Lysozyme particle formation during supercritical fluid drying: Particle morphology and molecular integrity

https://doi.org/10.1016/j.supflu.2006.07.005Get rights and content

Abstract

Studies have shown that diverse types of particles can be obtained when processing aqueous protein solutions into powders by using supercritical fluids, however, without identifying the mechanism behind these variations. Therefore, the particle formation of lysozyme by supercritical fluid drying was more systemically studied by varying the flow rates of protein solution, supercritical carbon dioxide and ethanol, co-currently sprayed through a coaxial nozzle. Three different morphologies were identified: agglomerated nanoparticles, microspheres and irregular microparticles. These morphologies could be related to the process conditions, in particular to the fraction of ethanol in the extraction medium: agglomerated nanoparticles were produced under anti-solvent precipitation conditions; microspheres under water extraction conditions; and microparticles under competitive rates of both mechanisms.

A slight increase in intermolecular β-sheets was observed in powders (<5% residual water content) produced under anti-solvent conditions. Nevertheless, the protein integrity was recovered after rehydration. In conclusion, the alcohol fraction in the extractant has shown to influence both the particle morphology and molecular integrity. The selection of ethanol fraction could be especially important when more labile proteins are to be processed using this technique.

Introduction

As the number of therapeutic proteins being approved and introduced to the market is growing [1], the need for appropriate long-term stabilisation methods for these labile compounds is increasing as well [2]. Proteins are often unstable in liquid formulations because of chemical and physical degradation reactions [2], and traditional drying processes such as freeze or spray-drying potentially cause harmful stresses on them [3]. The use of supercritical carbon dioxide (SC-CO2) as an anti-solvent for the precipitation of proteins has previously been suggested as an alternative because of its mild process conditions [4], [5], [6], [7], [8], [9], [10], [11], [12], [13], especially if considering the temperature, which can be set slightly above the critical point of CO2, but sufficiently low to avoid protein denaturation.

However, these techniques often use organic solvents because they are fully miscible in SC-CO2 (e.g., dimethylsulfoxide), or as solubility enhancers for water in SC-CO2 (e.g., ethanol) [14]. The techniques are very useful to produce very small particles, as the low or absent interfacial tension between solution and the CO2 facilitates the dispersion into a spray of fine droplets [15]. Despite the poor solubility of water in SC-CO2, the production of protein powders from aqueous solution is favoured over drying from organic solutions as organic solvents can affect protein stability [16] and are often poor at dissolving proteins.

Drying of aqueous protein solutions using SC-CO2 has been reported by several authors [14]. In these studies, ethanol was usually chosen as modifier in the SC-CO2 to enhance the water solubility (e.g., addition of 5 mol% of ethanol triples the solubility of water (at 37 °C and 100 bar)). Various process options, types of nozzles [5], [6], [8], [9], [10], and process conditions [11], [12], [13], were used. These studies focus on the morphology and size of the particles that are produced. SEM pictures of lysozyme showed more or less spherical micro-scale sized particles [4]. These previous studies further indicate that particle size could be modified by selection of the nozzle [5], the operating pressure [12] or temperature [11], the concentration of the protein solution [11] or the relative flow rate of SC-CO2 to the solvent and modifier [17], but no clear design rules could be derived. Although the bioactivity has been investigated in most cases [4], [9], [12], [17], [18], [19], [20], [21], [22], only Bustami et al. reported on the protein integrity. They showed using size exclusion chromatography that lysozyme remained monomeric to an extent of 97% [4], and using Raman spectroscopy that the protein structure was slightly affected by the process conditions [12]. However, the actual mechanisms causing the differences in the particle morphology and size were hardly addressed in these investigations. Also, despite the well known detrimental effects of organic solvents on the molecular integrity, the effect of their use as SCF modifier has hardly been verified.

To understand the effect of the process conditions on the particle formation, the various ways of protein precipitation should be considered. From an aqueous solution, protein precipitation can be induced via water removal by evaporation/extraction, reduction of dielectric constant by addition of miscible organic solvents, reduction of the protein charge by changing the pH, and addition of polymers or salts. Isoelectric precipitation can be excluded in this investigation as the high isoelectric point of lysozyme (pI 10.7) is not reached during the process [23]. Furthermore, as polymers and salts are absent in this present study, only precipitation via water removal by extraction in the SCF phase and via reduction of the dielectric constant by the anti-solvent within the droplet might be considered.

Another significant aspect of the particle formation process is the dynamics of water evaporation in SCF-drying. In a standard spray-drying process, the drying of a droplet is limited to the evaporation of its water. The process is significantly more complex in SCF-drying as the modifier is expected to first condense onto the droplet until saturation is reached, and then, water and the modifier are rapidly extracted into the SCF phase [24].

