FUNDAMENTAL AND APPLIED TOXICOLOGY 34, 2 5 - 3 5 (1996)
ARTICLE NO. 0172
Phenotypic Analysis of Lymphocyte Subpopulations in Lymph Nodes
Draining the Ear Following Exposure to Contact Allergens and Irritants
E. E. SIKORSKI, G. F. GERBERICK, C. A. RYAN, C. M. MILLER, AND G. M. RIDDER
Human Safety Department, The Procter & Gamble Company, Cincinnati, Ohio
Received October 30, 1995; accepted July 31, 1996
The murine local lymph node assay (LLNA)1 is a prePhenotypic Analysis of Lymphocyte Subpopulations in Lymph dictive test that utilizes in vivo cell proliferation in the DLN
Nodes Draining the Ear Following Exposure to Contact Allergens
for assessment of the contact sensitization potential of mateand Irritants. SIKORSKI, E. E., GERBERICK, G. F., RYAN, C. A.,
rials (Kimber et al, 1986; Kimber and Weisenberger, 1989;
MILLER, C. M., AND RIDDER, G. M. (1996). Fundam. Appl. Toxicol.
Kimber, 1993). This assay measures the proliferative re34, 25-35.
sponse in the draining auricular lymph nodes during the
induction
phase of a contact sensitization response after epiThe murine local lymph node assay (LLNA) measures in vivo
proliferation in draining lymph nodes (DLN) following topical cutaneous exposure to materials. It is based on the observaexposure to chemicals to assess contact sensitization potential. tion that the induction phase of a contact sensitization reHowever, proliferation has also been observed with some irritants. sponse to allergen is characterized by lymphocyte proliferaTo further characterize events in the DLN during the LLNA and tion and hyperplasia in the lymph nodes draining the site of
distinguish allergens from irritants, phenotypic analysis of lym- topical exposure (Oort and Turk, 1965; Parrott and de Sousa,
phocyte subsets was made following topical exposure. In prelimi- 1966; de Sousa and Fachet, 1972; Asherson and Barnes,
nary studies, mice were treated on the ears for 3 consecutive days, 1973). This assay has been evaluated extensively for its
and 48 hr following the final application, analysis of CD3, CD4, reliability and sensitivity in predicting the contact sensitizaCD8, and B220 expression was evaluated by flow cytometry. The tion potential of materials through both internal and interlaballergens oxazolone (OXAZ) and picryl chloride (TNCB) and the oratory validation studies (Kimber et al., 1991, 1995; Basirritant benzalkonium chloride (BC) increased cell number com- ketter et al, 1991; Scholes et al., 1992; Kimber and Baspared to vehicle. The increase in lymph node cellulariry for these
ketter, 1992) and has been found to correlate well with
materials was due to an increase in the total number of T and B
lymphocytes. Interestingly, even though contact sensitization is a guinea pig studies (Kimber et al., 1991, 1990; Basketter et
cell-mediated immune response (Thl), mice exposed to the contact al., 1991; Basketter and Scholes, 1992) and analyses in man
allergens showed a preferential increase in B lymphocytes in the (Basketter et al., 1994). However, one limitation is the ability
DLN as seen by an increase in the percentage of B220+ cells. The of some skin irritants to produce proliferation in the DLN
percentage of B220+ cells was 13.1 and 36.1% for OXA and TNCB, as well (Montelius et al, 1994; Gerberick et al., 1992).
respectively, compared to percentages of 7.4 and 9.3% for irritant An increased understanding of events in the DLN during
and vehicle, respectively. With some allergens, a concomitant de- induction of contact sensitization may lead to the developcrease in the percentage of CD3+ cells was seen. Time course ment of additional endpoints for the LLNA which improve
studies demonstrated the increase in the percentage of B220+ cells the predictability, selectivity, and sensitivity of the assay.
was seen in allergen treated mice by 24 hr after the final applica- To gain further insight into events occurring in the DLN
tion of material, plateaued by 48 hr, and was still elevated by 96 during the induction phase of the LLNA, flow cytometry was
hr. In allergen-treated mice, percentages of B220+ cells increased used to enumerate changes in the lymphocyte populations
dose dependency. Further studies were performed to evaluate ad- following exposure to various irritants and allergens.
ditional contact allergens and irritants and determine if evaluation
of flow cytometric parameters could potentially identify contact
METHODS
allergens and differentiate them from irritants. Analysis of data
Animals.
Female BALB/c mice (Charles River, Wilmington, MA, or
from these studies, which examined a total offivecontact allergens Harlan Industries, Indianapolis, IN), or female CBA/J mice (Harlan), 6 to
and six irritants, showed that the modifications to the LLNA improved the identification of irritants and allergens in individual
1
Abbreviations used: BC, benzalkonium chloride; BethC, benzethonium
experiments by including both phenotypic analysis of the DLN
chloride; DLN, draining lymph nodes; DNCB, dinitrochlorobenzene; EUG,
and cell number per node as endpoints rather than either endpoint eugenol; HCA, a-hexyl cinnamic aldehyde; LLNA, local lymph node assay;
alone.
© 1996 Sodety of Toxicology.
MS, methyl salicylate; NONA, nonanoic acid; OXA, oxazolone; SALA,
salicylic acid; TNCB, picryl chloride.
25
0272-0390/96 SI8.00
Copyright O 1996 by the Society of Toxicology
All rights of reproduction in any form reserved.
26
SIKORSKI ET AL.
8 weeks old, weighing approximately 17 to 23 g were used. Animals were
housed, fed, and handled in compliance with standards set forth by the U.S.
Animal Welfare Act or recommendations in the National Institutes of Health
"Guide for the Care and Use of Laboratory Animals." All procedures
performed on animals were reviewed and approved by a veterinarian and
our Institutional Animal Care and Use Committee.
Chemicals. Contact allergens used were l-chloro-2,4,6-trinitrobenzene
(TNCB; Polysciences Inc., Warrington, PA), l-chJoro-2,4-dinitrobenzene
(DNCB (99%); Sigma Chemical Co., St. Louis, MO), oxazolone (OXAZ
(99%), Aldrich Chemical Co., Inc., Milwaukee, WI), eugenol (EUG (99%),
Aldrich) and a-hexyl cinnamic aldehyde (HCA (85%), Aldrich). Skin irritants used were sodium lauryl sulfate (SLS (99%), Sigma), salicylic acid
(SALA (99%), Aldrich), benzalkonium chloride (BC, Sigma), benzethonium chloride (BethC (>99%), Sigma), and nonanoic acid (NONA (97%),
Sigma). Vehicles used were acetone (J. T. Baker, Inc., Phillipsburg, NJ),
30% ethanol (Aarper Alcohol and Chemical Co., Shelbyville, KY), and 1propanol (J. T. Baker). Allergens and irritants were used at concentrations
producing an increase in the number of cells per node that was greater than
1.5-fold over vehicle.
