Immunology 1991 73 398-406
ADONIS
001928059100172K
Maturation of the antibody response to a protein-coupled form of the organophosphorus toxin soman A. C. BUENAFE & M. B. RITTENBERG Department of Microbiology and Immunology, Oregon Health Sciences University, Portland, Oregon, U.S.A.
Acceptedfor publication I May 1991
SUMMARY
We have analysed the serum antibody response of BALB/c mice to the organophosphorus toxin soman coupled to the protein carrier keyhole limpet haemocyanin (So-KLH) and compared the specificity of the serum antibodies to that of hybridomas described previously. The relative inhibitory capacities of various soman analogues for serum antibodies correlated with those for the monoclonal antibodies. Our results also demonstrate that immune memory to this organophosphorus hapten is stable for greater than 1 year. Interestingly, maturation of the serum antibody response is accompanied by fine specificity changes resulting in increased binding to soman-protein conjugates but not in significant changes in binding to free hapten analogues of soman. This finding suggests that contributions made by the protein carrier or bridge structure, including those made by amino acid side chains involved in the linkage, may play a significant role in the maturation process of antibodies recognizing protein-coupled organophosphorus haptens such as So-KLH. Structurally related but charge-dissimilar organophosphate haptens such as nitrophenylphosphocholine were poorly recognized, even when conjugated to protein with the same diazophenyl linkage used to conjugate soman. This is consistent with maintenance of high specificity in the memory immune response to soman-coupled protein.
acetylcholinesterase for binding to soman in vitro was demonstrated and preliminary in vivo protection studies were shown to extend the survival time of mice exposed to soman.i We have produced 46 murine monoclonal antibodies against somankeyhole limpet haemocyanin (So-KLH)5'6 in order to define the structural elements important in antibody recognition of the soman-protein epitope; the nitrophenyl-soman (NPSo) hapten also has structural features resembling those of nitrophenylphosphocholine, another organophosphorus-containing hapten under study.7 '° We have reported that hybridomas generated at different times during the response to So-KLH display a large degree of heterogeneity in both fine structure recognition and V gene usage.5 6 We have not previously described the maturation ofthe serum antibody response to So-KLH which is the subject of this communication. In addition, we address the question of whether long-term antibody responses can be generated against organophosphorus haptens. Mechanisms directing maturational changes in memory antibody responses to hapten-protein conjugates are known to involve clonal recruitment, somatic mutation and antigen selection9"-'3 but are not yet clearly understood at the level of molecular interaction between antigen and antibody. We are interested in defining these interactions with organophosphoruscontaining ligands. For example, the immune response to phosphocholine-coupled keyhole limpet haemocyanin (PCKLH) is characterized by primary response antibodies which are
INTRODUCTION Since a large number of toxins and insecticides are organophosphorus compounds, the development and study of antibodies against such compounds could provide a sensitive means of detection in the environment or a means of novel therapy in the case of accidental poisoning. Soman (0-1,2,2 trimethylpropylmethylphosphonofluoridate) is an organophosphorus-containing neurotoxin which binds to acetylcholinesterase.' Lenz and co-workers have previously described the production and characterization of antibodies against soman-protein conjugates.24 The ability of one monoclonal antibody to compete with Abbreviations: APSo, p-aminophenyl-soman; DPMP, dipinacolylmethylphosphonofluroidate; ELISA, enzyme-linked immunosorbent assay; mAb, monoclonal antibody; NPDBP, p-nitrophenyl-3,3-dimethylbutylphosphonate; NPEP, p-nitrophenylethylphosphate; NPMPC, pnitrophenylmethylphosphocholine; NPPC, p-nitrophenylphosphocholine; NPSo, p-nitrophenyl-soman; OHSo, hydroxy-soman; PC-KLH, phosphocholine coupled to keyhole limpet haemocyanin; So-BSA, soman coupled to bovine serum albumin; So-KLH, soman coupled to keyhole limpet haemocyanin; soman, 0-1,2,2-trimethylpropylmethylphosphonofluoridate; V gene, variable gene. Correspondence: Dr M. B. Rittenberg, Dept. of Microbiology and Immunology-L220, OHSU, 3181 SW Sam Jackson Pk Rd, Portland, OR 97201, U.S.A.
398
Maturation of the antibody response to the toxin soman highly restricted in V gene usage and fine specificity.9'4 16 We have shown that following secondary challenge there is a dramatic shift to a more heterogeneous pool of antibodies, a large proportion of which has a unique fine specificity and represents the products of many different V gene combinations. 7 10-17-18 So-KLH is similar to PC-KLH in hapten structure and coupling to carrier but differs significantly in charge groups.5'6 We show here the temporal changes that occur in fine specificity
and antigen binding of anti-NPSo serum antibodies as the. immune response to So-KLH proceeds. Our findings are consistent with ongoing selection and expansion of memory Bcell clones specific for the immunogen and suggest that the protein carrier and/or bridge structure including the coupled amino acid side chains may influence this selection process, although use of this common linkage to the carrier (diazophenyl) is not sufficient to permit cross reactivity in the NPSo and NPPC systems. In general, the fine specificities found in the serum antibody population correlate well with those described for individual NPSo-specific hybridomas5'6 and thus are likely to be encoded by mutliple V gene combinations. MATERIALS AND METHODS Animals Female BALB/c mice were obtained from Jackson Laboratories, Bar Harbor, ME, and were first immunized at approximately 6 weeks of age. Immunization and serum collection Mice were immunized, as indicated in Fig. 2, and bled from the periorbital venous sinus on the days reported. Serum samples 200 and equal volumes of serum from each were stored at animal were pooled before testing. -
Quantification of NPSo-specific serum IgM, IgGi and K-bearing antibodies Serum antibodies were quantified by direct binding in an ELISA. Wells of ELISA microtitre plates were coated overnight at room temperature with 2 pg/ml So-BSA in carbonate-bicarbonate buffer, pH 9-6. Plates were blocked with PBS- I% BSA for 1 hr and washed with PBS-0-05 % Tween 20. Serum samples diluted in PBS- 1% BSA were added and incubated for 2 hr. Alkaline phosphatase-labelled rabbit anti-mouse IgM, IgGI or K light chain was added after washing and incubated for 2 hr. Isotypespecific reagents were column passaged against multiple mAb of other isotypes in order to remove cross-reactivity, and retained < 2% cross-reactivity with other isotypes under the conditions used. The amount of IgM, IgGI or IgK present in each serum sample was determined from standard curves using a mixture containing known concentrations of affinity-purified anti-NPSo mAb of various isotypes, as described previously.7 Plates were washed and p-nitrophenyl phosphate (I mg/ml; Sigma, St Louis, MO) in 0-9 M diethanolamine containing I mm MgCl, pH 9-8, was added. Absorbance was read at 405 nm after an approximate 2-hr incubation at room temperature. Measurements were calculated at serum dilutions derived from the linear portion of a titration curve.
