Cholesteryl-Conjugated Ribonuclease A Exhibits Enzyme Activity in Aqueous Solution and Resistance to Dimethyl Sulfoxide.

Using bovine pancreatic ribonuclease A (RNase A) and cholesterol, we synthesized cholesteryl-conjugated ribonuclease A (CHRNase A) to evaluate the influence of a conjugated hydrophobic moiety on protein function. Nuclear magnetic resonance and matrix-assisted laser desorption/ionization time-of-flight spectrometry suggested that one cholesteryl group was conjugated to RNase A. Differential scanning calorimetry indicated that CHRNase A was denatured in the solid state but was folded in phosphate buffer (0.05 mol/L, pH 6.5). CHRNase A resembled RNase A in its secondary structure, but circular dichroism (CD) spectra revealed that the helical content of CHRNase A was decreased and the tertiary structure of CHRNase A differed from that of RNase A. Furthermore, fluorescence measurements, CD spectra, an 8-anilino-1-naphthalenesulfonic acid ammonium salt-based assay, and surface tension measurements suggested that cholesterol was conjugated to a tyrosine residue on the protein surface. The relative activity of CHRNase A to RNase A was 79 ± 7%, and the enzyme activity of CHRNase A by adding β-cyclodextrin (β-CyD) increased to 129 ± 7%. Therefore, we considered that the cholesteryl group interacted with substrate (cytidine 2'3'-cyclic monophosphate monosodium salt) to inhibit the enzyme reaction. Finally, the environment around tyrosine residues in CHRNase A in dimethyl sulfoxide was similar to that of native RNase A in phosphate buffer (0.05 mol/L, pH 6.5). These results suggest that cholesterol conjugation to RNase A altered RNase A functionality, including improvement of RNase A resistance to dimethyl sulfoxide and modulation of the ability of β-CyD to control RNase A enzymatic activity.

Introduction

Recently, protein chemical modification (PCM) has emerged as a major technique in the chemical biology field and has found applications in industrial and pharmaceutical areas. The goal of PCM is to enhance the functionality of proteins by increasing their structural and functional diversity.[1,2] This technique has also been used in the development of new drugs, including antibody–drug conjugates.[3−5] Proteins can be chemically modified in several ways, including through conjugation to polyethylene glycol (PEG), which has proved useful in clinical applications. PEG conjugation not only improves the stability and solubility of proteins but also increases their retention time in the body.[6−9] Many proteins are modified with hydrophilic groups; for example, polysaccharides may be conjugated to proteins in this manner, which improves their thermal stability against stress conditions.[10−12] Although most proteins are modified with hydrophilic groups, some studies have introduced modifications with hydrophobic groups.[13,14]

In natively structured proteins, hydrophobic residues are burried;[15,16] however, protein denaturation, induced by physical stress, heat, or chemical exposure, allows these residues to become exposed.[13,17−19] Aggregation of proteins is mediated by hydrophobic interactions between the newly exposed hydrophobic residues of denatured proteins.[20,21] In addition, protein aggregation caused by denaturation can alter the biological activity of proteins, including changing their immunogenicity. Thus, protein aggregation is instead a factor to be avoided in the design and development of biological pharmaceuticals and must be appropriately controlled.[22,23] Consequently, very few studies have explored the use of hydrophobic groups and their chemical modifications in the field of PCM. Typically, organic solvents are required for the conjugation of hydrophobic groups, resulting in protein denaturation. Furthermore, hydrophobic interactions between modified hydrophobic groups and hydrophobic residues, which were exposed by protein denaturation, could drive protein aggregation.

However, some studies have reported the successful conjugation of hydrophobic groups to water-soluble polymers, such as cholesterol-bearing pullulan (CHP)[24] and cholesterol-bearing poly-l-lysine (CHPLL).[25] These cholesterol-based materials have been reported to self-assemble into nanogels in an aqueous solution via the hydrophobic interactions between the conjugated cholesteryl groups. Functionally, CHP mimics natural molecular chaperones,[26,27] and because of its ability to trap hydrophobic molecules, it has wide applications in drug delivery.[28] Furthermore, the α-helical structures of CHPLL are controlled by host–guest interactions between cholesterol and cyclodextrin.[25] Therefore, chemical modification of a hydrophobic group to water-soluble polymers may not only help enhance the existing functions of polymers but also introduce novel functions due to the formation of new intermolecular/intramolecular hydrophobic interactions.

Therefore, if hydrophobic modifications can be conducted to keep a protein activity, novel functions would be introduced by hydrophobic modification. Furthermore, we consider that the hydrophobic interactions associated with conjugated hydrophobic residues, in conjunction with the existing hydrophobic residues in a molecule, could accelerate the refolding of proteins.

