ISBT/ICSH Working Party 1603660

Vox Sang 1994;67:89-98

Marion L. Scotta
Douglas Voakb Review of the Problems Involved
Peter K. Phillipsc
P.Ann Hopped in Using Enzymes in Blood Group
Sheryl A. Kochmand

■' International Blood Group Reference

Serology - Provision of Freeze-
Laboratory, Bristol;
b Division of Transfusion Medicine,
Dried ICSH/ISBT Protease Enzyme
East Anglian Blood Transfusion Centre,
University of Cambridge; and Anti-D Reference Standards
c National Institute for Biological Standards
and Control, South Mimms, UK;
d Food and Drug Administration, Rockville,
Md„ USA

Abstract
Proteolytic enzyme preparations and techniques used routinely in blood group
serology for the detection of atypical patient antibodies prior to transfusion vary
widely and are often poorly standardised. Recent advances have been made in the
use of biochemical methods to standardise and stabilise the potency of the en­
zyme preparations used. A joint working party of the International Council for
Standardization in Haematology (ICSH) and the International Society of Blood
Transfusion (ISBT) has investigated possibilities for the provision of standards
for the protease preparations and techniques. The specification for these stan­
dards was that the performance of enzyme reference preparation in the reference
technique should be of equivalent sensitivity to the ICSH/ISBT LISS spin in­
direct antiglobulin test using a titration series of a reference weak anti-D, and be
free from false-positive reactions. The working party circulated materials for
evaluation in inter-laboratory trials, followed by a laboratory workshop meeting
to achieve agreement on the specification for reference materials and methods.
Reference freeze-dried papain at 0.6 azoalbumin units and weak anti-D prep­
arations (91/562) have been prepared and validated to meet these specifications.
The performance of a test enzyme preparation in the technique for which it is
recommended for use should be at least equal to that of the reference papain
preparation, by the reference two-stage technique in terms of sensitivity, using a
titration series of the reference anti-D, and freedom from false-positive reactions,
using six fresh inert sera. The reference papain and weak anti-D can also be used
to calibrate the level of proteolytic activity required in other procedures in blood
group serology, such as new technology methods for antibody detection, and
automated and microplate cell grouping procedures. These preparations and an
agreed method for their use are now available from listed centres as ICSH/ISBT
and Food and Drug Administration reference materials.
144.173.6.94 - 5/6/2020 3:28:22 AM

Received Dr Manon L Scott © 1994 S ICarger AG, Basel

February 14, 1994 International Blood Group Reference Laboratory 0042-9007/94/0671 -0089
University of Exeter

Accepted Southmead Road $5 00/0

Downloaded by:

February 18, 1994 Bristol BS105ND(UK)

