TOXICOLOGICAL SCIENCES 46, 3 0 8 - 3 1 6 (1998)
ARTICLE NO. TX982551
Neutrophil Activation by Polychlorinated Biphenyls:
Structure-Activity Relationship
Alan P. Brown,* Jesus Olivero-Verbel,* Wendy L. Holdan,* and Patricia E. Ganey*'t
*Department of Pharmacology and Toxicology and ^Department of Medicine, Institute for Environmental Toxicology, and National Food
Safety and Toxicology Center, Michigan State University, East Lansing, Michigan 48824
Received April 1, 1998; accepted July 23, 1998
along with the sum of chlorine atomic charges, are associated with
Neutrophil Activation by Polychlorinated Biphenyls: StructureActivity Relationship. Brown, A. P., Olivero-Verbel, J., Holdan,
W. L., and Ganey, P. E. (1998). Toxicol. Sci. 46, 308-316.
t h i s biological a c t i v i t y .
© 1998 Society of Toxicology.
Key Words: PCB; molecular model; superoxide anion; leukocyte.
Polychlorinated biphenyls (PCBs) rapidly stimulate polymorphonuclear leukocytes (neutrophils) in vitro to produce superoxide
anion (OJ). This response results from activation of various inPolychlorinated biphenyls (PCBs) are chemicals which have
tracellular signal transduction pathways and appears to occur in a been released into the environment, resulting in widespread
structure-specific fashion. Individual PCB congeners, varying in and persistent contamination. They exist as 209 different conpattern and extent of chlorination, were tested for their ability to
geners that vary with the chlorine substitution pattern of the
stimulate production of OJ and/or to enhance the response to
biphenyl
rings. Twenty of these congeners favor a coplanar
protein kinase C activation by phorbol myristate acetate (PMA).
Neutrophils were isolated from retired breeder, male, Sprague- configuration (Safe, 1984). Coplanar PCBs are structurally
Dawley rats and exposed to either vehicle, 10 or 50 /xM PCB for related to polychlorinated dibenzofurans and polychlorinated
30 min at 37°C. PMA (0 or 20 ng/ml) was added for an additional dibenzo-/?-dioxins, the most toxic of which is 2,3,7,8-tetrachlo10 min, and OJ generated during the incubation period was rodibenzo-p-dioxin (TCDD). Many of the biochemical changes
measured. 2,2'-Dichlorobiphenyl (2,2'-DCB), 2,4'-DCB, or 3,3'- and toxic responses elicited by TCDD, coplanar PCBs, and
DCB (50 jiM) stimulated neutrophils to produce OJ. Incubation structurally related compounds arise from interactions with the
of neutrophils with 4,4'-DCB, 3,3',4,4',5-pentachlorobiphenyl Ah receptor and subsequent alterations in gene expression
(3,3',4,4',5-PeCB), 3,3',4,5,5'-PeCB, or 2,2',3,3',4,4'-hexachlorobi(Whitlock, 1990). Structure-activity studies have shown that
phenyl (2,2',3,3',4,4'-HCB) did not result in generation of OJ. Of
the various congeners, 2,4'-DCB elicited the greatest production of the Ah receptor contains a hydrophobic ligand-binding site that
OJ. Exposure to 10 ftM 2,2'-DCB, 2,4'-DCB, 3,3'-DCB, or preferentially binds coplanar, nonpolar molecules (Landers and
2,2',3,3',4,4'-HCB prior to addition of PMA caused a significant Bunce, 1991).
Among the biological effects produced by PCBs is activaincrease in the amount of OJ produced, greater than that seen
with either compound alone. PMA-stimulated OJ production was tion of polymorphonuclear leukocytes (neutrophils). Neutrounaffected by prior exposure to 4,4'-DCB, 3,3',4,4',5-PeCB, or phils are usually the first leukocyte to arrive at sites of infection
3,3',4,5,5'-PeCB. In separate experiments, 3,3',4,4',5-PeCB inhib- or tissue injury and are a major cellular component of acute
ited the amount of OJ produced in response to activation with inflammatory responses (Sandborg and Smolen, 1988). As part
either 3,3'-DCB or 2,4'-DCB. Thus, it appears that congeners
of their function in inflammation and immunity, neutrophils
which are noncoplanar are capable of stimulating neutrophil OJ
become
activated and generate superoxide anion (OJ). PCBs
production. Coplanar congeners with high affinity for the Ah
activate
neutrophils
in vitro to produce OJ and alter the genreceptor do not activate neutrophils to produce OJ and may
inhibit this response. These results are consistent with the hypoth- eration of OJ in response to other stimuli (Ganey et ai, 1993).
esis that PCBs stimulate neutrophil OJ production by a mecha- Studies examining the mechanisms by which PCBs stimulate
nism that is structure-specific and dependent on the chlorine neutrophils have revealed roles for phosphoinositide hydrolysubstitution pattern of the biphenyl rings. Molecular modeling sis, activation of phospholipase A2 and tyrosine kinases, and
suggested that the sum of atomic charges on chlorine atoms is the calcium mobilization (Brown and Ganey, 1995; Tithof et ai,
most important descriptor for congeners which stimulate OJ pro- 1995, 1996, 1997, and 1998). Some PCB congeners are effecduction. The angle of rotation and the difference in energy betive activators of neutrophils and others are not.
tween the highest occupied molecular orbital and the lowest unThe ability of individual PCB congeners to stimulate OJ
occupied molecular orbital are integrative descriptors which,
production by neutrophils is likely to be related to structural
1096-6080/98 $25.00
Copyright © 1998 by the Society of Toxicology.
All rights of reproduction in any form reserved.
