Application of circular dichroism and magnetic circular dichroism for assessing biopharmaceuticals formulations photo-stability and small ligands binding properties

International Journal of Pharmaceutics 480 (2015) 84–91

Contents lists available at ScienceDirect

International Journal of Pharmaceutics journal homepage: www.elsevier.com/locate/ijpharm

Application of circular dichroism and magnetic circular dichroism for assessing biopharmaceuticals formulations photo-stability and small ligands binding properties Edoardo Longo, Rohanah Hussain * , Giuliano Siligardi ** Diamond Light Source Ltd., Diamond House, Harwell Science and Innovation Campus, Didcot, Oxfordshire OX11 0DE, United Kingdom

A R T I C L E I N F O

A B S T R A C T

Article history: Received 17 September 2014 Accepted 13 January 2015 Available online 14 January 2015

Synchrotron radiation circular dichroism (SRCD) is a powerful tool for photo-stability assessment of proteins. Recently our research has been interested in applying SRCD to develop screening methodologies for accelerated photo-stability assessment of monoclonal antibody formulations. Despite it was proven to be reliable and applicable within a wide range of salts and excipients containing solutions, the presence of far-UV (<260 nm) strong absorbing species (e.g., sodium chloride, histidine, arginine) in common formulations completely prevent the analysis. Herein, we propose a new method based on CD coupled with magnetic CD (MCD) to address the problem and offer an additional versatile tool for monitoring the photo-stability. This is done by assessing the stability of the samples by looking at the near-UV band, as well as giving insights in the denaturation mechanism. We applied this method to four mAbs formulations and correlated the results with dynamic light scattering data. Finally, we applied MCD in ligand interaction to key proteins such as lysozyme, comparing the human with the hen enzyme in the binding of N,N0 ,N0 0 -triacetylchitotriose. ã 2015 The Authors. Published by Elsevier B.V. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).

Keywords: Biopharmaceuticals Drug development Light stability assessment Ligand binding Magnetic circular dichroism Circular dichroism

1. Introduction Among the factors affecting protein stability, light exposure is a major issue for bio-pharmaceutics development. Light exposure leads to denaturation and degradation and eventually loss of activity. The optimisation of a methodology of stability assessment including fast and reproducible photo-stability is therefore a requirement. Light sources with constant and controlled photonflux and with precise and tuneable emission spectrum would be required to ensure reproducibility and reliability. In these respects, the high photon-flux available at synchrotron radiation circular dichroism B23 beamline (Diamond Light Source Ltd., Didcot, Oxfordshire) represents a resourceful innovation for bio-formulation stability assessment. Moreover, circular dichroism (CD) is able to give direct information on secondary structures which exhibits characteristic signatures in the farUV (below 260 nm) (Greenfield, 2006; Kelly et al., 2005). However, the farUV range is sometime un-

* Corresponding author. Tel.: +44 1235 778524. ** Corresponding author. Tel.: +44 1235 778425. E-mail addresses: [email protected] (R. Hussain), [email protected] (G. Siligardi).

accessible, particularly in presence of high salt (e.g., chloride) concentration and excipients, such as guanidinium derivatives (e.g., arginine) and histidine with high absorption in the far-UV (<250 nm). Nonetheless, proteins present other absorptions that could account for changes in the structure. The 260–320 nm CD bands due to aromatic side-chains (mostly tryptophans and tyrosines) and disulphide bonds are exploited to assess changes occurring in the tertiary structure due to either the presence or absence of rotational freedom of the aromatic side-chain causing decreased or increasing the optical activity respectively. Both can contribute to the overall intensity of the CD spectrum. Complementary to circular dichroism, magnetic circular dichroism (MCD) is a related technique where the optical transitions are induced by an external magnetic field (Faraday effect). MCD has been widely applied in the study of porphyrines, (Kalman et al., 1973; Kobayashi and Nakai, 2007) paramagnetic metal complexes, (Barrett et al., 1986; Modine et al., 1971) metal nano-particles, (Zaitoun et al., 2001) other biological relevant organics, (Voelter et al., 1968) indole derivatives, (Albinsson et al., 1989; Sprinkel et al., 1975) haem, (Brittain et al., 1978; Pond et al., 1999a,b,c,c) and proteins (Barth et al., 1971, 1972b; McFarland and Coleman, 1972). For proteins in particular, tryptophan residue has been widely targeted

http://dx.doi.org/10.1016/j.ijpharm.2015.01.026 0378-5173/ã 2015 The Authors. Published by Elsevier B.V. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).

