J. clin. Path., 24, suppl. (Ass. Clin Path.), 4, 22-30

Accuracy, precision, and quality control of enzymeassaysD. W. MOSS

From the Department of Chemical Pathology, Royal Postgraduate Medical School, London

Improvements in standards of analytical accuracyand precision are now firmly established goals inclinical biochemistry. A number of factors havecontributed to this. Among them are the appreciationof the value of sequential studies of changes in levelsof plasma constituents during the course of anillness, provided that these changes can be dis-tinguished reliably from the inherent variability ofthe analytical procedures. Furthermore, increasedaccuracy will lead to a higher degree of standardi-zation of clinical laboratory data: this is importantin view of the growing mobility of populations andthe likelihood that a patient will attend more thanone hospital. Improvements in analytical practicehave themselves served to indicate the need for stillgreater accuracy and precision, in that they haveshown the clinical significance of quite minorbiochemical changes. Most clinical chemistrylaboratories now operate quality control programmesroutinely, and the comparison of the analyticalperformance of different laboratories by interchangeof specimens of known composition is increasing.However, enzyme assays have so far benefited lessthan other procedures from advances in methods ofquality control. This is largely due to the physico-chemical nature and functional characteristics ofenzymes themselves. Efforts to improve the qualityof analysis have largely been directed towardsmonitoring and improving within-batch and be-tween-batch precision and improving accuracy. Inenzyme analysis the principal difficulty is to find asatisfactory method by which the accuracy of adetermination can be assessed. This paper discussespossible approaches to the calibration or standardi-zation of enzyme assays, as well as problemsencountered in assessing precision.Enzymes are catalysts, and in almost all cases are

assayed by measurements of their catalytic activities,ie, the extent to which they increase the rate of agiven chemical reaction. This is dependent on theconditions under which it is measured, including the

22

concentration of the reactants, temperature, andpH. The sensitivity of the rate of a chemical reactionto changes in such conditions is generally muchgreater in the presence of an enzyme than in itsabsence. As enzyme activities depend on the con-ditions under which they are determined, a particularunit ofenzyme activity must depend on the particularmethod of assay. It is not practicable at present todetermine absolute quantities or molar concen-tration of an enzyme except in a very few cases. Agiven enzyme is rarely present in a sufficiently pureform for its concentration to be estimated by adetermination of protein. Also, while a knowledgeof the specific molecular activity of the enzyme underthe conditions of assay would allow activity measure-ments to be translated into molar concentrations,such information is scanty; even if this were availableestimations of activity under strictly defined con-ditions would still be necessary. Since estimation ofenzymes for clinical purposes must continue todepend on activity measurements, two approaches tostandardization are possible. The first is the fomu-lation of agreed standard methods of assay whichdefine corresponding enzyme units. The second isthe provision of reference preparations of enzymes towhich are assigned agreed values of activity; thesepreparations may then be used as calibrationstandards in secondary methods of enzyme estima-tion, or as quality control preparations to assessanalytical performance. The two approaches are notmutually exclusive; indeed, the existence of referenceenzyme preparations presupposes that these havebeen assigned values on the basis of activity deter-mination, unless a completely arbitrary choice ofvalues is accepted. Therefore, standard methods ofanalysis and the availability ofreference preparationsof enzymes are both desirable in clinical enzymology.At present agreement on standard methods of analy-sis, at least for the more frequentlyestimated enzymes,seems to be the more easily attainable goal and onethat is being actively pursued by specialist groups.