 Introduction
In 1986 a joint working party on the standardisation of
protease enzyme reagents and techniques used in blood
group serology was re-established by the Expert Panel on
Serology of the International Council for Standardization in
Haematology (ICSH) and the International Society for
Blood Transfusion (ISBT). (In 1990 this became part of a
more general ICSH/ISBT working party on blood grouping
reagents.) The aims of the working party were to select and
provide a protease enzyme reference preparation) s) to be
used by a defined reference method to quality control the
protease preparations and techniques used in pre-transfu­
sion atypical antibody detection. The specification was that
the level of sensitivity of a reference preparation and refer­
ence technique with a reference weak IgG anti-D should at
least equal that of the LISS spin tube indirect anti-human
globulin test (IAT; for which ICSH/ISBT reference materi­
als and methods already exist [1]).
Many factors affect the results of enzyme serological
tests resulting in considerable variation between different
laboratories - reagent selection and definition in serological
terms is extremely difficult. Based on previous experience
[1], interlaboratory trials were used to give all workers expe­
rience with the standard methods to be used, and to select
the range of proteolytic activity of papain and bromelain
preparations which would then be assessed in a laboratory
workshop in one centre in which all participants used sam­
ples of the same master preparations of reagents and red
cells. Participants were also able to compare and standar­
dise reading techniques - a major source of variation in ser­
ological tests. At the workshop meeting the 29 participants
agreed on the selection and specification for a reference
freeze-dried papain preparation, a reference freeze-dried
weak IgG anti-D preparation and the reference method for
their use. This article describes how these reference prep­
arations and reference method were selected, how they were
produced and validated, and from where they are available
(see Appendix).
An ICSH/ISBT working party on enzyme reagents and
techniques had been in existence before 1986, but had made
little progress due to the lack of established objective meth­
ods for measuring the proteolytic activity of enzyme prep­
arations, lack of any methods for preparing stable reference
preparations and lack of comparative information on the
relative sensitivity of different techniques. The main prob­
lem areas (for full review see [2]) are summarised below.
Many different techniques for the use of the plant sul­
phydryl proteases (papain, bromelain and ficin) in blood
group serology have been described and there has been con­
90 Scott/Voak/Phillips/Hoppe/Kochman
siderable debate over the relative merits of the different
techniques and enyzmes used [e.g. 3-5]. However, it is diffi­
cult to evaluate most of the comparisons carried out, as
many workers have compared different enzyme prepara­
tions in different techniques, introducing variables that can­
not be separated. In the few studies where the enzymes have
been prepared to the same proteolytic activity level [6-8]
and compared in the same technique, there seems to be no
difference in the serological efficacy of papain, bromelain
and ficin; the serological efficacy is merely related to the
proteolytic activity of the enzyme preparation [6], When an
enzyme solution is prepared, depending on the source mate­
rial, variable amounts of crude material dissolve and the
specific proteolytic activity (proteolytic activity per g) of
different materials varies widely between manufacturers
and within manufacturers from batch to batch [2]; quoting
percentage weight/volume to define proteolytic potency is
meaningless.
Papain, bromelain and ficin are sulphydryl proteases re­
quiring a reduced cysteine residue within the active site. As
partially purified powdered extracts, any batch of enyzme
will have a variable degree of inactivation, caused by ox­
idation or blocking of the active site by divalent metal ions
[9], When the powdered material is dissolved in buffer, the
maximum proteolytic activity of that preparation will not be
available unless steps are taken to activate fully the enzyme
active sites. Full activation can be achieved using a combi­
nation of an added sulphydryl reducing agent (e.g. added
cysteine) and an added chelating agent such as EDTA. The
same proteolytic activity can be achieved from a particular
source material by either using a high concentration (w/v)
of incompletely inactivated enyzme, or a lower concentra­
tion of fully active enzyme. The presence of an optimal
amount of reducing agent has a greater effect on activity
than of a chelating agent. Traditionally in blood group serol­
ogy, papain has tended to be used in its activated form (e.g.
Tow’s formulation containing cysteine and sometimes
EDTA) whereas bromelain and ficin have been commonly
used unactivated. This is purely tradition; both bromelain
and ficin activities can be greatly increased by addition of
sulphydryl reducing agents.
In 1978, Lambert et al. [6] described a simple biochem­
ical assay (hydrolysis of azoalbumin) for the measurement
of the proteolytic activity of papain, bromelain and ficin
preparations that correlated well with serological efficacy.
This assay was later modified in 1984 by Phillips et al. [7]
andin 1988 by Scott et al. [8] for routine use. A similar assay
utilising the hydrolysis of casein was described in 1986 by
Mazda et al. [10]. Work by Scott and Whitton in 1988 [11]
showed that the molarity of protease active sites of papain
144.173.6.94 - 5/6/2020 3:28:22 AM
Problems Involved in Using Enzymes in
Blood Group Serology
University of Exeter
Downloaded by:
and bromelain preparations could be accurately determined
using E-64, a synthetic peptide inhibitor of the enzymes,
in combination with any of the above substrates and assay
methods. Experience with the azoalbumin assay by work­
ing party members has shown that different batches of
azoalbumin can give different assay results. Therefore
new lots must be calibrated against previous lots and an
adjustment factor determined (e.g. papain assay with previ­
ous azoalbumin lot is 0.7; assay with successor lot is 1.33;
future results should be multiplied by 0.7/1.33). Benzoyl-
L-arginine-/?-nitroanilide has also been proposed [12] as
an alternative substrate to azoalbumin or casein, but it has
been shown [13] that the hydrolysis of this substrate does
not correlate with the action of enzymes on red blood cell
membranes; therefore this assay is not a suitable means of
assessing the activity of enzyme reagents for serological
use.
The stability of liquid enzyme reagents is related to their
degree of activation. Activated enzyme will undergo auto­
lysis if stored in the liquid form. A liquid preparation of a
low specific activity due to high levels of incompletely
activated enzyme will be more stable than one of the same
activity but containing a small amount of fully activated
enzyme. In addition, cysteine, commonly used to activate
papain, is itself unstable in liquid form and readily auto-
oxidises. As the reducing activity of the cysteine in a prep­
aration falls, so will the proteolytic activity of the enzyme
preparation due to progressive inactivation of the active
sites. This latter effect can be reduced by storing pa-