In the present study, we used bovine pancreatic ribonuclease A (RNase A) as a model protein for conjugation with the hydrophobic group cholesterol and synthesized a novel cholesteryl-conjugated RNase A (CHRNase A). RNase A is a 124 amino acid, single-chain protein (molecular weight, 13.7 kDa) that catalyzes the cleavage of phosphodiester bonds in various types of single-stranded RNA.[29−32] RNase A is an extensively studied and homogeneous enzyme, and proteins with structural homology to RNase A belong to the RNase A superfamily.[33] Some RNases in this superfamily exhibit anti-HIV-1 activity and induce cytotoxic effects in tumor cells.[33] Notably, RNase mimics that can irreversibly cleave or otherwise elicit damage to targeted mRNA transcripts represent a particularly promising therapeutic approach to combat various diseases.[34] Cholesterol has been widely reported to improve the hydrophobicity, biocompatibility, and biodegradability of materials and is therefore frequently used in drug-delivery systems. Cholesterol has been found to be more efficient than high-molecular-weight polymers in this regard, as these polymers tend to gradually degrade.[35] Cholesterol is a vital building block for the structural integrity of cell membranes and maintains the fluidity and permeability of membranes. It is also important for intracellular transport and mediates signal transduction and cellular trafficking.[36−38]

Antisense oligonucleotides (ASOs) interacts with a target RNA. Michael et al. reported that cholesterol ASO conjugates showed a fivefold enhancement of ASO potency in the muscles of rodents compared with unconjugated ASOs.[39] Therefore, cholesterol conjugation to RNase A may affect the activity of the enzyme. We also conducted a structural characterization of CHRNase A using nuclear magnetic resonance (NMR) spectroscopy, matrix-assisted laser desorption/ionization time-of-flight mass spectrometry (MALDI-TOF MS), fluorescence, differential scanning calorimetry (DSC), and circular dichroism (CD) measurements. We also calculated the surface hydrophobicity of CHRNase A via a fluorescence assay using 8-anilino-1-naphthalenesulfonic acid ammonium salt (ANS) and surface tension. Furthermore, the enzyme activity was evaluated using cytidine 2′3′-cyclic monophosphate monosodium salt (cCMP) as a substrate. Finally, using fluorescence measurements, we investigated the environment of aromatic amino acids in dimethyl sulfoxide (DMSO). CHRNase A exhibited novel functions such as control of enzyme activity by addition of β-cyclodextrin (β-CyD) and resistance to DMSO. The results suggested that proteins can be imbued with novel functions by modification with hydrophobic moieties.

Results and Discussion

Confirmation of Cholesterol Conjugation to RNase A and the Number of Conjugated Cholesteryl Groups

NMR Spectroscopy

We first used NMR to confirm that cholesterol was successfully conjugated to RNase A. 1H NMR spectra of RNase A alone, CHRNase A, and a physical mixture of RNase A and cholesterol (RNase A [mol]/cholesterol [mol] = 1/12) in DMSO-d6 were acquired. Figure shows that the 1H NMR spectrum of CHRNase A differed from that of the physical mixture. A 1H signal from the cholesteryl group in CHP (0.6–2.4 ppm) was observed in DMSO-d6.[24] Therefore, the 1H signal in the CHRNase A spectrum at approximately 2 ppm likely corresponds to the cholesteryl group, as this 1H signal was not present in the spectra of either RNase A or the physical mixture. Cholesterol was conjugated to RNase A via the carbonyl carbon (C=O) of N,N′-carbonyldiimidazole (CDI). Figure shows the 13C NMR spectra of RNase A and CHRNase A in DMSO-d6. The 30 and 200 ppm signals in the 13C NMR spectrum of CHRNase A correspond to the carbon in the cholesteryl group and the C=O group that connects RNase A to cholesterol, as these signals were not seen for RNase A. In addition, a 1H–13C heteronuclear multiple-quantum correlation (HMQC) NMR spectrum demonstrated a correlation between the 1H signal at approximately 2.0 ppm and the 13C signal at approximately 30 ppm (Figure S1). We thus concluded that we successfully synthesized RNase A conjugated to cholesterol.

Figure 1

1H NMR spectra of the indicated samples in DMSO-d6 at 25 °C. (a) RNase A, (b) CHRNase A, and (c) physical mixture of RNase A and cholesterol. The red circle shows 1H signal from the cholesteryl group. RNase A and CHRNase A concentration: 21.4 mg/mL. Physical mixture of RNase A and cholesterol (RNase A [mol]/cholesterol [mol] = 1/12) concentration: 21.4 mg/mL.

Figure 2

13C NMR spectra of RNase A and CHRNase A in DMSO-d6 at 25 °C. (a) RNase A and (b) CHRNase A. The red circles at approximately 30 and 200 ppm show 13C signal from the cholesteryl group and carbonyl group connecting RNase A and cholesteryl group. RNase A and CHRNase A concentration: 21.4 mg/mL.

