Immunocytochemical Characterization of Human Blood Cells
LUNG T. YAM, M.D., MARY C. ENGLISH, ANTHONY J. JANCKILA, M.S.,
STEVE ZIESMER, B.S., AND CHIN-YANG LI, M.D.
A simple immunocytochemical method using calf intestinal
alkaline phosphatase as the enzymatic indicator to demonstrate
surface antigens on human blood cells has been developed. The
blood cells were labeled with cell specific monoclonal antibodies followed by linkage with an antiimmunoglobulin alkaline
phosphatase conjugate. Cytochemical demonstration of alkaline phosphatase activity on the blood cells reflects the presence of surface antigens on these cells. The effects on the cytochemical reaction of fixation, substrates, couplers, activators
and inhibitors, and storage of cytologic materials have been
examined systematically. The best staining conditions are to
incubate labeled smears in a 0.04 M barbital buffer at pH 7.6
containing 30 mg% naphthol AS-TR phosphate, 40 mg% fast
red ITR, and 1 mivi levamisole. This method is both sensitive
and specific and appears most practical for objective identification of the human blood cells. (Key words: Immunocytochemistry; Enzyme cytochemistry; Monoclonal antibodies;
Surface antigens; Human blood cells) Am J Clin Pathol 1983;
80: 314-321
Veterans Administration Medical Center and the Department
of Medicine, University of Louisville School of Medicine,
Louisville, Kentucky, and The Departments of Laboratory
Medicine and Surgical Pathology, Mayo Clinic
and Mayo Foundation, Rochester, Minnesota
ACCURATE IDENTIFICATION of the blood cells is
essential both for proper diagnosis and management of
hematologic diseases and for conducting meaningful cell
research. This has been done in the past by examining
tissues stained with either Wright-Giemsa stains or hematoxylin and eosin. More recently, the cytochemical
and immunologic cell-specific markers also have been
applied for cell identification.1,3,6,710,21,25 The cytochemical markers are most useful for the monocytes and granulocytes but not for the lymphocytes. 316,25 Accurate
identification of the lymphocyte subsets, however, can
be achieved in most instances by demonstration of certain cell-specific surface markers with immunologic
techniques1,6,7,10,21; the most powerful of these involving
the highly specific monoclonal antibodies.6,7,21
There are many monoclonal antibodies available as
specific markers for the human blood cells, particularly
for the various lymphocyte subsets.7,21 It is possible to
demonstrate these antigens through immunologic technics involving immunofluorescence and microscopy or
cytofluorographic analysis. Although these methods
have proven to be very useful in many instances, it is,
from a diagnostic clinical point of view, still most deReceived January 3, 1983; received revised manuscript and accepted
for publication March 21, 1983.
Supported in part by a research grant from the Research Service of
the Veterans Administration, Washington, D.C.
Address reprint requests to Dr. Yam: Veterans Administration
Medical Center, 800 Zorn Avenue, Louisville, Kentucky 40202.
sirable to be able to simultaneously identify the cells
through their cell-specific surface antigens and to appreciate the morphologic cellular changes by light microscopy. This can be accomplished by enzymatic immunocytochemistry. The immunoperoxidase technic is
employed most often for such immunodiagnostic purposes. 524 However, the clinical utility of this technic may
be limited in tissues containing many granulocytes and
monocytes with abundant endogenous peroxidase that
cause significant interpretive and technical difficulties.
Attempts made at inactivating the endogenous enzyme
activity invariably have proved to be deleterious to the
surface antigens. For these reasons, we believe that the
immunoalkaline phosphatase method 218 has definite
advantages over the immunoperoxidase method for this
purpose.
In this article, a simple immunochemical method
using alkaline phosphatase as an enzymatic indicator
has been developed for the light microscopic demonstration of cell-specific antigens on the various human
blood cells.
Materials and Methods
Preparation of Cells
Blood. A small volume (4 ml) of heparinized peripheral blood was diluted 2:1 with warm plasma gel (HTI
Corp., Buffalo, NY) and left at 37°C for 30 minutes to
allow the erythrocytes to settle. The leukocyte-rich
plasma was harvested, and the leukocytes pelleted and
washed three times with RPMI-1640 culture medium
(M. A. Bioproducts, Walkersville, MD). The leukocyte
count was adjusted to 107 cells/ml with this medium.