pain-cysteine formulations in airtight containers with mini­
mal air-space, which can be filled with nitrogen; however
this will not prevent autolysis. As a result of these mecha­
nisms, most liquid enzyme preparations cannot be stored at
4°C and retain constant proteolytic activity. Stable long­
term storage is best achieved in the frozen or freeze-dried
state.
In 1985, development of a stabilised liquid papain prep­
aration for storage at 4°C was described [14], The availabil­
ity of such stabilised material which could be prepared
standardised to different biochemically defined activity lev­
els, meant that the working party now had the means to cir­
culate stabilised, standardised enzyme preparation for in­
terlaboratory trials to establish the specifications for inter­
national reference preparations. Having established the
proteolytic activity level required, freeze-dried reference
preparation(s) at those level(s) could be prepared.
Even if different enzymes are prepared to the same prote­
olytic activity level, their efficacy will vary according to the
the technique and conditions used. Time and temperature of
incubation are important variables; in addition it has been
shown that the sensitivity and freedom from false-positive
reactions of enzyme serological techniques are also depend­
ent on the pH [15] and ionic strength [16] used.
Five basic different types of technique have been de­
scribed, from which the working party selected the refer­
ence technique.
Two-Stage Technique
The two-stage technique [17-19] involves incubation of
enzyme with test red cells, then washing the red cells free of
enzyme before incubating the treated cells with test serum
and evaluating the result. Countless minor variations exist
for this technique in terms of the ratio of cells to enzyme and
serum, incubation times and temperatures, using of centrif­
ugation and different reaction vessels.
The advantages of this technique are that it is cheap, sim­
ple and very sensitive for the detection of appropriate anti­
bodies. It is also very helpful to reference laboratories for
antibody identification, where the enzyme sensitivity of an
antigen(s) can be used to eliminate the reactivity of one or
more antibodies present in a serum sample containing a
mixture of antibodies. Enzyme-pretreated panel cells are
available commercially for use in antibody screening and
identification.
The major disadvantage of this technique is the time-
consuming two-stage process. Also, the technique must be
carefully controlled in terms of the proteolytic activity of
the enzyme preparation, as too much enzyme modification
causes false-positive reactions. Thus, it is not favoured for
the cross-match, where several donor red cell samples need
pretreatment and false-positive reactions could delay the is­
sue of blood. A modification of Stratton’s technique [18],
whereby 1 voi of saline is added to the test (introduced at the
Sheffield Transfusion Centre by Voak and Stapleton [un-
publ.], helps to reduce the problem of false-positive reac­
tions.
One-Stage Mix Technique
The one-stage mix technique was described by Low [20]
for Rh phenotyping cells. In this method, volumes of red
cells, enzyme and serum are mixed together, incubated and
the results read. This has provided a practical, feasible tech­
nique for Rh typing cells using incomplete IgG typing sera,
that would otherwise have to be used in a two-stage enzyme
test. Provided the proteolytic activity of the enyzme used is
controlled [21] and the typing sera used are potent polyclo­
nal reagents standardised for this method, this technique is
adequate, and has been extended for use in automated sys­
tems (continuous flow autoanalysers and microplate sys­
tems).
144.173.6.94 - 5/6/2020 3:28:22 AM
91
University of Exeter
Downloaded by:
 The one-stage method is not suitable for the detection of
low levels (i.e. <2 IU/ml anti-D) antibodies in compatibil­
ity testing. This is because mixing serum with the red cells
and enzyme provides alternative substrates for the enzyme
[8]. Antibody in the serum is attacked by the enzyme, and
low levels of specific antibody may be destroyed, rendering
them incapable of agglutinating cells. Serum albumin is al­
so degraded. Thus, this method is of insufficient sensitivity
to be used as a reference method for antibody detection.
This method is also not suitable for use with some pheno-
typing reagents e.g. enzyme-sensitive monoclonal antibod-
les.
One-Stage Layer Technique
It is possible to improve the sensitivity of simple one-
stage tests without the time-consuming cell washing stage.
The one-stage layer technique [22] achieves separation of
enzyme from serum by physical layering in small precipitin
(7 x 50 mm) or capillary tubes. Serum is added first, then
enzyme solution is layered on top, by placing the drop on
the side of the tube, followed last by a drop of 3% red cells.
During the incubation period (normally 50-60 min at 37°C
for sedimentation methods, or 25 min at 37 °C followed by
centrifugation), the red cells sediment through the enzyme
layer and become treated, then fall through the serum layer
and react with antibody. This technique can achieve sensi­
tivity almost as high as the two-stage technique if standar­
dised enzyme solutions are used and adequate layering is
achieved. Use of too strong an enzyme solution and/or too
long an incubation time can lead to false-positive reactions,
while mixing between layers will lead to low sensitivity due
to antibody digestion, as with the one-stage mix technique.
This technique does not work in standard size (10 or
12 x 75 mm) tubes as the layers mix [8]. The variability in
performance of this technique renders it unsuitable for use
as a reference technique.
Phased Mix Technique
Separating red cell treatment from serum reaction can
also be achieved with adequate sensitivity by phasing the
addition and mixing of reagents [23], In this type of tech­
nique, red cells and enzyme are mixed and incubated to­
gether. After a defined time (3-5 min at 37°C using 0.6
azoalbumin unit papain [8], serum is added and mixed. By
use of an enzyme preparation of correct activity, red cells
are adequately treated during the first phase, whilst the ad­
dition of serum provides sufficient albumin and enzyme in­
hibitors for the digestion of antibody not to be a significant
problem. This technique demands uninterrupted attention,
since too great a delay after adding the serum can result in a
92 Scott/Voak/Phillips/Hoppe/Kochman
test similar to the one-stage mix. If correctly performed, the
phased-mix technique can have equal sensitivity to a two-
stage procedure.
Phased Inhibitor Technique
This phased technique has been refined further by the
use of a specific papain inhibitor [24]. After the first phase
of incubation of red cells with papain, a specific papain in­
hibitor (antipain or E-64 [25] is added, serum is added and
mixed. This technique allows rapid tretment of red cells
with high-activity papain preparations, as the inhibitor pre­
vents enzyme digestion of subsequently added antibody
protein [8], The disadvantage of this technique is the neces­
sity to add a further reagent, the inhibitor, which is relatively
costly, and, as for the phased mix technique, the require­
ment for uninterrupted attention.
The working party decided to use a two-stage technique
as the reference method, but also to assess the relative effi­
cacy of one-stage, one-stage delay/phased, two-phase in­
