Search

Somatropin, human growth hormone (hGH), is an unstable protein, posing challenging problems for its formulation and long-term stability. Since hGH formed insoluble adducts with heparin our aim was to evaluate heparin as a stabilizing agent for the drug. These adducts were characterized by particle diameter, tertiary structure variations and release studies. Studies were also carried out to determine the stability of hGH in the presence and absence of heparin by an interfacial denaturation method and real-time stability studies by measuring hGH activity and particle diameter. Moreover, biological activity of hGH and hGH/UH (unfractionated heparin) adducts was identified by daily subcutaneous injections to hypophysectomized rats. There was a decrease in mean hydrodynamic particle diameter of hGH/UH adducts with increased pH (54.4 to 12.2 nm from pH 3 to pH 7) indicating that the adducts were either dissociating or dissolving at high pH. Furthermore, second-derivative spectroscopy indicated that complexation of hGH with heparin did not cause a major disruption in the tertiary structure of hGH but decreased the hydrophilic environment around the tyrosine residues. Release of hGH from hGH/UH adducts was pH and ionic strength dependent with the highest release at pH 8 (93%) and lowest release at pH 3 (0%) over the first hour. Interfacial denaturation methods indicated that vortex agitation over 120 s resulted in no change in the optical density of hGH/UH adducts compared with a substantial increase for hGH alone at pH 6.8. Real-time stability studies over 93 days demonstrated that hGH/UH adducts at both pH 3 and 7 with an excess of heparin produced the highest percent of active hGH remaining in the solution at 4 degrees C and 37 degrees C. The higher stability of hGH/UH adducts with excess heparin compared with the stoichiometric ratio was also confirmed by particle size measurements during storage. The biological activity of these adducts was comparable with hGH alone by weight-gain studies in hypophysectomized rats. The findings suggest the value of using hGH/heparin adducts to stabilize the protein.

PS4alpha is a high molecular weight peptidoglycan extracted from Mycobacterium vaccae, which has demonstrated considerable antineoplastic activity in-vivo without apparent toxicity. Available fortesting in only small quantities, a sensitive in-vivo method for measuring pulse and breathing rates in cannulated rats was applied to this compound at doses of 5, 50 and 500 microg kg(-1). Various parameters (mean arterial pressure, maximum transpulmonary pressure, compliance, heart rate, minute volume, respiratory rate and tidal volume) were followed for up to 1 h and demonstrated no significant deviation in the baseline values obtained before injection. This compound at doses up to 500 microg kg(-1) had no apparent acute toxicity in rats, but chronic effects at this and higher doses have to be determined by more conventional toxicological methods before proceeding to evaluate PS4alpha as an antineoplastic agent.

The release mechanism of high molecular weight fluorescein isothiocyanate dextrans (FITC-dextrans) from HPMC hydrogel matrices was studied. An anomaly was noted in the release behaviour of a series of high molecular weight FITC-dextrans from a tablet formulation designed to float in stomach contents. The tablets contained sodium bicarbonate and hydroxypropylmethyl cellulose (HPMC) in a cetyl alcohol matrix. When hydrated in an acid medium, this tablet consisted of a mixed solid with a viscous surface layer containing carbon dioxide bubbles through which the active ingredient (FITC-dextran) was released into the aqueous environment. However, it was observed that, above a critical molecular weight (approx. 65 kDa), the FITC-dextran was only released into the medium by an erosion-type mechanism, whereas, below this value, both diffusion and erosion processes took place. The key constraint appeared to be the apparent gel pore-size of the hydrated HPMC that was approximately 12 nm in diameter, irrespective of the molecular weight of the HPMC samples evaluated. It was concluded that FITC-dextran release was controlled by both FITC-dextran molecular weight and the HPMC hydrogel structure.

Although it has been claimed that nanoparticles can be produced from gelatin, a naturally occurring polypeptide, the commercial conversion of animal collagen to gelatin results in a heterogeneous product with a wide molecular-weight range. This is probably responsible for the widely observed variation in the experimental conditions required for nanoparticle formation. In this study, 0.2% w/v aqueous B225 gelatin solutions were incubated under various conditions of time, temperature, pH and ethanol concentration and characterized by both size-exclusion high-performance liquid chromatography (HPLC) and dynamic light scattering. Gelatin was shown to be denatured when the temperature was increased to 37 degrees C (approx.) and the rate of renaturation was optimized over the temperature range 7-20 degrees C at pH 5.0, equivalent to the isoelectric point (IEP). The molecular-weight profile remained unchanged at 37 degrees C (approx.) in the pH range 5-7. When the gelatin solutions were mixed with ethanol, higher-molecular-weight fractions (microgel, delta and zeta fractions, all with molecular weights > 700 kDa) precipitated at ethanol concentrations lower than those required to precipitate the lower molecular weight material ( < 700 kDa), with maximum precipitation occurring close to the isoelectric point (pH 5.0). The molecular weight profile of gelatin in solution is evidently critically affected in a time-dependent manner by both pH and temperature. These two factors influence the noncovalent interactions responsible for the molecular structure of gelatin. The molecular weight profiles, in turn, affect the phase behaviour of gelatin in hydroalcoholic solutions. Systematically investigating the effect of time, temperature, pH and ethanol concentration on the molecular-weight-distribution profile of a gelatin solution enabled a robust method to be developed for the preparation of colloidal dispersions of non-aggregated gelatin nanoparticles 220-250 nm in diameter. This contrasts with the multiparticulate aggregates produced by earlier literature methods.