Soman analogue and hapten inhibitors Haptens used in fine specificity analysis included the following soman derivatives. p-Amino-phenyl-soman (APSo), p-nitrophe-
399
nyl-soman (NPSo), dipinacolylmethylphosphonate (DPMP) and hydroxy-soman (OH-So) were provided by Dr D. Lenz (USAMRICD, Aberdeen Proving Ground, MD). Other inhibitors included p-nitrophenylphosphocholine (NPPC; Sigma), PCoBSA, So05-OVA, So7-BSA and So596-KLH; the synthesis of soman-protein conjugates has been described previously.5 pNitrophenyl-3,3-dimethylbutylphosphonate (NPDBP), p-nitrophenylmethylphosphocholine (NPMPC) and p-nitrophenylethylphosphate (NPEP) were synthesized by Drs U. Bruderer and J. Fellman (in this laboratory). Figure 1 illustrates the chemical structure of these compounds. Fine specificity analysis The fine specificity of NPSo-specific antibodies was determined in an inhibition ELISA as described previously.5 Percentage inhibition was calculated using a standard curve set up with each assay. The 150 value for each inhibitor is expressed as the mM concentration required for 50% inhibition of anti-NPSo antibody binding to So-BSA.
Hybridomas NPSo-specific hybridomas were generated after immunization with So-KLH and screening with So-BSA, as described elsewhere;5.6 mAb were affinity purified as described elsewhere.5-6 RESULTS Kinetics of the serum antibody response to So-KLH Levels of NPSo-specific IgM and IgGI were measured in serum samples pooled from 13-15 mice at various times after immunization with So-KLH. Mice received a primary injection of 100 pg So-KLH i.p. in complete Freund's adjuvant followed by boosting with 50 pg So-KLH i.p. in incomplete Freund's on Days 14 and 28. As shown in Fig. 2, the primary response consisted mainly of IgM antibodies. The peak IgM response, however, was recorded on Day 21 after a secondary boost. On Days 10 and 14 of the primary response, serum IgG I antibody was approximately 1/3 the level of IgM antibody. After secondary immunization IgGI levels increased rapidly (compare Day 14 to Day 21) and peaked at Day 35 before decreasing gradually; the level of IgG I was three- to fourfold that of IgM on Day 21. Mice were rested for 1 year before memory challenge with So-KLH. At this point, the remaining mice were divided into three groups: G 1 24A (n = 3) received 0-01 pg, G124B (n=4) received 0-1 pg, and G124C (n = 4) received 1 0 pg of So-KLH in saline, intravenously. Figure 2 shows the difference in anti-NPSo serum IgM and IgG I levels in response to boosting with different amounts of So-KLH. The dose-dependent IgG 1 memory response was rapid and vigorous in two (G I 24B and G 1 24C) of the three groups tested, ranging from 15- to 17-fold over that of IgM on memory Day 14, and 8- to 13-fold on memory Day 21. The relatively poor response of G124A animals was most likely due to administration of a suboptimal dose of So-KLH. The development of significant levels (up to 4000 pg/ml) of NPSo-specific IgG I serum antibody by Day 7 after memory boost (compare to primary response Day 7 where the average response was < 10 pg/ml) indicates that efficient priming of the NPSo-specific repertoire had taken place and that long-term memory lasting for at least a year had also been established. In contrast, the levels and kinetics of the IgM response after memory boost did not differ significantly from
400
A. C. Buenafe & M. B. Rittenberg OH
0U CH3 I CH3-C -CH-O-P -0CH3 CH3 CH3
M-N=NI'Protein carrier
Soman-protein PC derivatives
Soman derivatives 0 CH3 CH3-C - CH-O-P-O-Uo -N02 CH3 CH3 CH3 N02-0-So
0 CH3 CH3 - N -CH2-CH2-0-P-PO-(Q)N02 0 CH3
0 CH3 CH3-C - CH- -0P -0- -(NH2 CH3 CH3 CH3
0 CH3 - C - N02 P -P CH3 H3C-CH2-CH2-0 0 IC2-C2-O- -.Q.o
NPPC
- 103-fold better as inhibitors than the free hapten forms tested. This observation is unlikely to be attributable solely to the high valency of soman-protein conjugates. Rather, antibody recognition of hapten may be influenced significantly by the protein carrier or linkage. Consistent with contribution to recognition by the carrier is the observation that the immunogen, So-KLH
401
(hapten density=0 074/1000 MW protein), is 20-30-fold better as an inhibitor of NPSo-specific monoclonal56 and polyclonal serum antibodies than So-BSA (hapten density = 0 1/1000 MW
protein). Among the haptens, inhibition apppeared to correlate with structural homology to the immunogen So-KLH. NPSo and APSo most closely resemble this immunizing form since both bear a phenyl group which represents part of the linkage formed when soman is coupled via a diazophenyl conjugation to tyrosine or histidine residues of the protein carrier (Fig. 1). The hapten NPDBP also contains the phenyl linker group but possesses a negatively charged phosphate, has an additional carbon between the phosphate and the free end and lacks the carbon- I -branched methyl group found in soman analogues. These differences presumably account for the finding that NPDBP does not inhibit as well as APSo and NPSo. The soman analogue DPMP lacks the phenyl structure and is generally a less efficient inhibitor than NPDBP. NPPC is structurally similar to NPDBP except that NPPC has a positively charged nitrogen in place of carbon 3 in the tertiary butyl; NPPC appears to be the least effective inhibitor of the IgGI and IgG2a antibodies. The pattern of inhibition shown in Fig. 4 for Day 35 immune serum antibody (somanprotein > APSo/NPSo > NPDBP > DPMP > NPPC) was demonstrable at all time-points tested, as indicated by the 150 values in Table 1. Antibody maturation in the anti-So-KLH response Figure 5 demonstrates that shifts in fine specificity can be observed in anti-So-KLH IgG populations, but only if assessed by hapten-protein conjugates. Both IgGl and IgG2a serum antibodies showed a significant increase in sensitivity to So-BSA inhibition with time after immunization, as assessed by analysis of variance (Fig. 5, Table 2). In contrast, by the same analysis, inhibition by NPSo or APSo did not appear to change significantly in the IgGl antibody population (Fig. 5, Table 2); likewise, inhibition of IgG2a by NPSo did not change significantly with time (Fig. 5, Table 2). To determine if pooling of serum antibodies may somehow be masking significant changes in hapten inhibition, we assayed serum samples from individual mice for inhibition by So-BSA, APSo and NPSo. As shown in Fig. 6, the majority of individuals displayed an increase in SoBSA inhibition to the order of six- to 12-fold when Day 14 was compared to Day 290. On the other hand, increased inhibition by APSo and NPSo over the same time period (Fig. 6) varied widely among individuals. In summary, analysis of individual serum samples indicates a consistent increase in the binding of a somanprotein conjugate over time which is not observed with free
hapten. Finally, inhibition of NPSo-specific IgG2a or IgG I by NPDBP or NPPC, respectively, became less efficient with time (Fig. 5). These temporal alterations in recognition are consistent with maturation of the NPSo-specific serum antibody response towards recognition of the immunogen So-KLH. In addition, comparison of inhibition values obtained in the late primary/ early secondary response to those measured on comparable days after memory boosting revealed a divergence in fine specificity maturation among IgM and IgGI serum antibodies. Somanprotein inhibition of serum IgM 21 days after primary or memory boosting did not differ significantly, whereas soman-protein inhibition of serum IgG 1 was five- to 10-fold greater in the Day 14
402
A. C. Buenafe & M. B. Rittenberg IgGI
IgG2a
(A) SoKLH (l) SoBSA
100
(M) (0)
APSo NPSo (¢) NPDBP (X) DPMP (Ca) PCBSA (Q) NPPC
80 *c
60
al
40° 20 0
-10
-12
-8
-6
-4
-2
0
Log M hapten
Figure 4. Fine specificity of Day 35 anti-So-KLH
serum
IgG I and IgG2a. See legend to Fig. 3.