copyright. on M

ay 26, 2020 by guest. Protected by

http://jcp.bmj.com

/J C

lin Pathol: first published as 10.1136/jcp.s1-4.1.22 on 1 January 1970. D

ownloaded from

Accuracy, precision, and quality control of enzyme assays

The Nature and Origin of Diversity in Enzyme AssayProcedures

The prime consideration in the design of an enzymeassay method is that the amount of enzyme presentshould be the only variable which affects the reactionrate. This condition is more likely to be met if theassay procedure ensures optimal substrate con-centration, pH, temperature, etc, and thus many ofthe modifications that have from time to time beenmade to assay methods have been motivated bythe desire to achieve optimal conditions. Unfor-tunately, it has not always been appreciated thatsuch modifications almost inevitably result in analteration in the enzyme activity as measured by themethod.The use of suboptimal conditions does not

necessarily invalidate an assay procedure, particu-larly if only one type of enzyme or isoenzyme is to bemeasured. For example, the ratio of two enzymeactivities will be independent of substrate con-centration provided that the two samples havesimilar Michaelis curves, ie, the same Michaelisconstant, Km. However, when the Michaelisconstants of the two samples differ, the ratio of thetwo activities becomes increasingly dependent onsubstrate concentration as this falls progressivelybelow saturating levels (Fig. 1). One source ofvariation in Km from one enzyme sample to anotheris a difference in their isoenzyme composition, suchas is well recognized in the case of lactate dehydro-genase isoenzymes. It should be emphasized,however, that changes in Km of the magnitude ofthose shown for the alkaline phosphatase samples inFig. 1 can arise even when no change in the isoenzymeis involved. Alkaline phosphatase Km values areaffected by the composition and degree of purity ofthe enzyme solution (Moss, Campbell, Anagnostou-Kakaras, and King, 1961) so that different serumsamples do not necessarily show the same Km valueseven when they contain the same isoenzymes. TheMichaelis constant of alkaline phosphatases is alsovery markedly dependent on pH in the region ofpH0, at whichphosphatase assays areusually carriedout (Fig. 2), so that inadequate control ofpH cancause variation in the apparent Km. Vesell and othershave pointed out the effect of temperature on theshape of the Michaelis curves of lactate dehydrogen-ase isoenzymes (Vesell and Pool, 1966). Similar con-siderations apply to other variables in the enzymeassay system, notably the effects ofpH and tempera-ture. The shape of the pH-activity curve of anenzyme in the region of the optimum is largelygoverned by the ionization characteristics offunctional groups in the active centre, but at moreextreme pH values factors such as enzyme stability

z

0

C.)

cJ

LA.0

0-JU>'

2

[SUBSTRATE] mMM

L&J

.)

ocL&.0

C-

Fig. I Relationship between reaction velocity (j,molssubstrate hydrolysedper min per ml ofenzyme solution)and substrate concentration (mMp-nitrophenyl phosphate)atpH 10J0 for three specimens ofhuman alkalinephosphatase. A(O) and B (D) are both liver phosphataseand have the same Km value; C (A) is intestinalphosphatase with a slightly different Km. The ratio of theactivities ofA and B (0) is independent of the substrateconcentration at which the activities are measured, butthat ofA and C (A) is increasingly dependent onsubstrate concentration as this is decreased.

come into play to an extent which may affect enzymeconcentration. Thus, the ratio of the activities of twosamples of the same enzyme is independent of thepH of measurement when this is close to the pHoptimum, but may not be so when sub optimal pHvalues are used. Figure 3 shows this effect for alkalinephosphatase. The several variables in the assaysystem may be so interrelated that it may not bepossible to achieve optimal values for each and somedegree of compromise may be necessary. This isespecially true when fixed-time assays are used orwhen a single assay system is to be applied to sampleswhich contain mixtures of isoenzymes in varying

23copyright.

on May 26, 2020 by guest. P

rotected byhttp://jcp.bm

j.com/

J Clin P

athol: first published as 10.1136/jcp.s1-4.1.22 on 1 January 1970. Dow

nloaded from

D. W. Moss

8 9 10pH

S._

04

Fig. 2 Relationship between Km value (a; mMp-nitrophenyl phosphate), Vmax (0; ,umols substratehydrolysedper min per ml ofenzyme solution) and pH ofmeasurement for human bone alkaline phosphatase,showing the marked dependence ofKm on pH in theregion ofpH 10 (data from Moss, 1969).