Although heat-killed suspensions of Mycobacterium vaccae have been tested clinically against tuberculosis and cancer, from a pharmaceutical perspective it would be advantageous to utilize isolated active components rather than the heat-degraded bacterial materials. In our laboratory we have isolated from M. vaccae a number of high-molecular-weight proteoglycans with considerable immunological and antineoplastic activity. The structure of one of these, PS4A, obtained by extraction with boiling water, seems to consist of a basic unit with a 20-kDa protein core to which are attached glucans and O-methylated 4-kDa polysaccharides. The molecular weight is (approx.) 50 kDa, but because of self-association, that of the recovered high-molecular-weight fraction is greater than 150 kDa. A similar, but even larger, molecule (PS4alpha, MW approximately 20 MDa) is obtained by cold extraction with 8 M urea. Both are active in-vivo against an S-180 murine sarcoma model but have no activity in-vitro, suggesting an antitumour effect involving activated macrophages. For this reason gelatin nanoparticles are unsuitable as a vehicle but chitosan seemed to be a promising alternative. In this report we describe the production of stable 600-700-nm diameter nanoparticles of chitosan without organic solvents. Adsorption and release of bovine serum albumin seemed to be affected by the charge of the two reactants and at high doses not all adsorbate was released. PS4A, because of structural and compositional differences, had to be loaded on to the chitosan by freeze drying a suspension of the nanoparticles in a solution of the drug. After a rapid (burst) release phase, the rate of release into water was steady for the next 4 h, but not all the drug was released. In-vivo it was evident that PS4A and PS4alpha were equally active in solution or when formulated in the chitosan nanoparticles. These results show that chitosan nanoparticles, readily prepared without the use of organic solvents, are a suitable vehicle for the delivery of these immunostimulants from M. vaccae; the formulations might find application as antitumour agents.

Bacillus Calmette-Guérin (BCG) vaccine, developed originally for the prophylaxis of tuberculosis, is a potent immunostimulant used to treat superficial bladder carcinoma in man. The aim of this study was to compare the molecular weight and self-association properties of an antineoplastic glucan (PS1A1) extracted from BCG vaccine as determined by different techniques including diffusion, light-scattering and chromatographic methods. In the diffusion experiments, a semi-empirical relationship was derived between the effective diffusion coefficients, Dp, and the weight-average molecular weights, Mw, of several dextrans used as standards, according to the equation Dp = 2.233 x 10(-6) x Mw(-0.66). On the basis of this relationship, the molecular weight of PS1A1 was found to be 57.4 kDa, although, unexpectedly, membrane association was high, most probably because of molecular branching. In the light-scattering experiment it was observed that, unlike dextran, PS1A1 undergoes concentration-dependent multimerization in water. However, the molecular weight of PS1A1 in 0.1 M sodium chloride ranged from 60 to 68 kDa, with a mean of 65 kDa, over the same concentration range. This value was in agreement with the molecular weight determined for PS1A1 by gel-filtration chromatography in previous studies, suggesting that 65 kDa represents the approximate monomeric size of the unassociated molecule. Thus, it was evident that the aggregation was suppressed by electrolyte. Elemental analysis by X-ray fluorescence showed that PS1A1 contained carbon, oxygen, hydrogen and phosphorus, indicating that hitherto unobserved ionized phosphate groups might promote electrostatic interactions.

Phospholipid-stabilized emulsion properties were studied in the presence of lysine, a water-structure breaker, using two unrelated procedures, photon correlation spectroscopy and a light obscuration instrument. Commercial Intralipid was used as a control. Lysine 0.125, 0.25 and 0.5 M induced changes in the size distribution of a non-heated model emulsion system, irrespective of any changes produced by environmental pH. Some of the laboratory-prepared emulsions containing lysine were more stable than the corresponding commercial heat-sterilized product Intralipid, once heated. The results suggest that lysine is producing an effect on the nascent oil-water interface that controls the physical stability of the system. Once the heat-induced interfacial rearrangement of the individual phospholipid molecules occurs, the influence of lysine becomes diminished.