Table 1. 150 values (mM) for So-KLH-specific serum antibody
Inhibitor Serum IgM* DPMP NPSo NPDBP NPPC So-BSA So-KLH So-OVA PCBSA KLH Serum IgGI DPMP APSo NPSo NPDBP NPPC So-BSA So-KLH Serum IgG2a DPMP NPSo APSo NPDBP NPPC So-BSA So-KLH
Day 7
Day 14
Day 21
>lot >10 >10 >10 >0 1 0-0000082 0 00009 >0 1 >1
> 10 >10 >10 >10 >0 1 0-00052 0-0066 >0 1 >1
>10 >10 >10 >10 000061 0-0000061 0-000046 >0 1 >1
>10 0-0027 0 033 0 59
9-2 0-0048 0-049 0-46 4-11 0 00053 0-000023
965 0-0046 0-066 2-3 >10 0-00046 0 000011
>10 0-027 0-0067 25 >10 0-00015 0 0000079
>10 0 04 0 0085 8-0 >10
1-39 0-00052 0-000015
6-6 0 019 0-0027 091 99
0-000088 00000057
Day 35
Day 122
>10 >10 >10 >10 0015 0-00001 0-0008 >0 1 >1 >10 0 0043
0-051 2-2 >10 0-00032 0-000014
0-00006
>10 0 057 0.014 >10 >10 0 000055
0-0000024
0 0000019
Day 290 >10 >10 >10 >10 >0-1 0-0037 >0 1 >0 1 >1
>10 0-0027 0-050 2-4 >10 0-00010 0 000004
0 001 0-048 0-95 >10 0-000002
>10
>10
0 05 0 0073
0 034 0-0076
>10 >10 0 000011 0 00000065
>10 >10
0-00002 0 0000012
* Serum samples were collected and equal volumes pooled from 13-15 mice at each time-point represented and assayed for inhibition of anti-soman serum antibody binding to So-BSA-coated plates by ELISA. t Values represent mm concentration of inhibitor required to inhibit 500/ binding in the ELISA.
memory response compared
to Day 14 after primary immuniza-
tion (data not shown). Fine
specificity of
anti-So-KLH monoclonals
Table 3 shows the fine specificity profile of several monoclonals obtained after secondary and tertiary boosting with So-KLH. A comparison of Table 3, which provides the 15o values of various
inhibitors obtained for individual mAb, with Table 1, which provides similar information for the serum antibodies, demonstrates that the fine specificity of the monoclonals correlates well
with what is observed in NPSo-specific serum antibody populations. Although there is significant variation in relative avidity between monoclonals, the overall pattern of specificity appears to be: soman-protein > APSo 2 NPSo 2 NPDBP > NPPC. It is also noteworthy that 3/4 mAb obtained from a later fusion
Maturation of the antibody response to the toxin soman
403
SerLum IgGi SoBSA
NPSo
100 I00
-
80 80
cm .s
6
He
40 20
60
-
40
-
20
0
-'-
-10 -9
-8
-7
-6
-5
0
-4
-1
-7
-6
APSo
cm
.C
-5
-4
-3
-2
-3
-2
-I
NPPC
100
100
80
80
60
60
40
40
20
20
-
C
m
0
-I4
-9
-10
-8
-7
-6
-5
1-7
-4
-6
-5
-4
Serum IgG2a SoBSA
NPDBP
NPSo
100 100
100
-
80 80
-
80s
60
60
60-
40
40
40
20
20
20
C
-0
1
0
.
.,,
,.2i
.,
~o--
-
°04-
-2 -7 -6 -5 -4 -3 -2 -1 Log M hapten Log M hapten Figure 5. Maturation of fine specificity in response to So-KLH. Serum samples were pooled from 13-15 animals and tested in an ELISA for ability of So-BSA, NPSo, APSo, NPDBP, or NPPC to inhibit binding to So-BSA on Day 14 (0), Day 35 (A) and Day 122 (-) after primary immunization. Statistical significance of changes in Iso with time was evaluated by analysis of variance (ANOVA, Table 2).
-1O
-9
-8
-7
-6
-5
-4
-8
-7
(tertiary response) showed no inhibition by NPDBP or NPPC. Also, 10/13 IgM mAb were inhibited by soman-protein only, and were not inhibited by any free soman analogue tested.6 The analysis of hybridomas obtained after So-KLH priming and boosting is therefore likely to be representative of the in vivo SoKLH response.
DISCUSSION
Analysis of the
serum response to So-KLH demonstrates that true memory to this compound can be established and that
specificity is maintained throughout this response. Interestingly, inhibition of pooled IgGI serum antibody with the free haptens NPSo and APSo did not appear to vary significantly throughout the entire response to So-KLH. Among individuals, inhibition of IgG by free soman hapten was variable. The overall lack of maturation in relative affinity for NPSo/APSo suggests that
-6 -5 -4 Log M hapten
-3
carrier/linker influences or specific contacts may play an important selective role in maturation of the anti-So-KLH response. Such carrier- or linker-mediated effects on antibody maturation have been suggested by others.'920 However, it is important to note that no overall decrease over time was apparent in the ability of NPSo/APSo to inhibit serum antibody; these NPSo/APSo-binding antibodies were therefore maintained as part of the NPSo-specific antibody pool but apparently had not been subjected to a strong selective pressure that would have allowed for the expansion of higher avidity clones. This is in contrast to the observed temporal decrease in inhibition by NPPC and NPDBP, suggesting that these structures are less inhibitory because they lack the hydrophobic features of NPSo, which may play a dominant role in the selective process. The priming regimen utilizing a protein carrier led to a memory immune response which persisted for greater than 1 year, although significant affinity maturation was demonstrable
A. C. Buenafe & M. B. Rittenberg
404
Table 2. Summary table for analysis of variance (ANOVA) of IgG serum antibody inhibition at different time-points during the response to So-KLHt Serum Ab + inhibitor
IgG I + So-BSA
IgGI +NPSo
IgG 1 + APSo
IgG2a + So-BSA
IgG2a+NPSo
Source of variance
SST
dft
MST
F
Total Between groups Within groups Treatments Error
0-0000004 00000001 0-0000003 0-0000003 0-00000003
8 2 6 2 4
0-0000001 0-000000009
15.6**
Total Between groups Within groups Treatments Error
0-00133 0-00118 0-00016 0 00007 0 00009
8 2 6 2 4
0-00003 0-00002
Total Between groups Within groups Treatments Error
0-000002 0-0000006 0 000001 0-0000006 0 0000008
11 3 8 2 6
0 0000003 0 0000001
Total Between groups Within groups Treatments Error
0-0000001 0-00000002 0 0000001 0 0000001 0 00000002
11 3 8 2 6
0-00000005 0 000000003
Total Between groups Within groups Treatments Error
0-00877 0-00739 0 00138 0-00061 0-00077
11 3 8 2 6
0-0003 0 00013
1 4§
2-2§
16*
2 4§
t Analysis of variance29 was performed for 1so values obtained by So-BSA and NPSo inhibition of serum IgG at Days 14, 35 and 122 after primary immunization (Fig. 5). 1 SS, sum of squares; df, degrees of freedom; MS, mean square. §Not significant at a=0 1. * Significant at x=0-01; ** significant at a =005. SoBSA
APSo
NPSo v - -4 I* et
Mice no.