Furthermore, we obtained the nuclear Overhauser effect (NOE) difference spectrum of CHRNase A through radio frequency irradiation of the 1H signal at approximately 2 ppm, corresponding to the signal of the cholesteryl group in CHRNase A (Figure S2). NOE measurement provides information about spatial proximity of protons in a molecule. If the spatial distance between the cholesteryl group and amino acids was close, negative signals would be observed in the NOE difference spectra. Conversely, if the spatial distance was far, signals would not be observed. The results suggested that the spatial distance of conjugated cholesteryl group is close to many amino acid groups in DMSO.

MALDI-TOF MS

The NMR results suggested that cholesterol was conjugated to RNase A in our CHRNase A sample. However, we could not determine the number of conjugated cholesteryl groups in CHRNase A using this technique. Therefore, we next applied MALDI-TOF MS to confirm the number of conjugated cholesteryl groups in CHRNase A. The MALDI-TOF MS spectra of RNase A and CHRNase A, with α-cyano-4-hydroxycinnamic acid (CHCA) used as the matrix, were similar (Figure S3a). Therefore, the conjugated cholesteryl group may have been cleaved by an analytical process, such as sample preparation or irradiation by the laser during MALDI-TOF MS. Consequently, we conducted MALDI-TOF MS using another matrix (sinapinic acid, SA). Upon using SA as the matrix, the m/z of CHRNase A increased by approximately 400 Da compared with that of RNase A (Figure S3b), suggesting that the number of cholesteryl groups conjugated to RNase A was one.

Structural Characterization of CHRNase A

Evaluation of Protein State in Solid and Buffer Solution by DSC

Proteins that are denatured because of thermal stress undergo conformational collapse. DSC measures the thermal stability of proteins by detecting changes in their conformation. The endothermic peak observed in DSC is indicative of a cooperative structural transition, which classically corresponds to a native-denatured transition and occurs in both solid[40−42] and aqueous solutions.[43−47] Furthermore, there is an increase in hydrophobic hydration when a protein transitions from the native (folded) state to the denatured (unfolded) state. Therefore, to evaluate the state of hydrophobic residues in CHRNase A, we measured the DSC in both solid state and phosphate buffer (0.05 mol/L, pH 6.5). Phosphate buffer was composed of Na2HPO4 and KH2PO4.

Figure a,b shows the DSC thermograms for RNase A and CHRNase A in the solid and buffer solutions, respectively. In the solid state (Figure a), an endothermic peak was observed for RNase A and for the physical mixture of RNase A and cholesterol. The maximum peak of the endothermic point (Tm) of RNase A was observed at 110 °C, and that of the physical mixture was observed at 113 and 138 °C. Conversely, an endothermic peak for CHRNase A was not observed within the measurement range. Both acetone and DMSO were used during the synthesis of CHRNase A, which may have resulted in denaturation of the protein and hence the absence of the endothermic peak. We also evaluated whether CHRNase A in the solid state was denatured using other methods. As Fourier transform infrared (FT-IR) spectroscopy can be used to evaluate protein conformation using solid samples, using the potassium bromide (KBr) pellet method, we applied FT-IR measurements of RNase A and CHRNase A (Figure S4). In FT-IR measurements of proteins, amide band I (near 1650 cm–1) indicates C=O stretching and amide band II (near 1550 cm–1) indicates N–H bending and C–N stretching.[48,49] The intensity of the amide II band in proteins disappears upon unfolding, whereas the amide I band remains unchanged.[50] Therefore, we confirmed whether CHRNase A was unfolded by evaluating the ratio of transmittance (intensity) for amide II to amide I (Figure S4) and found that the ratios in RNase A and CHRNase A were 2.1 and 1.4, respectively. In agreement with the DSC measurements, these results suggested that CHRNase A was unfolded in the solid state.

Figure 3

DSC thermograms for the indicated samples in the solid state (a) and phosphate buffer (0.05 mol/L, pH 6.5) (b). Solid state: weight of RNase A, CHRNase A, and physical mixture of RNase A and cholesterol (RNase A [mol]/cholesterol [mol] = 1/12): 5 mg. Buffer solution: RNase A and CHRNase A concentration: 30 mg/mL.

Figure b shows that in buffer solution, the Tm’s of RNase A and CHRNase A were detected at 67 °C, indicating that CHRNase A was folded in buffer solution. Cholesterol-bearing polysaccharides and peptides, such as CHP[24] and CHPLL,[25] were previously shown to form hydrogel nanoparticles, suggesting that the conjugated cholesteryl group could affect the folding of CHRNase A in buffer solution. These results suggest that CHRNase A was folded in buffer solution, although it was denatured in the solid state.