In vivo treatment.
Mice were treated with allergen, irritant, or vehicle
once a day for 3 consecutive days unless otherwise indicated by applying
12.5 £tl of test material or vehicle topically to each side of both ears (25
/xl total/ear). With the exception of the studies which examined the response
kinetics, the draining auricular lymph nodes were removed 48 hr following
final application and pooled for each treatment group.
Single cell suspension preparation and staining. For each treatment
group, the draining auricular lymph nodes from the individual mice (n =
3) were pooled to obtain adequate cell numbers and single cell suspensions
were prepared by gently mashing the nodes on nylon mesh filter (100-/im
pore size; The Spectra Company, Los Angeles, CA). The cell suspensions
were washed once with PBS and resuspended in FACS buffer (10 mM Hepes
buffer (Sigma), 0.01% sodium azide (Sigma), and 1% heat-inactivated fetal
bovine serum (Gibco BRL Life Technologies, Grand Island, NY)) at 107
cells/ml. Cell numbers and viability were determined by counting with a
hemacytometer and trypan blue exclusion. Cells were labeled in 96-well,
round-bottom plates (106 cells/well) with optimum concentrations of fluorochrome-conjugated monoclonal antibodies and isotype controls. Antibodies used were phycoerythnn (PE)-conjugated hamster anti-CD3 (Pharmingen, San Diego, CA), fluoroscein isothiocyanate (FITC)-conjugated rat
anti-CD45R (Ly-5/B220) (Pharmingen), FTTC-conjugated goat anti-IgG and
IgM (Tago, Biosource International, Camarillo, CA), hiTC-conjugated rat
anti-CEM (Gibco), and PE-conjugated rat anti-CD8a(Ly-2) (Pharmingen)
The isotype control antibodies were anti-keyhole limpet hemocyanin hamster IgG PE-conjugated and rat I g G ^ FITC-conjugated. Briefly, cells were
incubated widi antibody on ice for 30 min, washed twice, and resuspended
in FACS buffer at I0 6 cells/ml for analysis.
criminate allergens from irritants. Several endpoints of the LLNA, namely,
cell number per node and the percentages of CD3, CD4, CD8, B220, and
IgG and IgM labeled lymphocytes, were evaluated and were normalized to
their vehicle controls to correct for day-to-day variability. To be able to
include data from both BALB/c and CBA/J mice, it was necessary to
compare the LLNA endpoints using the same concentrations of materials.
For this purpose, each of the endpoints from BALB/c mice was regressed
against the CBA/J endpoints for the same materials and concentrations to
correlate responses between the two strains. Because the correlations were
high, data from both mouse strains were included; however, measurements
for BALB/c mice were adjusted to compensate for interstrain differences
because BALB/c mice generally had less proliferative response to allergens
and irritants than CBA/J. These corrections of the raw data for day-to-day
variability and interstrain variations were not essential in obtaining the final
results but did improve reproducibility.
Data for each surface marker were inspected for distribution, correlation
with other surface markers, and ability to discriminate irritants from allergens alone or in combination with other surface markers Of the parameters
evaluated, the final endpoints chosen for statistical analysis of the LLNA
were cell number per node (as a measure of proliferation) and the ratio of
the % B220+/% CD3 + cells (designated B/T cell ratio throughout this report)
obtained by flow cytometric analysis. Based on these endpoints, nominal
logistic regression (SAS Institute, 1994) was used for classifying the materials tested as contact allergens or irritants. This is a standard statistical
method used when assessing a binary response outcome or classifying a
response
RESULTS
Cell Number and Flow Cytometric Analysis of
Lymphocytes in Draining Lymph Nodes Following
Epicutaneous Exposure to Contact Allergens and
Irritants
Flow cytometry and analysis. Cells were analyzed using a Coulter Elite
ESP flow cytometer and Elite Data Analysis software (Coulter Electronics,
Hialeah, FL). Fluorescence intensities were calibrated and compensated by
running Calibrite calibration beads (Becton Dickinson, San Jose, CA) and
adjusting gains and high voltages to meet target values. Nonviable cells
were excluded from analysis using a forward angle light scatter (FALS)
versus 90° light scatter (LS) gate established by backgating on cells which
excluded propidium iodide (Calbiochem, San Diego, CA). The FALS discriminator was adjusted to eliminate debris. Lymphocyte suspensions were
run to establish a lymphocyte gate in the FALS vs 90°LS histogram. Analysis thresholds were based on isotype FTTC and PE control samples At least
15,000 viable cells were analyzed per sample. Dual-parameter histograms
were analyzed with quad stats to determine percentages of fluorescently
labeled cells.
To characterize events in the DLN during the LLNA,
phenotypic analysis of lymphocyte subpopulations (CD3 + ,
B220 + , IgG and IgM + , CD4 + , and CD8 + ) in the auricular
lymph nodes following ear application of contact allergens
and irritants was made 48 hr after final application of test
materials. Cell number per node was determined and used
as a replacement for [3H]thymidine incorporation normally
used in the LLNA to evaluate proliferation. In the LLNA
developed by Kimber and Weisenberger (1989), the endpoint
of the LLNA is proliferation measured by [3H]thymidLne
uptake and the criteria for a positive response is a threefold
increase in proliferation over vehicle control. In studies presented here, we are using cell number per node as an indicator of proliferation. This endpoint is an appropriate indicator
for these studies as previous results (Gerberick et al., 1992)
have shown that when changes in cells/node are seen then
changes in [3H]thymidine incorporation are seen as well. A
representative experiment is shown in Table 1 following
application of the allergens oxazolone (OXAZ) and picryl
chloride (TNCB) and the irritant benzalkonium chloride
(BC) in acetone.