Marrow. Marrow aspirates were obtained in the conventional manner and anticoagulated with heparin. After they had served their clinical purpose, the residual
specimens from patients with anemia of chronic diseases
and minor hematologic abnormalities were used for
study. The marrow aspirates were pipetted repeatedly
0002-9173/83/0900/0314 $01.20 © American Society of Clinical Pathologists
314
Vol. 80 • No. 3
IMMUNOCYTOCHEMISTRY OF HUMAN BLOOD CELLS
with a Pasteur pipette to break up the marrow particles.
They were layered over a Ficoll-paque cushion (Pharmacia Fine Chemicals, Piscataway, NJ) and centrifuged
at 400 g for 30 minutes. Cells at the interface were harvested, washed three times as for blood cells, and adjusted to 107 cells/ml.
Recommended Staining Method
Reagents
1. Monoclonal antibodies and alkaline phosphatase
conjugated antiimmunoglobulins were purchased from
commercial sources and diluted with balanced salt solution (BSS) just before use. These antibodies included:
OKIa-1 specific for the framework DR antigen; OKT3 specific for peripheral blood T-cell antigens; OK.T-4
specific for T-helper cells; OK.T-8 specific for T-suppressor cells; and OKM-1 specific for myeloid cells.
These antibodies were purchased from Ortho-Pharmaceuticals, Raritan, New Jersey. Other antibodies used
occasionally for testing were monoclonal antibodies to
common ALL antigen including J5 (Coulter Corp.,
Hialeah, FL) and #94535A (Bethesda Research Laboratories, Gaithersburg, MD), Anti Leu-1, 3a, 2a, 4, and
HLA-DR (Becton-Dickenson, Sunnyvale, CA). Alkaline
phosphatase conjugated F(ab')2 fraction of goat
antimouse IgG was purchased from Tago, Inc. (Burlingame, CA).
2. Fixative: The citrate buffered 60% acetone solution
of Kaplow1' was used, except that the /?H of the fixative
was adjusted to 5.4. Thisfixative,when kept at 4-10°C,
is stable for at least a month.
3. Barbital buffer (0.04 M, p\\ 7.6): Prepared according to Pearse19 and adjusted to pH 7.6. This solution is
stable for at least three months when stored at 4-10°C.
4. Naphthol AS-TR phosphate, sodium salt.*
5. Fast red ITR salt.*
6. Levamisole.*
7. Mayer's hematoxylin: Self-prepared17 or purchased from a commercial source.
8. Glycerine jelly: Self-prepared18 or purchased from
a commercial source.
9. RPMI 1640 or other balanced salt solutions.
10. Fetal calf serum (FCS).
11. Normal goat serum.
315
a 1/20 dilution of normal goat serum for 10 minutes
at 4°C.
3. Centrifuge and discard supernate. Suspend in 0.2
ml of a 1/40 dilution of monoclonal antibody for 30
minutes at 4°C.
4. Centrifuge and discard supernate. Wash three
times in BSS by centrifugation.
5. Suspend in 0.2 ml 1/20 dilution of alkaline phosphatase conjugated goat antimouse IgG for 30 minutes
at 4°C.
6. Centrifuge and discard supernate. Wash three
times in BSS.
7. Suspend in 1 ml 10% FCS in BSS.
8. Prepare cytocentrifuge (Shandon Southern Instruments) smears using two drops of cell suspension and
four drops 10% FCS. Allow to dry thoroughly.
9. Fix smears with cold citrate buffered acetone for
30 seconds, and wash with distilled water.
10. Incubate smears in the following medium for 3045 minutes at 37°C. Filter the incubation medium before use.
Barbital buffer pH 7.6
50 ml
Naphthol AS-TR phosphate,
sodium salt
15 mg
Fast red ITR salt
20 mg
Levamisole
12mg(lmM)
11. Wash smears with water. Counterstain with
Mayer's hematoxylin for 1-5 minutes.
12. Wash in running tap water for 5-10 minutes.
13. Dry thoroughly and mount in glycerine jelly.
Controls were prepared by substituting nonimmune
mouse ascites for monoclonal antibody in step 3. When
it is desirable to label simultaneously the cells with both
cell-specific cytochemical and immunochemical markers, the smears prepared at step 9 are stained for cytochemical markers before proceeding for the immunochemical markers at step 10. The methods for the cytochemical markers such as peroxidase,12 chloroacetate
esterase,25 acid phosphatase,14 tartrate resistant acid
phosphatase,8 and nonspecific esterase" are those reported previously.