hibitor and one-stage mix methods with defined enzyme
preparations.
Following interlaboratory trials of the the serological
performance of papain and bromelain preparations at differ­
ent activity levels, and assessment of the various biochem­
ical techniques for measuring activity, members and ob­
servers of the working party met at the Jerome H. Holland
Laboratory of the American Red Cross Laboratories Blood
Services, Rockville, Md., USA, on the 16th—17th October
1989.
The aims of this workshop meeting were: (l)to decide
which non-serological method should be used to define the
proteolytic activity level of reference preparations; (2) to
select the proteolytic activity level (defined as agreed
above) required for freeze-dried papain and bromelain prep­
arations such that they meet the following specifications:
(a) sensitivity by the two-stage reference technique with a
standard weak anti-D at least equal to that obtained with this
serum and the same red cell suspensions in the ICSH/ISBT
LISS spin tube antiglobulin test, and (b) freedom from
false-positive results in the two-stage reference technique
using fresh inert sera; (3) to decide whether separate refer­
ence preparations are required for papain and bromelain,
and (4) to compare the performance of the selected mini­
mum potency reference preparation) s) in two-stage, phased
mix and one-stage mix methods.
144.173.6.94 - 5/6/2020 3:28:22 AM
Problems Involved in Using Enzymes m
Blood Group Serology
University of Exeter
Downloaded by:
 Materials and Methods
Freeze-dried papain and bromelain preparations at 0.7 azoalbumin
units [8] and freeze-dried diluent were prepared at IBGRL and freeze-
dried by the National Institute for Biological Standards and Control
(NIBSC) in the UK. Azoalbumin assays were performed using azoal­
bumin (Sigma Chemical Co. Ltd., Poole, Dorset, UK) lot No. 78F
9315. Lot No. 91F 9350 gave identical results, whereas lot No. 27F
9316 gave a value of 1.33 azoalbumin units for these preparations.
Whichever azoalbumin lot was used (or when azocasein or casein
were used), the E-64 inhibition assay [11] gave 11.3 pA/ active sites.
These freeze-dried preparations were reconstituted at the workshop,
and diluted with reconstituted diluent to provide papain and bromelain
preparations at 0.7,0.6,0.5,0.4 and 0.3 azoalbumin units. Pooled group
R,R2 red cells from 4 donors were washed and treated with each en­
zyme preparation in bulk, according to the agreed two-stage procedure
(see below). After washing, the treated red cells were resuspended to
3% in phosphate-buffered saline (PBS; 10 mM phosphate-buffered
0.15 M saline, pH 7.2) in bulk. Master doubling dilution series were
made of a 1.0 IU/ml anti-D serum and a weak anti-E serum in PBS pH
7.2 containing 2% BSA.
There were eleven workstations, with 2-3 working party members
ar each station. Centrifuges were calibrated for optimum sensitivity
with freedom from false-positive results and used on slow speed
(130 ref) for 25 s (Sorvall CW-1) or low speed for 45 s (Clay Adams).
Pasteur pipettes used for delivering volumes of reactants were cali­
brated at 28 pl/drop delivered from a vertical angle. Aliquots of the
master serum dilutions, four fresh inert sera and each suspension of
enzyme-treated red cells and untreated red cells and untreated red cells
were provided to each working station. Each workstation provided its
own anti-human globulin reagent.
Two-Stage Method
Enzyme was mixed with washed, packed pooled red cells in the
volume ratio 2:1 and incubated for 15 min at 37°C with periodic mix­
ing. The treated cells were washed three times with PBS (10 mM pH
7.2 phosphate buffer in 0.15 M NaCl) and resuspended to 3% in PBS.
Two drops of 3% treated cells were mixed with two drops of test serum
and two drops of PBS, and incubated for 15 min at 37°C. Tests were
centrifuged at the calibrated speed and time. The cells were gently re­
suspended by the tip and roll technique, or gentle agitation [1] and the
degree of agglutination assessed macroscopically. The grading
scheme used is shown in table 1.
Phased Mix Method
Two drops of 3% red cells in PBS were mixed with two drops of
enzyme preparation and incubated at 37°C for 5 min; two drops of se­
rum sample were added and mixed, incubated for a further 5 min at
37°C, centrifuged and read.
One-Stage Mix Method
Two drops each of 3% red cells, serum and enzyme were mixed
together, incubated for 5 min at 37°C, centrifuged and read.
Initially the workshop participants used one drop of all reagents in
the serological tests i.e. 28 pi, but found the tests difficult to read with
this low quantity of red cells. The use of 56-pl voi - two drops - pro­
duced more easily readable tests; this is perhaps why many laborato­
ries routinely use 5% cell suspensions for enzyme tests.
Table 1. Scoring system used throughout workshop and validation
Grade Titration Observation
score (macroscopic only)
C 12 cell button in one clump
+++ 10 cell button dislodges into several large clumps
++ cell button dislodges into many small clumps
5 cell button dislodges in granular but definite
clumps
(+) 3 cell button dislodges into fine small granules
0 0 negative result
The anti-D titration series was tested by the standard ICSH/ISBT
LISS IAT method [1], The four fresh inert sera, the anti-D and the an­
ti-E titration series (in this order) were tested against red cells pretreat­
ed with each different enzyme preparation, using the two-stage meth­
od above. As the work progressed, it became apparent that there was
little point in all workstations fully testing the red cells treated with 0.7
azoalbumin unit activity enzymes, due to excessive false-positive re­
actions. Similarly, by the time most workstations came to testing the
anti-E, it had become apparent that there was little point in testing cells
treated with enzyme activity levels less than 0.5 azoalbumin units. Af­
ter full discussion of all the results, tests were repeated by all work­
stations using cells treated with papain and bromelain at 0.6 and 0.5
azoalbumin activity levels only, as these preparations met the agreed
specification of equal sensitivity for the detection of anti-D to the IAT
method, with freedom from false-positive results.
In addition the sensitivity of the two-stage, phased mix and one-
stage mix techniques were compared, using the anti-D titration series,
the papain and bromelain at 0.6 and 0.5 azoalbumin activity levels, and
some commercial bromelain and ficin reagents.
Results
Initally there was considerable discrepancy between
workstations in reporting false-positive reactions between
enzyme-treated cells and the inert sera. Five workstations
reported no false positives with any serum or any treated
red cell samples, whereas the other six reported false posi­
tives with red cells which increased in frequency and ag­
glutination score with the proteolytic activity of the enzyme
preparation with which the red cells had been treated; par­
ticularly if the proteolytic activity was above 0.6 azoalbu­
min units.
Initial titres of the anti-D in the antiglobulin test ranged
from 8 to 64 with a mode of 32. Five different commercial
antiglobulin reagents were used; one workstation used the
FDA minimum potency standard and obtained a titre of 16.
Considering the different antiglobulin reagents used, these
results showed good agreement.
144.173.6.94 - 5/6/2020 3:28:22 AM
93
University of Exeter
Downloaded by:
 8-| 7 -,

