fnt. J. Cancer: 18, 116-121 (1976)
TUMOR BOUND IMMUNOGLOBULINS : THE RELATIONSHIP BETWEEN THE IN VIVO COATING OF TUMOR CELLS BY POTENTIALLY CYTOTOXTC ANTI-TUMOR ANTIBODIES, AND THE EXPRESSION OF IMMUNE COMPLEX RECEPTORS Gary BRASLAWSKYMaya RANand Isaac P. Wirz Department of Microbiology, George S. Wise Center for Life Sciences, Tel Aviv University, Tel Aviv, Israel
SUMMARY
In this study we investigated the relationship between binding of anti-tumor antibodies by polyomavirus-induced S E YF-a tumor cells and the expression of immune-complex receptors on these cells. It was shown that in vivo propagated cells became progressively coated with IgG. The increase in the fgG coating of SEYF-a cells, occurring during the second week of propagation, was directly correlated with an increase in the coating of cells with potentially cytotoxic anti-tumor antibodies. The activity of these antibodies was demonstrated by cell lysis following addition of exogenous complement. Cell populations propagated in vivo for longer periods (3 weeks or more) became less sensitive to exogenous complement although their IgC coating remained high, and although higher titers of anti-tumor antibodies could be eluted from them. This indicated that antibodies coating young tumor cell populations have the capacity to activate complement whereas those coating older tumor cell populations are incapable of complement activation. Previous findirigs that S E YF-a cell populations are able to bind unrelated immune complexes were confirmed in the present study. We also found that the capacity of these populations to bind such unrelated complexes decreased with propagation time in vivo. The involvement of anti-tumor antibodies in this phenomenon was indicated by the finding that such antibodies inhibited binding of unrelated immune complexes by young cells but not by old cells. Furthermore, treatments causing dissociation of Ig from the cell surface restored, to some extent, the capacity of anti-tumor antibodies to inhibit immune complex binding by old SE YF-a populations. The presence of a surface immunoglobulin (Ig) coat on some or all of the cells in human or animal tumors propagating in vivo is fairly well established (Witz, 1973). The presence o f anti-tumor antibodies as part of this surface Ig coat on cells of a polyomavirus-induced fibrosarcoma SEYF-a, has been demonstrated (Ran et al., 1976). These cell-bound antibodies were capable of mediating cell lysis upon addition of exogenous complement indicating
that they were attached to the surface in a typical antibody-antigen interaction. Data have also been obtained (Braslawsky et al., 1976a), indicating that SEYF-a tumor cells possess receptors for soluble immune complexes composed of antigens and antibodies clearly unrelated to the tumor cells such as ovalbumin-antiovalbumin. Binding of Ig to the surface of the SEYF-a cell could thus occur either through such complex receptors or by the binding of anti-tumor antibodies to their target antigen. In this paper the relationship existing between these two possible modes of binding of Ig by SEYF-a populations is investigated. Data are presented on the expression and availability of complex receptors in vivo and their relationship to the anti-tumor antibodies present on the in vivo propagating cells. MATERIAL A N D METHODS
Syngeneic antisera against S E YF-a cells Hyperimmune antisera directed against ascites SEYF-a cells were produced in syngeneic A.BY mice (Witz et al., 1976). The resulting antisera mediated complement-dependent cytotoxicity to the immunizing cells as well as to other murine tumors but not to normal A.BY spleen or lymphnode cells. Following absoprtions with several murine tumors these antisera reacted only with SEYF-a cells.
Quantitation of potentially antibodies
cytotoxic anti-tumor
The quantitation of potentially cytotoxic antitumor antibodies in syngeneic A.BY mice bearing SEYF-a tumors has been described (Ran et al., 1976). Essentially the same method of quantitation has been used in this study. Received: January 27, 1976 and in revised form May 12, 1976. Present address: Roswell Park Memorial Institute, 666 Elm Street, Buffalo, New York, USA.