08. -

-.kIN

A

N..N.~~~~~~~~~~~~~11

0'4 L.

9

N'.

10

.4

in

4.5

0

2

LI

pHFig. 3a.Fig. 3 A Activity-pH curves for two samples oJhuman intestinal alkaline phosphatase (O and @). Theratio of the activities (broken line) of the two samples is

- independent ofpH only over a narrow range around thepH optimum.

08

E

1-.a

E

04

4- %vo._

0

o o0 _

a >.

n

B

N

I

v rI I

9 10

I-0

0

4.

pHFig. 3b.B Thte same for two liver alkaline phosphatase samples;here the range over which the ratio is independent ofpHis even narrower. (Substrate in both A and B, 4mMp-nitrophenyl phosphate.)

24

S

S

2

E

ye

OL7

C

,-.

0E

>5

.)0

>0

0 v

_"20

I I I

copyright. on M

ay 26, 2020 by guest. Protected by

http://jcp.bmj.com

/J C

lin Pathol: first published as 10.1136/jcp.s1-4.1.22 on 1 January 1970. D

ownloaded from

Accuracy, precision, and quality control of enzyme assays

proportions, as is the case with human lactatedehydrogenase in serum, since optimal conditionsfor the activity of isoenzyme 1 are significantlydifferent from those for isoenzyme 5 (Gay, McComb,and Bowers, 1968).Another reason for the modification of methods of

enzyme estimation is to improve analytical sensitivityor specificity of the methods that are employed todetermine the amount of chemical change takingplace. The scope for this type of modification isgreater when enzyme specificity is low so that thenature of the substrate can be varied, as exemplifiedby the many substrates that have been used in theassay of alkaline phosphatase; but even withoutchanging substrate, sensitivity can be increased bychanging the method of estimating the product. Theextent to which the results of unmodified enzymeassays can be compared with those obtained aftermodification depends on the nature of the changeswhich have been made. If these do not includealterations in the conditions under which the enzymeacts, or in the nature of the substrate, the resultsshould be fully comparable. For example, the methodchosen for the estimation of phenol released fromphenyl phosphate by phosphatase does not affectthe level of enzyme activity. However, when theconditions of hydrolysis are changed, eg, by changingthe substrate from phenyl phosphate to p-nitrophenyl

600, Determined at 2S°C

phosphate or merely by altering the pH or nature ofthe buffer, the level of enzyme activity is also altered.The foregoing discussion of the complex relation-ships that obtain between the many variablesin an enzyme assay system indicates the difficultythat may then be encountered in relating unitsderived from the modified method to those fromthe original procedure.Not all alterations to methods of enzyme esti-

mation are intentional, however. When procedureshave been long established a gradual drift from thespecified protocol often takes place and may resultin considerable modification of the method. Thesignificance of the results consequently undergoeschange.The foregoing arguments may be summarized as

follows. First, the variables in all enzyme assayprocedures must be rigidly specified and controlled.Enzyme activities should not be reported in terms ofunits which imply a particular method unless thatmethod has been strictly followed. It is desirable thatthe methods on which enzyme units in common useare based should be re-examined and re-stated toremove ambiguity. Secondly, since many oldermethods are clearly unsatisfactory in the light ofrecent advances in enzymology, new definitive assaymethods are needed for enzymes of importance inclinical enzymology.

Determined at 35°C

5OoLC-0Ln

a 400--I:

-300

200

-' 100

0

_ X-X-X X-X-x-X- XX

_l I I I I I I I I I I I I I I I I I I1 2 3 4 5 6 7 8 9 2 3 4 5 6 7 8 9 10

DAYS

Fig. 4 Daily quality-control data for the spectrophotometric determination ofserum lactatedehydrogenase. X, pooled serum; 0, A, lO, A,**, reconstituted lyophilized materials.Considerable between-batch and within-batch variation is apparent at 250, particularly athigher enzyme activities. A change to 350 is followed by an improvement in precision.