7- 6 -
i
c 5-
i i
i o 4- I
i I 1 CO

2 3-
I I
I
5 2 I
I
I
I
i
S 2- [SM 1
1

si I g

:li 1
- sX

i i I ■ ■
0 0.6 0.5 0.4 0.3
Azoalbumin activity level
02 01
pap 0 6
M pap 0.5 IAT

Fig. 1. Laboratory workshop distribution of titres of anti-D ob­
tained using different activity papain preparations in the reference
two-stage enzyme technique. I = Titre 4; titre 8; D = titre 16;
titre 32.
Fig. 2. Laboratory workshop comparison of distribution of anti-D
titres obtained using 0.5 and 0.6 azoalbumin activity level papain (pap)
preparations in the reference two-stage enzyme technique and untreat­
ed cells in the LISS spin tube antiglobulin technique (IAT). I =
Titre 16; titre 32; D = titre 64; titre 128.

The distribution of titres obtained with the anti-D incu­ Table 2. Comparison of enzyme techniques with papain at two
bated with red cells pretreated with the different papain azoalbumin activity levels
preparations is shown in figure 1. Similar values were ob­
Two-stage Phased mix One-stage
tained with the bromelain preparations. For both enzymes,
as the proteolytic activity of the enzyme is reduced, the dis­ Anti-D
tribution of titres covers a lower range. To achieve ranges of Papain
titres similar to those recorded with IAT, papain and brome­ Azoalbumin 0.5 16 4
Azoalbumin 0.6 16 4 1
lain both must be used at 0.5-0.6 azoalbumin units (0.7 units
giving rise to high numbers of false-positive reactions - see Anti-E
above). Papain
Azoalbumin 0.5 2 0 0
To confirm these results, selected tests were repeated by Azoalbumin 0.6 4 1 0
all workstations on the next day of the meeting. Red cells
were treated with papain and bromelain at 0.5 and 0.6 azoal­ The values are titres obtained using a twofold dilution series of
bumin activity units, and the cells tested for sensitivity with anti-D and anti-E.
the titration series of anti-D and anti-E, and for false-posi­
tive reactions with four fresh inert sera. Untreated cells were
tested with the anti-D titration series by the 15-min LISS
spin antiglobulin method. The false-positive results observ­ The two-stage, phased mix and one-stage mix tech­
ed in these repeated tests were agreed to be acceptable; for niques were compared using a commercial bromelain prep­
papain, 1/10 workstations recorded one weak false-positive aration with similar serological activity to the 0.5-0.6 azoal­
reaction with one serum sample, and for bromelain, 2/10 bumin unit preparations, and the master anti-D titration se­
each recorded one weak false-positive reaction. The range ries. The two-stage technique was clearly the most sensitive
of titres recorded with the anti-D was similar to that of the (mode titre 32) and the one-stage mix the least sensitive
IAT when papain or bromelain at 0.5 azoalbumin units was (mode titre 4). The phased-mix test was more sensitive than
used. The distributions skewed to be slightly more sensitive the one-stage mix, with a mode titre of 8, but did not ap­
than the IAT titre range when papain or bromelain at 0.6 proach the sensitivity of the two-stage technique. The distri­
azoalbumin units was used (fig. 2). The range of titres rec­ bution of titres obtained is shown in figure 3. One worksta­
orded for the anti-E was higher when 0.6 azoalbumin unit tion also used the workshop 0.5 and 0.6 azoalbumin activity
papain or bromelain was used compared to 0.5 unit (see the papain preparations and the anti-D and anti-E titration se­
typical example in table 2). ries. The results are shown in table 1, and demonstrate the
144.173.6.94 - 5/6/2020 3:28:22 AM