308
NEUTROPHIL ACTIVATION BY POLYCHLOR1NATED B1PHENYLS
configuration, which is determined by the chlorine substitution
pattern on the biphenyl rings. Previous studies have demonstrated that congeners with ortho substitution such as 2,2',4,4'tetrachlorobiphenyl (2,2',4,4'-TCB) and 2,3,4,5-TCB activate
neutrophils, whereas congeners without ortho substitution
(e.g., 3,3',4,4'-TCB) do not (Ganey et al, 1993; Brown and
Ganey, 1995; Tithof et al, 1996). Since both 2,2\4,4'-TCB
and 2,3,4,5-TCB have low affinity for the/i/i receptor, whereas
3,3',4,4'-TCB has high affinity for the receptor (Safe et al,
1985), these results suggest that neutrophil activation occurs
through a mechanism independent of the Ah receptor. In the
present study, a more detailed examination of the relationship
between PCB structure and ability to activate neutrophils was
undertaken. PCB congeners with various chlorination patterns
were tested for their ability to stimulate production of OJ and
to enhance OJ generation in response to activation of protein
kinase C. Molecular modeling was conducted to identify molecular descriptors which associate with stimulation of neutrophil O^~ production.
MATERIALS AND METHODS
Chemicals
Individual polychlorinated biphenyl congeners (>99% pure) were purchased from ChemService (West Chester, PA). Glycogen (Type II from
oyster), phorbol myristate acetate (PMA), superoxide dismutase (SOD), dimethylformamide (DMF), and ferricytochrome C were purchased from Sigma
Chemical Company (St. Louis, MO).
Isolation of Rat Neutrophils
Neutrophils were isolated from the peritoneum of male, Sprague-Dawley,
retired breeder rats (CD-Crl:CD (SD)BR VAF/Plus; Charles River Laboratories, Portage, MI) by glycogen elicitation as described previously (Hewett and
Roth, 1988). Isolated neutrophils were resuspended in Hanks' balanced salt
solution (HBSS), pH 7.4. The final concentration of cells in all of the assays
was 2 X 106 cells/ml.
Superoxide Anion Production
Neutrophils (2 X 106 cells) were suspended in borosilicate glass test tubes,
12 X 75 mm (VWR, Chicago, IL), in a final volume of 1 ml HBSS and
exposed to either vehicle (1 fi\ DMF) or 10 or 50 /i.M PCB for 30 min at 37°C.
Stock solutions of the PCBs were prepared in DMF, and 1 (xl of the stock
solution of PCB was added to the cells to achieve the desired concentration.
These PCB concentrations were chosen based on results of previous studies
examining the ability of PCBs to activate neutrophils in the absence of
significant cytotoxicity (Ganey el al., 1993; Brown and Ganey, 1995). PMA (0
or 20 ng/ml) was then added for an additional 10 min (37°C). OJ generated
during this 40-min period was measured by the SOD-sensitive reduction of
ferricytochrome C (Babior et al., 1973). In a separate series of experiments,
neutrophils were exposed to PCB congeners alone or in combination for 40
min at 37°C, and OJ production was determined.
Cytotoxicity Assay
Release by neutrophils of the cytosolic enzyme, lactate dehydrogenase
(LDH), into the medium was used as a marker of cytotoxicity. Neutrophils
were exposed to either vehicle (DMF) or to PCB congeners alone or in
309
combination for 40 min at 37°C. The cells were centrifuged immediately at
4°C, and the cell-free supernatant fluids were collected. LDH activity in the
supernatant fluids was measured according to the method of Bergmeyer and
Bernt (Bergmeyer and Bernt, 1974). Release of LDH into the medium was
expressed as the percentage of total LDH activity that was present in an
equivalent number of cells lysed with 1% Triton X-100.
Statistical Analysis
All results are presented as means ± the standard error of the mean (SEM).
For all results presented, N represents the number of individual experiments
using neutrophils isolated from individual animals. Unless otherwise stated,
data were analyzed by a two-way analysis of variance (ANOVA), and individual comparisons were performed using the least significant difference test.
The criterion for statistical significance was p < 0.05.
Analysis of Structure-Activity Relationships
Descriptor generation. Chemical structures of the seven PCB congeners
were entered into a computer program which then searched for the conformation of least energy for each congener in aqueous solution. Based on these
structures, electronic, geometrical, and topological molecular descriptors were
calculated for each congener. To accomplish this, PCB congener structures
were entered into the Spartan Program (Wavefunction, Inc., Spartan Version 4,
1995, Irvine, CA) running on a Silicon Graphics Workstation. Conformer
search was conducted at the semiempirical level Austin Model 1 (AMI)
(Dewar et al, 1985) using water as a solvent, and this preliminary conformation was used to perform Ab Initio 3-21G(<) full optimization (Pietro et al,
1982) and single-point properties calculations. The following electronic descriptors were obtained from single-point calculations: the most positive
charge (MPC) and the most negative charge (MNC) in the molecule accounting for nonspecific interactions between molecules; the sum of the atomic
charges for all of the carbon atoms (ECAC) and the sum of the atomic charges
for all of the chlorine atoms (2C1AC) as a measure of localized electronic
density which may mediate electrostatic, intermolecular interactions; and the
energy of the highest occupied molecular orbital (HOMOE) and the energy of
the lowest unoccupied molecular orbital (LUMOE), which have been recognized as important descriptors in electron transfer from a donor to an acceptor
molecule through the formation of charge transfer complexes (Franke, 1984).
In general, charge-based descriptors are considered as chemical reactivity
indices which may mediate weak, intermolecular interactions (Karelson and
Lobanov, 1996). In addition, the difference between HOMOE and LUMOE,
the HOMO-LUMO gap, was included as a descriptor. The torsional angle
2,1,1 \2'(0) obtained from the Ab Initio calculation was used as a geometrical
descriptor.
