Potential effects of HP-β-CD on the release of IFN-α from tristearin implants

Therefore, the interactions of HP-β-CD with IFN-α and/or with PEG 6000 may affect the ability of PEG to precipitate IFN-α. In order to determine such effects the protein solubility in the presence of both excipients was investigated. However, the presence/absence of this co-lyophilisation agent did not alter the impact of PEG on the protein solubility (Figure 49).

This suggested that in-situ precipitation of IFN-α within PEG-containing tristearin implants occurs to the same extent regardless whether HP-β-CD is present or not.

0.00 0.50 1.00 1.50 2.00 2.50

0 5 10 15 20 25

PEG, % soluble fraction of IFN-α, mg/ml

without HP-ß-CD with HP-ß-CD

Figure 49: Effects of HP-β-CD on the precipitation of IFN-α in the presence of PEG (average +/- SD; n = 3).

Apart from the effect of cyclodextrins on solubility, the formation of inclusion complexes may modify the drug diffusivity and consequently the drug release from delivery systems [15]. De Rosa et al. applied Attenuated Total Reflection (ATR)-FTIR spectroscopy to investigate the influence of HP-β-CD incorporation on the properties of insulin-loaded PLGA microspheres prepared by spray drying. Upon addition of HP-β-CD, insulin embedded within microspheres revealed an extended content of β-sheet structures, which was ascribed to the formation of an insulin/CD complex within the microspheres. Since the mobility of this complex within the polymeric matrix was considered to be lower than the mobility of free protein, the overall release rate was slowed down by the addition of HP-β-CD [54].

In the case of tristearin implants the solubility of IFN-α was primarily affected by PEG (Figure 49). However, interactions between IFN-α and HP-β-CD might affect the diffusivity of IFN-α as well. Various techniques such as competitive spectroscopy

[101], isothermal titration calorimetry, and nuclear magnetic resonance [56] are proposed to gain insight into the interactions of cyclodextrins with peptides and proteins after dissolution. Furthermore, spectroscopic techniques like FTIR, circular dichroism, and fluorescence spectroscopy were shown to be suitable for characterising protein cyclodextrin interactions.

For instance, Chen et al. revealed that the interactions of HP-β-CD with the 40 kDa protein horseradish peroxidase were sufficient to increase the amount of α-helical structure detectable by FTIR spectroscopy [37].

The potential of steady state fluorescence for monitoring cyclodextrin-peptide interactions was outlined by Khajehpour et al. for melittin interacting with HP-β-CD. In the presence of HP-β-CD a blue shift of the fluorescence maxima and a slight increase in the intrinsic fluorescence emission maxima of melittin were observed.

The blue shift was ascribed to the reduction of the polarity around Trp. This might be a result of interactions with the cyclodextrin cavity or of an overall change of the peptide structure upon cyclodextrin addition [116].

In the course of this thesis fluorescence and FTIR spectroscopy were used to evaluate potential cyclodextrin-protein interactions. The protein to cyclodextrin mass ratio applied for lyophilisation and subsequent implant loading was 1 part protein to 3 parts cyclodextrin. This can be regarded as the highest protein to cyclodextrin ratio achievable within the implants, as HP-β-CD was found to leach out faster from the lipidic implant than IFN-α (see Figure 22). The following experiments were, therefore, consistently carried out with a protein to cyclodextrin mass ratio of 1 to 3.

Figure 50 A illustrates the vector-normalised spectra as well as the vector-normalised second derivative spectra of IFN-α in the presence and in the absence of HP-β-CD.

As it can be seen, no differences were detectable in the amid I region between both samples. Furthermore, the addition of HP-β-CD had no effect on the Trp fluorescence emission spectrum of IFN-α (Figure 50 B). In both spectra the emission maxima were detected at 336 nm with an identical intensity.

-0.3 -0.25 -0.2 -0.15 -0.1 -0.05 0 0.05 0.1 0.15 0.2

1600 1650

1700

wavenumber, cm-1

absorbance

with HP-ß-CD without HP-ß-CD

A

0 50 100 150 200

300 350 400 450

wavelenght, nm

fluorescence intensity, a.u.

with HP-ß-CD without HP-ß-CD

B

Figure 50: (A) Vector-normalised and second derivative transmission FTIR-spectra; (B) Trp fluorescence emission scans of IFN-α in the presence and in the absence of HP-β-CD (n=2).