To gain a better understanding of the particle formation process of protein drying by spraying into a SCF through a two-fluid coaxial nozzle, we systematically studied the effects of the different flow rates (CO2, ethanol, protein solution) on the particle morphology and protein molecular integrity. In Section 3.1, the different types of lysozyme particles produced are presented, followed by their characterisation. Effects of the different flow rates on the particle characteristics are then described. For reference, the precipitation of lysozyme by ethanol alone has been investigated. Following, is the description of the operational regimes leading to the formation of the different types of particles and the proposition of a mechanistic model of the process based on competitive driving forces. In Section 3.2, the effects of the process conditions on the lysozyme powder and solution are investigated. Finally, the last section correlates observations on the particle morphology and molecular integrity to the process conditions.

Section snippets

Materials

Lyophilised hen egg white lysozyme (∼70,000 U/mg) was purchased from Fluka (Buchs, Switzerland) and stored according to their instructions until use. Technical grade ethanol (100%) was used and carbon dioxide (grade 3.5) was purchased from Hoek Loos (Schiedam, The Netherlands). Lysozyme aqueous solutions (25 g per experiment) used in SCF drying were prepared by dissolving the lysozyme powder (2%, w/w) in ultra pure water.

Experimental set-up, operating procedure and conditions

The experimental set-up is illustrated in Fig. 1. The SC-CO2 was supplied by

Particle types

Depending on the process conditions, three types of particles were observed: (I) agglomerated nanoparticles, (II) microparticles or (III) microspheres. These particles differed in their morphology (Fig. 3) as well as their porosity (Fig. 4). From SEM pictures it could be observed that: (I) agglomerated nanoparticles were characterised by distinct nano-scale spheroids (primary particles, 200–300 nm), agglomerated into fragmented to spheroid clusters (5–50 μm). (II) Microparticles were ellipsoids

Conclusions and recommendations

The process variables systematically investigated – CO2, ethanol and protein solution flow rates – mainly affected the morphology of the lysozyme particles through their effect on the mass transfer in and out of the droplets. The process flow rates also have influenced other elements of the process, such as the atomisation and residence time, but that did not significantly affect the particle morphology. Three regimes of particle formation were identified and a specific type of particle is

Acknowledgments

The authors thank Paul Durville and Mies van Steenbergen for their technical assistance, Ángel Martín for his investigation on mass transfer rates, and Dr. Marc Sutter for helping with the interpretation of FT-IR results. This research was supported by the Technology Foundation STW, applied science division of NWO and the technology program of the Dutch Ministry of Economic Affairs.

References (40)

  • J.F. Carpenter et al.

    Rational Design of Stable Protein Formulations, Theory and Practice

    (2002)
  • R.T. Bustami et al.

    Generation of microparticles of proteins for aerosol delivery using high pressure modified carbon dioxide

    Pharm. Res.

    (2000)
  • R.T. Bustami et al.

    Generation of fine powders of recombinant human deoxyribonuclease using the aerosol solvent extraction system

    Pharm. Res.

    (2003)
  • M.H. Hanna, P. York, Methods and apparatus for the formation of particles. US Patent 6440337,...
  • S.D. Yeo et al.

    Formation of microparticulate protein powders using a supercritical fluid antisolvent

    Biotechnol. Bioeng.

    (1993)
  • S. Mawson et al.

    Coaxial nozzle for control of particle morphology in precipitation with a compressed fluid antisolvent

    J. Appl. Polym. Sci.

    (1997)
  • G. Del Re, M. Putrignano, G. Di Giacomo, C. Di Palma, Apparatus and method for micron and submicron particle formation...
  • R.B. Gupta, P. Chattopadhyay, Method of forming nanoparticles and microparticles of controllable size using...
  • R.T. Bustami et al.

    Generation of protein micro-particles using high pressure modified carbon dioxide

  • S. Moshashaee et al.

    Supercritical fluid processing of proteins: lysozyme precipitation from aqueous solution

    J. Pharm. Pharmacol.

    (2003)
  • Cited by (45)

    • Innovative Drying Technologies for Biopharmaceuticals

      2021, International Journal of Pharmaceutics
    • Preparation of PGX-dried gum arabic and its loading with coQ10 by adsorptive precipitation

      2020, Journal of Supercritical Fluids
      Citation Excerpt :

      Other researchers have used a similar approach to dry biopolymers, but employed conditions closer to the supercritical region. For example, Bouchard et al. [3] explored a range of conditions in the supercritical and two-phase regions and obtained agglomerated nanoparticles, microparticles or microspheres of lysozyme, depending on the position in the ternary diagram of CO2-ethanol-water. However, Bouchard et al. [3] did not explore the gas-expanded liquid region.

    View all citing articles on Scopus
    View full text