Topological descriptors included molecular connectivity values which describe molecular size and shape. These were calculated with a PC 586/66 MHz
using PCDM, a program written by Olivero et al. (1995). Calculated connectivity values included Randic (Randic, 1975) and Kier and Hall indices (Kier
and Hall, 1986).
RESULTS
In the absence of any stimulus, OJ production by rat neutrophils was negligible (Figs. 1-3), which is consistent with
previous reports (Ganey et al, 1993). Exposure to 50 /xM but
not 10 juM 2,2'-dichlorobiphenyl (2,2'-DCB) caused significant production of OJ in the absence of PMA (Fig. 1A).
Neutrophils exposed to vehicle (DMF) prior to stimulation
with PMA, an activator of protein kinase C (Tauber, 1987),
produced a significant amount of OJ, as demonstrated previously (Ganey et al, 1993). Exposure to 10 /i-M 2,2'-DCB prior
310
BROWN ET AL.
O2 in response to 10 /LtM but not to 50 /xM 3,3'-DCB was
elevated significantly by addition of PMA.
OJ generation was not stimulated by exposure of neutrophils to either 10 or 50 /xM 4,4'-DCB (Fig. 2A). Addition of
PMA resulted in significant generation of OJ which was
unaffected by prior exposure to 4,4'-DCB. Similar results were
seen with 3,3',4,4',5-pentachlorobiphenyl (3,3',4,4',5-PeCB;
Fig. 2B) and 3,3',4,5,5'-PeCB (Fig. 2C): neutrophils exposed
to either of these congeners alone were not stimulated to
produce OJ, and neither PCB congener increased the response
to PMA.
Exposure of neutrophils to either 10 or 50 fiM 2,2',3,3',4,4'hexachlorobiphenyl (2,2',3,3',4,4'-HCB) did not result in generation of O^ (Fig. 3). Prior exposure to 2,2',3,3',4,4'-HCB
significantly increased the amount of O^~ produced upon addition of PMA.
I
-o
2
O
~o
I
Pro
T3
CM
o
"o
70
60
50
40
30
20
10
0
0
T3
s
a.
CN
O
"o
E
10
20
30
40
50
70
60
50
40
30
20
10
0
2
CL
o
10
0
10
20
30
40
20
30
40
50
40
50
50
PCB
FIG. 1. Superoxide anion (O2 ) production during exposure to PCB congeners which activated neutrophils in the absence of PMA and increased the
response to PMA. Neutrophils were exposed to 0, 10, or 50 fiM of (A)
2,2'-DCB, (B) 2,4'-DCB, or (C) 3,3'-DCB in the presence (filled circles) and
absence (open circles) of PMA as described under Materials and Methods. OJ
produced during the incubation period was determined. N = 3-5. a, Significantly different from respective value in the absence of PMA. b, Significantly
different from respective value in the absence of DCB.
•o
§
3,3',4,4',5- Pentachlorobiphenyl
2
a.
O
"o
3,3',4,5,5'- Pentachlorobiphenyl
to addition of PMA resulted in a significant increase in the
amount of OJ produced, greater than that seen with either
compound alone. OJ production by neutrophils exposed to 50
juM 2,2'-DCB before addition of PMA was not different from
that seen with either stimulus alone.
Neutrophils exposed to 50 /xM but not 10 fjM 2,4'-DCB
produced a significant amount of OJ (Fig. 1B). Exposure to 10
fxM 2,4'-DCB prior to addition of PMA resulted in a significant increase in the amount of O^~ generated compared to that
produced by PMA alone. Exposure to 50 /xM 2,4'-DCB prior
to PMA produced an equivalent amount of OJ as with 50 JU,M
2,4'-DCB alone.
Exposure of neutrophils to 50 jtM but not 10 /tM 3,3'-DCB
elicited significant production of OJ (Fig. 1C). Generation of
2
a.
o
0
10
20
30
PCB
FIG. 2. Superoxide anion (O2 ) production during exposure to PCB congeners which did not activate neutrophils or alter the response to PMA.
Neutrophils were exposed to 0, 10, or 50 ^M of (A) 4,4'-DCB, (B) 3,3',4,4',5PeCB, or (C) 3,3',4,5,5'-PeCB in the presence (filled circles) and absence
(open circles) of PMA as described under Materials and Methods. OJ produced during the incubation period was determined. N = 3—6. a, Significantly
different from respective value in the absence of PMA.
311
NEUTROPHIL ACTIVATION BY POLYCHLORINATED BIPHENYLS
TABLE 2
Lactate Dehydrogenase Released (;% Total) from Neutrophils
T3
3,3 ',4,4',5-PeCB OxM)
o
a.
20
o
10
0
10
20
30
40
50
FIG. 3. Superoxide anion (OJ) production during exposure to a PCB
congener which increased the response to PMA but alone did not activate
neutrophils. Neutrophils were exposed to 0, 10, or 50 /xM 2,2',3,3',4,4'-HCB
in the presence (filled circles) and absence (open circles) of PMA as described
under Material and Methods. O7 produced during the incubation period was
determined. N = 4 - 6 . a, Significantly different from respective value in the
absence of PMA. b, Significantly different from respective value in the absence
of 2,2',3,3',4,4'-HCB.
The effect of a coplanar congener which alone did not
activate neutrophils on the response to two congeners which
did stimulate neutrophils was examined. Neutrophils were exposed to either 3,3'-DCB or 2,4'-DCB for 40 min in the
presence or absence of 3,3',4,4',5-PeCB (Table 1). 3,3',4,4',5PeCB inhibited the production of OJ in response to 3,3'-DCB
and 2,4'-DCB by 91 and 57%, respectively. Coincubation with
3,3',4,4',5-PeCB did not affect release of LDH from neutrophils exposed to either of the DCB congeners (Table 2).
