(566f) Mechanistic Insights into the Stabilization of Proteins By Saccharides

The drying of proteins might have detrimental impacts on their structure and function. Excipients stabilize proteins during drying mainly by the following mechanisms ¹: water replacement, glass dynamics, and interfacial competition. The first works by the saccharides replacing the hydrogen bonds to the protein. The second leads to reduced protein mobility within the glassy matrices, prohibiting aggregation. The third explains how surface-active excipients keep the protein away from the interface, thereby protecting it from denaturation.

This work aimed to develop a miniaturized screening platform easy to establish in every laboratory that offers fast screening of protein formulations without spending large amounts of material. Likewise, the drying kinetics of one single droplet can be followed in-situ, and enough material to analyze the protein activity can be generated. Moreover, by applying this approach to different protein formulations containing saccharides, it was possible to systematically analyze the impact of these excipients on the droplet evaporation kinetics.

This study shows how saccharides presenting distinct molecular weights affect the drying kinetics of catalase and trypsin, deriving further mechanistic insights on how some excipients successfully stabilize the proteins while others do not.

Materials and Methods:

Materials:
Trehalose dihydrate (TD) (Mw = 378.33 g/mol), Hydroxypropyl β-cyclodextrin, Kleptose® HP ORAL GRADE (HP) (Mw = 1501 g/mol), Hydroxypropyl β-cyclodextrin, Kleptose® HPB ORAL GRADE (HPB) (Mw = 1387 g/mol) and Dextran 40 EP (DEX) (Mw = 40000 g/mol) have been selected based on pre-screening experiments (data not shown). Milli-Q® (TKA Wasseraufbereitungssysteme GmbH) was used as solvent. Catalase from bovine liver (2,000-5,000 units/mg; ~250 kDa) and trypsin from bovine pancreas (≥2,500 units/mg; ~23.8 kDa) were used as model proteins. The concentrations were 0.04 mM for catalase and 0.42 mM for trypsin (10% w/w of enzymes in the formulations).

Miniaturized Screening (MS):
A single droplet drying method was developed based on the sessile drop principle. An EasyDrop equipment (EasyDrop, Krüss GmbH) equipped with Drop Shape Analysis (DSA1 v1.92, Krüss GmbH) was used. A 15μL-droplet was deposited onto a polypropylene film fixed to a Teflon membrane and placed on a hot plate at 75°C ± 2°C. Drying was performed for 15 min, and the resulting pellets were frozen until analysis.

Evaluation of evaporation rates and droplet shape:
As a droplet shape, a spherical cap was assumed, which describes the region of a sphere below or above a flat surface. According to this, the surface (S) and the Volume (V) were calculated using Eq. (1) ² and Eq. (2) ²:

S = Ï€(a² + h²) Eq. (1)

V = (1/6)Ï€h (3a² + h²) Eq. (2)

Here, h describes the height, and a represents the radius of the spherical cap. These parameters were obtained manually for each time point, at which pictures were taken with the DSA software during droplet drying.

Using Eq. 3, the aspect ratio (AR) was calculated:

AR = h/w Eq. (3)

Here, h describes the height, and w describes the width of the dried particle. Values closer to 1 indicate a more spherical shape, and values closer to 0 show a needle-like form.

A spherical droplet of the same volume as a spherical cap was assumed. Therefore, using the calculated volume, the radius was obtained:

r = (V/4.19)^1/3 Eq. (4)

Here, r describes the droplet radius and V the volume.

Eq. (5) describes the function used to obtain the evaporation rate. The droplet diameter (2*r) was plotted against the drying time, and the evaporation rate was obtained as the slope k ³:

d_t² = d₀² - kt Eq. (5)

Here, d_t is the droplet diameter at t, d₀ is the initial diameter of the droplet, k represents evaporation rate.

UV/vis spectroscopy:

A Shimadzu UV-Visible Spectrophotometer UV-2700 (Shimadzu GmbH) was used to gain information on the activity of catalase and trypsin after drying them with different saccharides. The spectrophotometer was operated at 25°C. The corresponding buffers were used as blanks. For catalase, decrease in absorbance at 240 nm (ΔA240) was measured for 3 minutes and for trypsin, increase in absorbance at 253 nm (ΔA253) was measured for 70 seconds and plotted against the recorded time (see Fig. 1a and 1b, n=3).

Statistical analysis:

A 1-way analysis of variance (ANOVA) was conducted using the concentrations of different saccharides and their related activities with catalase or trypsin (Microsoft Office Standard 2016). The goal was to find out if there were any differences between the distinct concentrations. A p-value > 0.05 indicates similarities of samples, whereas a p-value < 0.05 indicates that samples are not similar and that the null hypothesis (no difference) is rejected.

