsequently obtained from the raw data. G τ 31 33 4 \documentclass[12pt]{minimal}
\usepackage{amsmath}
\usepackage{wasysym} 
\usepackage{amsfonts} 
\usepackage{amssymb} 
\usepackage{amsbsy}
\usepackage{mathrsfs}
\usepackage{upgreek}
\setlength{\oddsidemargin}{-69pt}
\begin{document}$$\displaylines{G(\tau )=\frac{{\sum\nolimits_i {E_i^2 \langle N_i \rangle M_i (\tau )} }}{{\left( {\sum\nolimits_i {E_i \langle N_i \rangle } } \right)^2 }} \cdot (1\,{-}\, F\,{+}\, F \cdot \exp ( - \tau /\tau _f ))\cr
+\, G(\infty)}$$\end{document} E i i N i i M i τ 5 \documentclass[12pt]{minimal}
\usepackage{amsmath}
\usepackage{wasysym} 
\usepackage{amsfonts} 
\usepackage{amssymb} 
\usepackage{amsbsy}
\usepackage{mathrsfs}
\usepackage{upgreek}
\setlength{\oddsidemargin}{-69pt}
\begin{document}$$M_i (\tau ) = (1 + \tau /\tau _{D,i} )^{ - 1} \cdot (1 + (r_0^ * /z_0 )^2 \cdot (\tau /\tau _{D,i} ))^{ - {1 \mathord{\left/{\vphantom {1 2}} \right.\kern-\nulldelimiterspace} 2}}$$\end{document} τ D,i i \documentclass[12pt]{minimal}
\usepackage{amsmath}
\usepackage{wasysym} 
\usepackage{amsfonts} 
\usepackage{amssymb} 
\usepackage{amsbsy}
\usepackage{mathrsfs}
\usepackage{upgreek}
\setlength{\oddsidemargin}{-69pt}
\begin{document}$$r_0^*$$\end{document} z 0 e 2 −6 2 −1 F τ f 19 G τ 6 \documentclass[12pt]{minimal}
\usepackage{amsmath}
\usepackage{wasysym} 
\usepackage{amsfonts} 
\usepackage{amssymb} 
\usepackage{amsbsy}
\usepackage{mathrsfs}
\usepackage{upgreek}
\setlength{\oddsidemargin}{-69pt}
\begin{document}$$G(\tau ) = \frac{1}{{\langle N\rangle }} \cdot M(\tau ) \cdot (1 - F + F \cdot \exp ( - \tau /\tau _f )) + G(\infty )$$\end{document} 7 \documentclass[12pt]{minimal}
\usepackage{amsmath}
\usepackage{wasysym} 
\usepackage{amsfonts} 
\usepackage{amssymb} 
\usepackage{amsbsy}
\usepackage{mathrsfs}
\usepackage{upgreek}
\setlength{\oddsidemargin}{-69pt}
\begin{document}$$G(\tau ) \,{=} \left(\! {\sum\limits_i {f_i {\cdot} M_i (\tau )} }\! \right)\, {\cdot}\, (1 \,{-}\, F \,{+}\, F {\cdot} \exp ( \,{-}\, \tau /\tau _f )) \,{+}\, G(\infty)$$\end{document} 8 \documentclass[12pt]{minimal}
\usepackage{amsmath}
\usepackage{wasysym} 
\usepackage{amsfonts} 
\usepackage{amssymb} 
\usepackage{amsbsy}
\usepackage{mathrsfs}
\usepackage{upgreek}
\setlength{\oddsidemargin}{-69pt}
\begin{document}$$f_i = {{E_i^2 \langle N_i \rangle } \mathord{\left/{\vphantom {{E_i^2 \langle N_i \rangle } {\left( {\sum\limits_i {E_i \langle N_i \rangle } } \right)^2 }}} \right.\kern-\nulldelimiterspace} {\left( {\sum\limits_i {E_i \langle N_i \rangle } } \right)^2 }}$$\end{document} 34 9 \documentclass[12pt]{minimal}
\usepackage{amsmath}
\usepackage{wasysym} 
\usepackage{amsfonts} 
\usepackage{amssymb} 
\usepackage{amsbsy}
\usepackage{mathrsfs}
\usepackage{upgreek}
\setlength{\oddsidemargin}{-69pt}
\begin{document}$$G_{{\rm norm}} (\tau ) = \frac{{G(\tau ) - G(\infty )}}{{G(0) - G(\infty )}}$$\end{document} G G G G 34 10 \documentclass[12pt]{minimal}
\usepackage{amsmath}
\usepackage{wasysym} 
\usepackage{amsfonts} 
\usepackage{amssymb} 
\usepackage{amsbsy}
\usepackage{mathrsfs}
\usepackage{upgreek}
\setlength{\oddsidemargin}{-69pt}