117
COATING ANTIBODIES AND COMPLEX RECEPTORS
The activity of cell-bound potentially cytotoxic anti-tumor antibodies was expressed as the dilution of rabbit complement required to cause a cytotoxic index (CI) of 0.50 when added to ascites SEYF-a cells, propagated in vivo for different periods of time. The titer of circulating or elutable potentially cytotoxic anti-tumor antibodies was expressed as the dilution of tumor-bearer serum or of low PH eluate from tumor cells (Ran et al., 1976), causing a CI of 0.50 when added to SEYF-a cells, propagated in vivo for 7 days. Binding of soluble immune complexes to SEYF-a cells and its inhibition by an anti-SE YF-a antiserum
The binding capacity of soluble immune complexes by SEYF-a cells was assessed as described previously (Braslawsky et al., 19766). All binding studies were done in Cooke U-shaped microtiter plastic plates (Cooke Engineering Co., USA). Binding of purified antibodies against ovalbumin (PAOA) complexed with ovalbumin (comPAOA) by SEYF-a cells was determined by incubating 10 pl of 1251-labelled comPAOA (containing 0.2 pg labelled antibody protein) to lo6 tumor cells. Background levels were determined by incubating non-complexed 1251-PAOA(0.2 pg in 10 pl) with cells (PAOA control). Cells were incubated for 30 min at 4" C (with occasional shaking) and washed three times with Hank's balanced salt solution (HBSS). Cell-bound radioactivity was determined in a Packard model 5210 auto-gamma scintillation spectrometer. Inhibition of binding was performed by preincubation of SEYF-a cells with a hyperimmune anti-SEYF-a antiserum (see above). Two-fold dilutions (20 111) of antiserum were added to the cells (1 xlOs cells) and the mixture was left for 30 min at 4°C. Cells were washed three times in cold HBSS, and the comPAOA binding determined as described. Inhibition of binding was determined by the formula: A-B Percentage inhibition = -x 1 0 0 A-C where A = percentage of added comPAOA bound to untreated cells, B = percentage of added comPAOA bound to antiserum-treated cells, and C = percentage of added PAOA bound to untreated cells. Sinding of purified anti-mouse IgG by tumor cells
The total immunoglobulin coat of tumor cells was determined by the ability of tumor cells (5 x lo6 per reaction mixture) to bind goat purified antimouse IgG (PaMIgG, Witz et al., 1974).
RESULTS
The relationship between potentially cytotoxic antibodies and surface IgG on in vivo propagating tumor cell populations
The appearance of cytotoxic anti-SEYF-a antibodies in the sera of tumor-bearing mice has been previously reported (Ran et al., 1976). These antibodies were localized on in vivo propagating SEYF-a cells and occupied antigenic determinants on the cell surface. The antibodies were capable of mediating the lysis of tumor cells upon the addition of exogenous complement in vitro (Ran et al., 1976). Sensitivity of SEYF-a cells to the lytic activity of exogeneous complement correlated with propagation time in vivo. It increased during the second week of propagation time and reached peak values at the end of that week (Ran et al., 1976). During the third week of in vivo propagation, some decrease in the sensitivity of SEYF-a cells to complement was noted. This decrease was clearly apparent when the activity of cell-bound, potentially cytotoxic antibodies was evaluated. Figure 1 demonstrates the activity of cell-bound cytotoxic antibodies during the in vivo propagation period of SEYF-a cells and the capacity of the cells to bind 1251-labelledpurified antibodies directed against IgG) PaMIgG). In general there was a correlation between both parameters during the second week of propagation time, although binding of PaMIgG by the cells could be measured before any cell-bound potentially cytotoxic antibodies could be detected. With older tumor cell populations the ability to bind PaMIgG remained constant,and
A
iA '
'
t A
OA
P
A A
A
A
,
5
w
A
A
:
o
o
0
0
, o 10 Days m
15
20
25
30
V~YO
FIGURE 1
The IgG coat and the activity of cell-bound, potentially cytotoxic antibodies on SEYF-a cells propagated in vivo for various periods of time. IgG coat (A) is measured by the ability of cefls to bind purified anti-mouse IgG (PaMIgG) and is represented as percentage of added PaMIgG fixed. The activity of cell-bound potentially cytotoxic antibodies (0)is represented as the dilution of rabbit complement causing CI of 0.50 when added to the cells.