25

iir

copyright. on M

ay 26, 2020 by guest. Protected by

http://jcp.bmj.com

/J C

lin Pathol: first published as 10.1136/jcp.s1-4.1.22 on 1 January 1970. D

ownloaded from

D. W. Moss

Factors Affecting the Design of New ReferenceMethods

The theoretical advantages of obtaining a continuousrecord of enzyme reaction rates can now be realizedas a result of the greater availability of recordingspectrophotometers. Newer reference methods ofenzyme assay should be based on this principlewherever possible, rather than on a fixed-time or two-point basis. Apparatus designed to combine somedegree of automation with kinetic analysis is nowbecoming available' and the clinical enzymologistshould state clearly those features which he expectsto find in such instruments. Specifications for opticalcomponents, reagent dispensers, and sample trans-fer systems will, in general, be similar to thoserequired in other automated equipment. However,one feature of central importance is the provision ofa thermostatically controlled reaction cuvette, sothat it is possible to control the temperature at whichmeasurements are to be made.

Fixed-time assays have traditionally been carriedout at 37°C but with the introduction of kineticassays 25°C became widely accepted as a standardtemperature. However, whereas temperatures suchas 37°C that are well above the ambient temperaturealmost everywhere can be accurately maintained bysimple, inexpensive thermostats controlling a heating'See contribution by Dr Skillen on page 31.

15 _

L.

'4-

0

Em

IOL

5

0

0-5 6-10 11-15Difference between duplicates

as percentage of mean

16 -20

Fig. 5 Differences between duplicate serum lactatedehydrogenase estimations, expressed as a percentage ofthe mean value for each pair, determined spectrophoto-metrically at 25°C (open bars) and 35°C (solid bars).

circuit, 25°C is close to, or even below, the ambienttemperature in many laboratories so that this simpleform of temperature control becomes inefficient oreven totally ineffective. The Enzyme Commission ofthe IUB has therefore recommended an increase inthe standard temperature to 30°C. Nevertheless,some instruments now available are set to operate ateven higher temperatures, eg, 35°C, and require adifferential of at least 8°C above the ambienttemperature for temperature control to be effective.Figures 4 and 5 illustrate the great improvement inthe quality of results following a change of tempera-ture for a kinetic analysis of lactate dehydrogenasefrom 25° to 35°C with a consequent improvement incontrol of temperature. However, before accepting ahigher reaction temperature for reasons of analyticalconvenience the clinical enzymologist will wish to beassured that the selected temperatuie is suitable inother respects also: for example, that factors such asinstability of enzymes or isoenzymes during the pre-incubation or measuring periods do not affect theresults obtained. If the use of higher temperatures isunacceptable instruments which incorporate bothheating and cooling cycles will become necessary,and, in this case, apparatus with a liquid incubationbath or circulation system may prove more readilyadaptable than those depending on electrical heatingof metal racks or air chambers.

The Place of Stable Enzyme Preparations of KnownActivity in Enzyme Analysis

While standardization of assay methods will contri-bute greatly to uniformity in clinical enzymology it isnevertheless desirable that stable enzyme prepara-tions of known activity should be available fordetermining the relative accuracy of results obtainedin different laboratories. A number of preparations,usually freeze-dried sera or mixed tissue extracts, areavailable to which values of activity of variousenzymes have been assigned but, however usefulthese preparations have been, it is generally agreedthat they have not been accepted with the samedegree of confidence that has been bestowed onquality-control preparations for other, non-enzy-matic constituents. This is because clinical enzym-ologists have often experienced difficulty in obtainingresults which agree with the assigned values for thecontrol preparations, even though the laboratory'sown quality control programme suggests thatanalytical methods are operating satisfactorily.Doubts are therefore cast on the stability of thereference preparations, and on the accuracy of theiroriginal calibration.Some examples may help to illustrate these points.