The relationship among the absolute values for the H 0 M 0 LUMO gap, torsional angle, SC1AC, and the difference between OJ produced in response to 0 and 50 /iM PCB in the
presence of PMA (i.e., "biological response") is shown graphically in Fig. 4. As illustrated, the molecular descriptor SC1AC
TABLE 1
Inhibition of Neutrophil Superoxide Anion Production
(nmoles OJ Produced) By 3,3',4,4',5-Pentachlorobiphenyl
3,3'.4.4 '.5-PeCB (/J.M)
0
Vehicle
3,3'-DCB
2,4'-DCB
1.6
28.0
36.8
±0.2
± 5.1
±5.0
0
50
Vehicle
3,3'-DCB
2,4'-DCB
7.1 ± 1.5
15.6 ± 3.4
13.8 ± 1.9
8.4 ± 0.8
10.4 ± 0.2
13.9 ± 2.0
Note. Rat neutrophils were exposed to vehicle (DMF), 50 /xM 3,3'-DCB, or
50 juM 2,4'DCB in the presence or absence of 3,3',4,4',5-PeCB for 40 min at
37°C. LDH released into the medium was measured and reported as the
percentage of total cellular LDH activity as described under Materials and
Methods. N = 3. No significant differences were observed.
[jM 2,2',3,3',4,4'- Hexachlorobiphenyl
Stimulus
Stimulus
50
% Inhibition
2.9 ± 1.1
2.5 ± 1.0*
16.0 ± 3.7*
91
57
Note. Rat neutrophils were exposed to vehicle (DMF), 50 /xM 3,3'-DCB, or
50 jxM 2,4'-DCB in the presence or absence of 3,3',4,4',5-PeCB for 40 min at
37°C, and O^ produced during the incubation period was determined. N = 4.
* Significantly different from respective value in the absence of 3,3',4,4',5PeCB (Student's I test).
was inversely related to this biological response. Congeners for
which SC1AC is large (4,4'-DCB, 3,3',4,5,5'-PeCB, and
3,3',4,4',5-PeCB) produced little biological response, whereas
those congeners for which this molecular descriptor is small
(3,3'-DCB, 2,4'-DCB, and 2,2',3,3',4,4'-HCB) caused an increase in Oj production at 50 /uM in the presence of PMA. The
exception is 2,2'-DCB for which both the biological response
(at 50 JLIM) and the molecular descriptor were small. Torsional
angle did not appear to influence this response because the
value for this parameter was small for a congener which
produced a significant response (i.e., 3,3'-DCB) and for congeners (4,4'-DCB, 3,3',4,5,5'-PeCB, and 3,3',4,4',5-PeCB)
which did not elicit a response. Likewise, torsional angle was
large for congeners which did (2,4' DCB and 2,2',3,3',4,4'HCB) and did not (2,2'-DCB) induce OJ production under
these conditions. All of the molecules with small biological
activities also had small HOMO-LUMO gaps, whereas three
of the four congeners with large HOMO-LUMO gaps elicited
OJ production. Once again, the exception was 2,2'-DCB.
Relationships between biological activity (dependent variable)
and molecular descriptors (independent variables) were established by the stepwise multiple regression technique (Jandel
Corp.). The biological response used for modeling was the difference in the amount of OJ produced at 0 and 50 /uM PCB, and
models were generated for data obtained in the absence and in the
presence of PMA. The best regression models were selected on
the basis of the greatest multiple correlation coefficient (R) obtained between the experimental and predicted data, the lowest
standard error of estimation (SE) and the highest F value from the
variance analysis. The most important descriptor for models in the
absence or presence of PMA was the reciprocal of SC1AC. The
model for OJ production in the absence of PMA was improved
by the addition of torsional angle as a descriptor (R = 0.930 using
both descriptors compared to 0.538 with 1/SC1AC alone). The
addition of the HOMO-LUMO gap as a descriptor improved the
prediction model for data obtained in the presence of PMA (R =
0.964 using both descriptors compared to 0.574 with 1/SClAC
alone).
312
BROWN ET AL.
Biological
Response
HOMO
-LUMO
gap
2,2'-DCB
Torsional
angle
2,4'-DCB
SC1AC
4,4'-DCB
3,3',4,4',5'-PeCB
3,3 ',4,5,5 '-PeCB
2,2',3,3',4,4'-HCB
FIG. 4. Relationship among the specific response produced at 50 /xM PCB in the presence of PMA, HOMO-LUMO gap, 2C1AC, and torsional angle for
PCB congeners examined in this study. The length of the line from the center represents the absolute value for each parameter.
DISCUSSION
Much of the research into the biological effects of PCBs has
focused on TCDD-like activity of these chemicals such as
cytochrome P-450 induction, hepatotoxicity, thymic atrophy,
etc. (Safe, 1994). A primary structural requirement for TCDDlike effects of PCB congeners is the tendency toward coplanarity of the two phenyl rings, such that the PCB molecule
attains a structure with dimensions approximating 3 X 10 A.
Conditions favoring this structure include substitution with
chlorine at four meta and/or para sites and lack of substitution
at ortho positions. This structure provides steric fit and net
polarizability which increase binding affinity for the Ah receptor (McKinney and Singh, 1981). More recently, quantum
chemical calculations revealed that electronic-derived physicochemical properties such as electron affinities, entropies, and
electronic energy gaps are also important parameters governing
the interaction of PCBs with the Ah receptor (Kafafi et al,
1993a, b).