E. Longo et al. / International Journal of Pharmaceutics 480 (2015) 84–91

triacetylchitotriose was purchased by Sigma (90,251 Fluka). All measurements were recorded in water and concentrations measured either with Bradford assay and absorbance measurement (e280 nm = 150,000, 36,000, 35,700, 43,824, 13,980, 210,000, 35,034, 29,642, 94,460 mol1 L cm1 respectively) employing an IMPLEN nanophotometer P330. Salts for excipient preparation were bought from Sigma. Tween1 20 was bought from Thermo Scientific.

for MCD investigations. Notably, the 293 nm positive contribution in the near-UV region is mainly due to tryptophan contributions, with negligible opposite contributions due to the 275 nm tyrosine band, and it has been shown to display low dependency on the tertiary structure (Barth et al., 1971). Only above 15/1 ratio between tyrosine and tryptophan the 293 nm peak is distorted by a negative contribution from tyrosine (Barth et al., 1972b). However, corrections can be taken into account since all these contributions are linearly combined. The possibility to detect tryptophan in proteins independently by conformations can be applied using MCD for the evaluation of the overall protein concentration as reported by Barth et al. (1971, 1972a,b),),). A similar aim for MCD has been already employed with uro-porphyrines (Kalman et al., 1973). Here, we demonstrate the application of MCD to assess proteins photo-stability of cetuximab, a monoclonal antibodies, in combination with CD in the near-UV giving useful information on the behavior of bio-pharmaceutical upon light exposure. First, we will apply MCD for determining the tryptophan content and exposure to solvent, applying it on several commercialised model proteins, such as hen egg lysozyme, human serum albumin, bovine serum albumin, myoglobin and human serum immunoglobulin G, concanavalin A, a-lactalbumin and avidin, in comparison with the effects of denaturants. Then, we will apply it to the detection of the effects of light exposure, comparing the MCD results with the CD and dynamic light scattering. Finally, we will show how MCD could complement CD measurements to account for binding interactions between a protein and ligands. In particular, we chose the binding of hen lysozyme to N,N0 ,N00 -triacetylchitotriose (NAG), an inhibitor of enzymatic activity towards the cleavage of glycosidic bonds.

2.2. Bradford assay The Bradford test (BT) was performed on a Perkin Elmer Lambda 950 UV/VIS spectrometer equipped with a PTP 1 + 1 Peltier system with bovine serum albumin (BSA) and bovine g-globulin (BGG) standards purchased from Fisher Scientific. Coomassie Blue G-250 1X solution was purchased from Fisher Scientific and was added to each standard of required concentration, mixing vigorously, equilibrating for 5–10 min at room temperature, centrifuging and reading the absorbance at 595 nm (1 cm plastic cell, 25  C). Each measurement was repeated in quadruplicate with a 1 cm disposable plastic cell on four different solutions. All the spectra were subtracted from the baseline (Coomassie 1X at the same concentration in water). The results were compared with an absorbance reading at 280 nm. 2.3. CD and MCD Magnetic circular dichroism measurements were performed on the OLIS module B unit at B23 beamline at Diamond Light Source, Harwell Research and Innovation Campus, Didcot, Oxfordshire, UK. For UV-denatured samples the CD spectra were collected on an applied photophysics Chirascan Plus spectropolarimeter equipped with a fluorescence detector. The module B OLIS instrument for CD/ MCD was provided with a sample chamber room able to support a 1.4 T OLIS magnet. The spectra for model proteins were recorded in the 340–260 nm wavelength range, with a 1 cm quartz cell. For photo-denaturation samples, the spectra were collected in a 1 cm/ 100 mL cell. A 1 mm or 0.5 mm slit were employed with 1 sec of integration time. MCD spectra were recorded in both field directions (N/S and S/N), subtracted (S/N–N/S) ensuring a positive value at 293 nm. Spectra are reported un-corrected for the field magnitude.

2. Materials and methods 2.1. Samples preparation

IgG in H2O IgG in Guanidinium6M IgG in Urea 8 M

294 nm

2.4. Sample irradiation Controlled photo-denaturation of samples were performed with a Peqlab box, equipped with five 60 W lamps emitting at 254 nm. The mAbs were irradiated at 0.7 mg.mL1 concentration solution (concentration measured with Bradford test and

(a)

292.5 nm

Ellipticity (mdeg)

Ellipticity (mdeg)

5 mL of a solution of cetuximab (commercial name Erbitux) has been dialyzed using Pur-A-LyzerTM G2 (Sigma–Aldrich) dialysis tubes for 24 h in water. The solution was then concentrated using Vivaspin 500, 3–5 kDa MWCO (GE) in a centrifuge (21,000 rpm, 15 min). The concentrated solutions (about 40–50 mg mL1) were then diluted in the chosen buffers to a concentration of 1– 1.5 mg mL1. Hen egg lysozyme (Sigma L6876), human serum albumin (HSA, Sigma A3782), bovine serum albumin (BSA, Sigma A9085), equine skeletal muscle myoglobin (Sigma M0630) and human serum immunoglobulin G (IgG – Sigma I4506), concanavalin A (Type VI, Sigma L-7647), a-lactalbumin (Type III, Sigma L6010) and avidin (Sigma A-6128) and human lysozyme (Sigma L1667) were stored at 20  C before use. N,N0 ,N00 -