Evaluation of the Secondary and Tertiary Structures by CD Spectroscopy

The results of the DSC and FT-IR indicated that CHRNase A was denatured in the solid state but folded in phosphate buffer (0.05 mol/L, pH 6.5). We measured CD to evaluate the secondary and tertiary structures of CHRNase A, which may have been affected by hydrophobic interactions mediated by the cholesteryl group.[51−53]

RNase A (type α + β) comprised both α-helix and β-sheet secondary structures, with a greater percentage of the composition made up of β-sheets than that of α-helices.[54]Figure a shows the CD spectra of RNase A and CHRNase A in the far-ultraviolet (UV) (200–250 nm) region, providing information on protein secondary structure. Although the CD spectrum of CHRNase A was similar to that of RNase A, indicating that CHRNase A still exhibited a secondary structure comprising α-helices and β-sheets, the absolute value at 222 nm was lower than that of RNase A. As the ellipticity at 222 nm in CD spectra is correlated with a protein’s α-helical content,[51−53] this therefore indicates that the α-helical content in CHRNase A was lower than that of RNase A.

Figure 4

CD spectra of RNase A and CHRNase A in phosphate buffer (0.05 mol/L, pH 6.5) at 25 °C. (a) Far-UV (200–250 nm) region and (b) near-UV (250–320 nm) region. RNase A and CHRNase A concentration: 0.5 mg/mL.

The CD spectrum in the near-UV (250–320 nm) region reflects the environment of aromatic residues and provides information about the protein tertiary structure. Figure b shows the CD spectra of RNase A and CHRNase A in the near-UV region. The signal in the region around 275 nm, which arises from tyrosine (Tyr) residues, differed between CHRNase A and RNase A. RNase A has six Tyr residues: three are buried in the protein interior (i.e., Tyr25, Tyr92, and Tyr97) and three were partly exposed (i.e., Tyr73, Tyr76, and Tyr115), where the hydroxyl groups of the exposed Tyr interact with solvent.[47,54] It has previously been shown that the intensity around 275 nm decreases because of the unfolding of RNase A by thermal stress.[31,47] Therefore, the change in the CD spectrum of CHRNase A around 275 nm was likely derived from a change in the environment of the Tyr residues, such as cholesterol conjugation to Tyr or exposure of Tyr to the protein surface.

Altogether, the CD and DSC analyses suggest that CHRNase A existed in the folded state in buffer solution. Although the secondary structure of CHRNase A was similar to that of RNase A, its tertiary structure differed, which may have been induced by cholesterol conjugation to RNase A.

Evaluation of the Environment around Aromatic Residues and Surface Hydrophobicity Using Fluorescence Spectroscopy

Absorption and fluorescence emission at 280 nm are indicative of the environment around aromatic residues in proteins, especially tryptophan (Trp) and Tyr, which is associated with their high quantum yield and extinction coefficient. Therefore, we applied fluorescence measurements with excitation at 280 nm to investigate the environment around aromatic residues in CHRNase A.[55−57]

The fluorescence emission spectra of proteins shift to short wavelengths (blue shift) when the environment around aromatic residues is hydrophobic,[56−59] whereas there is a shift to longer wavelengths (red shift) when the environment is hydrophilic.[56−59] Thus, these changes indicate whether a protein is in the native or denatured state. Also, conjugation of chemical group may promote environmental shifts around aromatic residues.

Figure a depicts the fluorescence emission spectra of RNase A and CHRNase A. RNase A has six Tyr and three phenylalanine (Phe) residues but no Trp residues.[29,30] The maximum fluorescence emission spectra (λmax) for both RNase A and CHRNase A occurred at 306 nm, attributed to Tyr (Tyr λmax = 305 nm), and the intensity of the CHRNase A spectrum was similar to that of RNase A. These results indicated that the environment around buried and exposed Tyr residues in CHRNase A was highly similar to that of RNase A. Fluorescence emission spectrum excited at 280 nm is attributed to the emission of the aromatic ring in aromatic amino acid residues.

Figure 5

Fluorescence measurements of each RNase A and ANS in phosphate buffer (0.05 mol/L, pH 6.5) at 25 °C. (a) Fluorescence emission spectra of RNase A and CHRNase A in 0.05 mol/L phosphate buffer (pH 6.5) at 25 °C. (1) RNase A and (2) CHRNase A. RNase A and CHRNase A concentration: 0.5 mg/mL. Excitation wavelength: 280 nm. (b) Fluorescence emission spectra of ANS and ANS with RNase A or CHRNase A in 0.05 mol/L phosphate buffer (pH 6.5) at 25 °C. (1) ANS, (2) ANS with RNase A, and (3) ANS with CHRNase A. ANS concentration: 1.0 × 10–4 mol/L; RNase A and CHRNase A concentration: 0.5 mg/mL. Excitation wavelength: 365 nm.

