ANALYTICAL
BIOCHEMISTRY
202,331-336
(19%)
The Use of a Layering Technique for Enhancing Stability of Lyophilized Reagents J. R. Ramsay,*
G. D. J. Adams,?
H. C. Morris,$*3
R. S. Campbell,$l’
C. P. Price,$
*Division of Biotechnology, and fDivision of Biologics, PHLS Centre for Applied Microbiology Porton Down, Salisbury, Wiltshire, SP4 OJG; and *Department of Clinical Biochemistry, The London Hospital Medical College, Turner Street, London, El 2AD
Received
November
13, 1991
An enzyme-mediated assay has been developed for the measurement of salicylate usingsalicylate monooxygenase purified from Pseudomonas cepacia ATCC 29351. Two assay formulations were produced, based on either a multiple-reagent or a single-reagent formulation, to allow sufficient flexibility for automated use. The multiple-reagent formulation was especially suited to diagnostic laboratories performing infrequent manual salicyiate estimation where stability of the reconstituted reagent is of paramount importance. This was achieved by preparing the enzyme and color reagents in separate vials, so keeping the enzyme at a stable pH. For more frequent assay use where a reconstituted reagent shelf life was less important, the single-reagent system offers advantages of convenience. However, the working reagent required a pH of 10.0 upon reconstitution. Although the enzyme was sufficiently active at this pH to give a reliable assay, its storage stability was poor at pH 10.0, preventing lyophilization of the reagent at a pH suitable for immediate use on reconstitution. This incompatibility was overcome by use of a layering technique. The enzyme was separated from the buffering solution in the same vial by freezing the buffering solution and then overlayering with the enzyme reagent prior to a second freezing cycle and subsequent freeZe drying. 0 1992 Academic PRESS, h.
Freeze drying is a popular method of commercially processing and finishing enzymes, hormones, and simi1 Present address: Porton Cambridge Limited, Lanwade Business Park, Kennett, Newmarket CB8 7PW, UK. ’ To whom correspondence should be addressed. Fax: +44 (0) 980 610 898. ’ Present address: Dept. Clinical Biochemistry, Addenbrookes Hospital, Hills Road, Cambridge CB2 200 UK. 0003-2697192 Copyright All rights
and P. M. Hammond*12
& Research,
$3.00
0 1992 by Academic Press, of reproduction in any form
lar biological materials for clinical diagnostic use (1). Salicylate (aspirin) is a common nonsteroidal drug widely used for its analgesic and anti-inflammatory properties. Its wide availability without prescription has resulted in a high number of both accidental and deliberate overdoses of the drug leading to acute poisoning or death. Quantitation of serum salicylate levels is therefore important, both for the management of patients who have ingested excessive quantities of the drug, and for therapeutic monitoring (2). It has traditionally been measured by a calorimetric method (3) based on the interaction of salicylate and ferric ions. While the test is simple and rapid to perform, it requires a large amount of sample and initial precipitation and extraction steps. The reaction is also subject to interference, leading to false positive results (4,5). Reference methods, such as GLC (6) and HPLC (7), all require expensive instrumentation and a high level of technical skill and are not appropriate for emergency testing. It has been known for some time that salicylate monooxygenase (EC 1.14.13.1) in the presence of NADH and oxygen, will convert salicylate to catechol (8,9) as shown in Fig. 1. A method has been developed for the quantitation of salicylate using salicylate monooxygenase extracted from Pseudomonas cepacia ATCC 29351. Salicylate is quantified indirectly by the reaction of catechol with p-aminophenol or 4-aminophenazone under alkaline conditions to form a dye product which is measured spectrophotometrically (10,ll). However, salicylate monooxygenase has an inherent NADH oxidase activity, which interferes with assays based on direct quantitation of the NADH generated (12,13). In order to negate this activity toward NADH, a recycling system (Fig. 2) based on lactate dehydrogenase was incorporated into the formulation. For use of the assay on automated clinical analyzers, a single-reagent system is highly desireable. Although 331
Inc. reserved.
332
RAMSAY
NADH
COOH
OH + O2 + H+
ET
AL.
NAD
u
60H
+ CO2
+ Hz0
/
SALICYLATE FIG. 1.
CATECHOL Enzymic
conversion
of salicylate
all analyzers are, of course, capable of handling addition of one reagent to a sample, many cannot make further reagent additions. This rules out automation (on many machines) of any assay requiring two or more reagent additions. Although the enzymic assay described above performed well in clinical situations when carried out manually (14), formulation for automated used posed several problems. The reagents had acceptable shelf lives when stored separately, but the stability of the enzyme reagent was considerably reduced when stored at the working pH. Single-reagent presentation was not practical as mixing the assay components (prior to lyophilization as a single reagent) led to severe losses of enzyme activity. We describe the use of a multilayer technique for overcoming these losses, leading to a highly stable single-reagent formulation whereby incompatible reagents can be lyophilized together. The technique has general applicability to any liquid reagents which need to be lyophilized together in the same vial, but nevertheless need to be kept separated during the freeze-drying process. MATERIALS
AND
METHODS
Tris(hydroxymethyl)aminomethane (Tris) was purchased from Boehringer-Mannheim (UK). Dextran grade C (M, 60000-90000) was obtained from BDH (Poole, UK). All other reagents were purchased from Sigma Chemical Co. (Poole, UK) and were of the purest grade available. Enzyme Purification
Resuspended bacterial cell paste was passed through a Manton Gaulin homogenizer at 55 MPa. Disrupted
to catechol
by salicylate
monooxygenase.
