IVD Technology Magazine | IVDT Article Index
Originally published May, 1997
Meeting the challenge of stability
Knowing how to define and evaluate the causes of instability is an essential tool for companies seeking to supply more convenient and stable in vitro diagnostic products.
Craig M. Jackson(For related information, you may also want to read about general strategies for improving liquid IVD product stability.)
Today's demand is for in vitro diagnostic products--reagents, calibrators, and controls--that can be sold at a lower price to the end-user and yet are more convenient and can be used without error by less well educated laboratory personnel. One of the primary changes fueled by this demand for convenient, simple-to-use products is the change from dry powder or freeze-dried reagents to ready-to-use liquid reagents.
The benefits of this switch to the contemporary laboratory are evident. Pipetting errors during reconstitution are eliminated. Because the control does not need to be reconstituted, reagent deterioration as a result of too-vigorous shaking is no longer likely. Laboratory supervisory personnel lose far less time troubleshooting.
For the manufacturer, this all translates into the need to develop reagents that are stable under less favorable conditions (namely in liquid form), from what are likely to be more expensive raw materials (because of the higher purity required), if possible at a lower overall cost than that of previous-generation products.
At the same time, FDA's new quality system regulation--with its requirements for design control and product and process validation--will increase research and development costs. Reduced shelf life for such products will likely reduce batch size for some of them, also increasing costs.
Manufacturers seeking to meet these challenges should comprehensively examine the properties of the reagent components (raw materials) and formulations that are responsible for reagent stability. It is on this subject that this article will focus.
Stability and Instability Defined
Broadly defined, stability is the constancy of a property over time. In the case of an IVD product, unchanged performance throughout its claimed lifetime might be an idealized expression of such stability. But the more practical, realizable goal for IVD stability is for the product to meet or exceed performance specifications during its claimed shelf life. Manufacturers can improve stability by understanding the causes of instability and, from this understanding, taking actions that stabilize the components responsible for the unsatisfactory shelf life of the product.
Constructive, cost-effective action to increase the stability of a product or raw material requires that the source or sources of instability be correctly identified. An enzyme will be used as an example, although the same basic considerations can be applied to antibodies, other proteins, and, in many cases, polynucleotides and oligonucleotides.
The activity of an enzyme may decrease for a variety of reasons. The enzyme may be intrinsically unstable under the conditions--pH, ionic strength, and temperature--that are required for the reagent.1,2
The enzyme may be undergoing degradation by a contaminant in the reagent solution--for example, another enzyme such as a protease, glycosidase, or oxidase. The source of the contaminant might be a raw material, such as a particular batch of the enzyme, or the water or some other ingredient of the product.3
The enzyme might require a cofactor such as reduced nicotinamide-adenine dinucleotide (NADH), which is being oxidized. In such a case the reaction by which the enzyme is assessed slows, although the enzyme itself could be stable.4
Enzyme may be adsorbing to the surface of the container in which the product is packaged, or in which it is manufactured. The adsorbed enzyme could be undergoing more rapid degradation by a contaminant protease than is the enzyme in solution. Because of the dynamic nature of the adsorption/degradation/desorption process, both the adsorption and the subsequent degradation can with time inactivate or remove a significant fraction of the enzyme.5 Multisubunit enzymes might be slowly dissociating into their subunits at the low concentration at which they are present.6 Combinations of these possibilities must be considered also.
Similar processes occur with immunochemical reagents and oligonucleotide-based diagnostic reagents. In the former case additional considerations are relevant for ELISAs; in the latter case nucleases rather than proteases are the relevant degradative agents.
Proteases and nucleases frequently have a common origin--microbial contamination. Microbial contamination could readily occur as a result of conditions in the customer's laboratory, not at the manufacturing site.
This possibility introduces another consideration and distinction that is well understood by technical support personnel. If customer support personnel can exclude particular types of degradative processes because of the procedures used and understood in the manufacture of the product, complaint handling is easier and less costly.