Statistical analysis. Data were analyzed (SAS Institute, 1994) to determine if phenorypic changes observed in draining lymph nodes could dis-
The cellular composition of the auricular lymph nodes in
naive mice was 82.7% CD3 + cells (T lymphocytes) and
27
PHENOTYPIC ANALYSIS OF LYMPHOCYTE SUBPOPULATIONS
TABLE 1
Phenotypic Analysis of Lymphocytes in the Draining Lymph Nodes
Percentage positive
Test material
Cells/node"
B220
IgG, M
CD3
CD4
CD8
B/T cell ratio
Acetone
BC (2%)
OXA (0.005%)
TNCB (0.5%)
1.0
4.5
4.7
10.8
9.3
7.4
13.1
36.1
12.4
6.8
14.5
35.2
78.1
89.4
78 9
52 2
58.4
61.6
53.6
34.4
19.7
18.3
18.7
9.1
0.12
0.08
0.17
0.69
Note. Groups of three mice were treated on the ear once per day for 3 consecutive days. Forty-eight hours following the final application, the two
auricular lymph nodes from each mouse within a group were obtained and pooled. Determination of cell number per node and flow cytometric analysis
of the lymphocyte populations was made. This is a representative experiment in BALB/c mice.
° Cells/node X 106.
9.5% B220 + cells (B lymphocytes) (data not shown). Nodes
from acetone-treated mice consisted of 78% CD3 + cells and
9.3% B220 + cells (Table 1). All three test materials produced
an increase in cell number per node over the corresponding
vehicle control at the concentrations tested. For both allergens and irritants, the increase in cell number per node was
reflected as an increase in the total number of CD3 + , B220 + ,
IgG and IgM + , CD4 + and CD8 + cells. However, evaluation
of the percentage of the various subpopulations indicated
that allergen exposure altered the percentages of certain subpopulations to a greater extent than irritants. Most striking
was the preferential increase in the number of B lymphocytes
(as indicated by the increase in the percentage of B220 +
cells) consistently seen in the allergen-treated mice over vehicle control. The percentage of B220 + cells was 36.1 and
13.1% in 0.5% TNCB- and 0.005% OXAZ-treated mice,
respectively, compared to 7.4 and 9.3% in vehicle- and BCtreated mice. Analysis of IgG&M+ cells correlated well with
B220 + cells as the percentage of IgG and IgM + cells increased in mice treated with allergen when compared to
vehicle or irritant (Table 1). For the irritant benzalkonium
chloride, a slight decrease in the percentage of B220 + and
IgG and IgM + cells was seen. When analyzing the T cell
subsets, the total number of CD3 + , CD4 + , and CD8 + cells
increased in both the irritant- and the allergen-treated mice;
however, the percentages generally dropped in the allergentreated mice because of the concomitant increase in B lymphocytes. This decrease was more pronounced in the TNCBtreated mice where a greater increase in the number of B
lymphocytes was observed. The ratio of the % B220 + /%
CD3 + cells (B/T cell ratio) was increased in the allergentreated mice (0.17 and 0.69 in the OXAZ- and TNCB-treated
mice, respectively) compared to vehicle- and BC-treated
mice (0.12 and 0.08, respectively). Analysis of cells dual
labeled with anti-B220 and anti-CD3 antibodies showed the
percentage of the B220 + /CD3 + cells ranged between 1.2 and
1.5%, indicating that anti-B220+ antibody was not significantly binding to CD3 + cells (data not shown).
Dose-Dependent Effects of Contact Allergens on the
Phenotype of Lymphocytes in the Draining Lymph
Nodes
Experiments were performed to determine if changes in
percentage of B220 + cells in the allergen-treated mice were
dose-related. For the contact allergens TNCB and DNCB,
dose-related increases in the cell number per node and the
percentage B220 + cells were seen. For TNCB, the increase
in the percentage of B220 + cells appeared to peak at 0.1%
with a value of 39.5% (Fig. 1 A), while cell number per node
plateaued at 0.1%. For DNCB-treated mice, the percentage
of B220 + cells and the cell number per node were still increasing at a concentration of 0.25% with the % B220 + cells
reaching 21.6% (Fig. IB) at that concentration. For these
allergens, a concomitant dose-related decrease in the percentage of CD3 + cells was also noted (data not shown). For
the irritants, benzalkonium chloride and nonanoic acid, a
dose-dependent increase in cell number per node was seen
(Fig. 2). However, changes in the percentage of B220 + cells
were somewhat variable. As seen in earlier studies, BC produced a decrease in % B220 + cells which appeared to be
dose related (Fig. 2A). On the other hand, the irritant nonanoic acid produced slight increases in the % B220 + cells over
vehicle at all doses which were not related to dose (Fig. 2B).
Therefore, even at high concentrations, the cell number per
node and the % B220 + cells seen with irritants never approached the values seen with allergens.
Kinetics of Allergen and Irritant Treatment on B
Lymphocyte Percentages in the Draining Lymph Nodes
Kinetic studies were performed to assess alterations in the
percentage of B220 + cells at various times following the
28
S1K0RSKJ ET AL.
third application of irritants and allergens. For the allergen
DNCB (0.25%), cell number per node was elevated at 24
hr, peaked at 72 hr, and remained elevated at 96 hr (Fig. 3).
The percentage of B220 + cells was increased over vehicle
at 24 hr but reached a plateau at 48 hr and remained elevated
up to 96 hr. The time course for the irritant BC (2%) indicates
that cell number per node plateaued at 48 hr and remained
elevated at 96 hr (Fig. 4). The percentage of B220 + cells in
BC-treated mice was generally less than that of vehicle
treated mice at 24 and 48 hr with slight increases at 72
and 96 hr that approached values seen with the vehicle. To
determine if changes in the percentage of B22O+ cells oc-
45
40
35
30
25
20
15
10
5
0
0.0%
0.1%
0.5%
1.0%
2.0%
Benzalkonium Chloride in Acetone (w/v)
B
45
40 -3
35
30
25
20
15
10
0%
0.01% 0.05%
0.1%
0.5%
5
0
TNCB in Acetone (w/v)
20%
40%
60%
80%
Nonanoic Add in 1 -Propanol (v/v)
B
FIG. 2. Dose-dependent effects of the irritants benzalkonium chloride
(A) and nonanoic acid (B) on the cell number per node X 106 (open bars)
and the percentage of B220+ cells (hatched bars) in the draining lymph
nodes. Mice (n = 3) were treated on the ear once per day for 3 consecutive
days with various concentrations of allergens. Forty-eight hours following
the final application, two auricular lymph nodes were removed from each
mouse and pooled within each group (total of six nodes per group).