Gelatin, prepared commercially by degradation of animal collagen, was studied to see whether it had an affinity for fibronectin, which has a known affinity for collagen, and whether gelatin-based drugs could be used to target fibronectin-excreting tumours. Bacillus Calmette-Guérin (BCG) vaccine, an attenuated strain of Mycobacterium bovis, is currently the most effective treatment for superficial transitional cell carcinoma of the bladder. The living cells of the BCG vaccine associate with the fibronectin-bearing surfaces of the tumour. Using a multi-well culture plate technique, gelatin microparticles were shown to be adsorbed onto murine S180 sarcoma cells and this reaction was substantially inhibited by the addition of human plasma fibronectin. The avidities of various BCG substrains and gelatin microparticles for glass-bound fibronectin were measured and the association constants determined. The gelatin microparticles associated with the fibronectin with equal avidity as the BCG cells. The results suggest that this model system may allow the investigation of gelatin-based drug delivery devices capable of targeting fibronectin-bearing surfaces associated with some tumours.

The adsorption of D-arabinose onto gelatin microparticles demonstrated a Langmuirian adsorption pattern. Evaluation of the dissolution behaviour of D-arabinose-loaded gelatin microparticles suggested that the saccharide, loaded at a level below the adsorption saturation level, was released uniformly over a 14-h period after the loaded gelatin microparticles had been lyophilized for a second time. When dissolution curves were corrected for the initial burst effect seen after the gelatin microparticles had been loaded at higher levels of D-arabinose and lyophilized, steady-state release rates were also evident over prolonged periods. In addition, it was evident that the D-arabinose was adsorbed onto internal surfaces of the hydrated gelatin matrix. Calculation of this internal surface demonstrated the influence of the concentration of the glutaraldehyde used as a cross-linking agent and this parameter, in turn, influenced both the adsorption maxima and the subsequent equilibrium release rates. Application of this data base to a highly water-soluble complex polysaccharide antineoplastic agent, which has a higher molecular weight (22.4 kDa vs 150 Da), demonstrated similar behaviour in that a near zero-order release pattern over at least 16 h could be obtained by attention to the conditions under which the gelatin microparticles were made and subsequently loaded before lyophilization.

PS1, a complex polysaccharide derived from Mycobacterium bovis (Bacillus Calmette-Guérin, BCG) with considerable antitumor activity in-vivo, was loaded onto gelatin microparticles (mean diam. 1.45 micron) at a level shown to not produce the burst effect often seen with drug-loaded microparticulate systems. In-vitro dissolution experiments had demonstrated a sustained-release behaviour, with a half-life of approximately 8 h for what is an extremely water-soluble material. These PS1/gelatin systems had no measurable cytotoxicity against an S180 murine sarcoma cell in-vitro although fibronectin-mediated targeting of the microparticles for the tumour cells could be demonstrated. Injection into mice, with the S180 cells, of PS1 solutions or suspensions of PS1-loaded gelatin microparticles resulted in almost identical dose-related suppression for the tumour cell growth. When injected at intervals following injection of the tumour cells, however, for a period of 24-48 h there was a relatively enhanced activity of the formulated PS1, compared with the aqueous solution, after which both formulated and unformulated material became progressively less effective.

The Walden Product, the product of viscosity (eta 0) and conductivity at infinite dilution of a solution (lambda 0), provides a measurement of the water-structuring activity of the solute. Measuring the effect of concentration on viscosity of solutions of amino acids, together with the conductivity of solutions of sodium chloride containing increasing concentrations of the amino acids, enabled Walden Products to be determined. The classical form of the Walden Product (lambda 0 eta 0) was used, together with a modified form, lambda 0 eta c, in which eta c was the slope of the concentration/viscosity curve. Most amino acids demonstrated modest water-structure-breaking activity but L-lysine, L-glutamic acid and L-aspartic acid, and their respective salts, all showed relatively higher activity. Dextrose behaved as a classical water-structure maker and, when added progressively, reversed the breaking activity of L-lysine. It is speculated that effects seen in bulk water may also occur at emulsion droplet surfaces, thereby inducing structural changes associated with the occasional rapid instability experienced when making admixtures of phospholipid-stabilized emulsions and additives such as amino acids and dextrose.

Examination of conditions needed to form albumin microparticles (in the 1-5 microns diam. size range) by glutaraldehyde cross-linking, at low temperature (-15 degrees C), suggested that the particles formed very rapidly over a short time scale and at low concentrations of cross-linking reagent. Detailed analysis showed that the particles increased in size with time of reaction and with an increase in glutaraldehyde concentration. Evaluation of the particle matrix using a dilute trypsin digestion process suggested that the above factors may influence the internal composition since the particles dissolved less rapidly than might be anticipated from a consideration of particle diameter alone.