The following sections deal in depth with each of the major possible causes of instability--intrinsic instability, contamination, effects of cofactors, and adsorption.
Intrinsic Instability
In an idealized situation--that is, in the absence of the degradative processes caused by contaminating enzymes or adsorptive processes mentioned previously--intrinsic stability becomes the focus for consideration. Intrinsic stability is the stability of the component under the given conditions of pH, solution composition (including ionic strength), and temperature.
For multisubunit proteins, the total protein concentration is also a variable, because the multimeric protein and its subunits may differ in their intrinsic stabilities. Loss of functionality in the idealized situation may be attributed to the effects of thermal motion at the given temperature.
Intrinsic stability can be improved in some cases by addition of specific compounds to the solution. For example, a substrate, a substrate analog, a cofactor, or an allosteric effector added to the solution can bind to the enzyme and stabilize it, although destabilization is also possible. Divalent cations are also common stabilizing agents, because they shift the equilibrium between conformers of macromolecules to effect more stable forms.
Thrombin, shown here in a space-fill model, possesses a surface characterized by grooves that serve as binding sites for macromolecules on which thrombin acts or to which it binds. The residues shown in red are those of the inhibitor D-Phe-Pro-ArgCH2- bound to the active site histidine; the blue residues are the A or light-chain residues.
Thrombin, for example, aggregates (self-associates) at higher than physiological concentrations but is stabilized by high NaCl concentrations.7 In addition to this stabilization, a sodium ion binds specifically to residues Tyr184a, Arg221a, and Lys224 in thrombin. This sodium ion binding changes the preference of thrombin from fibrinogen to protein C, another of its biological substrates.10
The addition of cosolutes, such as nonionic detergents, that act primarily through their effects on solvent (water) structure or by excluded-volume effects may similarly increase or decrease the intrinsic stability.1 Most frequently, stabilization or destabilization by detergent is discovered only by experiment.
Excluded-volume effects arise as a result of the occupation of a volume by the added components, such as "inert" proteins and polyethylene glycol. These materials effectively concentrate the other components in the solution.
Altering solution composition may improve intrinsic stability. The availability of this option may be limited by the requirements imposed on the reagent, calibrator, or control material by the least-stable component, or by instrument or device limitations. However, knowledge of the intrinsic stability of each component permits the R&D scientist to focus on the "critical" component and to ignore components that can be considered entirely stable in the final composition.
If the component under consideration is the least stable component and is not adequately stable, then the manufacturer must select an alternative material from a different biological source. A good alternative source is one whose material is intrinsically more stable because of different molecular structure but is otherwise without important functional differences compared with those of the original raw material.
Data that may characterize intrinsic stability are frequently available from suppliers of raw materials, such as enzymes. Sometimes these data actually reflect the action of a contaminant in the raw material.
Proteins and enzymes prepared by recombinant technologies are available today. Those that derive from organisms that live at high temperatures or under extreme conditions and have evolved to possess extraordinarily stable components may ultimately be the best. Enzymes derived from these organisms, however, will not necessarily be more stable under the solution conditions of the diagnostic product. The considerations about stability discussed here are therefore as relevant to them as to conventional raw materials.
Immunological reagents will be modifiable for improved stability properties as well. For example, chimeric monoclonal antibodies now manipulated by recombinant techniques for decreased antigenicity in therapeutic applications may soon be available for use in IVDs.
Economic considerations may, however, negate any stability gains. Improved reagents may not be developed because of the pressure to keep reagent costs low.
The number of proteins for which both amino acid sequences and three-dimensional structures have been determined is rapidly increasing. This information is of some help in predicting the intrinsic stability of a compound. One prediction of stability can be obtained from the ExpAsy site at http://expasy. hcuge.ch/sprot/protparam.html. This prediction is based on the algorithm proposed by Guruprasad, Reddy, and Pandit.2 Given the current advances in protein structure analysis and methods for secondary and tertiary structure prediction, other approaches and methods are likely to emerge in the near future.