(o) I
(*) 3* b0-S
l005
to-?
10-60
10 -8
10?7
(U) 4 (°) 5 (a) 7* (a) 8 (A) 11 ( A) I2 (U) 14
2
E 0 S
( t) 15* 14
290
290
14
14
290
(days) Figure 6. Comparison of serum inhibition values for So-BSA, APSo and NPSo at an early time-point (Day 14) versus a late time-point (Day 290) in individual mice. For some mice (*) it was necessary to use serum collected on Day 17 rather than on Day 14. Time
against soman-protein conjugates only. The lack of improvement in binding free soman hapten analogues might be resolved by using non-immunogenic linkers which might elicit a maturational antibody response to the hapten structure alone. The extreme toxicity of soman precludes its general use for screening
purposes or as an inhibitor in the ELISA, although the toxin has been used in competitive inhibition assays performed by others under special laboratory conditions.2 3 The use of haptenprotein conjugates to detect hapten-specific antibodies is commonly used as a primary screening procedure9l" 13 and soman is
Maturation of the antibody response to the toxin soman
405
Table 3. Fine specificity profiles of secondary and tertiary response anti-So-KLH hybridomas Inhibitors So-BSA
So-KLH
3-8 8-0 1-07 7-73
0-00000008 0-0000005 0-0000002 0-0000003
0-000000003 0 00000006 0 000000008 0-00000004
2-18 1-05 2-15 0-58
1-0 0-9 0-5 > 10
0-0000001 0-0000003 0-00000001 0-000000008
0-00000001 0-00000002 0-000000009 0-00000003
ND
3-29
>10 >10 > 10
> 10
ND > 10
>10 4 03
>10 > 10
0 00000004 0-00000001 0-00000001 0-000000004
ND 0-000000004 0-000000001 0-0000000009
NPPC
NPMPC
NPEP
Secondary response IgG1 hybridomas 0-388 0.009* 0-26 SoGl-3 0-064 0-0003 0-002 SoGl-5 0 053 0-13 0-024 SoGl-9 0-365 0-0003 0-26 SoGI-13
> 10 9-9 > 10 > 10
> 10 > 10 > 10 > 10
> 10 2-98 1-57 3-47
Secondary response IgG2a hybridomas 0-018 0-015 0 001 SoG2a-3 0-014 0-0004 0-02 SoG2a-4 0-025 0-0009 0-026 SoG2a-6 0-129 0-18 0 04 SoG2a-7
1-02 0-58 0-75 3-76
0-63 0-63 0-88 > 10
Tertiary response hybridomas 0-0004 0-002 SoGl-1 0-033 0-033 SoG2a-1 0-0008 0-018 SoG2a-2 0 035 0-02 SoG2b-1
5 64 > 10 >10 > 10
APSo
NPSo
NPDBP
0-151 >10 > 10 > 10
OHSo
* Values represent mm concentration of inhibitor required to inhibit 50°/. binding to So-BSA-coated plates in ELISA. ND, not done.
inactivated as a toxin once it is coupled to the protein carrier. It remains possible, however, that the pattern of fine specificity maturation observed reflects the use of soman-protein as the screening antigen. In terms of fine structure recognition it appears that the phenyl linkage is important in the binding of NPSo-specific IgG serum antibody; this is consistent with what has been found for NPSo-specific mAb3-6 and polyclonal rabbit antisera.3 The generation of antibodies with specificity including recognition of linkage structures has been described for several hapten-protein conjugates.2"-23 The presence of the phenyl structure is also required for recognition by memory Group II antibodies generated against phenylphosphocholine conjugates.7'0 Antibodies produced in response to So-KLH resemble those generated in the memory response to PC-KLH in the diversity of fine specificity and V gene usage. However, in the anti-So-KLH response this molecular heterogeneity is observed as early as the primary response.6 The inability of PC-BSA and its hapten analogue NPPC to inhibit the binding of anti-NPSo antibodies to soman-protein illustrates the specificity of this response for a soman-protein epitope. Since the phenyl-phosphorus linkage is found in both PC-protein and soman-protein, the positively charged choline moiety of PC may play a significant role in preventing its recognition by NPSo-specific antibodies. In addition, we observed partial inhibition of NPSo-specific IgG antibody by NPDBP, which lacks this positive charge. The presence of the negatively charged phosphate in PC-BSA and NPPC may be less significant in preventing cross-reactivity since NPDBP also bears a negatively charged phosphate (Fig. 4, Table 1). Also, NPMPC, a methylphosphonate analogue of NPPC lacking this negative charge, was a very poor inhibitor of NPSo-specific IgG monoclonal antibodies (Table 3, and data not shown). The relative affinity, kinetics and levels of IgM antibody produced after memory boosting were similar to those of the
primary IgM response; this suggests that the majority of IgM produced to repeated challenge with So-KLH represents recruitment of new primary B cells into the ongoing response even after the memory response was well-established, as seen in other systems.24 Whether primary response IgM and secondary response IgG antibodies derive from the same B-cell population or whether they represent distinct B-cell populations is not known; it has recently been suggested that memory B cells may derive from a separate subpopulation.25 IgM-producing B-cell clones are commonly associated with the early response to protein antigens and it has been suggested that they often express , Ittiis therefore possible that the germline V gene sequences. 24.26.27t1 soman-protein-specific IgM antibodies detected here are products of newly stimulated B cells which have not yet been driven to mutate nor subjected to selective pressures, consequently demonstrating poor affinity for the antigen. On the other hand, IgG 1 serum antibodies produced after final boosting are likely to represent products of an antigen-selected memory B-cell population.' 1.13.28 We have found VH gene usage to be heterogeneous among IgM and IgG anti-NPSo hybridomas,6 although the J558 VH family appeared to be over-represented in both populations; members of the 36-60 and 7183 VH families were also present among both IgM and IgG hybridomas. Thus, although the individual V genes have not been identified, the results are consistent with the hypothesis that the IgG population arose selectively from antigen-specific IgM-producing B cells. Our studies show that a fine specificity profile representative of that observed in NPSo-specific hybridomas can be demonstrated among polyclonal serum antibodies produced in vivo. This suggests that in vitro culturing conditions have not altered this repertoire, whereas such a selection has been reported to take place during the in vitro culturing of some T-cell clones.29 The results also suggest that anti-NPSo antibodies produced in vito may be the products of a heterogeneous V gene population, as was found among NPSo-specific hybridomas. Since some of the
406
A. C. Buenafe & M. B. Rittenberg
V gene sequences encoding So-KLH-specific combining sites have recently been characterized,30 it may be possible to use these sequences together with current protein-ligand computer modelling techniques to design high affinity combining sites that would interact with the soman neurotoxin. Finally, the apparent lack of a restricted B-cell response at any point in the So-KLH response may reflect the absence of idiotypic regulation or evolutionary selection for an environmentally common structure such as appears to be the case in the response to PC.3233 Thus, the differences in molecular heterogeneity that characterize the early phases of the antibody responses to these two organophosphorus haptens, soman and PC, may reflect the fact that only one of them, soman, is new to the environment.