45
40
35
30
25
20
15
10
5
0
0%
0.025%
0.05%
0.10%
0.25%
DNCB in Acetone (w/v)
FIG. 1. Dose-dependent effects of the contact allergens TNCB (A) and
DNCB (B) on the cell number per node X 10* (open bars) and the percentage
of B220 + cells (hatched bars) in the draining lymph nodes. Mice (n = 3)
were treated on the ear once per day for 3 consecutive days with various
concentrations of allergens. Forty-eight hours following the final application, two auricular lymph nodes were removed from each mouse and pooled
within each group (total of six nodes per group).
curred prior to the final application of test materials, the DLN
composition was evaluated at various timepoints following a
single dose of TNCB. Increases in the % B220 + cells began
at 72 hr following a single dose (data not shown). Together
these data indicate that 48 hr following final application is
an appropriate time to evaluate changes in the lymphocyte
populations in the DLN.
Comparison of Phenotypes in CBA/J and BALB/c Mice
Following Application of Allergens and Irritants
A previous study by Kimber and Weisenberg (1989) demonstrated that CBA/J mice used in the LLNA exhibit a
PHENOTYPIC ANALYSIS OF LYMPHOCYTE SUBPOPULATIONS
24
48
72
96
Time Post Treatment (hours)
29
irritants was r = 0.93 (Fig. 5A). Interestingly, both strains
of mice exhibited similar alterations in the percentages of
B220 + and CD3 + populations in that contact allergens produced an increase in the B/T cell ratio over vehicle. The
correlation coefficient of this ratio between the two mouse
strains was 0.91 (Fig. 5B). Based on these correlations, data
from both strains of mice were incorporated in subsequent
model development. However, BALB/c data were corrected
for interstrain variation using the following equations of the
format y = mx + b, where m is the slope of the line and b
is the y intercept. For cell number per node, y = 1.36(cell
number per node from BALB/c) + 0.13. For the B/T ratio,
y = 1.08(B/T cell ratio from BALB/c mice) + 0.22.
18
16
iO
1
* 6
24
24
48
72
96
Time Post Treatment (hours)
48
72
96
Time Post Treatment (hours)
FIG. 3. Kinetics of cell number per node X 10* (A) and the percentage
of B220+ cells (B) in vehicle- (open bars) and DNCB-treated (hatched bars)
mice. Mice (n = 3) were treated on the ear once per day for 3 consecutive
days with various concentrations of allergens. At various times following
the final application, auricular lymph nodes were removed and pooled within
each group for flow cytometric analysis.
greater proliferative response to allergens in the draining
lymph nodes than BALB/c, C57BL710J, and Alpk/AP mice.
Initial studies presented here which demonstrated alterations
in cellular phenotypes in DLN were done in BALB/c mice.
To determine if data from studies in CBA mice could be
combined into the same dataset for analysis, each of the
endpoints from BALB/c mice was regressed against the
CBA/J endpoints for the same materials and concentrations
to correlate responses between the two strains. For contact
allergens tested at the same concentration, a greater increase
in the cell number per node was seen in CBA/J mice over
BALB/c mice, which is consistent with previously reported
data. The correlation coefficient of cell number per node
between the two mouse strains for two sensitizers and three
24
48
72
96
Time Post Treatment (hours)
FIG. 4. Kinetics of cell number per node X 10* (A) and the percentage
of B220 + cells (B) in vehicle- (open bars) and BC-treated (hatched bars)
mice. Mice (n = 3) were treated on the ear once per day for 3 consecutive
days with various concentrations of allergens. At various times following
the final application, auricular lymph nodes were removed and pooled within
each group for flow cytometnc analysis.
30
SIKORSKI ET AL.
12
0
2
4
6
8
Cells per Node
10
Baib/c
FIG. 5. Comparison of cell number per node and the B/T cell ratio in CBA/J and BALB/c mice. Three irritants (diamonds) and two sensitizers
(squares) were evaluated in the LLNA in both strains at the same concentrations and under the same exposure conditions. The correlation coefficients
for the two strains were calculated for each endpoint.
Statistical Analysis
To date, a series of five contact allergens (eugenol, ahexyl cinnamic aldehyde, DNCB, TNCB, and oxazolone)
and six irritants (SLS, benzalkonium chloride, benzethonium
chloride, nonanoic acid, methyl salicylate, and salicylic acid)
have been evaluated in CBA/J and BALB/c mice for effects
on the cellular composition of DLN in the LLNA. Two
parameters from these LLNA studies, namely, the ratio of
% B220 + /% CD3 + cells (B/T cell ratio) and cell number per
node, were analyzed for each test material and corrected for
day to day variability and interstrain variation. In all cases,
both parameters were corrected with the corresponding vehicle control. The data collectively shown in Figs. 6 and 7 are
from the highest concentration(s) evaluated for a particular
chemical. It is important to note that there are multiple points
for the same test material because each material was evaluated in two or more experiments (CBA/J and/or BALB/c)
at the concentration(s) listed in Table 2. Nominal logistic
regression was used to statistically analyze these data and
the mathematical equation which included data from both
cells/node and the B/T ratio is shown below. This is a standard statistical method used for predicting binary response
outcomes.
Estimated response = 9.23 - 0.745 X
(cells/node),,
(cells/node)vehldc
- 3.10 X
(B/T ratio)**
(BAT ratio) vchkk
The probability that a test material in our dataset would be
classified as a contact allergen was calculated by the equation
given below and the results are illustrated in Fig. 7 for the
irritants and allergens.
Probability test material is a sensitizer
1
1 + e
(estinuled response)
In Fig. 7, the three panels show results of the statistical
method comparing the endpoints (a) cells/node alone, (b) B/
T cell ratio alone, or (c) both endpoints to classify the various
test materials. In each of the three panels, known contact
allergens are listed on the right side and irritants are listed
on the left. Using 0.5 (50% probability) as a cutoff (line in
Fig. 7), data points above the line represents compounds that
would be classified as a contact allergen and those below
the line represent compounds that would be classified as
irritants based on the particular endpoint(s) evaluated. An
irritant that is misclassified as an allergen is a material on
the left side of the panel that falls above the line and an
allergen that is misclassified is a material on the right side
of the panel that falls below the line. On this basis, the
LLNA would have misclassified test materials 9 times using
cells/node data alone (Fig. 7a), 5 times using the B/T cell
ratio alone (Fig. 7b), but only 3 times using both endpoints
(Fig. 7c). This is illustrated in Fig. 7c, where the only allergen misclassified as an irritant was eugenol, a weak human
allergen. In Fig. 7a, three experiments with BC and one
with NONA would have classified these irritants as allergens
while four experiments with known allergens (HCA, OXAZ
31
PHENOTYPIC ANALYSIS OF LYMPHOCYTE SUBPOPULATIONS
•
6.0-
TNCB
•TNCB
•
TNCB
•TNCB
5.0 -
4.0 -
• DNCB
• DNCB
•TNCB
•DNCB
3.0 -
• DNCB
•QXAZ
2.0 -
0
•^ J W 0 *
s
•DNCBCQ<AZ
Vehicle SLjK^NONA
Controls
1.0 -
©
^TTBokOrfc
O%M-AOBelhC
OMS
OBC
Oec
lhC
Oec EUG
0.0 -
0.0
1
40
8.0
12.0
1
16 0
Cells/Node (Treated)
Cells/Node (Vehicle Control)
FIG. 6. Plot of the B/T ratio vs cell number per node X 106 for known allergens (squares) and irritants (diamonds) tested in the LLNA. These data
are from studies using three consecutive exposures and flow cytometric analysis at 48 hr following the final application of material. Each endpoint is
normalized with its corresponding vehicle control to correct for day-to-day variations. Data from both CBA/J and BALB/c mice are included with the
data from BALB/c mice corrected for interstrain variability. The mean values for the vehicles are shown for comparison purposes.