118
BRASLAWSKY ET AL.
high, while the activity of cell-bound cytotoxic antibody on the cell surface, expressed by the concentration (dilution) of complement required to cause a C.I. of 0.50, decreased markedly at the end of the third week and thereafter.
'/I281
..
The question whether the cell bound antibodies had lost their cytotoxic capacity towards old tumor cells or whether they disappeared altogether from the surface of these old cells, was investigated. Tumor cell populations harvested after various periods of propagation time in vivo were exposed to treatment with a low PH buffer. Antibodies elutable from those cells were tested against young (I-week old) SEYF-a cells for their ability to mediate complement-dependent lysis (CdL). The cytotoxic titer of these eluates is demonstrated in Figure 2. Also shown is the titer of circulating antibodies in the tumor-bearing host. The level of elutable cytotoxic antibodies increased throughout the entire propagation time of the tumor in parallel with that of circulating antibodies.
e
. .. .
.
.
me..
... 0
0
I
r
n
1,110.
5
L?.10
0
m
b a n
0
I
I
15
20
Days
111
I Q
25
vive
FIGURE 3 The titer of elutable potentially cytotoxic antibodies ( 0 ) on SEYF-a cells propagated in vivo for various periods (Fig. 2) is superimposed on the titer of cell-bound cytotoxic antibodies (0)present on the cell surface (see Fig. 1).
c
The situation is different when the level of elutable cytotoxic antibodies is compared to the activity of cytotoxic antibodies bound to the cell surface (Fig. 3). In this case the level of both cell-bound and elutable antibody increased during only 2-3 weeks of propagation time of the tumor. I n older cells (19 days and more) the level of elutable cytotoxic antibodies was still increasing while the activity of the potentially cytotoxic antibodies bound to the surface of cells decreased (Fig. 3). One possible explanation of this difference would be a cytophilic binding of the potentially cytotoxic antibodies by older cell populations. Cells coated in such a way might still be able to bind the anti-IgG reagent. This possibility was tested and the results are described in the following section.
a
a
a b
SA
b
a
b
9
D
a a
b
a0
The availability of receptors foi. irnmune-complexed antibody on in vivo-propagated tsmor populations
0
I\
8 m .
5
1
6
I
1
1
10
15
20
25
30
D a y s In vivo
FIGURE 2 The titer of circulating ( A ) and elutable ( 0 ) potentially cytotoxic antibodies in tumor-bearing mice. The titer of antibodies in the sera of tumor-bearing mice and in low PH eluates is represented as the dilution causing a CI of 0.50. Sera and eluates were drawn at various periods after tumor inoculation. The horizontal line indicates mean C1 value of elutable antibody.
In the next set of experiments we tested for the presence of cells capable of binding unrelated immune complexes in freshly harvested SEYF-a tumors. The soluble immune complexes used were composed of ovalbumin (OA) and the corresponding specifically purified antibodies. In each case, the binding of complexed antibody (comPAOA) was higher than that of uncomplexed antibody (PAOA). This confirmed previous findings (Braslawsky ef al., 1976~).
119
COATING ANTIBODIES AND COMPLEX RECEPTORS
U
amounts fixed by cells not subjected to the uncoating procedure. Thus, although sites were availabIe for binding of comPAOA on all tumor populations regardless of age, in older cells some of these sites became occupied in vivo (see also Fig. 4), and as a result of the in vifro incubation at 37" C became available once more for the complex.
J
J
20
10 Days
30
in vivo
FIGURE 4
Complex receptors of SEYF-a cells propagated in vivo for various periods of time. The percentage of added comPAOA fixed by the cells is plotted as a function of propagation time. Individual results are presented.