There is growing evidence that congeners which do not
favor a coplanar configuration also have biological activity
(Ganey ef a/., 1993; Shain et al, 1991; Kodavanti et al, 1995),
although the structural requirements for these effects are incompletely understood. We have demonstrated previously that
activation of neutrophils by PCBs does not appear to be mediated through the Ah receptor, despite the presence of these
receptors in rat neutrophils (Ganey et al, 1993). The current
study was undertaken to explore more extensively which PCB
congeners elicit neutrophil activation and to identify structural
requirements for these effects.
What emerged from examination of the effects of these congeners on neutrophil function in this and previous studies were
three patterns of response. In the first, PCBs (i.e., 2,2'-DCB,
2,4'-DCB, 3,3'-DCB, 2,2',4,4'-TCB, and 2,3,4,5-TCB (Brown
and Ganey, 1995)) stimulated neutrophils to produce O^~ in the
absence of PMA and increased the response to PMA (Table 3,
Fig. 1). The second pattern was characterized by 2,2',3,3',4,4'HCB, which alone did not activate neutrophils but potentiated
PMA-induced OJ production (Fig. 3). In the third pattern, PCBs
(i.e., 4,4'-DCB, 3,3',4,5,5'-PeCB, 3,3',4,4',5-PeCB, and 3,3',4,4'TCB (Ganey et al, 1993)) neither elicited OJ production nor
changed the response to PMA (Fig. 2).
Results of previous studies suggested that substitution at the
313
NEUTROPHIL ACTIVATION BY POLYCHLORINATED BIPHENYLS
TABLE 3
Physical Properties of Various PCB Congeners and Their Effects on Function of Neutrophils
Congener
2,2'-DCB
2,4'-DCB
2,2',4,4'-TCB
2,3,4,5-TCB
3,3'-DCB
2,2',3,3',4,4'-HCB
4,4'-DCB
3,3',4,5,5'-PeCB
3,3',4,4'-TCB
3,3',4,4',5-PeCB
Stimulates O 2
production?
Yes
Yes
Yesc
Yes"
Yes
No
No
No
Noe
No
Potentiates PMA
response?
*** (My
Log Kowb
0
2,1, l',2'
Yes
Yes
Yesc
Yesrf
Yes
Yes
No
No
Noc
No
6.9 X 10"'
1.9 x 10"3
5.2 x 10""
1.0 x 10""
9.4 x 10"5
2.7 X 10"5
1.3 X 10"6
1.0 x 10"7
7.1 x 10"8
3.3 X 10"8
4.9
5.1
5.9
5.9
5.3
7.0
5.3
NR
6.1
NR
90.4
73.2
90.2
59.9
52.3
90.2
51.7
53.0
52.4
52.4
Note. NR, not reported.
" ^caic (M) is the calculated in vitro dissociation constant of PCB-Ah receptor complexes in rat liver cytosols (from Kafafi et al, 1993b).
* ^ow ' s t r l e octanol/water partition coefficient (from Shiu and Mackay, 1986).
c
From Ganey et al, 1993.
''From Brown and Ganey, 1995.
ortho position may be a requirement for neutrophil activation
by PCBs because 2,2',4,4'-TCB and 2,3,4,5-TCB, but not
3,3',4,4'-TCB, influenced neutrophil function (Ganey et al.,
1993; Brown and Ganey, 1995). All of the ortho-substituted
congeners examined in this study stimulated neutrophils to
produce OJ and/or altered the response to PMA (Table 3). The
hypothesis that ortho substitution is required for activation of
neutrophils by PCBs was refuted, however, by results of studies in which 3,3'-DCB caused generation of OJ by itself and
increased the response to PMA (Fig. 1C).
Examination of octanol/water partition coefficients (Table 3)
for individual congeners revealed that the propensity to activate neutrophils did not correlate with hydrophobicity. For
example, the values for log Kow are the same for 3,3'-DCB
and 4,4'-DCB: The former both stimulated O^~ production and
increased the response to PMA, whereas the latter did not.
Based on the log A"ow values, at the concentrations used most
of the PCBs would be expected to partition into the lipid
membranes of the neutrophils. Thus, it is not surprising that
differences in biological activity among congeners are not
related to different partitioning within the cells. The relation of
the octanol/water partition coefficient to activity has been
examined in other systems. Similar to results observed here,
this parameter did not correlate with the decrease in dopamine
content of PC12 cells (Shain et al, 1991) and in rat hepatoma
cells hydrophobicity was found to modulate interactions of
PCBs with the Ah receptor (Safe et al., 1985). Alteration of
neutrophil function by PCBs also did not correlate with the
degree of chlorination because congeners with as few as two
and as many as six chlorines potentiated the response to PMA.
Activation of neutrophils by PCBs appears to be related to a
low affinity for the Ah receptor, based on the in vitro dissociation constant (Table 3). This suggests that congeners which
favor a noncoplanar structure are more likely than coplanar
congeners to activate neutrophils in vitro. Para substitution of
3,3'-DCB to yield 3,3',4,4'-TCB abolished activity to stimulate neutrophils; however, para substitution of 2,2'-DCB (to
yield 2,2',4,4'-TCB) did not diminish the potential to activate
neutrophils. 3,3',4,4'-TCB attains a coplanar structure more
readily than the diortho-substituted 2,2',4,4'-TCB because ortho substitution substantially raises the energy barrier to rotation about the bond between the phenyl rings (McKinney and
Waller, 1994). Thus, this observation supports the hypothesis
that noncoplanar structure is important in the ability of PCBs to
activate neutrophils. Similarly, ortho substitution of 3,3',4,4'TCB conferred the property to increase generation of O^" in
response to PMA (Fig. 3).