FTIR and Fluorescence spectroscopy provided information on the secondary and on the tertiary protein structure, respectively. Since both, FTIR and fluorescence spectra of IFN-α were not affected by the presence of HP-β-CD it could be assumed that potential interactions with the cyclodextrin had no influence on the overall protein structure. Since such an alteration of the three-dimensional protein structure as well as the formation of stable complexes would affect the protein diffusivity, it can be supposed that the mobility of IFN-α within the water-filled pores of tristearin implants was not affected by the presence of HP-β-CD.

In order to prove this experimentally the diffusion coefficients of IFN-α in the presence and in the absence of HP-β-CD were determined by the open-end capillary technique introduced in Chapter IV.1.2. The improved experimental setup illustrated in Figure 20 was applied. For matters of comparison HP-β-CD was replaced by trehalose as alternative stabilising agent.

This experiment was supposed to provide information on the diffusivity of IFN-α within the water-filled implant pores. In addition to IFN-α and HP-β-CD some implants were loaded with PEG as porogen and precipitant. Thus, after water penetration, IFN-α, HP-β-CD, and PEG will be dissolved within the implants pores. Hence, the effects of HP-β-CD on the diffusion coefficient of IFN-α were determined additionally with and without PEG. In agreement to implant formulation comprising 10 % IFN-α/HP-β-CD lyophilisate and 10 % or 20 % PEG, the IFN-IFN-α/HP-β-CD co-lyophilisate to PEG ratio was 1 to 1 or 1 to 2, respectively.

In Figure 51 the protein concentrations determined within the capillary over time are depicted. Obviously, the release of IFN-α out of capillaries containing HP-β-CD or

trehalose was comparable (Figure 51 A). The same tendency was observed when PEG was present (Figure 51 B and C).

0 20 40 60 80 100

0 100 200 300

time, h relative IFN-α concentration within the capillary, %

HP-β-CD trehalose

A

0 20 40 60 80 100

0 100 200 300

time, h relative IFN-α concentration within the capillary, %

HP-β-CD trehalose

B

0 20 40 60 80 100

0 100 200 300

time, h relative IFN-α concentration within the capillary, %

HP-β-CD trehalose

C

Figure 51: Determination of the diffusion coefficient of IFN-α: decrease of IFN-α concentration within the capillary.

Open symbols represent the IFN-α concentration in the presence of trehalose and closed symbols represent the IFN-α concentration in the presence of HP-β-CD. Protein concentration within capillaries containing no PEG (A), lyophilisate:PEG blend in the ratio 1:1 (B), and lyophilisate:PEG ratio of 1:2 (C) (average +/- SD; n = 3, experimental setup according to Figure 20).

The diffusion coefficients of IFN-α were calculated based on the liberation rates (as described in Chapter IV.1.2). The determined diffusivities of IFN-α are listed in Table 7. As the diffusivities of IFN-α in samples containing HP-β-CD and in samples containing trehalose were in the same order of magnitude, it could be concluded that HP-β-CD neither affected the release of IFN-α from tristearin implants by a modulation of the protein solubility nor by an alteration of the protein diffusivity.

However, as a soluble excipient HP-β-CD will promote the creation of water-filled pores and in this regard HP-β-CD will surely contribute to the formation of water-filled pores allowing protein release from lipidic implants.

Moreover, unlike natural cyclodextrins, HP-β-CD is surface active [106]. Recently, Koennings et al. revealed that the incorporation of surface active compounds accounted for increased water imbedding and consequently for increased release rates from lipidic implants [124].

Table 7: Diffusion coefficients of IFN-α within phosphate-buffer filled capillaries comprising various amounts of PEG. IFN-α was lyophilised either with HP-β-CD or with trehalose.

The diffusion coefficients were determined by fitting Equation {13} to the release data shown in Figure 51.

lyophilisate:PEG ratio D, x 10^-6 cm²/s

HP-β-CD 1:0 1:1 1:2

1.80 1.77 1.71

trehalose 1:0 1:1 1:2

1.48 1.29 1.72