Results and Discussion:

Catalase: The dried mass of protein formulations containing HP and HPB did not dissolve completely. It was observed that protein-saccharide formulations containing the same saccharides did not always seem to follow the same trend. Drying of catalase without saccharides did not result in an extended loss of activity. Thus, the addition of saccharides was not critical for protein stabilization. However, some trends could still be observed (see Fig. 1a): For HP, the activity increases with decreasing concentration until 1.5 mM and drops again. For HPB, no clear trend is observed, which indicates the potential difference the degree of substitution might have on how the mechanisms, by the tested cyclodextrins (CDs) might be affected. CDs could either stabilize the protein by substituting it at the interface (i.e., liquid-air, liquid-substrate) or by integrating it in their hydrophobic pocket. These mechanisms might be affected by the degree of substitution of CDs, leading to higher quantities of HPB being needed to stabilize catalase, whereas lower quantities of HP are enough. 1.875 mM DEX seems to stabilize catalase best. This is probably due to the high glass transition temperature of DEX and molecular bonding or steric hindrance driven by viscosity and rigidity. Further research is needed to fully elucidate these mechanisms. TD seems surprisingly variable, and no clear trend could be observed, although TD is well-known to have superior stabilizing effects on proteins. However, in this case, TD only seemed to impact the stability of catalase at 400 mM beneficially. This might be because catalase is a large protein, while TD is a small disaccharide. The evaporation rates increased with increasing concentrations of TD, HP, and HPB, naturally, as the solid content of the solutions increased. The evaporation rate was reduced with an increasing concentration of DEX. We hypothesize that this might be due to DEX binding water. DEX also resulted in more spherically shaped particles with higher volume.

Trypsin: All of the dried droplets dissolved completely. It was observed that protein-saccharide formulations containing the same saccharides behaved quite similarly (see Fig. 1b). Trypsin without saccharides loses activity, and the addition of saccharides seems to be beneficial. Because trypsin is smaller in size, we propose that more protein molecules be present at the interfaces during drying, leading to more degradation and instability compared to catalase. HP, surprisingly, did not have a stabilizing effect on trypsin. Although this has to be further elucidated, the stabilizing effect of CDs with different degrees of substitution is reported to depend on the nature of the protein, and distinct effects can be observed, as observed herein with trypsin and catalase ⁴. Compared to catalase, higher DEX concentrations could be used (7.5 mM) to stabilize the smaller enzyme trypsin. Probably, higher concentrations are needed to induce enough steric hindrance. TD stabilized trypsin at 200 mM, a lower concentration than the one required for catalase, again probably because of trypsin’s smaller size. The evaporation rate for trypsin formulations was also the highest for TD. The more HP used, the faster, but the more HPB used, the slower the evaporation. Increasing concentration of DEX also slowed down the evaporation rate and resulted in less elongated particles with higher volume.

Implications and outlook:

Trypsin seems more unstable than catalase during drying in absence of stabilizing excipients, possibly due to their different water solubility and size differences influencing the number of molecules present at the interface. This led to different stabilizing effects observed when distinct saccharides were tested, and some insights into their successful use to stabilize proteins with various molecular sizes could be gained. In future, it will be investigated how these observations align with the impact of distinct drying operations like spray drying on the enzyme formulations. Therefore, some relevant protein-saccharide formulations will be selected and spray dried at lab scale. If successful, this could be an essential pre-formulation tool to apply when producing dried biopharmaceuticals.

Literature:

(1) Ziaee, A.; Albadarin, A. B.; Padrela, L.; Femmer, T.; O’Reilly, E.; Walker, G. Spray Drying of Pharmaceuticals and Biopharmaceuticals: Critical Parameters and Experimental Process Optimization Approaches. Eur. J. Pharm. Sci. 2019, 127 (October 2018), 300–318. https://doi.org/10.1016/j.ejps.2018.10.026.

(2) MathWorld, W. Spherical Cap https://mathworld.wolfram.com/SphericalCap.html.

(3) Vehring, R. Pharmaceutical Particle Engineering via Spray Drying. Pharm. Res. 2008, 25 (5), 999–1022. https://doi.org/10.1007/s11095-007-9475-1.

(4) Samra, H. S.; He, F.; Bhambhani, A.; Pipkin, J. D.; Zimmerer, R.; Joshi, S. B.; Middaugh, C. R. The Effects of Substituted Cyclodextrins on the Colloidal and Conformational Stability of Selected Proteins. J. Pharm. Sci. 2010, 99 (6), 2800–2818. https://doi.org/10.1002/jps.22053.