(0.005%), DNCB, and EUG) would have misclassified them
as irritants. In contrast, Fig. 7c illustrates that use of both
endpoints has only one experiment misclassifying NONA as
an allergen and two experiments misclassifying eugenol as
an irritant. Other advantages of including both endpoints in
the model were the (1) increased reproducibility and (2)
decreased uncertainty of the classifications (seen as an increase in the distance of data points from the 50% probability
line). It is not possible to rank allergens using Fig. 7 because
they were tested at various concentrations and, therefore, not
at concentrations known to produce a maximal response.
DISCUSSION
The murine local lymph node assay (LLNA) measures the
proliferative response of lymphocytes in the draining lymph
nodes (DLN) during the induction of contact sensitization
by epicutaneous exposure to allergen. In studies presented
here, the response in the DLN was further characterized by
flow cytometric analysis of T and B lymphocyte subsets.
There exist few reports on the phenotypic characterization of
the DLN during the induction phase of a contact sensitization
response. A study by Morita et al. (1987) demonstrated that
T cells were preferentially accumulating in lymph nodes
draining the abdomen because L3T4 cells in BALB/c mice
increased from 51.7 to 79.3% on the 5th day after a single
application of DNFB in acetone:olive oil. This is in contrast
to studies presented here, though we did not evaluate DNFB
or abdominal lymph nodes. Two additional reports assessed
changes in T lymphocytes following application of contact
allergens that were, in general, consistent with our results.
In a study assessing cellular composition of abdominal DLN,
Gautam et al. (1991) observed an increase in cellularity of
the DLN in mice dosed with TNCB. They found the total
number of T cells increased although the percentage of T
lymphocyte subsets (Thy 1.2+, CD4 + , and CD8 + ) did not
change. In another study by deSilvaef al. (1993), an increase
in the total number of T lymphocytes in mice with the allergen DNCB was seen which was greater than that seen with
the irritant SLS. Only two reports to date contained informa-
32
SIKORSKJ ET AL
Classification of Irritants and Sensitizers by Cellular
Makeup of Draining Lymph Nodes
1.0-
Cells/Node
B/T
Cells/Node 4
B/T
0.9-
-0.9
0.8-
-0.8
<D
3
1 wil
a Sensiitize
-1.0
Eua
0.7-
-0.7
DNCB
OXAZ
0.6-
B
0.5
DOXAZ
OXAZ
DNCB
-0.6
hat
CO
10
CO
0.5
Prob abilit
llassi
0.4-
-0.4
8
0.3-
B
0.2-
NONA
SLS
-0.3
DNCB
RUQ
-0.2
SS"
0.10.0
Irritant
Sensitizer
Irritant
Sensitizer
Irritant
-0.1
Sensrtizer
0.0
Known Class
FIG. 7. Probability of classification of materials as allergens or irritants. A nominal logistic maximum model was developed using LLNA data with
known allergens (squares) and irritants (diamonds). The probability that a material would be classified as a sensitizer based on this model is plotted. The
three panels show results of the model using (a) the cell number per node alone, (b) the B/T cell ratio, and (c) both to classify the test materials. The
line represents the 50% probability of being classified a contact allergen. Materials falling above the line would be classified as allergens and those
below the line as irritants. Numbers in parentheses indicate the number of experiments plotted for the indicated chemical.
tion about changes in B lymphocyte populations during the
induction phase of contact sensitization in the draining
lymph nodes and both are assessing oxazolone. In a report
consistent with our observations, Kraal and Twisk (1984)
reported B cells were retained specifically in the stimulated
nodes when they compared unstimulated peripheral LN cells
with oxazolone-primed cells. A recent study by Kuhn et al.
(1995) demonstrated the phenotype of cells in lymph nodes
following oxazolone exposure (1%) in the LLNA (4-day
exposure) to consist of 37.8% CD4 + cells, 16.6% CD8 +
cells, and 44.7% B220 + cells. These two studies on oxazolone support observations reported here that B cells accumulate in DLN following topical application of contact allergens.
In studies presented here, total cell number in the draining
auricular lymph nodes was shown to increase following topical application of both irritants and allergens. This increase
in lymph node cellularity is due, at least in part, to proliferation of lymphocytes and is reflected by an increase in the
number of both T and B lymphocytes. It is not known if the
increase in T and B lymphocytes in the DLN during induction is due strictly to proliferation or if changes in lympho-
cyte trafficking are involved as well. A recent study by Kuhn
et al. (1995) has shown that oxazolone in the LLNA (4-day
exposure) predominantly increased proliferation of CD4 +
cells, though an increase was also seen in proliferation of
B220 + cells.
It was not surprising to see an increase in total numbers
of T cells (both CD4 + and CD8 + ) as delayed type hypersensitivity reactions are known to be T-cell-mediated immune
responses. More specifically, these reactions are mediated
by TH1 cells (Cher and Mosmann, 1987; Fong and Mosman,
1989; Diamanstein et al., 1988) and can have either the CD4
or CD8 phenotype (van Loveren et al., 1984; Cocinski and
Tigelaar, 1990).