Technical Parameters Tested
Staining Procedure
1. Add 0.2 ml (2 X 106) cells to a series of 1.5 ml
conical centrifuge tubes (one tube per monoclonal antibody).
2. Centrifuge (Brinkmann or Fisher) for 30 seconds
(at 3,200 g). Discard supernate. Suspend in 0.2 ml of
* Sigma Co., St. Louis, MO.
In order to establish the optimal staining conditions,
we examined the effects of various parameters on the
ability to demonstrate antigens on the cell surface. These
included (1) fixation; (2) substrates; (3) couplers; (4) inhibitors of endogenous alkaline phosphatase and activators of calf intestinal alkaline phosphatase; (5) buffer,
pH, temperature, and time of incubation; (6) counterstains and mounting media; and (7) storage of smears.
A.J.C.P. • September 1983
YAM ET AL.
316
Table 1. Surface Characteristics of the Human Blood Cells
Monoclonal Antibodies
Normal Mouse
Ascites
Cell Types
Myeloblasts
Promyelocytes
Myelocytes
Neutrophilic
Eosinophilic
Basophilic
Metamyelocytes
Neutrophilic
Eosinophilic
Basophilic
Band, neutrophilic
Neutrophils
Eosinophils
Basophils
Monocytes
Lymphocytes
Plasma cells
Pro-erythroblasts
Basophilic erythroblasts
Polychromatophilic erythroblasts
Orthochromatophilic erythroblasts
Megakaryocytes
OKMl
OKIal
OK.T-4
OKT-3
_
—
-l±
-l±
-
-/+
-/+
-l±
?
-l±
?
9
?
-
-/++
-/+
-l±
~l±
9
9
9
-
-/+ + +
-/+ + +
-/+
-/+
-/+
9
9
-/+ + +
-/+
-/+
-/+
~l±
-
9
-
~l±
—
"/±
OKT-8
_
—
_
_
-
-
-
-
-
-
-
-
-
-
~l±
-/±
-/±
-/+ +
-
-/+
-
-/+
-
9
9
9
9
9
?
9
9
•• negative: ± c very weak or questionable: + = weak. + + = moderate; + + + = strong staining: - / + = either negative or weakly positive: ? = unknown.
Results
Recommended Method
Presence of surface antigen is indicated as a bright red
granular deposit on the cells. Endogenous alkaline phosphatase in the neutrophils is effectively inhibited by levamisole20 and the relatively low pH of the medium. The
activity of calf intestinal alkaline phosphatase used in
the conjugated second antibody is unaffected by these
conditions. The occurrence of surface antigens on human marrow cells is shown in Table 1. Morphologic
detail is well preserved and a high degree of correlation
between morphology and surface antigen expression was
observed. The stain may be distributed evenly in a ringlike fashion around the surface of the cell or in one or
more patches over the cell surface. The staining pattern
depends greatly on the antigen detected and the cell type
possessing it. For example, the staining of OKM-1 on
monocytes frequently is seen as numerous vacuoles, suggesting an internalization of labeled membrane. Neutrophils may exhibit this same pattern (Fig. 1) with
OKM-1 but typically to a lesser extent and frequently
these cells appear capped (Figs. 1-3). OKT-8, on the
other hand, does not appear to be extensively modulated
by suppressor T cells and most often gives a ringlike
staining pattern or a strong, diffusely distributed pattern
over the entire cytoplasm (Fig. 4). For simultaneous
demonstration of the cell-specific cytochemical markers,
best results were achieved by staining these markers be-
fore the smears were stained for alkaline phosphatase
(Figs. 3 and 4).
Table 2 summarizes the results when leukocytes from
five normal subjects were stained with the panel of
monoclonal antibodies. The number of OKT-3+ cells
was roughly equal to the sum of OKT-4+ and OKT-8+
cells. Anti-DR antibody (OKIa-1) stained many monocytes and a small number of lymphocytes and eosinophils. Greater than 95% of the monocytes, granulocytes,
and eosinophils were OKM-l + . OKM-1 also stained a
small proportion of morphologically identifiable lymphocytes. It is presumed that these "lymphocytes" are
the null cells with natural killer activity described by
Breard and co-workers.4
Effects of Fixation
Several fixatives were tested. These included the fixative containing 3% paraformaldehyde, 0.2% glutaraldehyde, and 0.2% acrolein (PGA),3 formalin vapor, absolute methanol, phosphate buffered formyl acetone,25
unbuffered 60% acetone, 10% methanol in 60% acetone,14 and citrate buffered 60% acetone." The time of
fixation tested ranged from 30 seconds to five minutes.