As shown in Figure 4, the amount of comPAOA bound by young cell populations (7-13 days in vivo) was greater than that bound by the older tumor populations. The mean value for younger cells was significantly higher than that for older cells as determined by Student's t-test (p
TUMOR BOUND IMMUNOGLOBULINS : THE RELATIONSHIP BETWEEN THE IN VIVO COATING OF TUMOR CELLS BY POTENTIALLY CYTOTOXTC ANTI-TUMOR ANTIBODIES, AND THE EXPRESSION OF IMMUNE COMPLEX RECEPTORS Gary BRASLAWSKYMaya RANand Isaac P. Wirz Department of Microbiology, George S. Wise Center for Life Sciences, Tel Aviv University, Tel Aviv, Israel
SUMMARY
In this study we investigated the relationship between binding of anti-tumor antibodies by polyomavirus-induced S E YF-a tumor cells and the expression of immune-complex receptors on these cells. It was shown that in vivo propagated cells became progressively coated with IgG. The increase in the fgG coating of SEYF-a cells, occurring during the second week of propagation, was directly correlated with an increase in the coating of cells with potentially cytotoxic anti-tumor antibodies. The activity of these antibodies was demonstrated by cell lysis following addition of exogenous complement. Cell populations propagated in vivo for longer periods (3 weeks or more) became less sensitive to exogenous complement although their IgC coating remained high, and although higher titers of anti-tumor antibodies could be eluted from them. This indicated that antibodies coating young tumor cell populations have the capacity to activate complement whereas those coating older tumor cell populations are incapable of complement activation. Previous findirigs that S E YF-a cell populations are able to bind unrelated immune complexes were confirmed in the present study. We also found that the capacity of these populations to bind such unrelated complexes decreased with propagation time in vivo. The involvement of anti-tumor antibodies in this phenomenon was indicated by the finding that such antibodies inhibited binding of unrelated immune complexes by young cells but not by old cells. Furthermore, treatments causing dissociation of Ig from the cell surface restored, to some extent, the capacity of anti-tumor antibodies to inhibit immune complex binding by old SE YF-a populations. The presence of a surface immunoglobulin (Ig) coat on some or all of the cells in human or animal tumors propagating in vivo is fairly well established (Witz, 1973). The presence o f anti-tumor antibodies as part of this surface Ig coat on cells of a polyomavirus-induced fibrosarcoma SEYF-a, has been demonstrated (Ran et al., 1976). These cell-bound antibodies were capable of mediating cell lysis upon addition of exogenous complement indicating
that they were attached to the surface in a typical antibody-antigen interaction. Data have also been obtained (Braslawsky et al., 1976a), indicating that SEYF-a tumor cells possess receptors for soluble immune complexes composed of antigens and antibodies clearly unrelated to the tumor cells such as ovalbumin-antiovalbumin. Binding of Ig to the surface of the SEYF-a cell could thus occur either through such complex receptors or by the binding of anti-tumor antibodies to their target antigen. In this paper the relationship existing between these two possible modes of binding of Ig by SEYF-a populations is investigated. Data are presented on the expression and availability of complex receptors in vivo and their relationship to the anti-tumor antibodies present on the in vivo propagating cells. MATERIAL A N D METHODS
Syngeneic antisera against S E YF-a cells Hyperimmune antisera directed against ascites SEYF-a cells were produced in syngeneic A.BY mice (Witz et al., 1976). The resulting antisera mediated complement-dependent cytotoxicity to the immunizing cells as well as to other murine tumors but not to normal A.BY spleen or lymphnode cells. Following absoprtions with several murine tumors these antisera reacted only with SEYF-a cells.
Quantitation of potentially antibodies
cytotoxic anti-tumor
The quantitation of potentially cytotoxic antitumor antibodies in syngeneic A.BY mice bearing SEYF-a tumors has been described (Ran et al., 1976). Essentially the same method of quantitation has been used in this study. Received: January 27, 1976 and in revised form May 12, 1976. Present address: Roswell Park Memorial Institute, 666 Elm Street, Buffalo, New York, USA.