The CD spectrum of approximately 275 nm reflects the environment of Tyr residues, and the y-axis of the CD spectrum shows ellipticity. Ellipticity suggests the different extinction coefficient or absorbance of left and right circularly polarized light. If Tyr residues were exposed to the surface, λmax of CHRNase A reveals a longer shift (red shift) and the intensity changes. However, these changes were not observed in CHRNase A. Thus, we assumed that the alteration of CD spectrum of approximately 275 nm may have been induced by the conjugation of cholesteryl to the Tyr residue.

Buried hydrophobic residues in RNase A could have been exposed to the protein surface during the synthesis process and/or cholesterol conjugation. Additionally, because cholesterol is highly hydrophobic, its conjugation to RNase A could have increased the surface hydrophobicity of the protein. Therefore, we applied fluorescence measurements at 365 nm using ANS to evaluate the hydrophobicity on the surface of CHRNase A. ANS is a fluorescent probe that binds to hydrophobic regions on protein surfaces with high affinity and has therefore been used for the characterization of protein binding sites and the study of folding pathways.[47,60−63]

The λmax value of ANS shifts to shorter wavelengths (blue shift) when it transits from a polar to a nonpolar environment, which also increases its fluorescence intensity. Figure b depicts the fluorescence emission spectra of ANS and ANS with RNase A or CHRNase A. The λmax value and intensity for ANS in the presence of RNase A differed slightly from those for ANS alone. Conversely, both values for ANS with CHRNase A were significantly altered compared with those for ANS alone and ANS with RNase A. These results suggest that the hydrophobic regions on the surface of CHRNase A were higher than the hydrophobic regions on the surface of RNase A, which may correspond to an increased exposure of hydrophobic residues on the protein’s surface because of the synthesis process and/or cholesterol conjugation. However, on the basis of the fluorescence and CD measurements, we could not determine whether the cholesteryl group was located on the interior or surface of the protein.

The surface tension of proteins is decreased by denaturation, as hydrophobic residues typically buried in the interior of proteins are exposed to the surface.[64,65] Therefore, we measured the surface tension of RNase A and CHRNase A in phosphate buffer (0.05 mol/L, pH 6.5) and found that both RNase A and CHRNase A exhibited surface tensions of 59 mN/m. This suggests that the extent of hydrophobic residue exposure in CHRNase A was similar to that of RNase A. Thus, the apparent increase in hydrophobicity of CHRNase A likely corresponded to the presence of the cholesteryl group on the protein surface in buffer solution, indicating that the conjugated cholesteryl group was located on the surface of the protein. However, if this were the case, the cholesteryl group could induce protein aggregation. Therefore, we analyzed the particle size of RNase A and CHRNase A at a concentration of 1.0 mg/mL. The particle size for both RNase A and CHRNase A was approximately 119 nm, which indicated that the cholesteryl group was unlikely to drive the aggregation of the modified protein.

Enzyme Activity of CHRNase A

Relative Activity and Enzyme Kinetic Constants of CHRNase A

The changes we observed in the conformation and hydrophobicity on the protein surface of CHRNase A may have affected its enzyme activity. Therefore, using cCMP as a substrate and monitoring turbidity at 286 nm, we evaluated the enzyme activity of RNase A and CHRNase A.[32,66] This demonstrated that the relative activity of CHRNase A to RNase A was 79 ± 7%, implying that the enzyme activity of CHRNase A decreased compared with that of RNase A. We then measured the enzyme activity of RNase A and CHRNase A, again using cCMP (0.02–0.1 mg/mL), at 25 °C (Figure S5), and calculated the Michaelis constant (Km) and the maximum rate (Vmax) from the Lineweaver–Burk plots (Table S1). The Km value of CHRNase A was higher than that of RNase A, although the Vmax value of CHRNase A was almost the same as that of RNase A. The Km value indicates the affinity of an enzyme for a substrate, where a larger Km suggests that the enzyme substrate complex forms at higher concentrations. As previously indicated, Ostergaard et al. reported that cholesterol ASO conjugates showed a 5-fold enhancement of ASO potency in rodents relative to unconjugated ASOs.[39] RNA comprised cytidine monophosphate, and cCMP is a kind of structural isomer. Furthermore, ASOs interact with their target RNA. Therefore, we suggest that larger Km of CHRNase A was conformational change induced by cholesterol conjugation and interaction between conjugated cholesteryl group and cCMP.