12 10 8 6 4 2 0 -2 -4 -6 -8 -10 -12 -14

260 265 270 275 280 285 290 295 300 305

8 6 4 2 0 -2 -4 -6 -8 -10 -12 -14

BSA in H2O

Fig. 1. (a) MCD of IgG 1 mg mL

293 nm

BSA in Guanidinium6M BSA in Urea 8 M

(b)

292.5 nm

260 265 270 275 280 285 290 295 300 305

Wavelength (nm) 1

85

Wavelength (nm) 1

water, guanidinium 6 M or urea 8 M. (b) MCD of BSA 2.4 mg mL

water, guanidinium 6 M or urea 8 M. 1 cm cell, 4 scans. Spectra divided by 2.

86

E. Longo et al. / International Journal of Pharmaceutics 480 (2015) 84–91 Table 1 lmax for the MCD of selected proteins and conditions.

20

lmax MCD Ellipticity (mdeg)

293 293 293 292–293 292–293

Lysozyme Myoglobin BSA BSA (urea 8 M) BSA (guanidinium.HCl 6 M) HSA IgG IgG (urea 8 M) IgG (guanidinium.HCl 6 M)

293 294 292 292

10

HSA BSA Myoglobin Lysozyme IgG Concanavalin A α-Lactalbumin Avidin

0

-10 280

290

300

310

Wavelength (nm)

absorbance at 280 nm) in a 0.5 mm path length quartz cell, and the liquid exposed an overall surface of 3 cm2. 2.5. Dynamic light scattering DLS measurements have been performed on a Malvern Zetasizer Nano. The measurements have been performed on 0.68 mg mL1 samples with plastic cuvette, 50 mL of solution. The measurements size quality reports were always reported as good and therefore the evaluation of hydrodynamic radius acceptable. 3. Results and discussion 3.1. MCD spectra of proteins in the near-UV. Dependence on tryptophan exposure and measuring of tryptophan content As introduced earlier, MCD has widely been applied to the study of several systems, such as porphyrins, haems, metal complexes and proteins. Tryptophan, endowed with a MCD active probe (indole), has constituted a suitable target for MCD study of proteins. To prove the reliability of this residue as a probe, several proteins of known sequence were selected and their concentrations determined by measuring the MCD. MCD protein concentration evaluation relies on tryptophan residue in proteins. Water exposure or electron withdrawal moieties can affect the peak position and intensity, inducing blue-shift and reduce the spectral intensities. For example, we measured the effect on the MCD of IgG and BSA in two denaturant, urea and guanidinium (Fig. 1) widely used in protein biopharmaceutical expression and purification conditions.

Fig. 2. MCD spectra of some model proteins and peptides. Measurements recorded at 23  C, 1 cm path length, 4 scans with 0.5 nm increment and 1 s integration time. Concentrations evaluated by simple dilution of the commercial proteins. Spectra divided by 2.

The denaturants affect the MCD spectra with an overall loss of intensity and blue-shifting. Nonetheless, the spectra of the denatured samples are only slightly affected by the nature of the denaturants, with a major effect on the tyrosine band at 275 nm. The near-UV CD spectrum of BSA in presence of denaturants is also given for comparison (SI, Figure 1). The MCD spectrum is more informative, specifically in the 293 nm region. In the same conditions the MCD spectra are more intense and less noisy than CD, increasing the chance of application of MCD for analytical detection of tryptophan in protein even a lower concentrations. Moreover, for MCD the loss of conformation has a smaller influence on the spectra whereas the change is probably due mostly to the exposure of the tryptophan to the solvent only. Briefly, the lmax values for MCD of some of the investigated proteins are reported in Table 1. All the values are close to 293 nm for all the proteins under investigation, except for IgG that is even red-shifted to 294 nm. For more exposed tryptophan (with denaturants) the value shifts towards 290 nm, as had been found for N-Ac-(L)-Trp-OMe, which can be viewed as a completely water-exposed indole (Barth et al., 1971). Noticeably, all the proteins investigated here and in previous works show a lmax at 293–294 nm, suggesting that tertiary structure has little or even negligible influence on the spectrum. This is in agreement with most of the known protein sequences (Barth et al., 1972b; McFarland and Coleman, 1972; Vitale et al., 2000). To validate the method, several proteins such as hen egg lysozyme, human serum albumin, bovine serum albumin,

Table 2 MCD evaluation of Trp contents and concentrations.

Lysozyme BSA HSA Myoglobin IgG Concanavalin A a-Lactalbumin Avidin a b c d

Theoratical conc. (mg mL1)

Abs evaluated conc. (mg mL1)

Bradford assay (mg mL1)c

0.40 1.95c 3.00 0.98 0.50 0.60 0.40 0.60

0.34 1.27 1.53 0.77 0.44 0.53 0.36 0.54

0.11 1.95 2.42 0.87 0.22 0.10 0.26 0.37

Correction factor applied (see SI). Measured from a 1:4 based diluted solutions. Measured with the BT. Maximum at 292.5 nm.