Contamination
Real situations inevitably differ significantly from idealized cases. Intrinsic stability may not be alterable to any great extent because the overall assay reaction dictates a particular pH, ionic strength, and temperature.
Materials of biological origin are rarely, if ever, entirely free of all deleterious contaminants. Some contaminants originate from the source organism of the material of interest, others from the materials employed in the isolation and purification processes. Consequently, the practical evaluation of stability is likely to focus more on the nature and source of an agent responsible for instability than on intrinsic instability. The considerations of intrinsic stability, however, provide a logical starting point for stability assessment and guide actions to improve product stability.
Results of laboratory investigations do not always identify the specific cause of the instability, but they at least exclude some candidate causes. Exclusion of particular causes may be sufficient in some situations. Exclusion of microbial contamination as a cause for product deterioration in a specific manufacturing lot, for example, would enable product support personnel to focus on improper product handling by the end-user as the likeliest cause. In another example, laboratory results could eliminate the need for testing of a raw material for a contaminant that has been demonstrated to be irrelevant to the performance and stability of the final product.
Degradative processes unrelated or secondarily related to intrinsic stability are frequently responsible for inadequate or reduced stability. Proteases and glycosidases reduce product stability.
These contaminants can be of endogenous origin, that is, a contaminant of the particular component, or of exogenous origin, that is, acquired during the process of manufacture or present in a different component of the product. In the latter situation, the raw material could be stable, and instability would appear only in real-time product stability testing.
This possibility is a fundamental basis for the requirement for real-time stability data for individual lots of finished products. From real-time stability testing, the initial clue is usually simply deterioration of product performance with time. Optimal corrective action--changes in material specifications or manufacturing procedures that do not permit contamination--requires that the source of the agent be determined in the multicomponent product or the root cause in the procedure.
Accelerated stability testing--testing at temperatures higher than those specified for product storage--may aid in identification of specific lot instability. However, many reagents have complex compositions and complex intermolecular interactions with very different temperature dependencies. Kinetic activation energies and free energies of binding, for instance, affect results, so data from accelerated studies are potentially misleading.
Proteolytic (and sometimes glycolytic) degradation of chemically homogeneous enzymes or proteins can be detected by several electrophoretic procedures. Among these are acrylamide gel electrophoresis in the absence and presence of a denaturing agent, such as sodium dodecyl sulfate. Changes in migration behavior and quantity of a particular component and/or the appearance of new bands would be evidence for degradation. Manufacturers can also apply higher-resolution methods, such as capillary electrophoresis and high-performance liquid chromatography employing one of the various molecular-property-dependent separation procedures (e.g., ion exchange, gel filtration, or gel permeation with hydrophobic interaction or reverse-phase separation methods).
Mass detection methods such as measurement of absorbency or staining properties are commonly employed in the procedures just mentioned. If the unstable material or product is not sufficiently homogeneous to employ these methods, then alternative detection methods are required. Western blotting is a powerful approach in situations in which antibodies against the component of interest exist.36
An alternative and complementary approach to identifying proteolytic degradation is to employ "broad-spectrum" inhibitors of serine proteases of trypsinlike and chymotrypsinlike specificities, of thiol proteases, and of metalloproteases. Improved stability in the presence of such inhibitors implicates proteolytic degradation as the cause for instability. High-specificity and high-affinity inhibitors are not required to identify the cause of instability as proteolytic degradation.
Sometimes a high-affinity inhibitor is identified for a protease that is responsible for degradation of the protein of interest. Such an inhibitor may be attached as a ligand to a chromatographic matrix to produce an affinity column that can remove the culprit protease. This step is useful in the production of recombinant proteins.
In some situations, degradation may be prevented by the inclusion of an inhibitor in the reagent. High-affinity inhibitors are preferred for this application, because lower concentrations are required and therefore costs may be lower.
Inclusion of inhibitors in the reagent is the only approach that is available for some serum-based products, because serum contains both proteases and protease precursors. These together slowly generate proteolytic activity.