Effect of the Addition of β-CyD to CHRNase A on Enzyme Activity

β-CyD can trap cholesterol in a hydrophobic cavity because of its high binding constant with cholesterol.[67] We expected the inhibition by interaction with cholesteryl group and cCMP to be eliminated with the addition of β-CyD if the decrease of the affinity to cCMP was derived from the conjugated cholesteryl group. Therefore, we evaluated the enzyme activity of CHRNase A before and after adding β-CyD using cCMP.

The relative enzyme activity of CHRNase A with β-CyD to CHRNase A without β-CyD was 129 ± 7% (estimated relative activity to RNase A: about 100%). This suggested that the inhibition by interaction with cholesteryl group and cCMP could be eliminated by adding β-CyD. However, this experiment could not reveal the mechanism by which the enzyme activity of CHRNase A was altered with the addition of β-CyD. For example, β-CyD could have induced a conformational change of protein by trapping the cholesteryl group. Therefore, we conducted CD measurements of CHRNase A before and after adding β-CyD in the far- and near-UV regions (Figure S6). The results suggested that a conformational change did not occur after adding β-CyD. Thus, we assumed that the inhibition of enzyme reaction for CHRNase A occurred because of an interaction between the cholesteryl group and cCMP and the inhibition could be eliminated by binding of the cholesteryl group to the hydrophobic cavity in β-CyD (Figure ). This indicates that the enzymatic activity in CHRNase A can be controlled in the manner of an allosteric enzyme.

Figure 6

Schematic illustration of the possible enzymatic reaction of CHRNase A.

Evaluation of the Environment of Tyr Residues in DMSO

DMSO is used as protein denaturant at high concentrations[17,68] and can induce conformational changes in proteins by binding to newly exposed hydrophobic and aromatic side chains upon protein unfolding.[17]

We used 100% DMSO as a solvent in the synthesis of CHRNase A, and acetone was used to obtain the powdered product, which induced the denaturation of CHRNase A in the solid state. However, we observed that CHRNase A was folded in phosphate buffer (0.05 mol/L, pH 6.5). This suggested that hydrophobic interactions with the conjugated cholesteryl group may promote the folding of CHRNase A in organic solvents such as DMSO. Although the NOE result (Figure S3) indicated that the conjugated cholesteryl group is located near amino acid residues in the presence of 100% DMSO, it did not elucidate whether CHRNase A is in the native state in DMSO.

As fluorescence measurements can evaluate the environment surrounding Tyr residues and indicate the folding state of proteins, we applied fluorescence measurements of CHRNase A in DMSO at 25 °C. Figure shows the fluorescence emission spectra of RNase A and CHRNase A in phosphate buffer (in 0.05 mol/L, pH 6.5) and 100% DMSO. A red shift of λmax and an increase in the fluorescence intensity imply that aromatic amino acids are exposed to the protein surface.[56−59] The λmax of RNase A in DMSO (310 nm) showed a red shift compared with that of RNase A in 0.05 mol/L phosphate buffer (pH 6.5, 307 nm), with a concomitant increase in fluorescence. Conversely, the λmax of CHRNase A in DMSO (307 nm) was highly similar to that of the protein in 0.05 mol/L phosphate buffer (pH 6.5). However, the intensity of CHRNase A in DMSO was decreased compared with that of CHRNase A in 0.05 mol/L phosphate buffer (pH 6.5). The environment of Tyr in CHRNase A in both buffer and DMSO was similar to that of native RNase A, but RNase A was denatured in 100% DMSO. Thus, it was inferred that CHRNase A could form a folded state in DMSO and may have been resistant to DMSO.

Figure 7

Fluorescence emission spectra of each RNase A in phosphate buffer (0.05 mol/L, pH 6.5) and DMSO at 25 °C. (a) RNase A in (1) phosphate buffer (0.05 mol/L, pH 6.5) and (2) DMSO. (b) CHRNase A in (1) phosphate buffer (0.05 mol/L, pH 6.5) and (2) DMSO. RNase A and CHRNase A concentration: 0.5 mg/mL. Excitation wavelength: 280 nm.

Conclusions

We succeeded in synthesizing CHRNase A using CDI in DMSO. NMR measurements suggested that cholesterol was conjugated to RNase A through the carbonyl carbon (C=O) of CDI. Moreover, MALDI-TOF MS indicated that one cholesteryl group was conjugated to RNase A.

We conducted a structural characterization of CHRNase A by examining its folding state in a solid and in phosphate buffer (0.05 mol/L, pH 6.5) , the secondary/tertiary structure, the environment around aromatic residues, and the hydrophobicity on the protein surface. DSC results suggested that CHRNase A was unfolded following the synthesis process but folded in buffer solution. The CD spectra indicated that the secondary structure of CHRNase A in buffer solution was similar to that of RNase A. However, the helical content of CHRNase A was less than that of RNase A, and the tertiary structure of CHRNase A differed from RNase A. On the basis of fluorescence analysis excited at 280 nm and CD spectra at 275 nm, we considered the cholesterol was conjugated to a Tyr residue in RNase A. The conjugated cholesteryl group is likely located on the protein surface according to the ANS fluorescence assay and measurements of surface tension.