       

0.02 0.08b 0.01b 0.09 0.04 0.03 0.09 0.12

u293 nm (mdeg)

Evaluated number of Trpa

MCD conc. (mg mL1) (A)

18.20 6.46 2.49 9.73 6.09 8.46d 10.1 14.85

5.10 2 0.86 2.74 2.11 2.40 2.85 4.15

0.36 1.95 1.67 0.69 1.55 0.45 0.40 0.49

E. Longo et al. / International Journal of Pharmaceutics 480 (2015) 84–91

myoglobin and human immunoglobulin G was tested (Fig. 2). The concentrations employed were chosen to provide a UV absorbance close to 0.8 at the maximum of absorbance wavelength (usually 285–290 nm). All the main positive peaks are located at 293–294 nm, with the maximum red-shift for IgG. Conversely, concanavalin A shows a marked blue-shift, with the main peak position at 292 nm. In fact, the presence of exposed tryptophan residues in concanavalin A had already been reported (Mandal and Mandal, 2011; Pelley and Horowitz, 1976). In Table 2, tryptophan content was estimated

following a previously reported method (Barth et al., 1972a,b,b). Nonetheless, since the exposure of tryptophan to solvent can bias considerably the 293 nm peak position and intensity (a lmax of 290 nm is reported for N-Ac-(L)-Trp-OMe), we employed BSA as a reference sample, allowing also a further cross-checking with the Bradford test. As a correction for the negative contributions given by high tyrosine/tryptophan ratios, we corrected the evaluated tryptophan content as reported by Barth et al. (1972b). They evaluated a change of the 1% at 293 nm for a 1:1 ratio tyrosine/ tryptophan, and 40% for a 35/1 ratio. They also reported a

(b)

(a)

60

60

Ellipcity (mdeg)

40 20

0 0.5 2 4

0 -20 -40

MCD Cetuximab 0.68 mg/mL in His 40 Ellipcity (mdeg)

MCD Cetuximab 0.68 mg/mL in Acetate/PS20

20

0 0.5 2 4

0 -20 -40 -60

-60 -80 260

270

280

290

300

310

-80 260

320

(c)

(d)

40 20

0 0.5 2 4

0 -20 -40 -60 -80 260

300

60

MCD Cetuximab 0.68 mg/mL

20 0 -20 0 0.5 2 4

-40 -60

280

300

320

40 in Acetate/PS20/Sorbitol Ellipcity (mdeg)

60 MCD Cetuximab 0.68 mg/mL in PBS/Arg

280

Wavelength (nm)

Wavelength (nm)

Ellipcity (mdeg)

87

320

-80 260

Wavelength (nm)

280

300

320

Wavelength (nm)

(e)

MCD @ 294nm (mdeg)

30 20 10 0

-10

Acetate/PS20 Hisdine Phosphate/Arginine Acetate/PS20/Sorbitol

0 1 2 3 4 Energy Irradiated (Joule) @ 254nm Fig. 3. (a) Near-UV MCD spectrum of cetuximab 0.68 mg/mL in acetate buffer 20 mM pH 5.0 with 0.1% Tween 20; (b) near-UV MCD spectrum of Cetuximab 0.68 mg/mL in His 20 mM pH 6; (c) near-UV MCD spectrum of cetuximab 0.68 mg/mL in PBS 20 mM pH 6.0 with Arginine 200 mM. 4 scans averaged, 1 mm slit, 1 nm/s scan rate. 7 points-2nd polynomial order SG smoothing applied. (e) Plot of decay rates monitored at 294 m. No fitting applied.

E. Longo et al. / International Journal of Pharmaceutics 480 (2015) 84–91

correction factor dependent on the ratio between the intensities at 275 and 293 nm. The correction curve is reported in SI and it is explained thoroughly. Other contributions in this region (as for example the Myoglobin heme) prevent only few cases from a wide application of this method for concentration determination, in particular when neither the Bradford test nor the absorbance can be employed. We firstly applied it to correct the tryptophan content for BSA (|u293nm|/|u275 nm|). The full calculation along with the Bradford Test results is reported in SI. The calculation correlates the ellipticity at 293 nm with the protein concentration through a simple relation reported in SI. The application of this method implies that (i) the tryptophan contributions are simply additive, (ii) the average single tryptophan contribution for different folded proteins is the same, (iii) the BSA concentration has to be evaluated by other accurate mean (Bradford Test in our case). The first assumption is meant for solvent unexposed tryptophan residues. This is easily checked for our model proteins, since all have a lmax above 293 nm (in some cases even above 295 nm). The presence of strong denaturants that would expose the aromatic residues to solvent would shift the band down to 292 nm as showed before. The second assumption takes into account that the tryptophan contributions are the same for folded proteins since they are not affected by the conformation. The only correction made to the model was to introduce the correction based on the tyrosine/tryptophan ratio. Finally, we took the BSA ellipticity weighted for each tryptophan contribution and concentration (calculated for BSA only considering the Bradford assay measure for concentration) as a reference. Otherwise, for all the other proteins, the absorbance evaluated concentration is in