Manufacturers risk increasing a product's cost without increasing its value when they add materials such as inhibitors without clear evidence that they actually are directed to the cause of the instability problem.
If proteolytic degradation is responsible for the instability of a raw material and a specific, high-affinity inhibitor exists, the inhibitor could be profitably used in a raw material qualification procedure. For example, comparable stability in the presence and absence of the inhibitor, which would imply absence of any inhibitable protease, might be a criterion for acceptability of each raw material vendor lot. Ideally, the vendor would perform the test as a condition of the sale. By verifying the absence of proteolytic activity in the raw material, the vendor might ensure its preferred supplier status to the customer company.
If the source of the proteolytic agent is microbial contamination, then addition of an antimicrobial agent or microbe-removing filtration would be preferable to addition of a protease inhibitor. The use of antibiotics, however, carries the risk of promoting further development of antibiotic-resistant microorganisms. Use of inhibitors of aerobic metabolism, such as sodium azide, is also problematic because of regulatory limitations on azide concentrations in diagnostic reagents in some countries.
Identifying the root cause of instability almost always leads to a better solution. However, even if the actual agent were a protease and the stability problem could be solved with a protease inhibitor, cost considerations might make an antimicrobial the only economically feasible solution.
When an exogenous source of the degradative enzyme is suspected--for example, when it might have been introduced during the manufacturing process--the manufacturer may compare stability in samples with and without inhibitor added. Results of such a comparison will suggest the best way to prevent the unacceptable instability created by the acquired agent.
Such exogenous contamination has been encountered by the author in a research, not manufacturing, environment. In that situation, the contamination was seasonal and correlated with mold spore counts, which were provided by a local radio station as a service to allergy sufferers. Even fully compliant process validation, if done during the wrong season, would not and could not identify this type of problem. Similarly, nor-mally adequate cleaning procedures might be inadequate under these circumstances. Knowledge of cause, which leads to the elimination of a problem, not only reduces scrap losses but also gives assurance that the process is under control.
Approaches based on a well-defined rationale are not guaranteed to identify the cause or agent responsible for the instability. When they fail, empirical comparisons using different lots or suppliers of each raw material may be the only alternative. Root cause identification forms a better basis for constructive and cost-effective action than does supplier or lot rejection based only on comparisons. However, comparison may be the only means to test materials made unstable by agents below limits of detection or components that are not recognized, such as ions or small molecules that may be bound to the raw material.
Comparative analysis is straightforward with today's off-the-shelf software packages. Database and spreadsheet programs manage the data from empirical comparisons. Analysis and graph-ing tools facilitate the finishing steps, and transfer of data from database to analysis tools is convenient. Before the availability of such tools, these activities were cumbersome and tedious.
Effects of Cofactors
Many reagents employ secondary reactions to provide a readily measured light absorption or light emission signal. Secondary antibodies to which enzymes are coupled in ELISA methods are linked to substrates that absorb or emit light.
Coupling of dehydrogenases that oxidize NADH or reduce nicotinamide-adenine dinucleotide (NAD) to reactions that produce a substrate for a dehydrogenase is a classic general chemistry method. Such coupling is also used for detecting enzyme inhibition by drugs of abuse and therapeutic drugs in some drug assay methods.
Evaluation of stability in these systems must include consideration of the indicator reaction components as well as the primary antibodies or enzymes. Direct, spectrophotometric measurement of NADH/NAD makes this NADH oxidation or NAD reduction easily determinable. However, the nonenzymatic oxidation of NADH can be sufficiently great under some solution conditions, particularly at neutral pH, that it must be compensated for by additional reactions that slowly regenerate NADH from NAD.
Independent assessment would be straightforward if stability stemmed from only the overall performance of reagents that involve the NADH/NAD couple. In actuality, effects of reaction products of the indicator reaction on the assessment can be confusing.