Figures 6 and 7 show that, in assays of two different

'41.6copyright.

on May 26, 2020 by guest. P

rotected byhttp://jcp.bm

j.com/

J Clin P

athol: first published as 10.1136/jcp.s1-4.1.22 on 1 January 1970. Dow

nloaded from

Accuracy, precision, and quality control of enzyme assays

Fig. 6 Daily quality control datafor determination ofserum alkalinephosphatase by a colorimetric phenyl-phosphate method on the Auto-Analyzer. 0, A, D, reconstitutedlyophilized commercial controlpreparations, freshly prepared eachday. Activities are expressed as apercentage of the respective nominalvalues. X, pooled serum. The methodwas calibrated and the activity of thepooled serum determined by amanual King-Armstrong procedure.

L l I I I I1 5

I I I I I I I10

DAYS15

Fig. 7 Daily quality control data for acolorimetric aspartate aminotransferasemethod on the AutoAnalyzer. The procedurewas calibrated by reference to one commerciallyophilized material, and results for two others(0, A) are expressed as a percentage of theirrespective nominal values. X, pooled serum.

l I I I l5 7 9 11

DAYS

enzymes, control sera supplied from outside thelaboratory fail to reproduce their assigned activities.The deviation from the nominal value is a randomone and does not seem to be due to a systematic erroron the part of the user's laboratory, since if this werethe case the control materials should all show valuesof more or less the same percentage of their nominalvalues. Dobrow and Amador (1970) have recentlypublished data which also show that nominal valuesoften cannot be reproduced. Within a particular

batch of control material day-to-day reproducibilitywas generally good, with a few exceptions, and this isin agreement with previous reports (eg, Hanok andKuo, 1968).

Different batches of material from a single sourcedid not show a consistent deviation from the nominalvalue. It is not possible to decide on the basis of thepresent data whether the discrepancies are due todeviations in assay procedures between the cali-brating and using laboratories, or whether a relatively

I130_

a 50

>-

o

t., 140-J

C>_. 1202

0co

2 100

a

> 80-

cc

LI I1 3 13

-

27

P-I _%F

copyright. on M

ay 26, 2020 by guest. Protected by

http://jcp.bmj.com

/J C

lin Pathol: first published as 10.1136/jcp.s1-4.1.22 on 1 January 1970. D

ownloaded from

D. W. Moss

rapid change to a different level of activity takesplace after calibration, followed by a longer periodof stability. In this respect recent observationsthat the enzymatic activity of reconstituted lyophil-ized sera, as well as that of frozen serum pools,increases over a considerable period of time followingreconstitution or thawing may be relevant (B. Brojerand D. W. Moss, unpublished data).Whatever the explanation, a user would not feel

confident in accepting these preparations at theirnominal values of activity without re-calibratingthem himself and may prefer to rely on frozenserum pools prepared in the laboratory or repeatedanalysis of specimens of patients' serum for moni-toring within-batch or between-batch variation.

Required Properties of a Standard or ReferenceEnzyme Preparation

The most important characteristic of a standardenzyme preparation is stability. Recent studies ofthe three-dimensional structures of protein moleculeshave emphasized still further the dependence ofcatalytic activity on the integrity of the moleculararchitecture of enzyme molecules. Derangement ofthis structure, ie, denaturation, leads to loss ofactivity. When relatively concentrated and puresolutions of enzymes or other proteins are available,the occurrence of denaturation can be detected bychanges in a number of physical properties of thesolution. However, for the dilute solutions ofenzymesin use as controls or calibration standards in enzymeassay, with their admixture of other proteins, suchmeasurements are not possible and the only de-tectable consequence of denaturation is a decline incatalytic activity. In the absence of assured stabilitythere is thus no way in which a low result arisingfrom faults in analysis can be distinguished from onedue to deterioration of the reference preparation.Increased stability of enzyme reference preparationsmay require changes in procedures of production andstorage and even in the chemical constitution of theenzymes themselves, so that before these possibilitiesare discussed it is necessary to consider which of theproperties of the enzymes must be preserved andwhich characteristics can be permitted to undergochange, and to what extent, in the pursuance ofincreased stability. If enzyme reference preparationsare intended to provide a means of testing theaccuracy of a particular assay and to compare the re-sults obtained in one laboratory with those obtainedelsewhere, there is no reason why the referencecatalyst should be identical in all respects with theenzyme that the assay method is designed to measure.The reference catalyst may therefore be an enzyme ofsimilar function but of markedly different stability