However, an interesting observation from the studies presented here was an increase in B cells in the draining lymph
nodes with allergens in a dose-related manner in both B ALB/
c and CBA/J mice. This increase was consistently seen with
the five allergens evaluated but the magnitude of the increase
seemed to vary with the strength and concentration of the
sensitizer, with the stronger sensitizers and/or higher concentrations producing a greater increase in the percentage of
B220 + cells. When tested at optimal concentrations giving
PHENOTYPIC ANALYSIS OF LYMPHOCYTE SUBPOPULATIONS
TABLE 2
Chemicals and Concentrations Illustrated in Figs. 6 and 7
Chemical
Class
Benzalkonium chloride
Benzethonium chloride
Imtant
Irritant
Methyl salicylate
Nonanoic acid
Salicyclic acid
Sodium lauryl sulfate
Irritant
Irritant
Irritant
Irritant
Eugenol
Hexylcinnamic acid
Oxazolone
Allergen
Allergen
Allergen
TNCB
DNCB
Allergen
Allergen
Concentration (%)
2
One experiment each
with 2 and 4
20
60
40
One experiment each
with 20 and 40
25
50
One experiment with
0.005 and two
experiments with
0.01
0.50
0.25
Number of
experiments
5
2
1
4
2
2
2
2
3
5
6
Note. Groups of three mice were treated on the ear once per day for 3
consecutive days with the indicated materials and concentrations. Fortyeight hours following the final application, the two auricular lymph nodes
from each mouse within a group were obtained and pooled. Determination
of cell number per node and flow cytometric analysis of the lymphocyte
subpopulations was made and is presented in Figs. 6 and 7.
the highest responses, the strong allergens DNCB, TNCB,
and OXAZ produced greater increases in the percentage of
B220 + cells than the two moderate to weak allergens, HCA
and eugenol. The increase in the percentage of B220 + cells
likely represents an increase in the percentage of B lymphocytes since a second marker for B lymphocytes (IgG and
IgM + ) also showed the same trend. The increase in percentage of B220 + cells is not due to an increase in the population
of cells that coexpress B220 and CD3 because dual labeling
studies indicated that the population staining with both antibodies was generally between 0 and 2% (data not shown).
The increase in the % B220 + cells seems to be a consistent
and robust endpoint in the DLN following application of
allergen as it is seen in two mouse strains and via two different routes of allergen exposure, namely, the ear and buccal
mucosa (unpublished data). Regarding the local lymph node
assay, these observations suggest that proliferation measured
may not be due just to T cells but also to B cells. This
work lays the foundation for further studies investigating
activation and memory cell markers as well as the proliferative responses of the various lymphocyte subsets.
Speculation regarding the preferential accumulation of B
lymphocytes in the DLN with allergens suggests two possible explanations. First, this response might be attributed to
development of an antibody response to the hapten. Several
investigators have shown that epicutaneous application of
33
DNFB, TNCB, and oxazolone can elicit anti-hapten antibodies (Taylor and Iverson, 1971; Thomas etal., 1976; Askenase
and Hayden, 1974; Takahashi et al, 1978; Dearman and
Kimber, 1991, 1992). Still other investigators suggest that
anti-idiotypic antibodies may arise in animals as a consequence of contact sensitization and serve to regulate this
immune response (Sy et al, 1979). Second, the B cell response may be the outcome of T cell activation. It is known
that cytokines from both Thl and Th2 cells can stimulate B
lymphocytes. Data by Neuman et al (1992) indicate such a
B cell response may result from T cell activation. They
demonstrated a dose-dependent increase in the percentage
of B220 + cells in the popliteal lymph nodes after in vivo
anti-CD3 (method to stimulate T cells) treatment as well as
a decrease in the percentage of Thy 1.2, CD4, and CD8
positive cells. This increase seemed to reflect an influx of
B lymphocytes rather than increase in proliferation because
few of the B220+ cells were cycling in the S/G2M phase of
the cell cycle.
Interestingly, preferential accumulation of B lymphocytes
in draining lymph nodes has been seen in other T-cell-mediated immune responses. Fojtasek et al. (1993) studied the
CMI response (which is the principal host defense) to Histoplasma infection in the DLN of B6C3F1 mice after respiratory exposure. Seven days after infection, an increase in the
number of B lymphocytes caused the percentage to rise to
43% compared to 26% in control mice, and it remained
elevated throughout the course of infection. Constant and
Wilson (1992) also saw a proportionally greater increase in B
cells compared to T cells in draining lymph nodes following
immunization with the attenuated larvae of Schistosoma
mansoni for which the principal host defense is T-cell-mediated (Thl) (James et al, 1986; Aitken et al, 1988; Wynn
et al, 1994). This is in contrast to the Th2 response seen
with the eggs of S. mansoni (Chensue et al, 1994; Chensue
et al, 1994). To understand this unexpected observation,
they evaluated the proliferative response of the T and B
lymphocytes in the DLN. A striking feature of their results
was the preferential increase in the proportion of proliferating T, relative to B, cells. This supported their belief that
antigen-driven proliferation of T cells was important in this
immune response and also indicated that the significant increase in B cell numbers could not be accounted for by a
preferential increase in proliferation. Evaluation of lymphocyte migration patterns suggested that the observed accumulation of B cells in the DLN might be due to enhanced
recruitment as well as a failure of these cells to exit the
node. Similarly, Lynch et al. (1989) have shown that in
response to acute infection with choriomeningitis virus, massive recruitment of nondividing cells occurs from 3 days
after infection. The T:B cell ratio in that study was decreased, indicating preferential accumulation of B lymphocytes.
34
SIKORSKI ET AL.
Due to the observations showing changes in lymphocyte
subpopulations, several contact allergens and irritants were
evaluated to see if use of flow cytometric analysis as an
endpoint in the LLNA could enhance differentiation of allergens and irritants. A compilation of the data shown in Fig.
6 shows there is a clear correlation between cells/node and
the B/T cell ratio for the allergens tested, so, for any given
increase in proliferation, allergens will produce a greater
increase in the B/T cell ratio than will irritants. The data
obtained on five allergens and six irritants were analyzed by
a standard statistical method for binary response outcomes
known as nominal logistic regression (Fig. 7). Using this
statistical method, the predictive power of the LLNA to
classify a material as an irritant or an allergen was improved
when analysis included both cell number per node and B/T
cell ratio rather than either endpoint alone. Use of both endpoints also increased the reproducibility and accuracy of the
LLNA. This is illustrated most clearly by TNCB and DNCB
in Fig. 7c. In Fig. 7c, five experiments with TNCB and six
experiments with DNCB produced results that cluster tightly
at the top of the figure. In contrast, analysis of cells/node
alone shows scatter of the results of these 11 experiments
on the right side of Fig. 7a. Use of the two endpoints also
decreases the uncertainty of classification of materials. This
is illustrated by the increase in distance of the data points
from the 50% probability line. In Fig. 7a, several of the data
points are much closer to the probability line than those in
Fig. 7c. The farther away a data point is from this line,
the greater the certainty that the material will be correctly
classified as an allergen or irritant. It should be noted that
classification of these materials using the criteria of a threefold increase in cell number over vehicle control to identify
an allergen would misclassify 10 of the 16 irritants as allergens but would not have misclassified any of the allergens.