For example, many dehydrogenases catalyze reactions that are thermodynamically reversible. Lactate dehydrogenase (LDH) in the International Federation for Clinical Chemistry method for alanine aminotransferase couples the oxidation of NADH to the conversion of pyruvate to lactate for the forward reaction. The reverse reaction converts NAD to NADH in order to oxidize lactate to pyruvate. The affinities of the LDH for binding NADH and NAD are so similar that the presence of NAD in a reagent could significantly slow the reaction. Such an effect, if unrecognized, could produce a misleading interpretation of the cause of the instability.
Other dehydrogenases, such as glutamate dehydrogenase (GDH), exhibit complex dissociation behavior that is influenced by allosteric effectors, such as adenosine diphosphate (ADP).11 In cases such as this, loss of activity could be the result of degradation of the ADP with consequent changes in the GDH, rather than proteolytic degradation.
Understanding the reactions themselves is important for achieving optimal stabilization of systems such as these. The importance of substrate stability is also shown by products in which special conditions for the stabilization of p-nitrophenyl phosphate for alkaline phosphatase-coupled ELISA systems have been devised.
Stability was defined previously as constancy of property with time. Most frequently, instability presents itself as loss of a property with time. However, the author as a customer has had experience with a liquid control that was stored and shipped frozen by the manufacturer. Activity in this substance increased over several days after initial thawing, finally becoming constant.
Although the true cause for this behavior was never identified, one explanation might be stabilization of the relevant enzyme in the control material with a substance that inhibited the enzyme (a long-established method for increasing intrinsic stability). If the inhibitor were degraded slowly after thawing, decreased inhibition would be observed as increased activity.
This example, involving a run control material, illustrates that degradation of an inhibitor, which could have been enzymatic conversion of it to a noninhibitory substance, can result in enhanced activity. For a control material, this phenomenon is as unacceptable as loss of activity.
Adsorption
One mechanism by which loss of activity with time can occur is adsorption to the walls of the product container. Adsorption losses are most dramatic in solutions or reagents in which proteins are present in low concentrations. Such losses frequently occur quite rapidly, with an immediate large loss followed by a much more gradual one.
When adsorption is suspected, the dependence of the loss on the concentration can be monitored. Usually the percentage loss in activity decreases as protein concentration increases.
Plotting the amount lost (original amount minus the amount measured at equilibrium) creates a binding curve, or isotherm. This isotherm may be compared with those obtained using different container materials. The comparison provides a rational basis for selection of a container or vial for the product.
Researchers may vary the area of the container surface with which the product is making contact
to observe the dependence of performance loss on total surface area. Such assessments at a constant concentration and constant total amount of protein provide evidence diagnostic for adsorptive loss.
Adsorptive losses can also occur during the performance of assays that have very high sensitivity. Dilution to obtain a rate or a time interval suitable for the measurement process or instrumentation may be the culprit.
In such a case, the diluent should contain another component to prevent the adsorption loss. Such components (for example, polyethylene glycol solutions or solutions of inert proteins, such as serum albumin) act as competing adsorbates. By adsorbing preferentially to the surfaces of the containers or cuvettes, they reduce or eliminate the adsorption loss of the component of interest.
Comparison of the adsorption of thrombin in containers of different compositions under different conditions.
Stirring frequently increases the rate of adsorption loss, because diffusion may limit adsorption in unstirred containers. The ability of the container material to adsorb depends on the particular biological macromolecules in the solution. Relatively inert materials, such as polypropylene, adsorb substantial quantities of enzymes, such as thrombin. Coating the polypropylene surface with a high-molecular-weight polymer such as polyethylene glycol (PEG) 20,000, and adding a competing adsorbate such as PEG 8000 to the buffer, reduce the adsorption loss of thrombin to an undetectable amount for longer than one week, even at room temperature.12,13
Competitively adsorbing materials may actually not be inert. They may do more than promote stability by a competitive adsorption mechanism. Serum albumin, for instance, may contain proteases from the blood clotting cascade and thus introduce agents that proteolytically degrade the material of interest.