PGA often resulted in poor cellular morphology, with
cells appearing swollen and having no clearly defined
membranes. Fixatives containing formalin or methanol,
or both, are very good for preserving cellular morphology, however, they often resulted in decreased number
of positive cells, especially when the concentrations of
IMMUNOCYTOCHEMISTRY OF HUMAN BLOOD CELLS
Vol. 80 • No. 3
^» %di •
317
jTk
• %
# ^
W- ,; •
c
# *
• P 1 "r .
^^V
r>*mM
•*drf\
FIG. 1 (upper, left). OKM-1 antigen on peripheral blood monocytes and granulocytes. Monocytes frequently endocytose labeled antigen giving
a vacuolar staining pattern (arrow). Immunoalkaline phosphatase; naphthol AS-TR and fast red ITR method (X1,600).
FIG. 2 (upper, right). OKM-1 antigen on three myelocytes in the bone marrow. The labeled antigen displays a distinct surface pattern in one
cell and interrupted cytoplasmic pattern in two cells. One neutrophil is not labeled. Naphthol AS-TR-phosphate, fast red ITR (X1,600).
FlG. 3 (lower, left). Combined method for immuno-alkaline phosphatase (red) and chloroacetate esterase (blue), showing OKM-1 antigen (red)
in the bone marrow cells. The antigen present in the granulocytes (blue cytoplasm) is predominately in the mature cells. OKM-1 also is present
in one monocyte (red granular cytoplasm) but not in the lymphocyte (pale cytoplasm). Naphthol AS-TR phosphate, fast red ITR. Naphthol
AS-D chloroacetate, fast blue BBN. Counterstained with 1% methyl green (X 1,600).
FIG. 4 (lower, right). OKT-8 antigen on two suppressor T-Iymphocytes. Note strong antigen expression with ring or diffuse staining pattern.
Immunoalkaline phosphatase (red) by naphthol AS-TR and fast red ITR, and chloroacetate esterase (blue) by naphthol AS-D chloroacetate, fast
blue BBN, counterstained with 1% methyl green (X 1,600).
methanol were above 10%. Unbuffered acetone caused
cell shrinkage with poor morphologic preservation. Citrate buffered 60% acetone resulted in the best preservation of surface antigens and most often provided acceptable morphologic detail. At times, this fixative may
unpredictably fail to give optimal morphology with
some cells being lysed.
Effects of Substrates
Substrates examined included naphthol AS phosphate, naphthol AS-BI phosphoric acid, and naphthol
AS-TR phosphate in concentrations ranging from 5-50
mg per 50 ml of incubation medium. The sensitivities
of these substrates were naphthol AS-TR ^ naphthol AS
> naphthol AS-BI. The chromogenicity of the reaction
products of these three substrates varied with the couplers used, but naphthol AS-TR phosphate in general
yielded highly chromogenic reaction products with precise localization and little or no background stain.
Therefore, naphthol AS-TR phosphate was considered
as the substrate of choice in this study. It required only
30 minutes incubation for optimal staining. Optimal
substrate concentration for naphthol AS-TR phosphate
was 15 mg per 50 ml incubating medium. Increasing
the concentration did not enhance staining, while de-
YAM ET AL.
318
Table 2. Normal Values of Surface Markers in
Peripheral Blood Cells* (500 Cells
Counted for Each Antibody)
Monoclonal Antibodies
Cell Types
OKT-3
OK.T-4
OKT-8
OKM-1
OKJa-1
Lymphocyte
Monocyte
Granulocytest
(Neutrophils)
73 ± 4
42 ± 7
32 ± 3
13± 5
95 ± 2
17 ± 7
30 ± 14
99 ± 1
* Expressed as % positive cells ± SD (n = 5).
t Eosinophils may be positive or negative for OKM-I and OKIa-l.
creasing the concentration yielded a proportional decrease in intensity of the reaction product.
Effects of Couplers
The couplers that have been tested include hexazotized new fuchsin and hexazotized pararosaniline in concentrations of 0.25-2.5 ml per 50 ml, and fast garnet
GBC, fast red ITR salt, fast violet B salt, fast violet LB
salt, and fast blue BB and RR salts in concentrations
of 5-50 mg per 50 ml. The choice of an ideal coupler
is governed by its stability in alkaline pH, mountability
in aqueous or organic solvent media, and chromogenicity of the reaction product. Both fast garnet GBC and
hexazotized pararosaniline are not stable at pH 7.6 and
were excluded from consideration. Fast blue BB and RR
salts are good couplers in alkaline media, with reaction
products that are insoluble in organic or aqueous
mounting media. The fast blue salts, however, are less
sensitive than fast red ITR, fast violet LB salt, and fast
violet B salts when naphthol AS-TR is used as substrate.