3 2

CD at 294nm (medg)

88

1 0 A/T H P/A A/T/S

-1 -2 0

1 2 3 4 Energy Irradiated (Joule) @ 254nm

Fig. 4. CD plots of decay rates monitored at 294 nm.

good agreement with the MCD evaluation, so confirming the goodness of the Bradford test. A discrepancy of the Bradford Test with respect to the absorbance and MCD evaluation was found for concanavalin A. Similarly, disharmony was found in the results for IgG and myoglobin. This may be due to the underestimation of the tryptophan content in the case of IgG and contribution of the haem absorptions at 293 nm for myoglobin. The results, even restricted to few model proteins, are sufficient to envisage an expansion of MCD application to other protein systems. MCD could provide a complementary spectroscopic method to cross-check protein sample concentrations, in particular when other methods are un-applicable. It is material saving and does not require any particular sample preparation (as with the Bradford Test),

Fig. 5. MCD titration at increasing volumes on hen egg lysozyme (0.4 mg/mL) (left) and human lysozyme (0.4 mg/mL) (right). Titration points referred to equivalents of N,N0 , N0 0 -triacetylchetotriose (stock concentration 1.054 mg/mL) added to solution. 8 scans, 0.5 nm increments, 0.5 nm/s, 0.5 nm bandwidth. Inset: expansion of the 293–294 nm main positive peaks. (c) First derivative spectra of 5a, d) first derivative spectra of 5b. Spectra divided by 2 and corrected for dilution factor.

E. Longo et al. / International Journal of Pharmaceutics 480 (2015) 84–91

requiring only to detect by other means the concentration of a reference standard (BSA in our case) such as the standards provided for Bradford Test commercial kit. 3.2. Monitoring the photo-denaturation by MCD/CD comparison for three mAbs formulations The high photon-flux available at B23 beamline (Diamond Light Source Ltd., Didcot, Oxfordshire) allows direct measurement of proteins photo-denaturation by measuring synchrotron radiation circular dichroism (SRCD) spectra (Hussain et al., 2012a,b,b). The high photon-flux of the light beam generated by the synchrotron can be exploited to induce photo-denaturation of proteins and at the same time employed for spectroscopic analysis, such as recording CD spectra, allowing a versatile tool to monitor the change in secondary structure associated with the denaturation. Even though far-UV SRCD has been the major tool to assess photostability, in some cases this technique is not applicable (e.g., high salt concentration in the buffer). Moreover, the low dependence of the MCD upon small conformational changes may result useful to discriminate the protein unfolding due to back-bone denaturation vs photo-degradation of the aromatic content. We selected cetuximab, an IgG1 monoclonal antibody approved by FDA for treatment of some cancers such as colorectal cancer, head and neck cancer, in four different formulations, monitoring the effect of irradiation with MCD. We studied two formulations in particular that contains amino acids (as excipients) that absorbs light below 250 nm (Fig. 3b and c), preventing secondary structure analysis. Other formulations are transparent in the far-UV. Formulation 5a (Acetate 20 mM with PS20 0.1%) shows marked instability in the denaturation experiments. Conversely, the fourth formulation is the most stabilising for cetuximab. So, the comparison among the four formulations may lead to a comparative tool with far-UV denaturation and give more insights in photo-stability. In Fig. 3 the

89

MCD spectra for cetuximab in the four studied formulations have been reported. As expected, the four formulations display different denaturation rates. In particular, the fourth (Acetate/PS20/Sorbitol) shows slower degradation as already observed in the far-UV CD spectra. Notably, the overall higher stability of the fourth formulation is present in both the two major contributions (tryptophan 294 nm band and tyrosine 275 nm band). The fourth and the first formulations share many constituents except sorbitol which could induce stabilisation to cetuximab against the effect of light. Its role in preventing denaturation effects for biopharmaceuticals and biological materials in general had been already reported (Chang et al., 2005; Gonzalez et al., 1995), but it was not previously reported in particular for photo-stabilisation. However, aggregation/precipitation phenomena associated with the denaturation cannot be ruled out and the effect of the sorbitol may be preventing these phenomena. Dynamic light scattering (DLS) suggested that bigger but still stable aggregates are formed, as indicated by the increased hydrodynamic radius but very low polydispersity at 6 J irradiation (SI, Table 4–7). This may suggest that the antibodies tend to generate stable agglomerates, possibly in order to prevent the hydrophobic cores to be exposed, as confirmed lmax at 294 nm which does not change even after a long exposure. Sorbitol may play a major role in driving this process, since it is the only excipient present in the two most stable formulations, but this hypothesis will require to be tested. Parallel to MCD we monitored the decaying trends via near-UV CD measurements in the same range and in the same conditions employed for MCD. The near-UV CD decaying plot are reported in Fig. 4. Almost all the considerations made for the MCD decay plot can be observed here. However, the CD in the near-UV arises from the local tertiary structure surrounding the chromophores (tryptophans and tyrosines). Notably, the CD intensity at 294 nm for the