94 Scott/Voak/Phillips/Hoppe/Kochman Problems Involved in Using Enzymes in

Blood Group Serology
University of Exeter
Downloaded by:
same relative sensitivity of the different techniques. The an-
ti-E was reliably detected only by the two-stage technique,
and detection was better using the 0.6 azoalbumin activity
preparation.
The working party agreed that: (1) the reference enzyme
preparation should be prepared to 0.6 azoalbumin units (us­
ing azoalbumin Sigma lot Nos. 78F 9315 or 91F 9350) [8];
this gave sensitivity at least as high as an antiglobulin test,
without incurring significant false-positive reactions;
(2) given the similarity of results obtained with papain and
bromelain, only one enzyme (papain) was required as a ref­
erence preparation; (3) the reference preparation should be
prepared from the same source papain as the material as­
sessed in this workshop, to the same formulation and be
standardised in the same azoalbumin assay, using the same
lot of azoalbumin; (4) although the reference preparation
would be prepared to a defined azoalbumin activity level,
using the defined lot of azoalbumin, this would not be quot­
ed when issuing the preparation. Instead a reference freeze-
dried weak anti-D would be prepared and issued with the
reference papain. The serological activity of an unknown
enzyme preparation and method would be compared to the
reference enzyme in the reference two-stage technique with
the reference anti-D; (5) since the non-serological assay of
proteolytic activity would only be used in the preparation of
the reference material, it was felt that the azoalbumin assay
would be adequate for this purpose. Dr. Mazda agreed to
check the next batch of reference materials to see if those
with the same azoalbumin assay units also had the same ca­
sein units.
The Food and Drug Administration (FDA) offered to
prepare the freeze-dried papain. The UK agreed to supply,
select and freeze-dry the reference anti-D.
The FDA prepared a pilot batch of freeze-dried papain.
Samples were assessed by biochemical and serological
analysis in comparison to the original freeze-dried material
used at the workshop meeting. Adjustment of the pH and
osmolarity was required to equal the specification of the
material assessed at the workshop. A further batch was
freeze-dried after these adjustments and circulated to work­
ing party members for validation. Stability studies showed it
was stable over 1 year when stored at 20 °C or below. The
stability of the material will continue to be assessed yearly.
Anti-D sera containing less than 10 IU/ml were obtained
from UK Transfusion Centres. Quantitation on the samples
was performed by autoanalyzer against the WHO Interna­
tional Reference Preparation, 68/419, at IBGRF, UK: Those
containing less than 10 IU/ml were pooled to give a 4-litre
pool. This material undiluted gave a titre of 256 with R,R2
cells using the workshop and FDA freeze-dried papain
8-]
7-
c 6-
«2 5-
03 4-
CO
§
t
S 3"
I
2-
S-|
1- •Xx
0
0 1 2 4
I
8 16 32
litres
Fig. 3. Laboratory workshop comparison of anti-D titres obtained
in two-stage (I), two-phase (D) and one-stage mix (M) techniques,
using a commercial bromelain preparation.
preparations in the reference two-stage technique, in com­
parison to a titre of 32 with the anti-D used at the workshop.
The pool was also shown by titration with R^r^qdr and rr
cells to contain anti-C at a titre of 8. The material was dilut­
ed 1 in 8 with inert serum prior to freeze-drying at NIBSC,
UK. After this dilution stage the anti-C was no longer detec­
table. After freeze-drying and reconstitution, this anti-D
and the workshop anti-D gave a titre of 32 with R,r cells
treated with either the FDA papain or the workshop papain
preparations in the reference two-stage technique.
Samples of the FDA freeze-dried papain preparation and
the NIBSC freeze-dried anti-D preparation were circulated
in October 1992 to all working party members, and also
members of an EC working party on standards for blood
grouping reagents. Members were asked to reconstitute the
anti-D and prepare a twofold dilution series and then to
compare the sensitivity of detection with red cells treated
with the reconstituted papain preparation in the reference
two-stage technique with the LISS spin antiglobulin tech­
nique. They were also asked to assess false-positive reac­
tions using six fresh inert sera.
Twenty laboratories carried out the final assessment. In
addition to the serological assessment, Dr. Mazda checked
the proteolytic activity level of the papain in the casein as­
say in comparison to the original workshop material. The
FDA papain had 33.5 casein units/50 pi and the workshop
papain had 31.1 casein units/50 pi and this was deemed to be
acceptable.
The results of the serological assessments are shown in
table 3. Eighteen participants found that the reference pa­
pain in the reference two-stage technique gave titres equal to
or higher than those obtained in the FISS spin antiglobulin
144.173.6.94 - 5/6/2020 3:28:22 AM
95
University of Exeter
Downloaded by:
 Table 3. Ranked titres and false-positive scores obtained in each from false-positive reactions. Futher local reference prep­
participating laboratory in the final validation of papain and anti-D arations for papain may be prepared and tested by the
reference preparations in the reference two-stage enzyme technique in
agreed serological and azoalbumin methods to ensure
comparison to the LISS spin tube antiglobulin test (IAT)
agreement with the primary standard. Such local reference
Titres False-positive scores preparations can either be stored frozen at -20 °C or freeze-
(total of 6 sera) dried. It is possible that papain from different sources may
two-stage papain IAT two-stage papain IAT be required at a slightly different azoalbumin value to equa­
te to the serological performance defined by this reference
256 32 0 0
material. It is the serological performance that is paramount
256 64 6 0
in defining the reference material - the azoalbumin assay
128 32 15 0
128 64
merely provides a convenient method of ensuring batch-to-
128 128 0 0 batch consistency when using a particular source of papain.
128 64 0 0 Although these reference preparations have been de­
128 64 0 0
signed for controlling reagents and techniques for antibody
128 64 0 0
0
detection, the papain reference material may also be useful
64 32 0
64 32 0 0 as a reference point to determine levels of proteolytic activ­
64 32 0 0 ity required for other serological applications, e.g. red cell
32 32 0 0 typing, anti-D quantitation, etc. The reference papain and
32 32 0 0
anti-D may also prove useful in assessing the performance
32 16 0 0
0 0
of new technology systems [26],
32 16
32 16 0 0 Two-stage enzyme methods have been criticised and de­
32 16 0 0 clared unsuitable for routine use in antibody screening be­
32 64 0 0 cause of the level of false-positive reactions. Use of the ref­
16 0 0 erence preparations described in this report should enable
16 32 0 0 laboratories to standardise their working enzyme prepara­
tions and techniques to avoid false-positive results and pro­
vide an inexpensive, simple, sensitive, reliable adjunct to the
IAT for antibody screening. External quality assurance stud­
method. Three participants reported weak false-positive re­ ies in the UK have shown that although in theory an antiglob­
actions; one obtained the same ‘false-positives’ in their IAT ulin test alone should be of sufficient sensitivity for use as an
test, indicating that the serum in question was not inert. antibody screen, in practice the enzyme test provides an ex­
When the range of results is considered, it is apparent that tremely valuable back-up. In 1 published study [27] 742 par­
the two labs obtaining genuine false-positive reactions ob­ ticipants would have failed to detect antibodies of they had
tained the highest titres with the papain-treated cells; the used an antiglobulin test alone, whereas only 351 of these
two labs reporting lower titres for the papain tests than the participants would have failed to detect antibodies using
IAT tests obtained low titres in the enzyme tests. This prob­ both antiglobulin and enzyme tests. Whilst suboptimal per­
ably reflects differences in the reading techniques used in formance of the conventional antiglobulin test remains in
different laboratories with enzyme-treated cells. some laboratories the enzyme test serves a useful purpose as
a secondary method for antibody screening.