Protein "competitive adsorbates" are also alternative substrates for proteases that might contaminate the reagent. In this capacity they exert an additional stabilizing effect, which might be observed only if another material, such as polyethylene glycol, were substituted for the protein.
As noted previously, compounds such as supposedly inert proteins and polyethylene glycol also produce an excluded-volume effect. In so doing, they change the extent of association of multisubunit enzymes or complex formation between and among components of the test kit.1 For example, polyethylene glycol through its excluded-volume effects promotes association of antigen-antibody complexes in immunoturbidimetric and nephelometric assay procedures.
Occasionally, even though the manufacturer is aware of and has controlled an adsorption loss problem in the reagent or kit, end-users who dilute the reagents in an attempt to increase the number of tests per kit can reintroduce adsorption loss. When dilution causes a reduced concentration of a competitive adsorbate, it can result in quantitatively significant adsorption of a critical reaction component, and thus adversely affect test performance.
Adsorption is of course not universally bad. While in one situation it creates product loss that is seen as instability, in another situation it is the mechanism by which ELISA plates are coated with antibody or antigen in creation of the convenient microtiter plate and immunobead assays.
Conclusion
Although the discussion in this article has been directed to IVD product stability and has used as the primary examples proteins and proteases, most of the same considerations can be made for polynucleotides and oligonucleotides and nucleases.
The author's bias toward identification of root cause, and corrective action directed to root cause, is obvious. It reflects the underlying belief that this focus is the most definitive approach to solving any problem, not just instability. Root-cause solutions are in the long run the most cost-effective solutions.
The primary advice from this article is, "Know thy product." This goal includes knowing the raw materials and their properties, the probable contaminants in them, and the means by which contamination that might have no effect on the raw material in which it originates could have a deleterious effect on some other component of the product.
Product and process validation, as well as in-process testing, can be most effective as well as least expensive when knowledge of the product and its components allows focus solely on critical variables. The new product-design specification requirements, when based on the principles outlined in this article and implemented using personal computer database, spreadsheet, and data analysis tools that are available off the shelf, can produce long-term savings. These savings stem chiefly from reduced costs associated with manufacturing processes and with customer support.
In troubleshooting product manufacturing failures--that is, failures of a manufactured lot to meet release criteria--and in solving problems encountered by end-users, such detailed knowledge is power. Knowledge allows the manufacturer to keep support costs as low as possible.
The manufacturer who fails to resolve stability issues before a product hits the market will receive a flood of complaints when customers begin to use it. Resolving customers' complaints will be costly, product acceptance will be low, and the product may well turn out to be unprofitable. Ultimately, products that best solve customers' problems are those most likely to survive. Today's customer looks for products of lower cost that are more convenient to use, less prone to laboratory staff error, and so stable and reliable that they consume little of laboratory management's scarce attention.
Return to the IVDT home pageReferences
1. Winzor DJ, and Wills PR, "Effects of Thermodynamic Nonideality on Protein Interactions: Equivalence of Interpretations Based on Excluded Volume and Preferential Solvation," Biophys Chem, 25:243252, 1986.
2. Guruprasad K, Reddy BVB, and Pandit MW, "Correlation between Stability of a Protein and Its Dipeptide Composition: A Novel Approach for Predicting In Vivo Stability of a Protein from Its Primary Sequence," Protein Eng, 4:155161, 1990.