Fig. 6. CD titration experiment of 1.5 mL (upper-left). Hen egg lysozyme 0.4 mg mL1 in water (upper-right). Human lysozyme 0.4 mg mL1 in water, with NAG 1.054 mg mL1. Measurements corrected for the dilution factor and subtracted of the buffer/ligand contributions. CD titration plot expressed in DA (lower-left). Hen egg lysozyme 0.4 mg mL1 in water (lower-right). Human lysozyme 0.4 mg mL1 in water, with NAG 1.054 mg mL1. The data were analysed with CDApps software (Siligardi et al., 2002).

90

E. Longo et al. / International Journal of Pharmaceutics 480 (2015) 84–91

0.25

Fluorescence

0.20

0.15

0.10

Fluorescence

0.15

Hen Egg Lysozyme/NAG 1:0 1:0.5 1:1 1:2

Human Lysozyme/NAG 1:0 1:0.5 1:1 1:2

0.10

0.05

0.05

0.00

0.00 300

320

340

360

380

400

420

300

320

Wavelength (nm)

340

360

380

400

420

Wavelength (nm)

Fig. 7. Fluorescence titration experiment of 1.5 mL (left). Hen Egg Lysozyme 0.4 mg mL1 in water (right). Human Lysozyme 0.4 mg mL1 in water, with NAG 1.054 mg mL1. Measurements corrected for the dilution factor. lexc = 280 nm, Dlexc = 8 nm, 1 scans, 1 nm/s. Data are un-calibrated for external references.

Acetate/PS20 formulation initially increases slightly in consequence of the irradiation, whereas it decreases for the other two (histidine and phosphate–arginine formulations). This fact can be explained by an increased initial adjustment in the structure induced by the irradiation, followed by a loss of secondary structure that could be regarded as the reason for the further aggregation observed in the DLS. 3.3. MCD of Lysozyme titration with N,N0 ,N00 -triacetylchitotriose NAG is an inhibitor of the glycolysis activity of lysozyme. The interaction of the binder to the enzyme causes structural modifications to the protein and deactivation (Cheetham et al., 1992; Mine et al., 2000). MCD measurements have been exploited to determine whether the interaction of lysozyme with NAG inhibitor could induce any change to the MCD spectra. The results are reported in Fig. 5. The absence of dramatic changes in the MCD indicates that the binding does not affect the exposure to buffer of the aromatics involved. Conversely, the great change in the CD (Fig. 6) accounts for tertiary structure change caused by the binding that has already been described (Blake et al., 1967; Mine et al., 1999). It is relevant to note that the binding does not affect peak intensity of the 293 nm MCD band, confirming that conformational changes alone do not modify the MCD of buried residues. For the titration of hen egg lysozyme with NAG, the main positive peak shows a neat shift from 293.3 to 293.7 nm. This trend is absent with human lysozyme, apart from an initial red-shift upon addition of the first amount of ligand. The lmax shifting for hen egg lysozyme indicates that the binding to the ligand increases the shielding of the aromatic from the solvent. The curves for 1:1 and 1:2 ratio perfectly overlaps, suggesting that all the binding sites are already bound to NAG and the addition of ligand does not induce any further change. The discrepancies between the two spectra account mainly for the different amino acid composition in the binding site. In fact, NAG binds to a tryptophan in the hen egg lysozyme and to a tyrosine in the human lysozyme (Mulvey et al., 1973, 1974). The fluorescence and CD titration plot and spectra (Fig. 6 and Fig. 7) are reported. The evaluated dissociation constant for hen egg lysozyme is 5 mM. For human lysozyme the evaluated kd is 46.8 mM. The binding of NAG to human lysozyme does not involve a tryptophan (tryptophan-62) such as for hen lysozyme, as indicated by the different change in the CD and the fluorescence spectra. In fact, in the human version the site is point mutated with a tyrosine residue. In the CD the main changes are present at 293 nm for hen lysozyme and 275 nm for human lysozyme, in agreement with the