characteristics prepared from a non-human source.It may be an enzyme that has been modified chemi-cally to increase its resistance to denaturation at theexpense of some alteration in properties. It may,indeed, not be an enzyme at all but some other formof catalyst. For example, the cyclohexamylosesdescribed by Bender and others (Hennrich andCramer, 1965; Bender, van Etten, Clowes andSebastian, 1966) possess esterase activity towards anumber of organic esters, and, although their degreeof activity in this respect does not compare with thatof true enzymes, compounds of this nature may haveproperties that make them suitable for testing theaccuracy of reaction rate measurementsHowever, if the purpose of enzyme reference

preparations is to control the routine operation ofassay methods, or to calibrate one method in termsof another, then the reference sample should be asclosely similar as possible to the unknown samplesunder test. Neglect of this principle can result inerror if factors operating in the test system affect theenzymes in the unknown and reference samplesdifferently. For example, a change which results insuboptimal conditions, and therefore in reducedactivities, for the unknown enzyme samples may nothave the same effect on the reference preparation ifthis is of different isoenzymic composition orproperties, and instances in which preparations ofnon-human alkaline phosphatase have failed ad-equately to controlestimations of the human enzymehave been reported (MacWilliam, Moody, and Silk,1967; Bowers, Kelley, and McComb, 1967). Where adifferential effect of this nature exists, its results canbecome particularly significant when the referenceenzyme preparation is used in the calibration of oneenzyme assay method by comparison with a differentprocedure. A conversion factor established for thereference preparation may not be valid if samplescontaining a qualitatively different type of enzymeare analysed. Anomalies of this nature are morelikely to arise when enzymes of low specificity, suchas alkaline phosphatase, are being estimated, whena change in method may imply a change in thesubstrate. Human placental alkaline phosphatase isa valuable source of reference material for phos-phatase determinations, because of its remarkablestability to heat and other denaturing agents.However, placental alkaline phosphatase differs inimportant respects from the liver and bone iso-enzymes of which the alkaline phosphatase in serumusually consists. Among these differences arevariations in the relative rates at which differentsubstrates are hydrolysed (Moss, 1969; Wolf,Dinwoodie, and Morgan, 1969). Placental phos-phatase, and also intestinal phosphatase, arerelatively less effective in their action on some

28copyright.

on May 26, 2020 by guest. P

rotected byhttp://jcp.bm

j.com/

J Clin P

athol: first published as 10.1136/jcp.s1-4.1.22 on 1 January 1970. Dow

nloaded from

Accuracy, precision, and quality control of enzyme assays

chromogenic substrates such as p-nitrophenylphosphate than on non-aromatic phosphate esterssuch as fl-glycerophosphate. There are alsodifferences in relative rates of breakdown of p-nitrophenyl phosphate and thymolphthalein mono-phosphate by liver and intestinal phosphatases.These differences imply that if, for example, aconversion factor for translating p-nitrophenylphosphatase units into thymolphthalein mono-phosphatase units is established by comparisonsbased on samples which contain intestinal phos-phatase, this factor will not be valid when samplescontaining liver or bone phosphatases are analysed.It follows that some substrates provide a moresensitive assay than others for certain isoenzymes.