ACKNOWLEDGMENTS We wish to thank Drs U. Bruderer, J. Fellman, and D. Lenz for providing the haptens used in this study and M. Fuller for technical assistance. We also thank Dr C. Chen and M. Nutt for helpful discussion and critical review of this manuscript. This work was supported by National Institutes of Health Grants Al 14985 and Al 26827, and by Contract DAMD-17-83-C-3246 from the U.S. Army Medical Research and Development Command.
REFERENCES 1. KOELLE G.B. (1981) Organophosphate poisoning-an overview. Fund. Appl. Toxicology, 1, 129. 2. HUNTER K.W., LENz D.E., BRIMFIELD A.A. & NAYLOR J.A. (1982) Quantification of the organophosphorus nerve agent soman by competitive inhibition enzyme immunoassay using monoclonal antibody. FEBS Lett. 149, 147. 3. LENz D.E., BRIMFIELD A.A., HUNTER K.W., BENSCHOP H.P., DEJONG L.P.A., VAN DIJK C. & CLOW T.R. (1984) Studies using a monoclonal antibody against Soman. Fund. Appl. Toxic. 4, 5156. 4. BRIMFIELD A.A., HUNTER K.W., LENz D.E., BENSCHOP H.P., VAN DIJK C. & DEJONG L.P.A. (1985) Structural and stereochemical specificity of mouse monoclonal antibodies to the organophosphorous cholinesterase inhibitor soman. Molec. Pharmacol. 28, 32. 5. BUENAFE A.C. & RITTENBERG M.B. (1987) Combining site specificity of monoclonal antibodies to the organophosphate hapten Soman. Molec. Immunol. 24, 401. 6. BUENAFE A.C., MAKOWSKI F.F. & RITTENBERG M.B. (1989) Molecular analysis and fine specificity of antibodies against an organophosphorus hapten. J. Immunol. 143, 539. 7. CHANG S.P., BROWN M. & RITTENBERG M.B. (1982) Immunologic memory to phosphorylcholine II. PC-KLH induces two antibody populations that dominate different isotypes. J. Immunol. 128, 702. 8. CHANG S.P., BROWN M. & RITTENBERG M.B. (1982) Immunologic memory to phosphorylcholine III. IgM includes a fine specificity population distinct from TEPC 15. J. Immunol. 129,1559. 9. STENZEL-POORE M.P., BRUDERER U. & RITTENBERG M. B. (1988) The adaptive potential of the memory response: Clonal recruitment and epitope recognition. Immunol. Rev. 105, 113. 10. BRUDERER U., STENzEL-POORE M.P., BACHINGER H.P., FELLMAN J.H. & RITTENBERG M.B. (1989) Antibody combining site heterogeneity within the response to phosphocholine-keyhole limpet hemocyanin. Molec. Immunol. 26, 63. 11. WYSOCKI, L., MANSER T. & GEFTER M. (1986) Somatic evolution of variable region structures during an immune response. PNAS 83, 1847.
12. BEREK C. & MILSTEIN C. (1987) Mutation drift and repertoire shift in the maturation of the immune response. Immunol. Rev. 96, 23. 13. SIEKEVITZ M., KOCKS C., RAJEWSKY K. & DILDROP R. (1987) Analysis of mutation and class switching in naive and memory B cells generating adoptive primary and secondary responses. Cell, 48, 757. 14. GEARHART P.J., SIGAL N.H. & KLINMAN N.R. (1975) Heterogeneity of the BALB/c antiphosphorylcholine antibody repsonse at the precursor cell level. J. exp. Med. 141, 56. 15. QUINTANS J. & COZENZA H. (1976) Antibody response to phosphorylcholine in vitro II. Analysis of T-dependent and T-independent responses. Eur. J. Immunol. 6, 399. 16. PERLMUTTER R.M., CREWS S.T., DOUGLAS R., SORENSEN G., JOHNSON N., NIVERA N., GEARHART P.J. & HOOD L. (1984) The generation of diversity in phosphorylcholine-binding antibodies. Adv. Immunol. 35, 1. 17. TODD I., BROWN M. & RITTENBERG M.B. (1985) Immunologic memory to phosphocholine. VI. Heterogeneity in light chain gene expression. Eur. J. Immunol. 15, 177. 18. STENZEL-POORE M.P. & RITTENBERG M.B. (1989) Clonal diversity, somatic mutation, and immune memory to phosphocholine-keyhole limpet hemocyanin. J. Immunol. 143,4123. 19. FISH S. & MANSER T. (1987) Influence ofthe macromolecular form of a B cell epitope on the expression of antibody variable and constant region structure. J. exp. Med. 166, 711. 20. ALLEN D., SIMON T., SABLITZKY F., RAJEWSKY K. & CUMANO A. (1988) Antibody engineering for the analysis of affinity maturation of an anti-hapten response. EMBO J. 7, 1995. 21. RODWELL J.D. & KARUSH F. (1980) Restriction in IgM expression. I. The VH regions of equine anti-lactose antibodies. Molec. Immunol. 17, 1553. 22. HOPPNER W., FISHER K., POSCHMANN A. & PAULSEN H. (1985) Study on the carbohydrate specificity of antibodies formed in rabbits to synthetic glycoproteins with the carbohydrate structure of asialoglycophorin A. Molec. Immunol. 22, 1341. 23. CHOUCHANE L., STROSBERG A.D. & HOEBEKE J. (1988) Stereospecific immuno-recognition of the tetracyclic anti-depressant oxaprotiline. Molec. Immunol. 25, 1299. 24. BEREK C., JARVISJ.M. & MILSTEIN C. (1987) Activation of memory and virgin B cell clones in hyperimmune animals. Eur. J. Immunol. 17, 1121. 25. KLINMAN N.R. & LINTON P.J. (1990) The generation of B cell memory: a working hypothesis. Contemp. Topics Microbiol. Immunol. 159, 19. 26. KARUSH F. (1978) The affinity of antibody: range, variability and the role of multivalence. In: Immunoglobulins (eds G. Litman and R. A. Good), Vol. 5, p. 85 Plenum Press, New York. 27. RODWELL J.D., GEARHART P.J. & KARUSH F. (1983) Restriction in IgM expression. IV. Affinity analysis of monoclonal antiphosphorylcholine antibodies. J. Immunol. 130, 313. 28. EISEN H. & SISKIND G. (1964) Variations in affinities of antibodies during the immune response. Biochemistry, 3, 996. 29. GAMMON G., KLOTZ J., ANDO D. & SERCARZ E. (1990) The T cell repertoire to a multideterminant antigen. Clonal heterogeneity ofthe T cell response, variation between syngeneic individuals, and in vitro selection of T cell specificities. J. Immunol. 144, 1571. 30. BUENAFE A.C. (1989) Specificity of antibodies binding organophosphorus haptens. PhD thesis, Oregon Health Sciences University. 31. PHILLIPS D.S. (1978) Basic Statisticsfor Health Science Students. W. H. Freeman, New York. 32. BOTTOMLY, K. (1984) 1984: All idiotypes are equal, but some are more equal than others. Immunol. Rev. 79, 45. 33. BRILES D.E., FORMAN C., HUDAK S. & CLAFLIN J.L. (1982) Antiphosphorylcholine antibodies of the T15 idiotype are optimally protective against Streptococcuspneumoniae. J. exp. Med. 156, 1177.