Furthermore, we measured the enzymatic activity of CHRNase A to evaluate the effects of cholesterol conjugation on protein function. The relative activity of CHRNase A to RNase A was 79 ± 7%, and the Lineweaver–Burk plots implied that a decrease of the affinity to cCMP had occurred. We then evaluated the enzyme activity of CHRNase A before and after adding β-CyD and found that it increased to approximately 129 ± 7% (estimated relative activity to RNase A: approximately 100%); however, we observed no conformational change in CHRNase A before and after adding β-CyD. Therefore, we assumed that the enzyme activity of CHRNase A would be decreased by inhibition by interaction with cholesteryl group and cCMP, which could be eliminated by adding β-CyD, thus recovering enzyme activity. Additionally, although RNase A was denatured in 100% DMSO, fluorescence measurements indicated that the environment of Tyr residues in CHRNase A in 100% DMSO was similar to that of native RNase A. Thus, CHRNase A may form a folded state in DMSO and may be resistant to DMSO.

This study suggests that novel functions, including β-CyD-dependent control of enzyme activity and improved endurance to organic solvents like DMSO, can be added by cholesterol conjugation to RNase A. Such changes in protein function may be more widely applicable. Hence, we conclude that the techniques used in the present study have the potential to enhance industrial and pharmaceutical methods, including those used in the development of immobilized enzymes, allosteric enzymes, biosensors, and protein formulations.

Experimental Section

Materials

Bovine pancreatic RNase A, cCMP, CHCA, SA, and ANS were purchased from Sigma-Aldrich (St. Louis, USA). Cholesterol (Wako special grade), CDI (for organic synthesis), DMSO (guaranteed reagent), dehydrated DMSO (for organic synthesis), DMSO-d6 (for NMR), and acetone (guaranteed reagent) were purchased from FUJIFILM Wako Pure Chemical Industries, Ltd. (Osaka, Japan). β-CyD was purchased from Nihon Shokuhin Kako Co., Ltd. (Tokyo, Japan). All other reagents were commercial products of analytical grade. Phosphate buffer (0.05 mol/L, pH 6.5) was prepared by Na2HPO4 and KH2PO4.

Synthesis of CHRNase A

CDI is used for conjugation with many organic compounds and polymers such as polysaccharides, peptides, and proteins.[69−74] The imidazole in CDI is highly reactive, which allows it to be converted to an ester by reacting with alcohols or amines. Therefore, we used the following procedure and synthesized CHRNase A by conjugating cholesterol with hydroxyl and/or amino groups in RNase A.

For conjugation to RNase A, cholesterol was agitated with CDI in dehydrated DMSO at 50 °C for 3 h and cooled at room temperature (RT). Then, RNase A was added based on the ratio mentioned here: CDI [mol]/RNase A [mol] = 10 and cholesterol [mol]/RNase A [mol] = 12. This solution was agitated at RT for 20 h to cause conjugation of the cholesteryl group to the hydroxyl and amino groups of RNase A. CHRNase A powder was obtained by first adding acetone, then washing with acetone several times, and finally, drying under reduced pressure at RT. The product was stored under −20 °C (Figure S7).

Confirmation of Cholesterol Conjugation to RNase A and the Number of Conjugated Cholesteryl Groups

RNase A (15 mg) and CHRNase A (15 mg) were separately dissolved in DMSO-d6 (0.7 mL). Following filtration of the solutions through membrane filters (0.45 μm), they were analyzed using an NMR spectrometer (JNM-LA500, JEOL, Tokyo, Japan). NMR (500 MHz 1H–, 125 MHz 13C–, and 1H–13C HMQC) spectra were all recorded at 25 °C. DMSO (2.49 ppm) was used as an internal standard in DMSO-d6.

Structural Characterization of CHRNase A

Evaluation of Protein State in a Solid State and Buffer Solution Using DSC

In the solid state, RNase A (5 mg), a physical mixture of RNase A and cholesterol (RNase A [mol]/cholesterol [mol] = 1/12) (5 mg), and CHRNase A (5 mg) were separately placed in aluminum sample pans, sealed, and analyzed by DSC (DSC8240D, Rigaku, Tokyo, Japan). RNase A and CHRNase A were thawed and dried in a desiccator before analyzing. The scan rate and scan range were set at 2 °C/min and 40–160 °C, respectively; alumina (Al2O3) was used as a reference material, and the analysis was conducted in air atmosphere.