3. Bergmeyer HU (ed), Methods of Enzymatic Analysis, 3rd ed, Deerfield Beach, FL, Verlag Chemie, 1983.
4. Walsh G, and Heanon D, Protein Biotechnology, New York, Wiley, p 210, 1994.
5. Bollag DM, Rozycki MD, and Edelstein SJ, Protein Methods, 2nd ed, New York, Wiley-Liss, p 415, 1996.
6. Creighton TE, Proteins, Structures and Molecular Properties, 2nd ed, New York, W.H. Freeman, p 505, 1993.
7. Fenton JW, Fasco MJ, Stackrow AB, et al., "Human Thrombin: Production, Evaluation, and Properties of Alpha-Thrombin," J Biol Chem, 252:35873598, 1977
8. Bode W, Mayr I, Baumann U, et al., "The Refined 1.9-Crystal Structure of Human a-Thrombin: Interaction with D-Phe-Pro-Arg Chloromethylketone and Significance of the Tyr-Pro-Pro-Trp Insertion Segment," Embo J, 8:3467 3475, 1989.
9. Bernstein FC, Koetzle TF, Williams GJB, et al., "The Protein Data Bank: A Computer-Based Archival File for Macromolecular Structures," J Molec Biol, 112:535542, 1977.
10. Di Cera E, Guinto ER, Vindigni A, et al., "The Na+ Binding Site of Thrombin," J Biol Chem, 270:2208922092, 1995.
11. Huang CY, and Frieden C, "The Mechanism of Ligand-Induced Structural Changes in Glutamate Dehydrogenase: Studies of the Rate of Depolymerization and Isomerization Effected by Coenzymes and Guanine Nucleotides," J Biol Chem, 247:36383646, 1972.
12. Latallo ZS, and Hall JA, "Reaction of Thrombins with Human Antithrombin III: Enzyme Activity Losses Unrelated to the Inhibitory Reaction and Their Prevention," Thromb Res, 43:507521, 1986.
13. Wasiewski W, Fasco MJ, Martin BM, et al., "Thrombin Adsorption to Surfaces and Prevention with Polyethylene Glycol 6000," Thromb Res, 8:881886, 1976.
14. Soriano-Garcia M, Padmanabhan K, De Vos AM, et al., "The Ca2+ Ion and Membrane Binding Structure of the Gla Domain of Ca-Prothrombin Fragment 1," Biochem, 31:25542566, 1992.
15. Soriano-Garcia M, Park CH, Tulinsky A, et al., "Structure of Ca2+ Prothrombin Fragment 1 Including the Conformation of the Gla Domain," Biochem, 28:68056810, 1989.
Craig M. Jackson, PhD, is a consultant to the in vitro diagnostics manufacturing industry and a member of the IVD Technology editorial advisory board.
A general strategy for improving
liquid IVD product stability
Although the focus of the accompanying article is biochemical investigation and evaluation, judiciously narrowing the scope of the laboratory investigation makes sense. Particularly if instability involves an existing product, reviewing the history of the problem and records from the development of the product can provide clues otherwise likely to be overlooked. Document review and prior thought are less costly than laboratory activity.
On the other hand, if the stability problem to be solved is for a new product, then a review of the scientific literature might be most important. A review of the development and manufacturing history of that product would be irrelevant, because no knowledge base exists from previous in-house work. Following is a step-by-step evaluation procedure:
1. Review what is known about the properties of the biological components.
External Sources and Literature. Review relevant manufacturer specifications, technical descriptions, scientific monographs, review articles, and original research papers. One goal is to evaluate the advantages and disadvantages of different animal, tissue, microbial, and recombinant biological sources. Another is to assess possible deleterious contaminants that may be present in different source preparations. The latter may require asking difficult questions of the prospective supplier. Information from such external sources should enable the manufacturer to prepare a summary of the conditions--solution compositions, pH ranges, and temperatures--at which individual components can be expected to be stable.
Internal Sources. Review existing research and manufacturing data, when available. Observations made during original R&D and transfer to manufacturing may identify some components likely to be unstable. For existing products, look for lot-to-lot differences from raw material qualification data files. Obtain insights from colleagues with experience inside your organization, and outside it if appropriate. Don't waste time rediscovering what is already known.
2. Define the most stable product that can be expected based on the current knowledge of the formulation's intrinsic stability.
This is the reference situation unless the goal is to discover new, better conditions for stability or for stabilizing the material. Summarize the maximum stabilities to be expected, from the sources noted above, for each component or reactant alone. Note the benefits associated with alternative sources (supplier, animal source, tissue source, recombinant product, etc.).