117
COATING ANTIBODIES AND COMPLEX RECEPTORS
The activity of cell-bound potentially cytotoxic anti-tumor antibodies was expressed as the dilution of rabbit complement required to cause a cytotoxic index (CI) of 0.50 when added to ascites SEYF-a cells, propagated in vivo for different periods of time. The titer of circulating or elutable potentially cytotoxic anti-tumor antibodies was expressed as the dilution of tumor-bearer serum or of low PH eluate from tumor cells (Ran et al., 1976), causing a CI of 0.50 when added to SEYF-a cells, propagated in vivo for 7 days. Binding of soluble immune complexes to SEYF-a cells and its inhibition by an anti-SE YF-a antiserum
The binding capacity of soluble immune complexes by SEYF-a cells was assessed as described previously (Braslawsky et al., 19766). All binding studies were done in Cooke U-shaped microtiter plastic plates (Cooke Engineering Co., USA). Binding of purified antibodies against ovalbumin (PAOA) complexed with ovalbumin (comPAOA) by SEYF-a cells was determined by incubating 10 pl of 1251-labelled comPAOA (containing 0.2 pg labelled antibody protein) to lo6 tumor cells. Background levels were determined by incubating non-complexed 1251-PAOA(0.2 pg in 10 pl) with cells (PAOA control). Cells were incubated for 30 min at 4" C (with occasional shaking) and washed three times with Hank's balanced salt solution (HBSS). Cell-bound radioactivity was determined in a Packard model 5210 auto-gamma scintillation spectrometer. Inhibition of binding was performed by preincubation of SEYF-a cells with a hyperimmune anti-SEYF-a antiserum (see above). Two-fold dilutions (20 111) of antiserum were added to the cells (1 xlOs cells) and the mixture was left for 30 min at 4°C. Cells were washed three times in cold HBSS, and the comPAOA binding determined as described. Inhibition of binding was determined by the formula: A-B Percentage inhibition = -x 1 0 0 A-C where A = percentage of added comPAOA bound to untreated cells, B = percentage of added comPAOA bound to antiserum-treated cells, and C = percentage of added PAOA bound to untreated cells. Sinding of purified anti-mouse IgG by tumor cells
The total immunoglobulin coat of tumor cells was determined by the ability of tumor cells (5 x lo6 per reaction mixture) to bind goat purified antimouse IgG (PaMIgG, Witz et al., 1974).
RESULTS
The relationship between potentially cytotoxic antibodies and surface IgG on in vivo propagating tumor cell populations
The appearance of cytotoxic anti-SEYF-a antibodies in the sera of tumor-bearing mice has been previously reported (Ran et al., 1976). These antibodies were localized on in vivo propagating SEYF-a cells and occupied antigenic determinants on the cell surface. The antibodies were capable of mediating the lysis of tumor cells upon the addition of exogenous complement in vitro (Ran et al., 1976). Sensitivity of SEYF-a cells to the lytic activity of exogeneous complement correlated with propagation time in vivo. It increased during the second week of propagation time and reached peak values at the end of that week (Ran et al., 1976). During the third week of in vivo propagation, some decrease in the sensitivity of SEYF-a cells to complement was noted. This decrease was clearly apparent when the activity of cell-bound, potentially cytotoxic antibodies was evaluated. Figure 1 demonstrates the activity of cell-bound cytotoxic antibodies during the in vivo propagation period of SEYF-a cells and the capacity of the cells to bind 1251-labelledpurified antibodies directed against IgG) PaMIgG). In general there was a correlation between both parameters during the second week of propagation time, although binding of PaMIgG by the cells could be measured before any cell-bound potentially cytotoxic antibodies could be detected. With older tumor cell populations the ability to bind PaMIgG remained constant,and
A
iA '
'
t A
OA
P
A A
A
A
,
5
w
A
A
:
o
o
0
0
, o 10 Days m
15
20
25
30
V~YO
FIGURE 1
The IgG coat and the activity of cell-bound, potentially cytotoxic antibodies on SEYF-a cells propagated in vivo for various periods of time. IgG coat (A) is measured by the ability of cefls to bind purified anti-mouse IgG (PaMIgG) and is represented as percentage of added PaMIgG fixed. The activity of cell-bound potentially cytotoxic antibodies (0)is represented as the dilution of rabbit complement causing CI of 0.50 when added to the cells.