ADONIS
001928059100172K
Maturation of the antibody response to a protein-coupled form of the organophosphorus toxin soman A. C. BUENAFE & M. B. RITTENBERG Department of Microbiology and Immunology, Oregon Health Sciences University, Portland, Oregon, U.S.A.
Acceptedfor publication I May 1991
SUMMARY
We have analysed the serum antibody response of BALB/c mice to the organophosphorus toxin soman coupled to the protein carrier keyhole limpet haemocyanin (So-KLH) and compared the specificity of the serum antibodies to that of hybridomas described previously. The relative inhibitory capacities of various soman analogues for serum antibodies correlated with those for the monoclonal antibodies. Our results also demonstrate that immune memory to this organophosphorus hapten is stable for greater than 1 year. Interestingly, maturation of the serum antibody response is accompanied by fine specificity changes resulting in increased binding to soman-protein conjugates but not in significant changes in binding to free hapten analogues of soman. This finding suggests that contributions made by the protein carrier or bridge structure, including those made by amino acid side chains involved in the linkage, may play a significant role in the maturation process of antibodies recognizing protein-coupled organophosphorus haptens such as So-KLH. Structurally related but charge-dissimilar organophosphate haptens such as nitrophenylphosphocholine were poorly recognized, even when conjugated to protein with the same diazophenyl linkage used to conjugate soman. This is consistent with maintenance of high specificity in the memory immune response to soman-coupled protein.
acetylcholinesterase for binding to soman in vitro was demonstrated and preliminary in vivo protection studies were shown to extend the survival time of mice exposed to soman.i We have produced 46 murine monoclonal antibodies against somankeyhole limpet haemocyanin (So-KLH)5'6 in order to define the structural elements important in antibody recognition of the soman-protein epitope; the nitrophenyl-soman (NPSo) hapten also has structural features resembling those of nitrophenylphosphocholine, another organophosphorus-containing hapten under study.7 '° We have reported that hybridomas generated at different times during the response to So-KLH display a large degree of heterogeneity in both fine structure recognition and V gene usage.5 6 We have not previously described the maturation ofthe serum antibody response to So-KLH which is the subject of this communication. In addition, we address the question of whether long-term antibody responses can be generated against organophosphorus haptens. Mechanisms directing maturational changes in memory antibody responses to hapten-protein conjugates are known to involve clonal recruitment, somatic mutation and antigen selection9"-'3 but are not yet clearly understood at the level of molecular interaction between antigen and antibody. We are interested in defining these interactions with organophosphoruscontaining ligands. For example, the immune response to phosphocholine-coupled keyhole limpet haemocyanin (PCKLH) is characterized by primary response antibodies which are
INTRODUCTION Since a large number of toxins and insecticides are organophosphorus compounds, the development and study of antibodies against such compounds could provide a sensitive means of detection in the environment or a means of novel therapy in the case of accidental poisoning. Soman (0-1,2,2 trimethylpropylmethylphosphonofluoridate) is an organophosphorus-containing neurotoxin which binds to acetylcholinesterase.' Lenz and co-workers have previously described the production and characterization of antibodies against soman-protein conjugates.24 The ability of one monoclonal antibody to compete with Abbreviations: APSo, p-aminophenyl-soman; DPMP, dipinacolylmethylphosphonofluroidate; ELISA, enzyme-linked immunosorbent assay; mAb, monoclonal antibody; NPDBP, p-nitrophenyl-3,3-dimethylbutylphosphonate; NPEP, p-nitrophenylethylphosphate; NPMPC, pnitrophenylmethylphosphocholine; NPPC, p-nitrophenylphosphocholine; NPSo, p-nitrophenyl-soman; OHSo, hydroxy-soman; PC-KLH, phosphocholine coupled to keyhole limpet haemocyanin; So-BSA, soman coupled to bovine serum albumin; So-KLH, soman coupled to keyhole limpet haemocyanin; soman, 0-1,2,2-trimethylpropylmethylphosphonofluoridate; V gene, variable gene. Correspondence: Dr M. B. Rittenberg, Dept. of Microbiology and Immunology-L220, OHSU, 3181 SW Sam Jackson Pk Rd, Portland, OR 97201, U.S.A.
398
Maturation of the antibody response to the toxin soman highly restricted in V gene usage and fine specificity.9'4 16 We have shown that following secondary challenge there is a dramatic shift to a more heterogeneous pool of antibodies, a large proportion of which has a unique fine specificity and represents the products of many different V gene combinations. 7 10-17-18 So-KLH is similar to PC-KLH in hapten structure and coupling to carrier but differs significantly in charge groups.5'6 We show here the temporal changes that occur in fine specificity
and antigen binding of anti-NPSo serum antibodies as the. immune response to So-KLH proceeds. Our findings are consistent with ongoing selection and expansion of memory Bcell clones specific for the immunogen and suggest that the protein carrier and/or bridge structure including the coupled amino acid side chains may influence this selection process, although use of this common linkage to the carrier (diazophenyl) is not sufficient to permit cross reactivity in the NPSo and NPPC systems. In general, the fine specificities found in the serum antibody population correlate well with those described for individual NPSo-specific hybridomas5'6 and thus are likely to be encoded by mutliple V gene combinations. MATERIALS AND METHODS Animals Female BALB/c mice were obtained from Jackson Laboratories, Bar Harbor, ME, and were first immunized at approximately 6 weeks of age. Immunization and serum collection Mice were immunized, as indicated in Fig. 2, and bled from the periorbital venous sinus on the days reported. Serum samples 200 and equal volumes of serum from each were stored at animal were pooled before testing. -
Quantification of NPSo-specific serum IgM, IgGi and K-bearing antibodies Serum antibodies were quantified by direct binding in an ELISA. Wells of ELISA microtitre plates were coated overnight at room temperature with 2 pg/ml So-BSA in carbonate-bicarbonate buffer, pH 9-6. Plates were blocked with PBS- I% BSA for 1 hr and washed with PBS-0-05 % Tween 20. Serum samples diluted in PBS- 1% BSA were added and incubated for 2 hr. Alkaline phosphatase-labelled rabbit anti-mouse IgM, IgGI or K light chain was added after washing and incubated for 2 hr. Isotypespecific reagents were column passaged against multiple mAb of other isotypes in order to remove cross-reactivity, and retained < 2% cross-reactivity with other isotypes under the conditions used. The amount of IgM, IgGI or IgK present in each serum sample was determined from standard curves using a mixture containing known concentrations of affinity-purified anti-NPSo mAb of various isotypes, as described previously.7 Plates were washed and p-nitrophenyl phosphate (I mg/ml; Sigma, St Louis, MO) in 0-9 M diethanolamine containing I mm MgCl, pH 9-8, was added. Absorbance was read at 405 nm after an approximate 2-hr incubation at room temperature. Measurements were calculated at serum dilutions derived from the linear portion of a titration curve.