Possible Approaches to Enzyme Stabilization

As already mentioned, the constancy of the activityof an enzyme depends on the preservation of thethree-dimensional structure of its molecules. Alsoimportant is the protection of vulnerable aminoacid side chains which may have an important rolein the catalytic mechanism. Reactive groups on theexterior of the molecule probably play only a smallpart in determining overall conformation butmodification of external charged residues maynevertheless affect stability. Ways of increasing thestability of enzyme molecules may be divided intothose designed to protect the molecules from theaccess of denaturing or inactivating factors, andthose designed to increase the resistance of moleculesto these factors (Fig. 8).Examples of the first approach, ie, the prevention

ENZYME STABILIZATION

BY INCREASING RESISTANCETO DENA TURA TION \

Chemical _ CH3CO.NHmodification ofvulnerable groups r

of the access of inactivating factors, are the additionof substances to the enzyme preparation to ensure afavourable environment, eg, an acid pH for acidphosphatase or reducing conditions for enzymesthat depend for their activity on sulphydryl groups.The enclosure of enzyme molecules within semi-permeable membranes (Chang, 1964) also increasesenzyme stability, presumably by the same mech-anism. Changes in the resistance of enzyme moleculesto inactivation as a consequence of the chemicalmodification of reactive groups are examples of thesecond approach. For instance, acetylation of somehuman alkaline phosphatases increases their re-sistance to denaturation at alkaline pH values butdecreases it at more acid values (Moss, 1970). Asimilar effect is noted when enzyme molecules areattached to an insoluble matrix, commonly by theirterminal and side-chain amino groups (Silman andKatchalski, 1966). This latter procedure probablyalso increases stability by making the enzymemolecule more rigid, but it must be rememberedthat slight conformational changes appear to be anessential feature of the mechanism of enzymecatalysis and, while these intramolecular movementsare probably of a lower order of magnitude thanthose involved in denaturation, the need for somedegree of molecular flexibility will ultimately limitthe extent to which the enzyme molecule can bestiffened. Similar considerations apply to the intro-duction of additional cross-links between adjacentsegments of the polypeptide chain. Covalent cross-linkages, except for disulphide bridges, do not seemto be a major feature in stabilizing the structures ofnative enzyme molecules which depend for their

Fig. 8 Diagrammaticrepresentation ofpossible

BY REGULA TION approaches to stabilizationOF ENIVRONMENT ofenzyme molecules for

the preparation ofreference or controlsamples.

29copyright.

on May 26, 2020 by guest. P

rotected byhttp://jcp.bm

j.com/

J Clin P

athol: first published as 10.1136/jcp.s1-4.1.22 on 1 January 1970. Dow

nloaded from

30

conformation mainly on the existence of multiplehydrophobic interactions and hydrogen-bonding.However, the attachment of enzymes to solublematrices seems to be an approach to stabilizationwhich merits exploration.These several ways of increasing the stability of

enzyme preparations differ in the degree to which thecatalytic properties of the enzymes are changed.Such changes must be taken into account whenmodified enzymes are put to use in standardizationand control of enzyme assays, for the reasonsdiscussed earlier. The preservation of an acid pH orreducing conditions in the environment should notsignificantly alter functional characteristics, but thechemical modification of reactive groups may beexpected to alter catalytic properties to an extentwhich will depend on the relationship between thesegroups and the active centre. In the example of theacetylation of human alkaline phosphatases referredto previously, the modified enzymes are unchangedin their Michaelis constants and not significantlydifferent from the native forms in substrate specificity,as judged by their relative orthophosphatase andpyrophosphatase activity. There is a slight alkalineshift in the pH optimum of intestinal alkalinephosphatase but this is due to the changes in pHstability characteristics rather than to a modificationof the active centre; it would thus be expected to bemore pronounced in assay methods depending on anextended incubation period. Unfortunately, liverphosphatase is affected less than intestinal phos-phatase by acetylation, while the increase in stabilityof the intestinal isoenzyme is not such as to meet therequirements of a reference enzyme preparation.While the application of these principles to the

production of enzymes possessing long-term stabilityis at the moment in an elementary state, knowledgeof enzyme structure is growing rapidly and enzymetechnology is consequently also expanding quickly.The prospects for the production of stabilizedenzyme preparations of defined characteristics as aresult of systematic research programmes thereforeare encouraging.