In the buffer solution, RNase A (30 mg/mL) and CHRNase A (30 mg/mL) were separately prepared using phosphate buffer (0.05 mol/L, pH 6.5). The solution (30 μL) was placed in stainless steel sample pans and sealed for analysis by DSC (DSC8240D). The scan rate and scan range were set at 2 °C/min and 40–100 °C, respectively. Phosphate buffer (0.05 mol/L, pH 6.5) was used as a reference material, and the analysis was conducted in air atmosphere.

Evaluation of the Secondary and Tertiary Structures by CD Spectroscopy

RNase A and CHRNase A were separately dissolved in phosphate buffer (0.05 mol/L, pH 6.5) at a concentration of 0.5 mg/mL and analyzed using a CD spectrometer (J-820, JASCO, Tokyo, Japan). The CD spectra were recorded at 25 °C from 200 to 320 nm.

To evaluate the environment around aromatic amino acids, RNase A and CHRNase A were separately dissolved in a phosphate buffer (0.05 mol/L, pH 6.5) at a concentration of 0.5 mg/mL and analyzed using a fluorescence spectrometer (FP-750, JASCO, Tokyo, Japan) with the following parameters: excitation wavelength = 280 nm, band pass = 10 m, and emission band pass = 10 nm. Fluorescence emission spectra were recorded at 25 °C from 260 to 400 nm.

RNase A, CHRNase A, and ANS were each dissolved in phosphate buffer (0.05 mol/L, pH 6.5) to evaluate the hydrophobic regions on the protein surface. Each RNase A and ANS solution were mixed to achieve concentrations of 0.5 mg/mL and 1.0 × 10–4 mol/L, respectively, and then analyzed using a fluorescence spectrometer with the following parameters: excitation wavelength = 365 nm, band pass = 10 nm, and emission band pass = 10 nm. Fluorescence emission spectra were recorded at 25 °C from 400 to 600 nm.

Enzyme Activity of CHRNase A

Relative Activity and Enzyme Kinetic Constants of CHRNase A

Relative Activity of CHRNase A to RNase A

RNase A and CHRNase A were separately dissolved in phosphate buffer (0.05 mol/L, pH 6.5) to a concentration of 0.5 mg/mL (enzyme solution). cCMP was dissolved in phosphate buffer (0.05 mol/L, pH 6.5) to a concentration of 0.1 mg/mL (substrate solution); 0.3 mL of enzyme solution (3 h post-preparation) was added to 2.7 mL of the substrate solution. The change in absorbance was analyzed using a UV spectrometer (V-560, JASCO) for 10 s at 286 nm at 25 °C. The relative activity of CHRNase A was calculated by the following formula

Enzyme Kinetic Constants of CHRNase A

RNase A and CHRNase A were prepared similarly as for measuring the relative activity, but the cCMP substrate solution was formulated to have a concentration range of 0.02–0.1 mg/mL. Enzyme solution (0.3 mL, 3 h post-preparation) was added to 2.7 mL of the substrate solution, and the change in absorbance was recorded as for the relative activity at 286 nm. The initial reaction rate for each RNase A was computed using the initial slope. Using a Lineweaver–Burk plot, the Michaelis constant (Km) and the maximum rate (Vmax) for each RNase A were calculated.

Effect of the Addition of β-CyD on the Enzyme Activity of CHRNase A

CHRNase A was dissolved in phosphate buffer (0.05 mol/L, pH 6.5) to achieve a concentration of 0.5 mg/mL (CHRNase A solution without β-CyD). β-CyD was dissolved in phosphate buffer (0.05 mol/L, pH 6.5) to a concentration of 0.2 × 10–3 mol/L; 0.1 mL of 0.2 × 10–3 mol/L β-CyD or the phosphate buffer (0.05 mol/L, pH 6.5) was added to 3.0 mL of CHRNase A solution (CHRNase A solution with or without β-CyD). cCMP was freshly dissolved in phosphate buffer (0.05 mol/L, pH 6.5) to achieve a concentration of 0.1 mg/mL in each measurement (substrate solution); 0.3 mL of CHRNase A solution after 24 h post-preparation was added to 2.7 mL of the substrate solution. The change in absorbance was analyzed using a UV spectrometer for 10 s at 286 nm at 25 °C. The relative activity of CHRNase A before and after the addition of β-CyD was calculated using the following formula

Evaluation of the Environment of Tyrosine Residues to DMSO

RNase A and CHRNase A were separately dissolved in DMSO at a concentration of 0.5 mg/mL. The solutions were analyzed using a fluorescence spectrometer with the following parameters: excitation wavelength = 280 nm, band pass = 10 nm, and emission band pass = 10 nm. Fluorescence emission spectra were recorded at 25 °C from 260 to 400 nm.