Hexazotized new fuchsin gives a moderately chromogenic reaction product. Smears stained with this coupler
can be mounted in either aqueous or organic mounting
media; however, its reaction product at the light microscopic level appears diffuse, making proper interpretation of weak staining particularly difficult. It is also less
sensitive than fast red ITR or fast violet B salts. Both
fast violet B salt and fast violet LB salt produce a very
fine product with little or no crystallization, but the precipitates are often too small to be appreciated under low
power microscopy. Fast red ITR salt provides a granular
and highly chromogenic precipitate that is easily visible
under low power light microscopy and gives very little
background color at pH 7.6. Therefore, it is used as the
coupler in our immunocytochemical procedures. The
optimal concentration for this coupler is 20 mg per 50
ml incubation medium. Fast red ITR must be mounted
in aqueous medium and has a tendency to form nonspecific crystalline precipitates in cells several days after
mounting.
A.J.C.P. • September 1983
Effects of Inhibitors
Both levamisole20 and ethylene diamine tetraacetate
(EDTA) were examined for their inhibitory effect on
alkaline phosphatase. Addition of 1 mM levamisole to
the incubation medium minimized background staining
by inhibiting the endogenous enzymes, in neutrophils,
vascular endothelial cells, and in tissues other than the
small intestine; it did not appreciably suppress the activity of the intestinal phosphatase. Ethylene diamine
tetraacetate (EDTA) at concentrations of 0.05 mM to
5 mM had limited inhibitory effect on endogenous alkaline phosphatase and no effect on the calf intestinal
alkaline phosphatase label.
Divalent cations including Zn++, Mn++, and Mg++
were examined for their ability to activate calf intestinal
alkaline phosphatase. Biochemical assay of 0.22 IU of
enzyme (Sigma type VII) was carried out according to
described methods22 with minor modifications. Barbital
buffer, 0.04 M, pH 7.6, was used for incubation. Cations
were added at concentrations of 10"6 to 10-2 M in the
form of chloride or sulfate salts. Absorbance was measured at 410 nm and compared with that in the absence
of activator. Maximal activation by Zn ++ , Mn ++ , and
Mg++ was achieved at 10~4 M, 10"4 M, and 10~2 M, respectively (Fig. 5). These cations then were added at
these concentrations to the incubating medium in efforts
to enhance the cytochemical sensitivity. The addition
of Zn++ proved deleterious to surface antigen detection.
Staining became diffuse rather than granular and prevented the detection of a majority of cells with weak
antigens.- The addition of Mg++ at 10~2 M or Mn ++ at
10~4 M enhanced the intensity by producing larger granular deposits. This enhancement, however, did not, provide for an increase in the number of positive cells.
Magnesium caused the unwanted effect of reactivating
endogenous enzyme activity in the leukocytes or abrogating the inhibitory effect of levamisole.
Effects of Other Staining Conditions
Tris, 2-amino-2 methyl-propanol and barbital buffers
of pH range of 7.0 to 9.8 and variable ionic strengths
were examined. Four-hundredths molar veronal buffer
is the most satisfactory. Although increased pH of the
incubating medium will enhance the activity of the alkaline phosphatase label, it also becomes difficult to sufficiently inhibit the endogenous enzyme activity. A decrease in the pH of the incubating medium greatly reduces the endogenous enzyme activity, while it suppresses
the calf intestinal enzyme only moderately. Optimal pH
for the immunocytochemical reaction appears at pH 7.6.
Further decrease of pH will cause a significant decrease
in enzyme label staining. The use of Tris buffer significantly reduced the sensitivity of the stain, resulting in
Vol. 80 • No. 3
IMMUNOCYTOCHEMISTRY OF HUMAN BLOOD CELLS
decreased numbers of positive cells. This was particularly noticeable for those antigens that are weakly expressed. The propandiol buffer is unsuitable for use at
/?H 7.6 because of poor buffering power at this pH.