Soman analogue and hapten inhibitors Haptens used in fine specificity analysis included the following soman derivatives. p-Amino-phenyl-soman (APSo), p-nitrophe-
399
nyl-soman (NPSo), dipinacolylmethylphosphonate (DPMP) and hydroxy-soman (OH-So) were provided by Dr D. Lenz (USAMRICD, Aberdeen Proving Ground, MD). Other inhibitors included p-nitrophenylphosphocholine (NPPC; Sigma), PCoBSA, So05-OVA, So7-BSA and So596-KLH; the synthesis of soman-protein conjugates has been described previously.5 pNitrophenyl-3,3-dimethylbutylphosphonate (NPDBP), p-nitrophenylmethylphosphocholine (NPMPC) and p-nitrophenylethylphosphate (NPEP) were synthesized by Drs U. Bruderer and J. Fellman (in this laboratory). Figure 1 illustrates the chemical structure of these compounds. Fine specificity analysis The fine specificity of NPSo-specific antibodies was determined in an inhibition ELISA as described previously.5 Percentage inhibition was calculated using a standard curve set up with each assay. The 150 value for each inhibitor is expressed as the mM concentration required for 50% inhibition of anti-NPSo antibody binding to So-BSA.
Hybridomas NPSo-specific hybridomas were generated after immunization with So-KLH and screening with So-BSA, as described elsewhere;5.6 mAb were affinity purified as described elsewhere.5-6 RESULTS Kinetics of the serum antibody response to So-KLH Levels of NPSo-specific IgM and IgGI were measured in serum samples pooled from 13-15 mice at various times after immunization with So-KLH. Mice received a primary injection of 100 pg So-KLH i.p. in complete Freund's adjuvant followed by boosting with 50 pg So-KLH i.p. in incomplete Freund's on Days 14 and 28. As shown in Fig. 2, the primary response consisted mainly of IgM antibodies. The peak IgM response, however, was recorded on Day 21 after a secondary boost. On Days 10 and 14 of the primary response, serum IgG I antibody was approximately 1/3 the level of IgM antibody. After secondary immunization IgGI levels increased rapidly (compare Day 14 to Day 21) and peaked at Day 35 before decreasing gradually; the level of IgG I was three- to fourfold that of IgM on Day 21. Mice were rested for 1 year before memory challenge with So-KLH. At this point, the remaining mice were divided into three groups: G 1 24A (n = 3) received 0-01 pg, G124B (n=4) received 0-1 pg, and G124C (n = 4) received 1 0 pg of So-KLH in saline, intravenously. Figure 2 shows the difference in anti-NPSo serum IgM and IgG I levels in response to boosting with different amounts of So-KLH. The dose-dependent IgG 1 memory response was rapid and vigorous in two (G I 24B and G 1 24C) of the three groups tested, ranging from 15- to 17-fold over that of IgM on memory Day 14, and 8- to 13-fold on memory Day 21. The relatively poor response of G124A animals was most likely due to administration of a suboptimal dose of So-KLH. The development of significant levels (up to 4000 pg/ml) of NPSo-specific IgG I serum antibody by Day 7 after memory boost (compare to primary response Day 7 where the average response was < 10 pg/ml) indicates that efficient priming of the NPSo-specific repertoire had taken place and that long-term memory lasting for at least a year had also been established. In contrast, the levels and kinetics of the IgM response after memory boost did not differ significantly from
400
A. C. Buenafe & M. B. Rittenberg OH
0U CH3 I CH3-C -CH-O-P -0CH3 CH3 CH3
M-N=NI'Protein carrier
Soman-protein PC derivatives
Soman derivatives 0 CH3 CH3-C - CH-O-P-O-Uo -N02 CH3 CH3 CH3 N02-0-So
0 CH3 CH3 - N -CH2-CH2-0-P-PO-(Q)N02 0 CH3
0 CH3 CH3-C - CH- -0P -0- -(NH2 CH3 CH3 CH3
0 CH3 - C - N02 P -P CH3 H3C-CH2-CH2-0 0 IC2-C2-O- -.Q.o
NPPC
- 103-fold better as inhibitors than the free hapten forms tested. This observation is unlikely to be attributable solely to the high valency of soman-protein conjugates. Rather, antibody recognition of hapten may be influenced significantly by the protein carrier or linkage. Consistent with contribution to recognition by the carrier is the observation that the immunogen, So-KLH
401
(hapten density=0 074/1000 MW protein), is 20-30-fold better as an inhibitor of NPSo-specific monoclonal56 and polyclonal serum antibodies than So-BSA (hapten density = 0 1/1000 MW
protein). Among the haptens, inhibition apppeared to correlate with structural homology to the immunogen So-KLH. NPSo and APSo most closely resemble this immunizing form since both bear a phenyl group which represents part of the linkage formed when soman is coupled via a diazophenyl conjugation to tyrosine or histidine residues of the protein carrier (Fig. 1). The hapten NPDBP also contains the phenyl linker group but possesses a negatively charged phosphate, has an additional carbon between the phosphate and the free end and lacks the carbon- I -branched methyl group found in soman analogues. These differences presumably account for the finding that NPDBP does not inhibit as well as APSo and NPSo. The soman analogue DPMP lacks the phenyl structure and is generally a less efficient inhibitor than NPDBP. NPPC is structurally similar to NPDBP except that NPPC has a positively charged nitrogen in place of carbon 3 in the tertiary butyl; NPPC appears to be the least effective inhibitor of the IgGI and IgG2a antibodies. The pattern of inhibition shown in Fig. 4 for Day 35 immune serum antibody (somanprotein > APSo/NPSo > NPDBP > DPMP > NPPC) was demonstrable at all time-points tested, as indicated by the 150 values in Table 1. Antibody maturation in the anti-So-KLH response Figure 5 demonstrates that shifts in fine specificity can be observed in anti-So-KLH IgG populations, but only if assessed by hapten-protein conjugates. Both IgGl and IgG2a serum antibodies showed a significant increase in sensitivity to So-BSA inhibition with time after immunization, as assessed by analysis of variance (Fig. 5, Table 2). In contrast, by the same analysis, inhibition by NPSo or APSo did not appear to change significantly in the IgGl antibody population (Fig. 5, Table 2); likewise, inhibition of IgG2a by NPSo did not change significantly with time (Fig. 5, Table 2). To determine if pooling of serum antibodies may somehow be masking significant changes in hapten inhibition, we assayed serum samples from individual mice for inhibition by So-BSA, APSo and NPSo. As shown in Fig. 6, the majority of individuals displayed an increase in SoBSA inhibition to the order of six- to 12-fold when Day 14 was compared to Day 290. On the other hand, increased inhibition by APSo and NPSo over the same time period (Fig. 6) varied widely among individuals. In summary, analysis of individual serum samples indicates a consistent increase in the binding of a somanprotein conjugate over time which is not observed with free
hapten. Finally, inhibition of NPSo-specific IgG2a or IgG I by NPDBP or NPPC, respectively, became less efficient with time (Fig. 5). These temporal alterations in recognition are consistent with maturation of the NPSo-specific serum antibody response towards recognition of the immunogen So-KLH. In addition, comparison of inhibition values obtained in the late primary/ early secondary response to those measured on comparable days after memory boosting revealed a divergence in fine specificity maturation among IgM and IgGI serum antibodies. Somanprotein inhibition of serum IgM 21 days after primary or memory boosting did not differ significantly, whereas soman-protein inhibition of serum IgG 1 was five- to 10-fold greater in the Day 14
402
A. C. Buenafe & M. B. Rittenberg IgGI
IgG2a
(A) SoKLH (l) SoBSA
100
(M) (0)
APSo NPSo (¢) NPDBP (X) DPMP (Ca) PCBSA (Q) NPPC
80 *c
60
al
40° 20 0
-10
-12
-8
-6
-4
-2
0
Log M hapten
Figure 4. Fine specificity of Day 35 anti-So-KLH
serum
IgG I and IgG2a. See legend to Fig. 3.