Conclusions
The FDA freeze-dried papain preparation and the
NIBSC freeze-dried anti-D preparations meet the specifica­
tions for these reference materials agreed by the working
party. They are suitable for adoption as ICSH/ISBT refer­
ence materials to be used according to the defined protocol.
When used in the reference two-stage technique, the papain
reference preparation gives sensitivity equivalent to the IAT
for the detection of the reference weak anti-D, with freedom
Acknowledgements
We wish to thank the staff of the Jerome H. Holland Laboratory,
and the FDA, Rockville, Md., USA, for their help and co-operation in
the organisation and running of the laboratory workshop meeting, the
staff at the FDA, Rockville, Md., USA and NIBSC, Potters Bar, UK for
their help in preparing the freeze-dried preparations and the staff at
IBGRL, Bristol, UK, for their assistance in the preparation and circu­
lation of materials to working party members and Barry Dawes for
anti-D quantitations.
144.173.6.94 - 5/6/2020 3:28:22 AM

96 Scott/Voak/Phillips/Hoppe/Kochman Problems Involved in Using Enzymes in

Blood Group Serology
University of Exeter
Downloaded by:
 Appendix The reference materials are themselves free of charge, but each
centre will charge a locally determined fee to cover their administra­
Availability ofreference materials tive and postage costs. Reference materials will be provided with a
The ICSH/ISBT freeze-dried papain and anti-D reference materi­ detailed Standard Operating Procedure for their use.
als and protocols for use can be obtained from the following centres:
Participants in the workshop
Dr Toshio Mazda
Japanese Red Cross Tokyo Metropolitan Blood Center R Aerts, Beerse, Belgium
Kyonan-cho, Musashino City J. Andre/C Lockhart. Mames La Coquette, France
Tokyo 180 (Japan) D. Brazier, Elstree, UK
Tel: 422-32-1995 J Case, Houston, Tex., USA
Fax: 422-32-2685 A. Chung, Ottawa, Canada
R. Collins/D. Mallory/S. Kochman, Rockville, Md , USA
Dr. Pierre-Yves Le Pennec D. Davies, Raritan, N.J., USA
Centre National de Référence pour les Groupes Sanguins D. Ford, Parkville, Victona, Australia
53, Bd Diderot H. Gunson, Manchester, UK
75571 Pans Cedex 12 (France) P.A. Hoppe/P. Jones/J. Manak/C. Santos, FDA, Bethesda, Md., USA
Tel: 1-45-12-75-75 C Jersild/B. Langeland, Aalborg, Denmark
Fax:1-45-12-75-05 T Mazda, Tokyo, Japan
B. P.L. Moore, Toronto, Canada
National Institute for Biological Standards and Control M. Nichols, Chicago, 111., USA
PO Box 1993 M. Overbeeke/T. Frame, Amsterdam, Netherlands
Potters Bar, Herts. EN6 3QH (UK) P. Phillips, NIBSC, Potters Bar, UK
Tel: 707-646977 M. Scott, Bristol, UK
Fax: 707-646977 H. Seyfried, Warsaw, Poland
D Tills, Horsham, UK
(NIBSC also have a local papain reference material available, prepared to be D Voak/D.M. Dowme, Cambridge, UK
equivalent to the ICSH/ISBT reference preparation) M.-T. Weber, Dreieich, Germany

Dr. Amy Chung Final assessment ofreference preparations

The Canadian Red Cross Society Assessments from EU countries
1800 Alta Vista Drive
Ottawa, Ont K1G 4J5 (Canada) S. Maia, Porto, Portugal
Tel: 613-739-3000x2445 C Martin-Vega, Barcelona, Spain
Fax 613-731-1411 F. Morelati, Milano, Italy
A Peters, Tumhout, Belgium
Derek S. Ford
National Blood Group Reference Laboratory
NSW Red Cross Blood Transfusion Service Other members ofthe working party
153 Clarence Street
Sydney, NSW 2000 (Australia) D Anstee, Bristol, UK (to 1988)
Tel: (02) 229 4330 H Gerber, Berne, Switzerland
Fax: (02) 229 4487 G. Inglis, Glasgow, Scotland, UK (to 1988)
P.-Y Le Pennec, Pans, France (from 1990)
John Case W. Mayer, Vienna, Austria (from 1990)
Gamma Biologicals Inc. L. Messeter, Lund, Sweden (from 1990)
3700 Mangum Road J. Moulds, Houston, Tex., USA (to 1988)
Houston, TX 77092-4597 (USA) R. Nordhagen, Oslo, Norway
Tel: (713) 681 8481 A. Pirkola, Helsinki, Finland (from 1990)
Fax:(713)956 3333 H. Sonnebom, Dreieich, Germany (from 1990)
D. Van Rhenen, Rotterdam, Netherlands (from 1990)
Mrs. Carla van Dalen
Central Laboratory of the Netherlands Red Cross
Blood Transfusion Service
Plesmanlaan 125
1066 CX Amsterdam (Netherlands)
Tel: 020 512 3377
Fax: 020 512 3474