Despite the apparent importance of noncoplanarity in the
ability of PCBs to influence O^" production, results of molecular modeling argue against a structure-activity relationship
based solely on coplanarity of the molecules. Examination of
Table 3 reveals that the value for torsional angle (0) for
3,3'-DCB, a congener which activated neutrophils, is similar to
the value for a number of congeners which did not. Similar
results are depicted in Fig. 4 for the specific responses observed at a PCB concentration of 50 jxM in the presence of
PMA. Although torsional angle alone did not correlate with OJ
production, inclusion of this parameter with the reciprocal of
SC1AC in a regression model for the response in the absence
of PMA significantly improved the regression coefficient.
Thus, angle of rotation about the bond between phenyl rings
may be an important parameter for PCB-induced OJ production under some, but not all, conditions, and other factors
influence this response as well.
In fact, results of regression analysis suggest that the reciprocal of 2C1AC is the most important descriptor for the re-
314
BROWN ET AL.
sponse to PCBs in the presence or absence of PMA. The sum
of atomic charges is an important descriptor in other models of
toxicity (Katritzky and Gordeeva, 1993) as well as in some
pharmacological activities (Cocchi et al, 1995). The regression model for biological activity in the presence of PMA was
strengthened by addition of the value for the HOMO-LUMO
gap, which correlates strongly (/? = 0.953, p = 0.001) with
torsional angle. The HOMO-LUMO gap is a measure of the
capacity of a molecule to transfer one electron from the HOMO
to the LUMO, and its magnitude is related directly to molecular stability (Mekenyan et al, 1994). A model described by
SC1AC and the HOMO-LUMO gap suggests that PCB-induced O^~ production may be mediated through electrostatic
interactions which involve an electron transfer complex between the PCB congener and a cellular molecule(s) and that
this interaction is favored for PCB congeners with greater
molecular stability.
Studies in vitro investigating the neurotoxicity of individual
PCB congeners have demonstrated similar structure-activity relationships as seen in the current study. Congeners with chlorine
atoms in the ortho position decreased cellular dopamine content in
PC 12 cells (Shain et al, 1991) and increased binding of phorbol
esters in rat cerebellar granular cells (Kodavanti et al, 1995),
whereas congeners without ortho substitution that are coplanar
were inactive. Similar to our studies in which 3,3'-DCB stimulated O^~ production by neutrophils, 3,3'-DCB decreased PC 12
dopamine content and increased phorbol ester binding in cerebellar granular cells (Shain et al., 1991; Kodavanti et al., 1995).
Studies with isolated rat cerebellar microsomes and mitochondria
demonstrated that congeners with chlorine substitution at the
ortho positions were the most potent in inhibiting Ca +2 sequestration (Kodavanti et al., 1996). Similarly, in sarcoplasmic reticulum from skeletal and cardiac tissue ortho-substituted, and not
coplanar, PCB congeners increased Ca + 2 release and binding of
ryanodine to ryanodine-sensitive Ca + 2 release channels (Wong
and Pessah, 1996).
The mechanisms by which noncoplanar PCBs induce biological effects are a subject of current investigation. In studies
presented here, both the ability to stimulate OJ production and
the capacity to alter the response to activation of protein kinase
C by PMA were examined. Generation of OJ by neutrophils is
effected by an NADPH oxidase which is assembled and activated in the plasma membrane, and this process involves a
number of intracellular signal transduction pathways (Sadler
and Badwey, 1988). PMA initiates OJ production through
activation of protein kinase C (PKC) (Tauber, 1987). The
observation that 3,3',4,4',5-PeCB decreased the response to
2,4'-DCB and 3,3'-DCB but not to PMA suggests that the
mechanism by which these two DCBs activate neutrophils
differs from the mechanism of activation by PMA. In support
of this hypothesis, previous studies have demonstrated that
PCBs activate OJ production by a mechanism which is depen-
dent on calcium, whereas PMA stimulates O 2 generation by a
calcium-independent mechanism (Brown and Ganey, 1995).
The fact that some congeners (e.g., 2,2',3,3',4,4'-HCB) did
not alone stimulate O^~ production but did increase the response to PMA raises the possibility that these two events
occur by different mechanisms. 2,2',3,3',4,4'-HCB may invoke a signal transduction pathway which does not activate
NADPH oxidase but enhances O^~ generation once the enzyme
is assembled and activated. Alternatively, the same mechanism
may be responsible for both responses. For example, in the
case of 2,4'-DCB, a signal transduction pathway may be activated which is sufficient at large PCB concentrations (50 yM)
to activate NADPH oxidase in the absence of PMA but at
smaller concentrations (10 fxM) requires additional pathways
to be activated by PMA to cause generation of OJ. It has been
demonstrated that noncoplanar PCBs increase phorbol ester
binding in cerebellar granular cells (Kodavanti et al., 1995), so
that one mechanism by which these congeners potentiate the
response to PMA in neutrophils may be by increasing the
accessibility or affinity of PKC for PMA.
In the current study, 3,3',4,4',5-PeCB was inactive in stimulating Oj production by neutrophils but inhibited the response to stimulation of neutrophils with either 3,3'-DCB or
2,4'-DCB. This inhibitory effect was unrelated to cytotoxicity,
and the mechanism(s) involved is unknown. Previous studies
have shown that [3H]TCDD bound specifically to cytosol from
rat neutrophils, suggesting the presence of Ah receptors (Ganey
et al., 1993). It is possible that congeners with dioxin-like
characteristics and high binding affinities for the Ah receptor
may elicit Ah receptor-dependent responses which downregulate activation of neutrophils. For example, activation of neutrophils is dependent on the sequential phosphorylation and
dephosphorylation of several proteins in the signalling cascade.