Identify restrictions likely to preclude obtaining stability as good as that intrinsic to the component. Composition restrictions appropriate for one component may be adverse for another component in the final product. Use of stabilizing agents may be restricted in countries in which the product will be sold.
Relate the anticipated stability properties to the proposed product design specifications. This step is an early reality check on specifications from marketing.
Describe the stability potentially achievable under the best possible situation. Contrast this scenario with alternative situations (e.g., using a more expensive component from a different source).
Note and communicate possible limitations that may be encountered in manufacturing the product. Include the potential inability to manufacture the product if instability considerations create the need for unusually tightly controlled processes or conditions.
3. Conduct laboratory investigations to identify causes of inadequate stability as well as ways to improve stability.
Distinguish reversible from irreversible activity loss. Return the component to the pH and solution composition at which activity was previously stable. A recovery of activity suggests reversibility; failure to regain activity suggests an irreversible inactivation process.
Macromolecule Association or Dissociation. Activity loss from macromolecule association or dissociation is frequently reversible. If this cause is suspected, evaluate dependence on solution composition and pH and total protein concentration. Association and dissociation are both influenced by these variables. However, if higher concentrations favor inactivation, the increased concentration of a contaminating degradative enzyme could also be responsible. Alter allosteric or enzyme cofactor concentrations to compensate for binding constant changes that occur with changes in pH, ionic strength, and other variables. Look for changes in light scattering (turbidity). Also look for precipitation at the point of sample application to electrophoresis gels under nondenaturing conditions, especially if precipitation was not observed with other solution compositions.
Adsorption. Activity loss from adsorption is usually irreversible. If this cause is suspected, first determine if the loss is dependent on container material. Multiple transfers of solution to new containers can be used to "amplify" this type of adsorption loss. Second, determine if the loss is influenced by the presence of competing adsorbates (e.g., polyethylene glycol, nonionic detergents, "inert" proteins). Be aware that polyethylene glycol, detergents, and other proteins may promote enzyme self-association. Measure mass loss as well as biological activity loss, if possible. Adsorption produces mass loss from solution.
Proteolysis. If activity loss is suspected to be due to proteolysis, determine if the loss is reduced by a protease-inhibitor cocktail. Use a combination cocktail to implicate protease-related degradation. Use protease type-specific inhibitors to distinguish trypsinlike proteases, chymotrypsinlike proteases, thiol proteases, and metalloproteases. Determine if decreases in molecular size are occurring by applying molecular sizedependent separation procedures, such as gel electrophoresis. Sodium dodecyl sulfate polyacrylamide gel electrophoresis is well suited to this application. Use Western blotting techniques, if antibodies are available, to increase sensitivity. Gel permeation or gel filtration chromatography may be useful for detecting size changes.
Determine if other alterations in the component are occurring. By use of ion-exchange and reverse-phase chromatography, look for elution behavior differences between freshly prepared and "aged" components. Compare the stability of chromatographed and unchromatographed raw materials to implicate removal of an undetectable contaminant that might be responsible for degradation.
Microbial Contamination. Microbial enzymes, without obvious microbial growth, may cause activity loss. When this cause is suspected, one option is to add an antimicrobial agent. Chloroform effectively disrupts cell membranes and by doing so prevents microbe growth, but it may not be usable in the product. Sodium azide is effective against aerobic organisms only, but its use is controlled in some countries. Antibiotics may increase stability but may also promote development of resistant strains of microbes. Another option is to ameliorate the situation by sterile filtration.
Other Contaminants. Sometimes a contaminant in one component makes another component unstable. When this process is suspected, prepare single-component solutions and two-component mixtures of otherwise identical composition in all combinations. Evaluate single-component and two-component mixtures by the laboratory approaches described above. Compare combinations to identify the sources of the component causing degradation.
Return to the IVDT home page