cells were centrifuged, and the cell-free extract was purified by chromatography on a DEAE-Sepharose ion exchange column, followed by hydrophobic interaction chromatography on phenyl-Sepharose (15). The purified enzyme was dialyzed against the appropriate buffer at pH 8.6, and excipients were dissolved in the dialyzed enzyme preparation prior to freeze drying. Freezing Analysis
Freezing analysis was carried out using the Edwards Kniese freezing analyzer. Freezing and melting characteristics of experimental formulations were determined by measuring the changes in electrical resistance which occur when solutions undergo a phase change from solid to liquid (or vice versa). Using the data from the freezing analysis, freeze drying trials were completed on the Edwards Minifast 3400 two-shelf freeze drier (Edwards High Vacuum, Manor Royal, Crawley, Surrey, UK). For freezing analysis, a 4.5-ml volume of the solution was pipetted into the analyzer measuring cells, which was equipped with two stainless steel electrical resistance measuring probes and a platinum resistance thermometer (type Pt 100). The cell, enclosed within a cooling chamber, can be cooled to approximately -5O’C and heat then applied to slowly raise the sample temperature; heating capacity was held at approximately 10 W. Changes in electrical resistance and temperature of electrolyte solutions were logged on a three-point chart recorder. Freezing and melting temperatures were determined from the trace, shown in Fig. 3. Definitive measurements of melting temperatures were determined on the thawing cycle (16) since complicated precrystallization phenomena (such as supercooling) resulted in poor
NADH oxidasc 02
b
H20
f SALICYLATE
+ NADH
PYRUVATE
4
Salicylate
monooxygenasc
b
NAD
+ CATECHOL LACTATE
Lactate dehydrogenasc FIG. 2. oxidase.
NADH
recycling
system
used in multiple-reagent
formulation
of salicylate
monooxygenase
to counteract
the presence
of NADH
ENZYMATIC
ASSAY Log
Centigrade
FOR
reslrtance 19
60 50. 40.
Eutestlc c----------I< Resistance
30
,V (ohms)
region
\ \
I
/ 38
ld
1. 0
12
3
4
5
6
7
6
9
Time (hours) FIG. 3. Typical freezing analyzer trace for ase in 200 mM glycylglycine/NaOH, pH 8.6, ium lactate. The Figure shows eutectic region perature (“C), and electrical resistance over
salicylate monooxygencontaining 500 mM lithand freezing point, temthe range lOa-10’ ohms.
reproducibility during the freezing cycle. The melting points of nonelectrolyte solutions were determined by evacuating the chamber containing the sample. The behaviour of the frozen mass or thawed solution was observed through the plastic chamber cover using an inbuilt enlarging lens. Foaming of the warming, frozen mass under vacuum was equated with the melting point of nonconducting solutions. The use of electrical resistance as a measure of melting, eutectic, and freezing characteristics of aqueous solutions when cooled has been described in detail elsewhere (16-18). Reagent Formulation Freeze-drying studies were carried out using vials of 46 mm height X 23 mm external diameter. For multiplereagent formulation, the purified enzyme was dialyzed against either Tris/HCl buffer, pH 8.6, or glycylglycine/ NaOH buffer, pH 8.6. Lithium lactate (250-500 mM), bovine serum albumin (BSA,4 2 mg/ml), NaCl (100 InM), lactate dehydrogenase (4 U/ml), and NADH (5 mM) were added to the dialysate. Lactose (5%, w/v) and dextran (2%, w/v) were added as bulking excipients to improve the structure of the freeze-dried cake. Aliquots of 5 ml of this solution were frozen for 2 h within the Minifast prior to commencing the freeze-drying cycle. For single-reagent formulation, reagents were all added to one vial, using a layering technique to keep noncompatible components separate during freezing and freeze drying. The first reagent was added to the vial and frozen, before addition of the next reagent. In this way, layers of frozen reagent were built up in the ’ Abbreviations used: BSA, bovine serum albumin; Taps, 3-[Tris (hydroxymethyl) methyl] amino propane sulfonic acid; Tricine, NTris (hydroxymethyl) methylglycine; Ches, 2-(N-cyclohexylamino) ethanesulfonic acid.