Time of incubation is dependent on the temperature
and the substrate used. A 30-minute incubation at 37°C
is optimal. Increasing the temperature causes inactivation of the enzyme and decomposition of the azo dye,
resulting in nonspecific precipitation. Incubation times
exceeding 30 minutes also generate nonspecific background staining resulting from dye decomposition. Less
specific substrates require longer staining times, again
increasing the occurrence of nonspecific precipitation.
Effects of Counterstain and Mounting Media
Several nuclear counterstains including hematoxylin,
methyl green, methylene blue, toluidine blue, and neutral red in concentrations of 0.1-1% were examined.
Hematoxylin is the counterstain of choice, producing
excellent nuclear detail and fine contrast to the red color
of the surface label. Smears can be mounted permanently in either aqueous or organic mounting media.
Methyl green and neutral red yielded poor nuclear detail
and rapidly faded when mounted in aqueous media.
Methyl green, however, did not stain the cells deeply
and therefore exerted only minimal masking effect on
the alkaline phosphatase activity. It may be of some use
in reactions with weak alkaline phosphatase activity
(Figs. 3 and 4). Methylene blue and toluidine blue yield
discouraging results, because they may deeply tint the
cytoplasm, thus masking enzymatic activity. In addition,
these stains must be mounted in organic solvents,
thereby minimizing their application in this method for
the immunoalkaline phosphatase reaction. Glycerine
jelly was chosen as the aqueous-based medium for enzyme stains with a reaction product soluble in organic
solvents. Diatex® (Scientific Products) in xylol is the
organic medium of choice for cytochemical stains with
a reaction product that is insoluble in organic solvents.
Methacrylate and Permount yield results similar to
Diatex.
Effects of Storage
In order to examine the effects of storage on the sensitivity of the immunoalkaline phosphatase stain, cell
suspensions were labeled as described in the "Materials
and Methods" section. Smears either were left unfixed
or fixed in citrate buffered 60% acetone or buffered formal acetone and stored at 4°C or at room temperature
for up to one week. Unfixed smears werefixedjust before
staining with citrate buffered acetone. The percentage
of positive cells were determined after 1, 3, and 7 days
of storage and compared with freshly stained controls.
319
l.5-i
o
o
CO
K
O
CD
/
/
^^*T/
*-
"^-
/
/
»
<0.5-
0
Mn*+
Mgft
•
•
•
•
*
*• Z n * +
10-6
io-5
l0-«
10-3
ION CONCENTRATION (Moles/L)
10-2
FIG. 5. Activation of calf intestinal alkaline phosphatase by Mn ++ ,
Mg ++ , and Zn ++ ions. This was achieved best with 10"4 M, Mn ++ .
Mn ++ above 10"4 M precipitated in alkaline solution. Zn ++ was less
effective as an activator.
Immediately apparent was the differential "stability"
of the various labeled antigens and of a given antigen
that might be present on several cell types. In general,
OKM-1 in monocytes and OKT-8 in lymphocytes are
stable. OKT-3, OKT-4, OKJa-1 in lymphocytes, and
OKM-1 in neutrophils are somewhat labile markers and
OKM-1 in lymphocytes and OKJa-1 in monocytes are
very sensitive to storage. The findings are demonstrated
in a typical experiment as shown in Figure 6. This general trend was seen, regardless of fixation or storage conditions. However, storage of unfixed smears more consistently provided better preservation of enzyme activity
as shown in another typical experiment in Figure 7.
While the quantitative differences in sensitivity (i.e., the
number of positive cells detected) may not be significant,
fixation resulted in a weaker and more diffuse reaction
product in stored preparations. This impaired quality
of stain is even more pronounced when cells are fixed
with formalin-containing solutions. While storage at
4°C may provide slightly better quantitative results for
some cell types, the staining is more diffuse, compared
A.J.C.P. • September 1983
YAM ET AL.
320
The results of the surface antigens on the human
blood cells as demonstrated immunocytochemically in
this study are very similar to those by immunofluorescence and flow cytometry. The additional use of cytochemistry along with the immunoalkaline phosphatase
technic undoubtedly will help to identify accurately leukocytes with increased precision.
While many substrates and couplers may be used, the
most sensitive and chromogenic combination is the
naphthol AS-TR phosphate and fast red ITR. Other substrate/coupler combinations such as naphthol AS-TR
phosphate/fast violet B or fast violet LB salt, naphthol
AS phosphate/fast violet B salt or fast violet LB salt also
may be suitable, although these combinations are less
sensitive than the recommended combination. One
drawback of fast red ITR salt is its tendency to salt out
o
o
o
Q
z
o
tr
UJ
UJ
0.