118
BRASLAWSKY ET AL.
high, while the activity of cell-bound cytotoxic antibody on the cell surface, expressed by the concentration (dilution) of complement required to cause a C.I. of 0.50, decreased markedly at the end of the third week and thereafter.
'/I281
..
The question whether the cell bound antibodies had lost their cytotoxic capacity towards old tumor cells or whether they disappeared altogether from the surface of these old cells, was investigated. Tumor cell populations harvested after various periods of propagation time in vivo were exposed to treatment with a low PH buffer. Antibodies elutable from those cells were tested against young (I-week old) SEYF-a cells for their ability to mediate complement-dependent lysis (CdL). The cytotoxic titer of these eluates is demonstrated in Figure 2. Also shown is the titer of circulating antibodies in the tumor-bearing host. The level of elutable cytotoxic antibodies increased throughout the entire propagation time of the tumor in parallel with that of circulating antibodies.
e
. .. .
.
.
me..
... 0
0
I
r
n
1,110.
5
L?.10
0
m
b a n
0
I
I
15
20
Days
111
I Q
25
vive
FIGURE 3 The titer of elutable potentially cytotoxic antibodies ( 0 ) on SEYF-a cells propagated in vivo for various periods (Fig. 2) is superimposed on the titer of cell-bound cytotoxic antibodies (0)present on the cell surface (see Fig. 1).
c
The situation is different when the level of elutable cytotoxic antibodies is compared to the activity of cytotoxic antibodies bound to the cell surface (Fig. 3). In this case the level of both cell-bound and elutable antibody increased during only 2-3 weeks of propagation time of the tumor. I n older cells (19 days and more) the level of elutable cytotoxic antibodies was still increasing while the activity of the potentially cytotoxic antibodies bound to the surface of cells decreased (Fig. 3). One possible explanation of this difference would be a cytophilic binding of the potentially cytotoxic antibodies by older cell populations. Cells coated in such a way might still be able to bind the anti-IgG reagent. This possibility was tested and the results are described in the following section.
a
a
a b
SA
b
a
b
9
D
a a
b
a0
The availability of receptors foi. irnmune-complexed antibody on in vivo-propagated tsmor populations
0
I\
8 m .
5
1
6
I
1
1
10
15
20
25
30
D a y s In vivo
FIGURE 2 The titer of circulating ( A ) and elutable ( 0 ) potentially cytotoxic antibodies in tumor-bearing mice. The titer of antibodies in the sera of tumor-bearing mice and in low PH eluates is represented as the dilution causing a CI of 0.50. Sera and eluates were drawn at various periods after tumor inoculation. The horizontal line indicates mean C1 value of elutable antibody.
In the next set of experiments we tested for the presence of cells capable of binding unrelated immune complexes in freshly harvested SEYF-a tumors. The soluble immune complexes used were composed of ovalbumin (OA) and the corresponding specifically purified antibodies. In each case, the binding of complexed antibody (comPAOA) was higher than that of uncomplexed antibody (PAOA). This confirmed previous findings (Braslawsky ef al., 1976~).
119
COATING ANTIBODIES AND COMPLEX RECEPTORS
U
amounts fixed by cells not subjected to the uncoating procedure. Thus, although sites were availabIe for binding of comPAOA on all tumor populations regardless of age, in older cells some of these sites became occupied in vivo (see also Fig. 4), and as a result of the in vifro incubation at 37" C became available once more for the complex.
J
J
20
10 Days
30
in vivo
FIGURE 4
Complex receptors of SEYF-a cells propagated in vivo for various periods of time. The percentage of added comPAOA fixed by the cells is plotted as a function of propagation time. Individual results are presented.
As shown in Figure 4, the amount of comPAOA bound by young cell populations (7-13 days in vivo) was greater than that bound by the older tumor populations. The mean value for younger cells was significantly higher than that for older cells as determined by Student's t-test (p