Recent evidence suggests that binding of TCDD to the Ah
receptor elicits protein tyrosine phosphorylation through activation of a protein tyrosine kinase, pp60src (Enan and Matsumura, 1996; Blankenship and Matsumura, 1997; Ma et al,
1992). It is conceivable that activation of a protein tyrosine
kinase may interfere with the signaling required for PCBinduced OJ production. If this is the case, the inhibitory effect
may not extend to agents which stimulate neutrophils via direct
activation of PKC, such as PMA (Fig. 2B). These results, in
which a PCB congener inhibited the biological effect of another, are similar in nature to those observed for estrogenic
activity of PCBs (Connor et al, 1997; Jansen et al, 1993). For
example, whereas Aroclor 1242 and 2,2',5,5'-TCB demonstrated estrogenic effects in rats, 3,3',4,4'-TCB did not and
inhibited this effect of Aroclor 1242 (Jansen et al, 1993).
In summary, the results of this study suggest that chlorines
in the ortho or meta position are required for stimulation of
neutrophil OJ production in vitro. This pattern of substitution
may provide characteristics of atomic charge distribution on
chlorine atoms, torsional angle, and HOMO-LUMO gap that
NEUTROPHIL ACTIVATION BY POLYCHLORINATED BIPHENYLS
result in a configuration capable of activation of neutrophils.
By contrast, coplanar congeners may act in an inhibitory fashion to downregulate O^~ production in the presence of a second
neutrophil stimulus.
ACKNOWLEDGMENTS
The authors acknowledge the technical assistance of Henry Ng. This research was supported by NIH Grant ES04911. A.P.B. was supported in part by
a National Research Service Award from the National Institute of Environmental Health Sciences, NIH Grant ES05684.
REFERENCES
Babior, B. M., Kipnes, R. S., and Cumutte, J. T. (1973). Biological defense
mechanisms: The production by leukocytes of superoxide, a potential bactericidal agent. J. Clin. Invest. 52, 741-744.
Bergmeyer, H. V., and Bernt, E. (1974). Lactate dehydrogenase UV assay with
pyruvate and NADH. In Methods of Enzymatic Analysis (H. V. Bergmeyer,
Ed.), pp. 524-579. Academic Press, New York.
Blankenship, A., and Matsumura, F. (1997). 2,3,7,8-Tetrachlorodibenzo-pdioxin-induced activation of a protein tyrosine kinase, pp60sre, in murine
hepatic cytosol using a cell-free system. Mol. Pharmacol. 52, 667-675.
Brown, A. P., and Ganey, P. E. (1995). Neutrophil degranulation and superoxide production induced by polychlorinated biphenyls are calcium dependent. Toxicol. Appl. Pharmacol. 131, 198-205.
Cocchi, M., Menziani, C , Fanelli, F., and de Benedetti, P. G. (1995). Theoretical quantitative structure-activity relationship analysis of congeneric and
non-congeneric a, adrenoreceptor antagonists: A chemometric study. J.
Mol. Struct. (Theochem.) 331, 79-93.
Connor, K., Ramamoorthy, K., Moore, M., Maustain, M., Chen, I., Safe, S.,
Zacharewski, T., Gillesby, B., Joyeux, A., and Balaguer, P. (1997). Hydroxylated polychlorinated biphenyls (PCBs) as estrogens and antiestrogens: Structure-activity relationships. Toxicol. Appl. Pharmacol. 145, 111123.
Dewar, M. J. S., Zoebisch, E. G., Healy, E. F., and Stewart, J. J. P. (1985).
AMI: A new general purpose quantum mechanical molecular model. J. Am.
Chem. Soc. 107, 3902-3909.
Enan, E., and Matsumura, F. (1996). Identification of c-Src as the integral
component of the cytosolic Ah receptor complex, transducing the signal of
2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD) through the protein phosphorylation pathway. Biochem. Pharmacol. 52, 1599-1612.
Franke, R. (1984). Theoretical Drug Design Methods. Elsevier, New York.
Ganey, P. E., Sirois, J. E., Denison, M., Robinson, J. P., and Roth, R. A.
(1993). Neutrophil function after exposure to polychlorinated biphenyls in
vitro. Environ. Health Perspect. 101(5), 430-434.
Hewett, J. A., and Roth, R. A. (1988). Dieldrin activates rat neutrophils in
vitro. Toxicol. Appl. Pharmacol. 96, 269-278.
Jansen, H. T., Cooke, P. S., Porcelli, J., Liu, T. C , and Hansen, L. G. (1993).
Estrogenic and antiestrogenic actions of PCBs in the female rat: In vitro and
in vivo studies. Reprod. Toxicol. 7, 237-248.
Kafafi, S. A., Afeefy, H. Y., Ali, A. H., Said, H. K., Abd-Elazem, I. S., and
Kafafi, A. G. (1993a). Affinities for the aryl hydrocarbon receptor, potencies
as aryl hydrocarbon hydroxylase inducers and relative toxicities of polychlorinated biphenyls: A congener specific approach. Carcinogenesis 14,
2063-2071.
Kafafi, S. A., Afeefy, H. Y., Ali, A. H., Said, H. K., and Kafafi, A. G. (1993b).
Binding of polychlorinated biphenyls to the aryl hydrocarbon receptor.
Environ. Health Perspect. 101, 422-428.
315
Karelson, M. and Lobanov, V. S. (1996). Quantum-chemical descriptors in
QSAR/QSPR studies. Chem. Rev. 96, 1027-1043.
Katritzky, A. R., and Gordeeva, E. V. (1993). Traditional topological indices
vs electronic, geometrical, and combined molecular descriptors in QSAR7
QSPR research. J. Chem. Inf. Comput. Sci. 33, 835-857.
Kier, L., and Hall, H. L. (1986). Molecular Connectivity in Structure-Activity
Analysis. Wiley, New York.