DAYS
FIG. 6. Stability of the various antigens in monocytes, lymphocytes,
and neutrophils. Cells were labeled as described in the "Materials and
Methods" section on day 0. Cytocentrifuge smears were prepared and
stored unfixed. Smears were then fixed and stained on days 0, 1,3,
and 7. The percentages of lymphocytes (
), monocytes (
), and
neutrophils (• • •), which stained positively, with the listed monoclonal
antibodies, were determined on each day. Percentage of day 0 control
= % positive cells on day 1, 3, or 7/% positive cells on day 0 X 100.
with that of cells kept at room temperature. Cold storage
of fixed smears does not provide results as consistent as
room-temperature storage of unfixed smears. The addition of manganese as an activator did not enhance the
quantitative or qualitative staining in stored smears.
Discussion
Although the immunoalkaline phosphatase technic
has been used elsewhere for diagnostic purposes,2,9'8-23
a systematic examination of the technical aspects of this
technic has not been reported. The method described
herein is relatively simple, sensitive, and specific. Endogenous alkaline phosphatase activity in cells other
than those of small intestinal mucosa can be inhibited
adequately by the combined use of a relatively low pH
and the inclusion of levamisole in the final disclosure
reaction.
25 -
— unfixed
- - fixed
0KT3
OKT-8
OK la-1 in monocytes
~i
7
DAYS
FIG. 7. Correlation between stability of labeled surface antigens and
fixation. Cells were labeled as described in the "Materials and Methods" section on day 0. Cytocentrifuge smears were prepared and stored
unfixed (
) or fixed (
). Unfixed smears werefixedjust before
staining for alkaline phosphatase on days 0, 1,3, and 7. Fixed smears
were stained without further treatment. The percentages of OKT-3+
lymphocytes (•), OKT-8+ lymphocytes (A), and OKIa-1 positive
monocytes (D) were determined on each day. Percentage of day 0
control = % positive cells on day 0, 1,3, or 7/% positive cells on day
0 X 100.
Vol. 80 • No. 3
IMMUNOCYTOCHEMISTRY OF HUMAN BLOOD CELLS
in mounted slides, causing some crystalline nonspecific
precipitates. For this reason, smears should be evaluated
the same day they are prepared, and review of older
slides should be made with caution. If smears cannot
be evaluated the day of staining, then they may be kept
unmounted until it is convenient to do so.
The inhibition of endogenous leukocyte alkaline
phosphatase by the inclusion of levamisole in the final
disclosure reaction is essential for accurate interpretation of the results. The addition of magnesium either
re-activates the enzyme or abrogates the effect of levamisole, and, therefore, should not be used as an activator. While the addition of manganese at 10"4 M does
not result in increased numbers of stained cells, it does
enhance the intensity of staining in fresh smears. The
addition of this activator is optional, depending on the
preference of each individual.
The sensitivity of the method is affected by the conditions under which the smears are prepared and stored.
Best results are obtained when the entire procedure is
conducted the same day. If this cannot be accomplished,
then the cells should be labeled and smears prepared
and left unfixed at room temperature for subsequent
fixation, staining, and evaluation. It should be kept in
mind that under these circumstances, the detection of
certain antigens (therefore, cell types) may be less reliable than others. The monoclonal antibodies tested in
this study were purchased mainly from a single commercial source. Those from other manufacturers are
likely to vary with respect to specificity, sensitivity, and
stability when applied as immunocytochemical reagents.
It may be necessary for each investigator to evaluate his
preferred antibodies to suit his own needs. Technical
parameters that require careful attention in conventional cytochemical methods apply as well to immunocytochemical procedures.81316 Deleterious effects of
various manipulations always should be considered
when interpretations are made.
Ultimately, a method to identify cell types through
the demonstration of specific surface antigens in airdried bloodfilmsor tissue imprints would be the optimal
diagnostic tool in clinical practice. Our preliminary studies in this area have revealed a number of stumbling
blocks. We currently are examining a number of modifications to circumvent these problems and to determine the optimal conditions for immunocytochemical
demonstration of cell surface antigens in commonly
available clinical material. In addition, studies are under
way to increase our experience with the technic herein
and to document its correlative reliability with other
established procedures for cell identification.
Acknowledgment. The authors thank Miss Pat Hagan for typing the
manuscript.