Therefore, statistical evaluation of the data using nominal
logistic regression in combination with additional endpoints
improves the ability of this assay to classify allergens and
irritants at the concentrations evaluated in these studies.
Further development of this model may be useful by some
laboratories where use of radioactive materials is a problem.
In addition, such a modification may be important for a
specific class of chemicals such as the quaternary ammonium
compounds. It was interesting to note that both quaternary
ammonium compounds (BC and BethC) are irritants and
actually produced a decrease in the B/T cell ratio compared
to increases seen with allergens. In the past, BC has been
misclassified in the LLNA as a contact allergen (Gerberick
el al., 1992), but it is clearly separated from the allergens
in these studies by inclusion of the flow cytometry data.
In studies presented here, characterization of phenotypic
changes in the DLN during the induction phase of a contact
sensitization response have been performed for several con-
tact allergens and irritants. These studies contribute information regarding cellular events during the induction phase
of a contact sensitization response. These changes in the
percentages of T and B lymphocytes seen in the DLN have
been evaluated as additional endpoints in the LLNA which,
along with statistical analysis of the data, appear to improve
the predictability of the assay for the compounds evaluated
here. Further evaluation and validation would be required
using a number of other allergens and irritants to determine
the usefulness of this approach.
ACKNOWLEDGMENTS
We gratefully acknowledge review of the manuscript by Dr. Michael K.
Robinson and Dr. Stephen I. Katz. We also thank Dr. Greg Carr for assistance in statistical analysis of the data
REFERENCES
Aitken, R., Coulson, P. S., and Wilson, R. A. (1988). Pulmonary leukocytic
responses arc linked to the acquired immunity of mice vaccinated with
irradiated cercariae of Schisotsoma mansoni. J. Immunol. 140, 3573.
Asherson, G. L., AJlwood, G. G., and Mayhew, B. (1973). Contact sensitivity in the mouse. XI. Movement of T blasts in the draining lymph nodes
to sites of inflammation. Immunology 25, 485.
Asherson, G. L., and Barnes, R. M. R. (1973). Contact sensitivity in the
mouse. XII. The use of DNA synthesis in vitro to determine the anatomical location of immunological responsiveness to picryl chloride. Immunology 25, 495.
Askenase, P. W., and Hayden, B. J. (1974). Cytophilic antibodies in mice
contact-sensitized with oxazolone: Immunochemical characterization and
preferential binding to a trypsin-sensitive macrophage receptor. Immunology 27, 563.
Basketter, D. A., Scholes, E. W., Kimber, I., Botham, P. A., Hilton, J.,
Miller, K., Robbins, M. C , Harrison, P. T. C , and Waite, S. J. (1991).
Inter-laboratory evaluation of the local lymph node assay with 25 chemicals and comparison with guinea pig test data. Toxicol. Methods 1, 30.
Basketter, D A., and Scholes, E. W. (1992). Comparison of the local lymph
node assay with the guinea pig maximization test for the detection of a
range of contact allergens. Food Chem, Toxicol. 30, 63.
Basketter, D. A., Selbie, E., Scholes, E. W., Lees, D., Kimber, I., and
Botham, P. A. (1993). Results with OECD recommended positive control
sensitizers in the maximization, Buehler and local lymph node assays.
Food Chem. Toxicol. 31, 63.
Basketter, D. A., Scholes, E. W., and Kimber, I. (1994). The performance
of the local lymph node assay with chemicals identified as contact allergens in the human maximization test. Food Chem. Toxicol. 32, 543.
Chensue, S. W., Warmington, K. S., Ruth, J., Lincoln, P. M., and Kunkel,
S L. (1994). Cross-regulatory role of interferon-gamma (IFN-gamma),
IL-4 and IL-10 in schistosome egg granuloma formation: In vivo regulation of Th activity and inflammation. Clin. Exp. Immunol. 98, 395.
Chensue, S. W., Warmington, K., Ruth, J., Lincoln, P., Kuo, M. C , and
Kunkel, S. L. (1994). Cytokine responses during mycobacterial and schistosomeal antigen-induced pulmonary granuloma formation. Production
of Thl and Th2 cytokines and relative contribution of tumor necrosis
factor. Am. J. Pathol. 145, 1105.
Cher. D. J.. and Mosmann, T. R. (1987). Two types of murine helper T
PHENOTYPIC ANALYSIS OF LYMPHOCYTE SUBPOPULATIONS
cell clones. II. Delayed type hypersensitivity is mediated by TH1 clones.
J. Immunol. 138, 3688.
Cocinski, B L., and Tigelaar, R. E. (1990). Roles of CD4+ and CD8+ T
cells in murine contact sensitivity revealed by in vivo monoclonal antibody depletion. J. Immunol. 144, 4121.
Constant, S. L., and Wilson, R. A. (1992). In vivo lymphocyte responses
in the draining lymph nodes of mice exposed to Schistosoma mansoni:
Preferential proliferation of T cells is central to the induction of protective
immunity. Cell. Immunol. 139, 145.
Dearman, R. J., and Kimber, I. (1991) Differential stimulation of immune
function by respiratory and contact chemical allergens. Immunology 72,
563.
Dearman, R. J., and Kimber, I. (1992). Divergent immune responses to
respiratory and contact chemical allergens: Antibody elicited by phthalic
anhydride and oxazolone. Clm. Exp. Allergy 22, 241
De Silva, O., Perez, M. J., Pineau, N., Rougier, A., and Dossou, K. G.
(1993). Local lymph node assay: Study of the in vitro proliferation and
control of the specificity of the response by FACScan analysis. Toxic, in
Vitro 7, 299.
35
and Loveless, S. E. (1995). The local lymph node assay: An international
trial and comparison of modified procedures. Toxicologist 15, 100.
Kimber, I. (1993). The murine local lymph node assay: Principles and
practice. Am J. Contact Derm. 4, 42.
Kraal, G., and Twisk, A. (1984). The interaction of high endothelial venules
with T and B cells in peripheral lymph nodes after antigenic stimulation.
Eur. J. Immunol. 14, 586.
Kuhn, U., Lempertz, U., Knop, J., and Becker, D. (1995). A new method
for phenotyping proliferating cell nuclear antigen positive cells using
flow cytometry: Implications for analysis of the immune response in vivo.