The mechanism of the stabilization of triglyceride emulsions by phospholipids has been studied using an HPLC-FID method to determine phosphatidylcholine, lyso-phosphatidylcholine and phosphatidylethanolamine in the oil and aqueous phases of a model emulsion consisting of soybean oil 20 g, egg phospholipid 1.2 g, glycerol 2.25 g and water to 100 mL. It was shown that, on heat sterilization of the emulsions, the phospholipids rapidly relocate from the aqueous phase to the oil phase. It is suggested that the phospholipids concentrate in the oil/water meso phase, forming a cubic liquid crystalline phase, the bulk of which is converted to a lamellar phase on cooling, and that this organization of interfacial material accounts for the enhanced stability of phospholipid emulsions after heat sterilization.

A model 20% w/v emulsion, prepared with either a commercially available pharmaceutical grade soy oil or a highly purified grade of oil from the same origin and stabilized with a commercially available mixture of egg yolk phospholipids was passed through a Microfluidics homogenizer until the mean particle size fell below 500 nm diameter. Samples stored in sealed all-glass ampoules were thermally stressed over a temperature range of 5-90 degrees C and samples taken at appropriate intervals for analysis by HPLC. Hydrolysis degradation kinetics were in conformation with the Arrhenius equation. The energy of activation for phosphatidylcholine was virtually identical for emulsions prepared with either pharmaceutical or purified oil (65 and 63 kJ mol-1, respectively). For phosphatidylethanolamine itself the respective activation energies were 53 and 54 kJ mol-1, suggesting that the source of the oil used in preparing the emulsions had no significance in the degradation processes of the resulting systems.

Dehydration following non-specific diarrhoea may be prevented by oral administration of a simple glucose/salt mixture. A solution tablet of this mixture would have advantages of stability under environmental exposure and transport if the costs could be held within reasonable limits. The moisture adsorption and compression characteristics of Oral Rehydration Salts (ORS) ingredients have been studied. Combinations of ingredients resulted in a moisture adsorption higher than that of the individual components. This may be explained in terms of critical relative humidity, RHo, and environmental relative humidity RHi. Preparation of a stable ORS solution tablet therefore requires protection of moisture adsorbing components from the environment. The present UNICEF ORS mixture compacted easily by direct compression but gave fragile tablets, which were hygroscopic. This can be reduced by film coating the electrolyte component as granules with a resin (Eudragit L), or by simulating direct compression of the glucose as a compression-coating around the precompressed electrolytes. The packaging of compression-coated solution tablets in inexpensive polyethylene bags may lengthen the shelf-life and make the preparation less costly than the currently supplied ORS powders packed in laminated aluminium sachets. The increased dissolution lag time for the compacted tablet is a disadvantage that can be overcome by instructions to crush the product immediately before use.

The solubility of 17 beta-oestradiol (E2) in aqueous solutions of polyethylene glycol (PEG) 400 has been measured at 35 degrees C. Up to 80% w/w PEG the solubility data conform to a log linear equation ln S = ln Sw + f sigma where S is the E2 concentration in the water/cosolvent mixture, Sw is E2 solubility in water, f is the weight fraction of PEG 400 and sigma is a parameter representing the solubilizing power of the cosolvent for the drug. Above 80% PEG the relationship becomes less convincing, with significant deviation from anticipated values. Reasons for these aberrations are discussed. It is suggested that conformational changes may be induced in the PEG by the addition of small quantities of water. Deviations noted for the melting point, viscosity and density data of PEG 400-water solutions may also confirm this suggestion.

An improved column-type dissolution apparatus is described. The column of the apparatus is built from standard screw-cap connected "Sovirel" glass tubes which should enable inter-apparatus variability to be reduced. The main problem with this type of device is the build up to back pressure as fragments from the disintegrating solid dosage form progressively block the filtration system, itself necessary to stop the solution process and present the solution for analysis. A simple type of light-activated switch is therefore used which works across the indicator float of a flow meter to drive a motor connected to a needle valve. As the float falls the motor is switched on to open the valve and increase the flow rate until the float rises again and causes the motor to switch off. Observation on the dissolution characteristic of dyed lactose tablets and the injection of dye solutions into the flowing stream of liquid around tablets showed that there was back-flow of solution at Reynolds Numbers below 10 but visible instability to flow at Re greater than 100. Over a range of flow rates between Re 10-70 the mass of drug released at a given time was a direct function of flow rate for a film coated product but an inverse function for a sugar coated tablet. The results are discussed in relation to the main problems which can be experienced with a column-type dissolution apparatus.