321
References
1. Aisenberg AC: Cell-surface markers in lymphoproliferative disease. N Engl J Med 1981; 304:331-336
2. Avrameas S: Coupling of enzymes to proteins with glutaraldehyde.
Use of the conjugates for the detection of antigens and antibodies. Immunochemistry 1969; 6:43-52
3. Beckstead JH, Halverson PS, Ries CA, Bainton DF: Enzyme histochemistry on biopsy specimens of pathologic human bone
marrow. Blood 1981; 57:1088-1098
4. Breard J, Reinherz EL, Kung PC, Goldstein G, Schlossman SF:
A monoclonal antibody reactive with human peripheral blood
monocytes. J Immunol 1980; 124:1943-1948
5. DeLellis RA, Sternberger LA, Mann RB, Banks PM, Nakane PK:
Immunoperoxidase techniques in diagnostic pathology. Report
of a workshop sponsored by the National Cancer Institute. Am
J Clin Pathol 1979; 71:483-488
6. Diamond BA, Yelton DE, Scharff MD: Monoclonal antibodies.
A new technology for producing serologic reagents. N Engl J
Med 1981; 304:1344-1349
7. Foon KA, Schroff RW, Gale RP: Surface markers on leukemia
and lymphoma cells. Recent advances. Blood 1982; 60:1-19
8. Janckila AJ, Lam KW, Li CY, Yam LT: The cytochemistry of
tartrate resistant acid phosphatase, I. Technical considerations.
Am J Clin Pathol 1978; 70:45-55
9. Janckila AJ, Stelzer GT, Wallace JH, Yam LT: Phenotype of the
hairy cells of leukemic reticuloendotheliosis defined by monoclonal antibodies. Am J Clin Pathol 1983; 79:431-437
10. Janossy G, Bollum FJ, Bradstock KF, Ashley J: Cellular phenotypes of normal and leukemic hemopoietic cells determined
by analysis with selected antibody combinations. Blood 1980;
56:430-441
11. Kaplow LS, Burstone MS: Acid buffered acetone as a fixative for
enzyme cytochemistry. Nature 1963; 200:690-691
12. Kaplow LS: Simplified myeloperoxidase stain using benzidine
dihydrochloride. Blood 1965; 26:215-219
13. Kaplow LS: Leukocyte alkaline phosphatase cytochemistry: Applications and methods. Ann NY Acad Sci 1968; 155:911-947
14. Li CY, Yam LT, Lam KW: Acid phosphatase isoenzyme in human leukocytes in normal and pathologic conditions. J Histochem Cytochem 1970; 18:473-481
15. Li CY, Lam KW, Yam LT: Esterases in human leukocytes. J
Histochem Cytochem 1973; 21:1-12
16. Li CY, Yam LT, Crosby WH: Histochemical characterization of
cellular and structural elements of the human spleen. J Histochem Cytochem 1972; 20:1049-1058
17. Luna LG: Manual of histologic staining methods of the Armed
Forces Institute of Pathology. Third edition. New York,
McGraw-Hill, 1968, pp 33, 141
18. Mason DY, Sammons R: Alkaline phosphatase and peroxidase
for double immunoenzymatic labelling of cellular constituents.
J Clin Pathol 1978; 31:456-460
19. Pearse AGE: Histochemistry, theoretical and applied. Third edition. Vol 1. Boston, Little, Brown, 1968, p 584
20. Ponder BA, Wilkinson MM: Inhibition of endogenous tissue alkaline phosphatase with the use of alkaline phosphatase conjugates in immunohistochemistry. J Histochem Cytochem
1982;29:981-984
21. Reinherz EL, Schlossman SF: Regulation of the immune response-inducer and suppressor T-lymphocyte subsets in human beings. N Engl J Med 1980; 303:370-373
22. Sigma Technical Bulletin 104. Phosphatase, acid, alkaline, and
prostatic; Colorimetric method for serum or urine. Sigma
Chemical Company, St. Louis, Missouri, 1982
23. Sigma Technical Bulletin 95. Immunohistochemical method for
detecting surface immunoglobulins of mature lymphocytes.
Sigma Chemical Company, St. Louis, Missouri, 1981
24. Taylor CR: Immunoperoxidase techniques: Practical and theoretical aspects. Arch Pathol Lab Med 1978; 102:113-121
25. Yam LT, Li CY, Crosby WH: Cytochemical identification of granulocytes and monocytes. Am J Clin Pathol 1971; 55:283-290