different residue that they bind. In the case of the fluorescence, the binding to the hen lysozyme induces a peak blue-shift and increasing intensity that is compatible with the binding of a tryptophan and an increasing hydrophobicity (Mulvey et al., 1973). Conversely, in the case of the human lysozyme there is only a raising of the spectrum intensity, probably due to the conformational rearrangement. This approach would eventually allow selective investigations of protein–ligands interactions involving aromatic residues in the binding. 4. Conclusions We showed a coupled CD/MCD application for the investigation of biopharmaceuticals formulation photo-stability and binding interactions. MCD had previously been as an alternative method for concentration assessment as demonstrated showing independence to conformational arrangement. This allows an accurate evaluation of tryptophan content and concentration from evaluation of 294 nm tryptophan contribution. The method showed few discrepancies from the classic absorption and Bradford Test concentration evaluation. MCD can be a useful supplementary tool for assessing the concentration of protein samples with highly absorbing additives or excipients that are not suitable for the Bradford Test analysis. Furthermore, MCD is useful in UV denaturation rate assessment or photostability pertinent in biopharmaceutical characterisation, to assess protein stability such as mAbs formulations. This allows correlation of effects of different buffers/excipients to changes in MCD with the aggregation effects supported by DLS which can further be extended to correlate with immune-activity assay and other aggregation assays. MCD can also be employed in discriminating aromatic residues involve in ligand binding as shown in aromatic residue preference of NAG to human and hen lysozyme. For human lysozyme the change is mainly located in the tyrosine band (275 nm) and no neat shift is observed for the 293 peak. For hen lysozyme, a 1 nm red shift in the maximum peak position tryptophan (293–294 nm) is an indication that the ligand binds the residue. In summary, MCD coupled with CD and fluorescence or other biophysical techniques can be a useful tool in biopharmaceutical characterisation. Acknowledgment The authors thank Dr. Renato Schiesari (University of Padova – Italy) for the kind support in the acquisition of MCD spectra of lysozyme.

E. Longo et al. / International Journal of Pharmaceutics 480 (2015) 84–91

Appendix A. Supplementary data Supplementary data associated with this article can be found, in the online version, at http://dx.doi.org/10.1016/j. ijpharm.2015.01.026. References Albinsson, B., Kubista, M., Norden, B., Thulstrup, E.W., 1989. Near-ultraviolet electronic transitions of the tryptophan chromophore: linear dichroism, fluorescence anisotropy, and magnetic circular dichroism spectra of some indole derivatives. J. Phys. Chem. 93, 6646–6654. Barrett, C.P., Peterson, J., Greenwood, C., Thomson, A.J., 1986. Optical detection of paramagnetic resonance by magnetic circular dichroism: applications to aqueous solutions of met alloproteins. J. Am. Chem. Soc. 108, 3170–3177. Barth, G., Records, R., Bunnenberg, E., Djerassi, C., Voelter, W., 1971. Magnetic circular dichroism studies: XII. The determination of tryptophan in proteins. J. Am. Chem. Soc. 93, 2545–2547. Barth, G., Bunnenberg, E., Djerassi, C., 1972a. Magnetic circular dichroism studies. XIX. Determination of the tyrosine: tryptophan ratio in proteins. Anal. Biochem. 48, 471–479. Barth, G., Voelter, W., Bunnenberg, E., Djerassi, C., 1972b. Magnetic circular dichroism studies: XVII. Magnetic circular dichroism spectra of proteins. A new method for the quantitative determination of tryptophan. J. Am. Chem. Soc. 94, 1293–1298. Blake, C.C., Mair, G.A., North, A.C., Phillips, D.C., Sarma, V.R., 1967. On the conformation of the hen egg-white lysozyme molecule. Proceedings of the Royal Society of London. Series B, Containing papers of a Biological character. Royal Society 167, 365–367. Brittain, T., Greenwood, C., Springall, J.P., Thomson, A.J., 1978. Magnetic-circulardichroism studies of haem a and its derivatives. Biochem. J. 173, 411–417. Chang, L.L., Shepherd, D., Sun, J., Tang, X.C., Pikal, M.J., 2005. Effect of sorbitol and residual moisture on the stability of lyophilized antibodies: implications for the mechanism of protein stabilization in the solid state. J. Pharm. Sci. 94, 1445–1455. Cheetham, J.C., Artymiuk, P.J., Phillips, D.C., 1992. Refinement of an enzyme complex with inhibitor bound at partial occupancy. Hen egg-white lysozyme and tri-Nacetylchitotriose at 1.75A resolution. J. Mol. Biol. 224, 613–628. Gonzalez, M., Murature, D.A., Fidelio, G.D., 1995. Thermal stability of human immunoglobulins with sorbitol. A critical evaluation. Vox Sang. 68, 1–4. Greenfield, N.J., 2006. Using circular dichroism spectra to estimate protein secondary structure. Nat. Protoc. 1, 2876–2890. Hussain, R., Javorfi, T., Siligardi, G., 2012a. 8.23 Spectroscopic analysis: synchrotron radiation circular dichroism. Comprehensive Chirality 438–448. Hussain, R., Javorfi, T., Siligardi, G., 2012b. Circular dichroism beamline B23 at the diamond light source. J. Synchrotron Radiat. 19, 132–135. Kalman, S.M., Barth, G., Linder, R.E., Bunnenberg, E., Djerassi, C., 1973. Magnetic circular dichroism spectroscopy: a definitive method for the determination of urinary porphyrins. Anal. Biochem. 52, 83–95. Kelly, S.M., Jess, T.J., Price, N.C., 2005. How to study proteins by circular dichroism. Biochim. Biophys. Acta 1751, 119–139.