Kodavanti, P. R. S., Ward, T. R., McKinney, J. D., and Tilson, H. A. (1995).
Increased [3H]phorbol ester binding in rat cerebellar granule cells by polychlorinated biphenyl mixtures and congeners: Structure-activity relationships. Toxicol. Appl. Pharmacol. 130, 140-148.
Kodavanti, P. R. S., Ward, T. R., McKinney, J. D., and Tilson, H. A. (1996).
Inhibition of microsomal and mitochondrial Ca2+-sequestion in rat cerebellum by polychlorinated biphenyl mixtures and congeners: Structure-activity
relationships. Arch. Toxicol. 70, 150-157.
Landers, J. P., and Bunce, N. J. (1991). They4/! receptor and the mechanism of
dioxin toxicity. Biochem. J. 276, 273-287.
Ma, X., Mufti, N. A., and Babish, J. G. (1992). Protein tyrosine phosphorylation as an indicator of 2,3,7,8-tetrachloro-p-dioxin exposure in vivo and in
vitro. Biochem. Biophys. Res. Commun. 189, 59-65.
McKinney, J. D., and Singh, P. (1981). Structure-activity relationships in
halogenated biphenyls: Unifying hypothesis for structural specificity. Chem.-Biol. Interact. 33, 271-283.
McKinney, J. D., and Waller, C. L. (1994). Polychlorinated biphenyls as
hormonally active structural analogues. Environ. Health Perspect. 102,
290-297.
Mekenyan, O. G., Ankley, G. T., Veith, G. D., and Call, D. J. (1994). QSARs
for photoinduced toxicity. I. Acute lethality of polycyclic aromatic hydrocarbons to Daphnia magna. Chemosphere 28, 567-582.
Olivero, J., Vivas, R., Payares, P., Dfaz, D., and Mercado, J. (1995). PCDM
Version 2, Universidad de Cartagena, Cartagena, Columbia.
Pietro, W. J., Francl, M. M., Hehre, W. J., DeFrees, D. J., Pople, J. A., and
Binkley, J. S. (1982). Self-consistent molecular orbital methods. 24. Supplemented small split-valence basis sets for second-row elements. J. Am.
Chem. Soc. 104, 5039-5048.
Randic, M. (1975). On characterization of molecular branching. J. Am. Chem.
Soc. 97,6609-6615.
Sadler, K. L., and Badwey, J. A. (1988). Second messengers involved in superoxide production by neutrophils. HematoUOncol. Clin. N. Am. 2, 185-200.
Safe, S. (1984). Polychlorinated biphenyls (PCBs) and polybrominated biphenyls (PBBs): Biochemistry, toxicology, and mechanism of action. CRC Crit.
Rev. Toxicol. 13, 319-395.
Safe, S., Bandiera, S., Sawyer, T., Bozena, Z., Mason, G., Romkes, M.,
Denomme, M. A., Sparling, J., Okey, A. B., and Fujita, T. (1985). Effects of
structure on binding to the 2,3,7,8-TCDD receptor protein and AHH induction—Halogenated biphenyls. Environ. Health Perspect. 61, 21-33.
Safe, S. (1994). Polychlorinated biphenyls (PCBs): Environmental impact,
biochemical and toxic responses, and implications for risk assessment. CRC
Crit. Rev. Toxicol. 24, 87-149.
Sandborg, R. R., and Smolen, J. E. (1988). Biology of disease: Early biochemical events in leukocyte activation. Lab. Invest. 59(3), 300-320.
Shain, W., Bush, B., and Seegal, R. (1991). Neurotoxicity of polychlorinated
biphenyl congeners on phosphoinositide hydrolysis and protein kinase C
translocation. Toxicol. Appl. Pharmacol. I l l , 33-42.
Shiu, W. Y., and Mackay, D. (1986). A critical review of aqueous solubilities,
vapor pressures, Henry's law constants, and octanol-water partition coefficients
of the polychlorinated biphenyls. J. Phys. Chem. Ref. Data 15, 911-929.
Tauber, A. I. (1987). Protein kinase C and the activation of the human
neutrophil NADPH-oxidase. Blood 69(3), 711-720.
316
BROWN ET AL.
Tithof, P. K., Contreras, M. L., and Ganey, P. E. (1995). Aroclor 1242
stimulates the production of inositol phosphates in polymorphonuclear neutrophils. Toxicol. Appl. Pharmacol. 131, 136-143.
Tithof, P. K., Peters-Golden, M , and Ganey, P. E. (1998). Distinct phospholipases A 2 regulate the release of arachidonic acid for eicosanoid
production and superoxide anion generation in neutrophils. J. Immunol.
160, 953-960.
Tithof, P. K., Schiamberg, E., Peters-Golden, M., and Ganey, P. E. (1996).
Phospholipase A2 is involved in the mechanism of activation of neutrophils
by polychlorinated biphenyls. Environ. Health Perspecl. 104, 52-58.
Tithof, P. K., Watts, S., and Ganey, P. E. (1997). Protein tyrosine kinase
involvement in the production of superoxide anion by neutrophils exposed
to Aroclor 1242, a mixture of polychlorinated biphenyls. Biochem. Pharmacol. 53, 1833-1842.
Whitlock, J. P., Jr. (1990). Genetic and molecular aspects of 2,3,7,8-tetrachlorodibenzo-p-dioxin action. Annu. Rev. Pharmacol. Toxicol. 30, 251-277.
Wong, P. W., and Pessah, I. N. (1996). Ortfto-substituted polychlorinated
biphenyls alter calcium regulation by a ryanodine receptor-mediated mechanism: Structural specificity toward skeletal- and cardiac-type microsomal
calcium release channels. Mol. Pharmacol. 49, 740-751.