Sheryl A. Kochman
Center for Biologies Evaluation and Research
Food and Drug Administration
1401 Rockville Pike Suite 200N
HFM-355
Rockville, MD 20852-1448 (USA)
Tef 301 594 6487
Fax: 301 594 6431
144.173.6.94 - 5/6/2020 3:28:22 AM

97
University of Exeter
Downloaded by:
 References
1 Engelfriet CP, Voak D: International reference
polyspecific anti-human globulin reagents Vox
Sang 1987;53.241—247
2 Scott ML, Bird GWG: Some contributions of
the plant kingdom to transfusion medicine -
Lectins and plant enzymes. Transfus Med Rev
1992;6:103-115.
3 Cullen EC, Knight RC: A simple technique us­
ing bromelain for enzyme treating red cells
XXth Congr Int Soc Blood Transfus, London,
1988, pp 146
4 LewisAJ: Papain, ficin and bromelain in the de­
tection of incomplete rhesus antibodies. Br J
Haematol 1957:3.332-339.
5 Moore BPL. Tests with protease extracts; in
Moore BPL (ed) Serological and Immunolog­
ical Methods of the Canadian Red Cross Blood
Transfusion Service, ed 8. Toronto, Canadian
Red Cross Society, 1980, pp 65-72.
6 Lambert R, Edwards J, Anstee DJ: A simple
method for the standardisation of proteolytic
enzymes used in blood group serology. Med
Lab Sci 1978,35:233-243
7 Phillips PK, Prior D, Dawes B: A modified
azoalbumin technique for the assay of proteo­
lytic enzymes for use in blood group serology.
J Clin Pathol 1984:37 329-331.
8 Scott ML, Voak D, Downie DM. Optimum en­
zyme activity, and a new technique for antibody
detection: An explanation for the poor perform­
ance of the one-stage mix technique. Med Lab
Sci 1988;45:7-18.
98
9 Glazer AN, Smith EL: Papain and other plant
sulfhydryl proteolytic enzymes; in Boyer PD
(ed): The Enzymes. New York, Academic
Press, 1971, voi 3.
10 Mazda TM, Ogasawara K, Nakata K, Shimizu
M: Standardization of bromelain for use in rou­
tine one-stage antibody screening. Vox Sang
1987;52:63-66.
11 Scott ML, Whitton CM: Standardisation of pa­
pain reagents by measurement of active sites
using a synthetic inhibitor, E-64. Transfusion
1988,28-24-28
12 Damtree K, Scott N' Papain standardisation us­
ing benzoyl-L-argimne-p-nitroanilide (BAP-
NA) Vllth Meet Br Blood Transfus Soc, Dur­
ham, 1989, p 93.
13 Scott ML, Guest AG, Mushens RE: Assessment
of components of papain and bromelain in ser­
ological and biochemical assays. Transfus Med
1990;l(suppl 1)17.
14 Scott ML, Phillips PK. A standardised stable
liquid papain preparation. Illrd Meet Br Blood
Transfus Soc, Oxford, 1985, p 74
15 Scott ML, Johnson CA, Phillips PK: The pH op­
tima for papain and bromelain treatment of red
cells. Vox Sang 1987;52:223-227.
16 Campbell E, Scott ML: Enzyme techniques in
blood group serology: The effect of ionic
strength. Med Lab Sci 1991;48:52-58.
17 Kuhns WJ, Bailey A: Papain treatment of red
cells. Am J Clin Pathol 1950;20:1067-1070.
18 Stratton F- A slide test for the detection of Rh
antibodies using papam-treated red cells. Vox
Sang 1953;3:43-50.
Scott/Voak/Phillips/Hoppe/Kochman
19 Goldsmith K: Papain-treated red cells in the de­
tection of incomplete antibodies. Lancet 1955;i:
76-77
20 Löw B : A practical method using papain and in­
complete Rh-antibodies in routine Rh blood
grouping. Vox Sang 1955,5:94-98.
21 Allan JC, McMillan J, Whitelaw S, et al: The
effect of crude papain on anti-D typing. Vllth
Meet Blood Transfus Soc, Durham, 1989, p 97.
22 Dodd BE, Eeles DA- Rh antibodies detectable
only by enzyme technique. Immunology 1961;
4:337-345.
23 Odell WR, Roxby DJ, Ryall RG, et al' A LISS
spin enzyme method for the detection of red cell
antibodies and its use in routine antibody screen
procedures Transfusion 1983;23:373-376.
24 Scott ML, Phillips PK: A sensitive two-stage
papain technique without cell washing. Vox
Sang 1987;52:67-70
25 Barratt AJ, Kembhavi AA, Brown MA: L-trans
epoxy succinyl leucylamido(4-guanidino )bu-
tane (E64) and its analogues as inhibitors of
cysteine proteases including cathepsms B, H
andL.BiochemJ 1982:201-189-198.
26 Voak D: Validation of new technology for anti­
body detection by antiglobulin tests. Transfus
Med 1992;2-177-179.
27 Phillips PK- External quality assessment of
blood grouping, antibody screening and cross­
match procedures within the United Kingdom,
in Rock G, Seghatchian MJ (eds): Quality As­
surance in Transfusion Medicine Boca Raton,
CRC Press Inc. 1992, voi. 1.
144.173.6.94 - 5/6/2020 3:28:22 AM
Problems Involved in Using Enzymes in
Blood Group Serology
University of Exeter
Downloaded by: