4 REVIEW OF THE ENZYMOLOGY LITERATURE

WHILE THE STUDIES cited above place appropriate emphasis on the speed, specificity and apparent safety of enzyme treatments, some misapprehensions about enzymes do appear in the conservation literature.

First, the importance of pH and temperature to the efficacy of enzyme reactions is generally overemphasized. Published treatment procedures usually include the prescribed pH and temperature optima for the enzyme employed and state that the enzyme cannot function efficiently unless it is used in an environment controlled by these optima.25 For example, articles discussing the use of the S. griseus protease and the B. subtilis α amylase have always included instructions for the preparation and maintenance of solutions that meet the optima of 37�C and pH 7.5.26

However, for the enzymologist temperature and pH optima are not absolute but relative terms.27 It is important to note that while a so-called temperature optimum may increase the rate of the reaction between enzyme and substrate, it may also increase the rate of the enzyme's inactivation.28 The pH also affects several factors simultaneously because it alters the ionization state, or charge, of the enzyme and substrate.29 The factors affected by pH include the affinity of the substrate for the enzyme, the velocity of the catalyzed reaction, and the stability of the enzyme.30 However, such effects are not important if there is an excess of enzyme or substrate.31 Since an excess of substrate (old glue or starch paste for the conservator) often prompts the decision to use an enzyme, close observance of pH optima is not essential. It should be remembered that enzymes are stable and thus will remain active over a wide range of pH. For example, α amylase from B. subtilis is stable from pH 5–12; protease from S.griseus from pH 5–10.32 Similarly, although each 10�C rise in temperature can quadruple an enzyme's reaction rate, enzymes will function within a broad temperature range and can be used effectively at room temperature.33

The purpose of the above discussion is to stress that enzymes lend themselves to experimentation. There are a number of variables available to the conservator in a system of enzyme, water and starch paste or animal glue. In many cases, a satisfactory working environment for the enzyme can be produced by the pH of the moistened accretion or the pH of the deionized water used to make the enzyme solution. A slightly elevated temperature can be used to offset a pH that is far from the optimum but which the conservator is not at liberty to change.34 Finally, although the velocity of the reaction would doubtless be enhanced by strict maintenance of one or the other if not both optima, satisfactory results can be obtained even when ignoring both conditions.35

The second point to be discussed is the safety of enzyme treatments. The safety of such treatments is dependent upon a number of factors pertaining to (1) the purity of the enzyme, (2) the possibility of removing the enzyme from the paper and (3) the possibility of the irreversible inactivation of any residual enzyme.

The specifications for enzyme purity established by most purchasers will usually be far less stringent than those required by conservators. An unadulterated enzyme would be useless to the majority of a chemical company's customers. Salts, acids, preservatives and fillers such as sawdust must be added to any enzyme which is to be subjected to the industrial environment.36 An enzyme that one buys “off the shelf”, such as digestive aids or contact lens cleaners, has been similarly treated and would not be suitable for our purposes.37 The conservator should also be aware that preparations designed for enzyme research often contain stabilizers.38 Consequently, a careful reading of the product description is essential even when purchasing enzymes from companies specializing in reagent grade materials. One should choose enzymes described as “free from foreign extenders”.Preparations described as “crude” and preparations that already include buffers should be avoided. Mixtures of several enzymes sold as one reagent have been found useful in some situations.39 However, one must be certain that all the components of such mixtures are listed in the product description, since many enzymes have properties that could necessitate changes in one's treatment procedure. For example, the protease from S. griseus is stabilized by calcium against heat inactivation (a property to be discussed below).40

There is a potential danger of an impurity of cellulase, the enzyme that breaks down cellulose, in any enzyme preparation that the conservator might buy, since the microorganisms which produce protease and α amylase also produce cellulase. However, purification techniques have become so sophisticated that the realization of this potentiality is quite improbable. Indeed, methods are currently being developed whereby the DNA of the enzymesynthesizing microorganism is adjusted so that the desired enzyme is the only one produced.41 Even more reassuring is the resistance of cellulose to cellulase attack. Commercial cellulases work well on cellulose derivatives like carboxymethylcellulose, but do not work or work very slowly on native cellulose.42 A pulp of cotton linters must be fine ball milled for an hour to enable cellulase to attack it, and even then complete breakdown of the cellulose takes ten days.43 A pulp containing lignin would require even more vigorous conditions, since lignin blocks an enzyme's accessibility to cellulose.44 Finally, a sample of cellulose will resist enzymatic breakdown even after it has been hydrolyzed by dilute acids,45 so a degraded, acidic paper should not easily be attacked by a chance cellulase impurity.46

There are three other characteristics which the conservator should evaluate before purchasing an enzyme preparation. Enzymes derived from thermophilic bacteria may not be desirable in some cases, since these are unusually heat stable (a property to be discussed below).47 Such bacterial properties are found in Bergey's Manual of Determinative Bacteriology,48 a reference book that is available in any college library. Some scientists also speculate that an enzyme of low molecular weight would be more easily washed out of the paper.49 Although this proposition has yet to be proven, conservators might consider purchasing a low molecular weight reagent, which for enzymes would be about 15,000 atomic mass units,50 rather than one of a high molecular weight, which would be over 50,000 atomic mass units.51 Finally, it is important to consult an enzyme's product description for data describing the rate of the enzyme's hydrolysis reaction. This information appears under headings like “activity” and “unit definition”.If several preparations satisfy the conservation specifications discussed above, one should choose the enzyme which offers the highest number of units for a given weight of reagent (often referred to as “solid”). As a result, one should increase the possibility of removing the enzyme preparation from an artifact during a rinse step, since a more powerful reagent can be used in more minute percentages.

As implied above, the safety of an enzyme treatment not only depends upon the proper choice of enzyme preparation but also upon an effective rinsing step. The conservation literature reveals a prevalent mistrust of the adequacy of a rinse step to effect the complete removal of enzymes from an artifact. Recommendations for the inactivation of any residual reagent thus appear in published treatment procedures. Application of alcohol to areas that have been treated with enzymes and subsequently rinsed has been suggested as a method for the inactivation of any residual enzymes.52 However, if this inactivation step is to be effective, the area treated with enzymes must be flooded with alcohol. Studies made by Cooper, King and Segal led them to propose the following reaction:

Fig. .

The behavior of enzymes in organic solvents is not necessarily due to the creation of an enzyme-solvent complex. Enzymologists have explained the interaction of enzymes and solvents in terms of the ionization of a solvent-enzyme as opposed to a water-enzyme mixture. Ethanol has a lower dielectric constant than water. Since a decrease in dielectric constant increases the attractive force between two opposite charges, ethanol decreases the ionization of proteins and thus promotes their coalescence.56 An enzyme can be seen, then, as opened up and free to interact with the substrate when in water, and closed, and therefore not as accessible to the substrate, when in an organic solvent. The coalescence of such an inactivated enzyme has a practical implication for the conservator: as Segal has noted,57 a coagulated enzyme would be difficult to remove from paper. Consequently, if inactivation is deemed necessary, it must follow the rinsing step.

It is noteworthy that an enzyme inactivated by alcohol could possibly reactivate and again serve as an efficient catalyst if it were returned to a temperature and pH within its stability ranges.58 The likelihood of this possibility is diminished by using a minimal amount of enzyme and by rinsing the object thoroughly before the solvent inactivation step.59

Inactivation is also caused by denaturation, which has been defined as the unfolding of a protein's folded, native structure to an open coil.60 Inactivation by denaturation can be reversed,61 but this possibility is again unlikely if one is careful to reduce the amount of enzyme present by using the minumum required and by thoroughly rinsing the treated area before the denaturation step.62

Immersion in a water bath maintained at 50�C has been prescribed to denature an enzyme that might remain in a treated, rinsed paper,63 but this method is not universally applicable. For example, such a denaturation step would have no effect on the known α amylases (derived from B. subtilis, Bacillus stearothermophilus, Aspergillus oryzae and Aspergillus niger).64 These enzymes are metalloenzymes including calcium in their structure, and researchers have found that the bound calcium enables these enzymes to withstand temperatures up to 100�C.65 Denaturation of the α amylases at temperatures of 100�C or below would require the removal of the protective calcium with a chelating agent like EDTA.66

The protease derived from S. griseus enjoys a similar immunity to elevated temperatures.67 Although calcium is not part of the native structure of the S. griseus protease, this enzyme will take up calcium, magnesium, and a number of other divalent cations from solution and thereby be protected against inactivation by heat.68 The amounts of calcium or magnesium needed for these effects is minute, on the order of 0.0015 grams of these divalent cations per six liters of enzyme solution.69 Conservators should be alerted to this property of the S. griseus protease, since many of us have used some form of calcium or magnesium to buffer solutions of this enzyme to its optimum pH of 7.5. We have thereby unwittingly stabilized this enzyme against changes in its environment. It should be noted that a final deacidification step would also stabilize any residual enzyme.

Many enzymes are not protected by divalent cations against changes in their environment. For example, the protease derived from Aspergillus saitoi (to be presented below) would be inactivated by a bath of 55�C, and if enough calcium or magnesium ions were present to raise the pH to eight, this inactivation would be irreversible.70 It has become apparent to this author that consultation of the enzymology literature is imperative when choosing an enzyme for conservation purposes. Before ordering an enzyme, the conservator should refer to The Enzymes, edited by Paul Boyer.71 This set of fifteen volumes is available in any college library. The spines of each volume identify the categories of enzymes discussed according to the reaction they catalyze, such as “Hydolysis. Glycosides”.The index is well cross-referenced, and one can research enzymes not only by their general name (such as “α amylase”) but by one of the microorganisms which synthesizes them (such as “Bacillus subtilis”). Pertinent chapters present summaries of all the currently known facts about each enzyme's stability ranges, optima, and useful and damning peculiarities.

If not stored in cool, dry conditions, enzyme preparations suitable to conservation use72 will eventually lose their catalytic properties. Unfortunately, it is unclear just how long the enzymes remain in a condition that makes reactivation possible. According to one microbiologist, an enzyme dried on paper could retain the ability to be reactivated for several years.73

An unsophisticated experiment performed by the author suggests that enzymes reactive for longer periods than one might suspect and emphasizes the importance of sufficient rinsing and some type of inactivation step. A piece of paper that had been coated with a thick layer of gelatine and oven aged for three days at 100�C was immersed in a 0.5% solution of the protease derived from Aspergillus saitoi (hereafter referred to as A.saitoi) until the gelatine dissolved. The treated paper was given a cursory rinse with tapwater and remained uncovered for five weeks on an exposed shelf in the Paper Conservation Lab of the National Gallery of Art. The paper was then cut in half. One half received an application of warm gelatine, was folded, and left under pressure to dry. After drying, the paper came apart easily: it had not adhered. The A.saitoi protease is active in the acid pH range and at body temperatures,74 and the warm gelatine, which had a pH of four, apparently created an environment in which the enzyme could reactivate. The author is reasonably certain that this phenomenon was due to residual enzyme, since running the above-described “adhesion test” on a desized, but otherwise untreated paper produced a well-adhered sample.

The remaining half of the enzyme treated paper was cut in two. One piece received a forty-five minute soak in deionized water raised to pH eight with calcium and magnesium. These conditions are supposed to irreversibly inactivate the A.saitoi protease,75 and after being subjected to the “adhesion test”, the paper did indeed adhere to itself. The other piece received a fifteen minute soak in ethanol. This treatment apparently inactivated the enzymes as well, for the paper remained bonded after undergoing the “adhesion test”.

The author recogizes that this experiment is only a crude preliminary study. A 0.5% enzyme solution was used, and initially, the paper was rinsed only briefly. Treatments designed for works of art on paper require very different procedures. The least possible percentage of enzyme is employed and the sheet is rinsed as extensively as the nature of the object permits. Nonetheless, it is significant that in this experiment the enzyme apparently survived five weeks of exposure to dust, light and room temperature. Since unfavorable storage conditions do not easily destroy enzymes, it seems worthwhile to seek enzymes that can be irreversibly inactivated by methods that are safe for art objects.

It should be noted that some scientists present a strong case for omitting the inactivation step from enzyme treatments when the least possible percentage of enzyme is used and when the object can be rinsed extensively. Scientists who hold this view stress the importance of keeping the enzyme stable during the initial rinsing period to eliminate any danger of the enzyme coagulating while still in the paper. To this end, a rinse in any buffers used for treatment should precede rinsing in the wash-water appropriate for a final bath. In addition, solvents should not be applied to a sheet that has been treated with enzymes and has yet to be rinsed.76

The author concurs that careful rinsing must precede any attempts to inactivate an enzyme. Moreover, her research indicates that the minute percentages of enzyme used by conservators should not harm paper, even if the enzyme is not completely removed or inactivated. Nonetheless, until data for a wide variety of papers can be obtained regarding the susceptibility of enzymes to rinsing,77 it can only be advantageous to seek an enzyme which is not more resistant to removal by rinsing than other reagents and for which the inactivation procedure is actually considered beneficial to many works of art on paper. Consultation of Boyer's opus uncovered a protease that is appealing for these reasons.78

Presentation of the protease “Aspergillopeptidase A”.The “new” protease is derived from Aspergillus saitoi (mentioned above). Its properties make it preferable to the S.griseus protease that many of us have used in the past. Since the A.saitoi protease is an acid protease, it is active when the pH of its solution is below seven and is inactivated when the pH is above seven.79 Indeed, unlike the protease from S.griseus, A.saitoi protease is not stabilized by calcium, magnesium or any other divalent cation, and will be completely and irreversibly inactivated if subjected to a bath raised to pH eight using whatever alkaline solution the conservator prefers.80

For glue accretions which are acidic, even mildly acidic at pH 6, the A.saitoi protease can be used without a buffer.81 However, when not employing a buffer, it is suggested that the optimum temperature of 37�C be maintained to derive the full benefit of using the enzyme.