\begin{document}$$\displaylines{G_{{\rm norm}} (\tau ) = G_{{\rm norm}} (0) \cdot M(\tau ) \cdot (1 - F + F \cdot \exp ( - \tau /\tau _f ))\cr +\, G_{{\rm norm}} (\infty )}$$\end{document} G norm G norm 11 \documentclass[12pt]{minimal}
\usepackage{amsmath}
\usepackage{wasysym} 
\usepackage{amsfonts} 
\usepackage{amssymb} 
\usepackage{amsbsy}
\usepackage{mathrsfs}
\usepackage{upgreek}
\setlength{\oddsidemargin}{-69pt}
\begin{document}$$\displaylines{G_{{\rm norm}} (\tau ) = \left( {\sum\limits_i {f_{i,{\rm norm}} \cdot M_i (\tau )} } \right) \cr
\cdot\, (1 - F + F \cdot \exp ( - \tau /\tau _f )) + G_{{\rm norm}} (\infty )}$$\end{document} 12 \documentclass[12pt]{minimal}
\usepackage{amsmath}
\usepackage{wasysym} 
\usepackage{amsfonts} 
\usepackage{amssymb} 
\usepackage{amsbsy}
\usepackage{mathrsfs}
\usepackage{upgreek}
\setlength{\oddsidemargin}{-69pt}
\begin{document}$$\displaylines{f_{i,{\rm norm}} = \frac{{{{E_i^2 \langle N_i \rangle }\big/{\left( {\sum\nolimits_i {E_i \langle N_i \rangle } } \right)^2 }}}}{{{{\left( {\sum\nolimits_i {E_i^2 \langle N_i \rangle } } \right)} \big/{\left( {\sum\nolimits_i {E_i \langle N_i \rangle } } \right)^2 }}}}\cr
= {{E_i^2 \langle N_i \rangle } \mathord{\left/{\vphantom {{E_i^2 \langle N_i \rangle } {\left( {\sum\limits_i {E_i^2 \langle N_i \rangle } } \right)}}} \right.\kern-\nulldelimiterspace} {\left( {\sum\limits_i {E_i^2 \langle N_i \rangle } } \right)}}}$$\end{document} f i i σ 34 35 τ D D 13 \documentclass[12pt]{minimal}
\usepackage{amsmath}
\usepackage{wasysym} 
\usepackage{amsfonts} 
\usepackage{amssymb} 
\usepackage{amsbsy}
\usepackage{mathrsfs}
\usepackage{upgreek}
\setlength{\oddsidemargin}{-69pt}
\begin{document}$$D = (r_0^ * )^2 /4\tau _D$$\end{document} R h 14 \documentclass[12pt]{minimal}
\usepackage{amsmath}
\usepackage{wasysym} 
\usepackage{amsfonts} 
\usepackage{amssymb} 
\usepackage{amsbsy}
\usepackage{mathrsfs}
\usepackage{upgreek}
\setlength{\oddsidemargin}{-69pt}
\begin{document}$$D = kT/6\pi \cdot \eta \cdot R_h$$\end{document} Light scattering Detection of β-galactosidase aggregates in solution was done with static light scattering measurements using a fluorometer (Fluorolog FL3-21, Jobin Yvon–Horiba, Edison, NJ). The excitation and emission wavelengths were set to 400 nm, and measurements were performed at 25°C. R h Size exclusion chromatography (SEC) 3 6 7 λ ex λ em λ ex λ em Results and discussion Heat denaturation of β-galactosidase 36 25 25 1a Fig. 1 a dotted line b c  1b 1b Fig. 2 a broken line λ max dotted line λ max solid line λ max b broken lines circles c  1c 1c λ ex λ em Nile red steady-state fluorescence 2a 2a 14 2b 2b λ max λ max 2a 2c 2c 1b λ ex λ em Fig. 3 I t τ 1 τ 2 τ 3 α 1,norm α 2,norm α 3,norm χ 2  Table 1 Lifetime analysis Sample 2 a b buffer 1.039 0.20 non-heated 1.011 0.42 15–49°C 1.062 1.30 60–49°C 1.056 1.89 120–49°C 1.064 2.48 240–49°C 1.000 2.10 5–62°C 1.032 1.50 a χ 2 b \documentclass[12pt]{minimal}
\usepackage{amsmath}
\usepackage{wasysym} 
\usepackage{amsfonts} 
\usepackage{amssymb} 
\usepackage{amsbsy}
\usepackage{mathrsfs}
\usepackage{upgreek}
\setlength{\oddsidemargin}{-69pt}