In summary, therefore, two initiatives in the

D. W. Moss

standardization and quality control of enzyme assaysare needed, and are beginning to be implemented.The first of these is the re-definition, rationalization,and standardization of enzyme assay procedures,while the second is the provision of stabilized enzymereference preparations. Progress in both thesedirections can be expected to increase the value ofclinical enzymology.

References

Bender, M. L., Van Etten, R. L., Clowes, B. A., and Sebastian, J. F.(1966). A pictorial description of the 'lock and key' theory.J. Amer. chemn. Soc., 88, 2318-2319.

Bowers, G. N., Jr., Kelley, M. L., and McComb, R. B. (1967).Precision estimates in clinical chemistry. I. Variability ofanalytic results in a survey reference sample related to the use ofa non-human serum alkaline phosphatase. Clin. Chemti., 13,595-608.

Chang, T. M. S. (1964). Semipermeable microcapsules. Science, 146,524-525.

Dobrow, D. A., and Amador, E. (1970). The accuracy of commercialenzyme reference sera. Anmer. J. c/in. Path., 53, 60-67.

Gay, R. J., McComb, R. B., and Bowers, G. N., Jr. (1968). Optimumreaction conditions for human lactate dehydrogenase iso-enzymes as they affect total lactate dehydrogenase activity.Clin. Chemn., 14, 740-753.

Hanok, A., and Kuo, J. (1968). The stability of a reconstituted serumfor the assay of fifteen chemical constituents. Clint. Chern., 14,58-69.

Hennrich, N., and Cramer, F. (1965). Inclusion compounds. 18. Thecatalysis of the fission of pyrophosphates by cyclodextrin. Amodel reaction for the mechanism of enzymes. J. Attmer. chempi.Soc., 87, 1121-1126.

MacWilliam, K. M., Moody, A. H., and Silk, J. (1967). Variation inalkaline phosphatase results using the method of Bessey, Lowry,and Brock. Clin. chitm. Acta, 17, 514-515.

Moss, D. W. (1969). Biozhemical studies on phosphohydrolaseisoenzymes. Ann. N. Y. Acad. Sci., 166, 641-652.

Moss, D. W. (1970). Chemical modification of alkaline phosphatases.The effects of amino-group reagents on the electrophoreticmobilities and activities of phosphatases from several humantissues. Enzymnologia, 39, 319-330.

Moss, D. W., Campbell, D. M., Anagnostou-Kakaras, E., and King,E. J. (1961). Characterization of tissue alkaline phosphatasesand their partial purification by starch-gel electrophoresis.Biochem. J., 81, 441-447.

Silman, 1. H., and Katchalski, E. (1966). Water-insoluble derivativesof enzymes, antigens, and antibodies. Ann. Rev. Biochem., 35,873-908.

Vesell, E. S., and Pool, P. E. (1966). Lactate and pyruvate concen-trations in exercised ischemic canine muscle: relationship oftissue substrate level to lactate dehydrogenase isozyme pattern.Proc. nat. Acad. Sci. (Wash.), 55, 756-762.

Wolf, M., Dinwoodie, A., and Morgan, H. G. (1969). Comparison ofalkaline phosphatase isoenzymes activity using five standardmethods. Clin. chitt2. Acta, 24, 131-134.

copyright. on M

ay 26, 2020 by guest. Protected by

http://jcp.bmj.com

/J C

lin Pathol: first published as 10.1136/jcp.s1-4.1.22 on 1 January 1970. D

ownloaded from