91

Kobayashi, N., Nakai, K., 2007. Applications of magnetic circular dichroism spectroscopy to porphyrins and phthalocyanines. Chem. Commun. (Camb.) 4077–4092. Mandal, P., Mandal, D.K., 2011. Localization and environment of tryptophans in different structural states of concanavalin A. J. Fluoresc. 21, 2123–2132. McFarland, T.M., Coleman, J.E., 1972. Magnetic circular dichroism of tryptophanyl residues in proteins. Azurin and carbonic anhydrases. Eur. J. Biochem./FEBS 29, 521–527. Mine, S., Tate, S., Ueda, T., Kainosho, M., Imoto, T., 1999. Analysis of the relationship between enzyme activity and its internal motion using nuclear magnetic resonance: 15 N relaxation studies of wild-type and mutant lysozyme. J. Mol. Bio. 286, 1547–1565. Mine, S., Ueda, T., Hashimoto, Y., Imoto, T., 2000. Analysis of the internal motion of free and ligand-bound human lysozyme by use of 15 N NMR relaxation measurement: a comparison with those of hen lysozyme. Protein Sci. 9, 1669–1684. Modine, F.A., Izen, E.H., Kemp, J.C., 1971. The magnetic circular dichroism of bands associated with the F+ center in SrO. Phys. Lett. A 34, 413–414. Mulvey, R.S., Gaultieri, R.J., Beychok, S., 1973. Spectral properties of human lysozyme and its inhibitor complexes: fluorescence and difference spectra. Biochemistry 12, 2683–2690. Mulvey, R.S., Gualtieri, R.J., Beychok, S., 1974. Composition, fluorescence, and circular dichroism of rat lysozyme. Biochemistry 13, 782–787. Pelley, R., Horowitz, P., 1976. Fluorimetric studies of tryptophyl exposure in concanavalin A. Biochim. Biophys. Acta 427, 359–363. Pond, A.E., Roach, M.P., Sono, M., Rux, A.H., Franzen, S., Hu, R., Thomas, M.R., Wilks, A., Dou, Y., Ikeda-Saito, M., Ortiz de Montellano, P.R., Woodruff, W.H., Boxer, S.G., Dawson, J.H., 1999a. Assignment of the heme axial ligand(s) for the ferric myoglobin (H93G) and heme oxygenase (H25A) cavity mutants as oxygen donors using magnetic circular dichroism. Biochemistry 38, 7601–7608. Pond, A.E., Sono, M., Elenkova, E.A., Goodin, D.B., English, A.M., Dawson, J.H., 1999b. Influence of protein environment on magnetic circular dichroism spectral properties of ferric and ferrous ligand complexes of yeast cytochrome c peroxidase. Biospectroscopy 5, S42–52. Pond, A.E., Sono, M., Elenkova, E.A., McRee, D.E., Goodin, D.B., English, A.M., Dawson, J.H., 1999c. Magnetic circular dichroism studies of the active site heme coordination sphere of exogenous ligand-free ferric cytochrome c peroxidase from yeast: effects of sample history and pH. J. Inorg. Biochem. 76, 165–174. Siligardi, G., Panaretou, B., Meyer, P., Singh, S., Woolfson, D.N., Piper, P.W., Pearl, L.H., Prodromou, C., 2002. Regulation of Hsp90 ATPase activity by the co-chaperone Cdc37p/p50cdc37. J. Biol. Chem. 277, 20151–20159. Sprinkel, F.M., Shillady, D.D., Strickland, R.W., 1975. Magnetic circular dichroism studies of indole, DL-tryptophan, and serotonin. J. Am. Chem. Soc. 97, 6653–6657. Vitale, D.J., Goldbeck, R.A., Kim-Shapiro, D.B., Esquerra, R.M., Parkhurst, L.J., Kliger, D. S., 2000. Near-ultraviolet magnetic circular dichroism spectroscopy of protein conformational states: correlation of tryptophan band position and intensity with hemoglobin allostery. Biochemistry 39, 7145–7152. Voelter, W., Records, R., Bunnenberg, E., Djerassi, C., 1968. Magnetic circular dichroism studies. VI. Investigation of some purines pyrimidines, and nucleosides. J. Am. Chem. Soc. 90, 6163–6170. Zaitoun, M.A., Mason, W.R., Lin, C.T., 2001. Magnetic circular dichroism spectra for colloidal gold nanoparticles in xerogels at 5.5 K. J. Phys. Chem. B 105, 6780–6784.