\begin{document}$$\langle \tau \rangle = \sum {\alpha _{i,{\rm norm}} \cdot \tau _i }.$$\end{document} Time-resolved fluorescence spectroscopy with Nile red 3 I t χ 2 1 τ τ τ 2c τ 2a r t Fluorescence correlation spectroscopy (FCS) measurements with Nile red 4a 4a χ 2 τ f F τ D 2 37 −6 2 −1 τ D G norm 2 χ 2 τ D τ f F G norm 2 τ D τ f F 2 Table 2 Coefficients from autocorrelation function analysis Sample with           Nile red Model τ D a f 1,norm a τ D a f 2,norm a τ f a F a G norm a G norm a \documentclass[12pt]{minimal}
\usepackage{amsmath}
\usepackage{wasysym} 
\usepackage{amsfonts} 
\usepackage{amssymb} 
\usepackage{amsbsy}
\usepackage{mathrsfs}
\usepackage{upgreek}
\setlength{\oddsidemargin}{-69pt}
\begin{document}$$\chi ^2$$\end{document} Phosphate buffer One-Component 0.309±0.008 – – – 0.022±0.002 0.451±0.010 0.920±0.016 −0.0049±0.0003 1.03 β-Galactosidase non-heated One-Component 0.291±0.009 – – – 0.018±0.002 0.434±0.015 0.915±0.021 −0.0047±0.0004 1.08 β-Galactosidase 62°C/5 min Two-Component 0.287±0.017 0.335±0.005 45.7±1.02 0.690±0.007 0.019±0.001 0.246±0.008 – −0.0368±0.0021 1.44 a Fig. 4 a black F τ f τ D G norm G norm \documentclass[12pt]{minimal}
\usepackage{amsmath}
\usepackage{wasysym} 
\usepackage{amsfonts} 
\usepackage{amssymb} 
\usepackage{amsbsy}
\usepackage{mathrsfs}
\usepackage{upgreek}
\setlength{\oddsidemargin}{-69pt}
\begin{document}$$\chi _r^2$$\end{document} b black F f 1,norm τ D f 2,norm τ D G norm \documentclass[12pt]{minimal}
\usepackage{amsmath}
\usepackage{wasysym} 
\usepackage{amsfonts} 
\usepackage{amssymb} 
\usepackage{amsbsy}
\usepackage{mathrsfs}
\usepackage{upgreek}
\setlength{\oddsidemargin}{-69pt}
\begin{document}$$\chi _r^2$$\end{document}  Fig. 5 a b τ D  Fig. 6 upper curve  4b 4b τ f τ D 2 F 2 19 τ D 2 D −8 2 −1 R h D −7 2 −1 R h 4b χ 2 2 35 5a 5b τ f τ D R h τ D 25 6 1a 1c 1a 1c 2c 1a 6 Conclusions Using the model protein β-galactosidase, it was shown that the presence small amounts of large, denatured protein aggregates in solution can be detected by Nile red fluorescence. Aggregates detected by Nile red had hydrodynamic radii around 130 nm with a broad size distribution. Native protein and small aggregates thereof had no substantial effect on Nile red fluorescence intensity. By steady-state fluorescence measurements, it was possible to detect 1 nM denatured and highly aggregated β-galactosidase in solution. The spectroscopic detection of protein aggregates by Nile red is potentially useful for formulation screening or quality control of protein pharmaceutics. The presence of even minute fractions of aggregates in protein formulations needs to be avoided, because this can cause immune reactions in patients. Since large aggregates are particularly potent for breaking immune tolerance, their analytical detection is very important. SEC, which is the standard method for studying protein aggregation, is not reliable for detecting large aggregates, because they may be excluded from the separation column. In this work we showed that an analytical detection method with Nile red is a possible approach to overcome this shortcoming of SEC. After establishing the method with time-resolved fluorescence spectroscopy and FCS, as was done for β-galactosidase, steady-state fluorescence measurements, which can be performed in most labs, may enable an experimentally simple and sensitive detection of large protein aggregates.