J. Immunol. Methods 179, 215-222.
Lynch, F., Doherty, P. C , and Ceredig, R. (1989). Phenorypic and functional
analysis of the cellular response in regional lymphoid tissue during an
acute virus infection. J. Immunol. 142, 3592.
Montelius, J., Wahlkvist, H., Boman, A., Femstrom, P., Grabergs, L., and
Wahlberg, J. E. (1994). Experience with the murine local lymph node
assay: Inability to discriminate between allergens and irritants. Acta.
Derm. Venereol. 74, 22.
De Sousa, M , and Fachet, J. (1972). The cellular basis of the mechanism
of action of cortisone acetate on contact sensitivity to oxazolone in the
mouse. Clin. Exp. Immunol 10, 673.
Morita, H., Kuramoto, M., Oishi, M., Yu, M., Yamagata, M., and Sagami,
S. (1987). L3T4-positive cells in the lymph nodes during the induction
phase of contact hypersensitivity reaction of mice: Flow cytometry analysis. Arch. Dermatol. 279, 491
Diamanstein, T , Eckert, R., Volk, H-D., and Kupeir-Weglinski, J-W.
(1988) Reversal by interferon-g inhibition of delayed-type hypersensitivity induction by anti-CD4 or anti-interleukin 2 receptor (CD25) monoclonal antibodies. Evidence for physiological role of the CD4 + Thl subset
in mice. Eur. J. Immunol. 18, 2101.
Neumann, C. M., Oughton, J. A., and Kerkvliet, N. I. (1992). Anti-CD3induced T-cell activation in vivo. I. Flow cytometnc analysis of doseresponsive, time-dependent, and cyclosponn A-sensitive parameters of
CD4 + and CD8 + cells from the draining lymph nodes of C57B1/6 mice.
Int. J. Immunopharmacol. 14, 1295.
Fojtasek, M. F., Sherman, M. R., Garringer, T., Blair, R., Wheat, L. J., and
Schmzlein-Bictc, C. T. (1993). Local immunity in lung-associated lymph
nodes in a murine model of pulmonary histoplasmosis. Infect. Immun.
61, 4607.
Oort, J., and Turk, J. L. (1965). A histological and autoradiographic study
of lymph nodes dunng the development of contact sensitivity in the
guinea pig. Br. J. Exp. Pathol. 46, 147.
Parrott, D. M. V., and de Sousa, M. A. B. (1966). Changes in the thymus
dependent areas of the lymph nodes after immunological stimulation.
Nature 212, 1316.
SAS Institute (1994). JMP Users Guide, Version 3.0.2. SAS Institute, Cary,
NC.
Scholes, E. W., Basketter, D. A., Sarll, A. E., Kimber, I., Evans, C. D.,
Miller, K., Robbins, M. C , Harrison, P. T. C , and Waite, S. J (1992)
The local lymph node assay: Results of a final inter-laboratory validation
under field conditions. J. Appl. Toxicol. 12, 217.
Sy, M. S., Moorhead, J. W., and Claman, H. N. (1979). Regulation of cell
mediated immunity by antibodies: Possible role of anti-receptor antibodies in the regulation of contact sensitivity to DNFB in mice. J. Immunol.
123, 2593
Takahashi, C , Nishikawa, S., Katsura, Y., and Izumi, T. (1978). AntiDNP antibody response after the topical application of DNFB in mice.
Immunology 22, 589
Taylor, R. B., and Iverson, G. M. (1971). Hapten competition and the nature
of cell cooperation in the antibody response. Proc. R. Soc. London Biol.
176, 393.
Thomas, W. R., Asherson, G. L., and Watkins, M. C. (1976). Reaginic
antibody produced in mice with contact sensitivity. J. Exp. Med. 144,
1386.
van Loveren, H., Kato, K., Meade, D., et al. (1984). Characterization of
two different Lyl T cell populations that mediate delayed type hypersensitivity. J. Immunol. 133, 2402.
Wynn, T. A., Oswald, I. P., Eltoum, I. A., Caspar, P., Lowenstein, C. J.,
Lewis, F. A., James, S. L., and Sher, A. (1994). Elevated expression of
Thl cytokines and nitnc oxide synthase in the lungs of vaccinated mice
after challenge infection with Schistosoma mansoni. J. Immunol. 153,
5200.
Fong, T. A. T., and Mosmann, T. R. (1989). The role of IFN--y in delayedtype hypersensitivity is mediated by T h l clones. J. Immunol. 143, 2887.
Gautam, S. C , Matriano, J. A., Chikkala, N. F., Edinger, M. G., and Tubbs,
R. R. (1991). L3T4 (CD4 + ) cells that mediate contact sensitivity to tnnitrochlorobenzene express I-A determinants. Cell. Immunol. 135, 27.
Gerberick, G. F., House, R V., Fletcher, E. R., and Ryan, C. A. (1992).
Examination of the local lymph node assay for use in contact sensitization
risk assessment. Fundam. Appl. Toxicol. 19, 438.
James, S. L., Deblois, L. A., AJ-Zamel, F., Glaven, J., and Langhome, J.
(1986). Defective vaccine-induced resistance to Schistosoma mansoni in
P strain mice. III. Specificity of the associated defect in cell-mediated
immunity. J. Immunol. 137, 3959.
Kimber, I., Mitchell, J. A., and Griffin, A. C. (1986). Development of a
munne local lymph node assay for the determination of sensitizing potential. Food Chem. Toxicol. 6/7, 585.
Kimber, I., Hilton, J., and Botham, P. A. (1990). Identification of contact
allergens using the murine local lymph node assay: Comparisons with
the Buehler occluded patch test in guinea pigs. J. Appl. Toxicol. 10, 173.
Kimber, I., Hilton, J., and Botham, P. A. (1991). The munne local lymph
node assay: Results of an inter-laboratory trial. Toxicol. Lett. 55, 203.
Kimber, I., and Weisenberger, C. (1989). A murine local lymph node assay
for the identification of contact allergens. Assay development and the
results of an initial validation study. Arch. Toxicology 63, 274.
Kimber, I., and Basketter, D. A. (1992) The munne local lymph node
assay: A commentary on collaborative studies and new directions. Food
Chem. Toxicol. 30, 165.
Kimber, I., Hilton, J., Dearman, R. J., Gerberick, G. F., Ryan, C. A.,
Basketter, D. A., Scholes, E. W., House, R. V., Guy, A., Ladies, G. S.,