MEASURING
SALICYLATE
333
same vial, keeping incompatible reagents separate. Double- and triple-layered vials were experimentally prepared as follows. In two-layer vials, the reagent layer contained 2-(N-cyclohexylamino)ethane sulfonic acid/ NaOH, pH 10.0 (181 mM), cobalt acetate (0.31 mM), and 4-aminophenazone (1.61 mM). The enzyme layer contained NADH (3 mM), salicylate monooxygenase (>0.75 U/ml), and Tris/HCl, pH 8.6 (30 mM). In three-layer vials, an interposing buffer layer containing Tris/HCl, pH 8.6 (30 mM), was added to ensure no contact between the two reagents. The concentration of components of the reagent layer was increased by 33% to account for the change in total volume. All layers contained 2 mg/ml BSA and 2% (w/v) dextran. The following volumes of different layers were investigated: (i) 1 ml enzyme layer overlayered with 4 ml reagent layer; (ii) 4 ml reagent layer overlayered with 1 ml enzyme layer; (iii) 1 ml enzyme layer overlayered with 1 ml interposing buffer, then overlayered with 3 ml reagent layer; (iv) 3 ml reagent layer overlayered with 1 ml interposing buffer, then overlayered with 1 ml enzyme layer; (v) Control: 4 ml of reagent layer and 1 ml of enzyme layer, mixed prior to freezing. Each layer was frozen to -40°C for at least 2 h in the freeze drier prior to overlaying with successive layers. After all layers had been completely frozen the chamber was evacuated. When a vacuum of 10 Pa had been attained, the shelf temperature was raised to 4°C over a period of 3 h with maximum heat input.
Salicylate
Determination
For the multiple reagent formulations, solutions were freeze dried individually in separate vials; the product was reconstituted to 10 ml with distilled water and salicylate concentrations were determined spectrophotometrically at 565 nm according to the method of Chubb et al (14). Single-reagent formulations, freeze dried as separate layers in the same vial, were reconstituted with 5 ml of distilled water to give final constituent concentrations of 145 mM 2-(N-cyclohexylamino)ethanesulfonic acid/ NaOH, pH 10.0,&0.15 U/ml salicylate monooxygenase, 0.6 mM NADH, 0.25 mM cobalt acetate, 1.29 mM 4aminophenazone, 2 mg/ml BSA, and 2% (w/v) dextran. Using the single-reagent formulation, salicylate was determined according to the method of Atkinson et al. (11). A l-ml volume of the above reagent was added to 25 ~1 of sample. The mixture was incubated at room temperature for 6 min and the absorbance read at 520 nm.
334
RAMSAY
ET
TABLE Multiple-Reagent
(Single
Layer)
AL. 1
Freeze
Drying
of Salicylate Reconstituted
Buffer
system
Lithium lactate (mm
200 mM Tris/HCl 150 mM Tris/HCl 100 mM Tris/HCl
200 mM
glyCylglyCine
200 mM glycylglycine
500 250 250 500
250
Lactose % (w/v) 5 5 5 0 0
Dextran % (w/v)
Melting point P.3
2 2 2 2 2
-57 -51 -43 -41 -39
Cake Poor Poor Satisfactory Satisfactory Excellent
Monooxygenase product
acceptibility,
Enzyme (U/vial) 0.0 0.0 10.8 10.6 11.1
Note. The table describes the various formulations used in the development of an acceptable reagent composition. levels were 12.5 units per vial and the initial NADH concentration was 5.0 mM. With regard to shelf life, the reagent if the surviving enzyme level was 110.0 units per vial and NADH levels were >0.6 mM.
Salicylate
Monooxygenase
Determination
Salicylate monooxygenase activity was determined by monitoring the oxidation of NADH to NAD spectrophotometrically at 340 nm. One unit of activity is defined as that amount of enzyme which catalyzes the hydrolysis of 1 pm01 salicylate per minute at 30°C. RESULTS
Freezing analysis indicated that a multiple-reagent formulation developed for a liquid assay technique, which included high levels of sodium chloride (100 KIM), lithium lactate (500 mM), and Tris/HCl (200 mM), displayed a melting point of -57”C, a value below the minimum shelf temperature of the freeze drier. Freeze drying trials, using the protocol described, support the freezing analyzer data since the product foamed under vacuum. Reducing the Tris concentration to 150 mM and lithium lactate to 250 mM raised the melting point to -5l’C. No improvement in cosmetic acceptability was observed after freeze drying even when the shelf temperature was maintained at the minimum value of -40 to -44°C throughout the cycle and the total drying time extended to 8 days to ensure complete drying. A marked improvement in cosmetic acceptability was observed using the prolonged cycle when the Tris concentration was further reduced to 100 mM, thus raising the melting point to -43°C. These results are shown in Table 1. Using a more conventional drying cycle with a shelf temperature of 4”C, the low melting point of this formulation produced a cosmetically inferior cake to that obtained using a prolonged drying cycle. On reconstitution, the use of a larger volume of water reduced ingredient concentrations by a factor of two. The reduced level of lithium lactate (125 InM in the reconstituted vial) was insufficient to maintain NADH stability in the recycling system, even though all formulations contained 100 mM sodium chloride to suppress NADH oxidase activity. As a result, the shelf life of the reconstituted product was limited to 14 days.
postlyophilization
NADH 21 days
after (mM)
10.60