Table 1. 150 values (mM) for So-KLH-specific serum antibody
Inhibitor Serum IgM* DPMP NPSo NPDBP NPPC So-BSA So-KLH So-OVA PCBSA KLH Serum IgGI DPMP APSo NPSo NPDBP NPPC So-BSA So-KLH Serum IgG2a DPMP NPSo APSo NPDBP NPPC So-BSA So-KLH
Day 7
Day 14
Day 21
>lot >10 >10 >10 >0 1 0-0000082 0 00009 >0 1 >1
> 10 >10 >10 >10 >0 1 0-00052 0-0066 >0 1 >1
>10 >10 >10 >10 000061 0-0000061 0-000046 >0 1 >1
>10 0-0027 0 033 0 59
9-2 0-0048 0-049 0-46 4-11 0 00053 0-000023
965 0-0046 0-066 2-3 >10 0-00046 0 000011
>10 0-027 0-0067 25 >10 0-00015 0 0000079
>10 0 04 0 0085 8-0 >10
1-39 0-00052 0-000015
6-6 0 019 0-0027 091 99
0-000088 00000057
Day 35
Day 122
>10 >10 >10 >10 0015 0-00001 0-0008 >0 1 >1 >10 0 0043
0-051 2-2 >10 0-00032 0-000014
0-00006
>10 0 057 0.014 >10 >10 0 000055
0-0000024
0 0000019
Day 290 >10 >10 >10 >10 >0-1 0-0037 >0 1 >0 1 >1
>10 0-0027 0-050 2-4 >10 0-00010 0 000004
0 001 0-048 0-95 >10 0-000002
>10
>10
0 05 0 0073
0 034 0-0076
>10 >10 0 000011 0 00000065
>10 >10
0-00002 0 0000012
* Serum samples were collected and equal volumes pooled from 13-15 mice at each time-point represented and assayed for inhibition of anti-soman serum antibody binding to So-BSA-coated plates by ELISA. t Values represent mm concentration of inhibitor required to inhibit 500/ binding in the ELISA.
memory response compared
to Day 14 after primary immuniza-
tion (data not shown). Fine
specificity of
anti-So-KLH monoclonals
Table 3 shows the fine specificity profile of several monoclonals obtained after secondary and tertiary boosting with So-KLH. A comparison of Table 3, which provides the 15o values of various
inhibitors obtained for individual mAb, with Table 1, which provides similar information for the serum antibodies, demonstrates that the fine specificity of the monoclonals correlates well
with what is observed in NPSo-specific serum antibody populations. Although there is significant variation in relative avidity between monoclonals, the overall pattern of specificity appears to be: soman-protein > APSo 2 NPSo 2 NPDBP > NPPC. It is also noteworthy that 3/4 mAb obtained from a later fusion
Maturation of the antibody response to the toxin soman
403
SerLum IgGi SoBSA
NPSo
100 I00
-
80 80
cm .s
6
He
40 20
60
-
40
-
20
0
-'-
-10 -9
-8
-7
-6
-5
0
-4
-1
-7
-6
APSo
cm
.C
-5
-4
-3
-2
-3
-2
-I
NPPC
100
100
80
80
60
60
40
40
20
20
-
C
m
0
-I4
-9
-10
-8
-7
-6
-5
1-7
-4
-6
-5
-4
Serum IgG2a SoBSA
NPDBP
NPSo
100 100
100
-
80 80
-
80s
60
60
60-
40
40
40
20
20
20
C
-0
1
0
.
.,,
,.2i
.,
~o--
-
°04-
-2 -7 -6 -5 -4 -3 -2 -1 Log M hapten Log M hapten Figure 5. Maturation of fine specificity in response to So-KLH. Serum samples were pooled from 13-15 animals and tested in an ELISA for ability of So-BSA, NPSo, APSo, NPDBP, or NPPC to inhibit binding to So-BSA on Day 14 (0), Day 35 (A) and Day 122 (-) after primary immunization. Statistical significance of changes in Iso with time was evaluated by analysis of variance (ANOVA, Table 2).
-1O
-9
-8
-7
-6
-5
-4
-8
-7
(tertiary response) showed no inhibition by NPDBP or NPPC. Also, 10/13 IgM mAb were inhibited by soman-protein only, and were not inhibited by any free soman analogue tested.6 The analysis of hybridomas obtained after So-KLH priming and boosting is therefore likely to be representative of the in vivo SoKLH response.
DISCUSSION
Analysis of the
serum response to So-KLH demonstrates that true memory to this compound can be established and that
specificity is maintained throughout this response. Interestingly, inhibition of pooled IgGI serum antibody with the free haptens NPSo and APSo did not appear to vary significantly throughout the entire response to So-KLH. Among individuals, inhibition of IgG by free soman hapten was variable. The overall lack of maturation in relative affinity for NPSo/APSo suggests that
-6 -5 -4 Log M hapten
-3
carrier/linker influences or specific contacts may play an important selective role in maturation of the anti-So-KLH response. Such carrier- or linker-mediated effects on antibody maturation have been suggested by others.'920 However, it is important to note that no overall decrease over time was apparent in the ability of NPSo/APSo to inhibit serum antibody; these NPSo/APSo-binding antibodies were therefore maintained as part of the NPSo-specific antibody pool but apparently had not been subjected to a strong selective pressure that would have allowed for the expansion of higher avidity clones. This is in contrast to the observed temporal decrease in inhibition by NPPC and NPDBP, suggesting that these structures are less inhibitory because they lack the hydrophobic features of NPSo, which may play a dominant role in the selective process. The priming regimen utilizing a protein carrier led to a memory immune response which persisted for greater than 1 year, although significant affinity maturation was demonstrable
A. C. Buenafe & M. B. Rittenberg
404
Table 2. Summary table for analysis of variance (ANOVA) of IgG serum antibody inhibition at different time-points during the response to So-KLHt Serum Ab + inhibitor
IgG I + So-BSA
IgGI +NPSo
IgG 1 + APSo
IgG2a + So-BSA
IgG2a+NPSo
Source of variance
SST
dft
MST
F
Total Between groups Within groups Treatments Error
0-0000004 00000001 0-0000003 0-0000003 0-00000003
8 2 6 2 4
0-0000001 0-000000009
15.6**
Total Between groups Within groups Treatments Error
0-00133 0-00118 0-00016 0 00007 0 00009
8 2 6 2 4
0-00003 0-00002
Total Between groups Within groups Treatments Error
0-000002 0-0000006 0 000001 0-0000006 0 0000008
11 3 8 2 6
0 0000003 0 0000001
Total Between groups Within groups Treatments Error
0-0000001 0-00000002 0 0000001 0 0000001 0 00000002
11 3 8 2 6
0-00000005 0 000000003
Total Between groups Within groups Treatments Error
0-00877 0-00739 0 00138 0-00061 0-00077
11 3 8 2 6
0-0003 0 00013
1 4§
2-2§
16*
2 4§
t Analysis of variance29 was performed for 1so values obtained by So-BSA and NPSo inhibition of serum IgG at Days 14, 35 and 122 after primary immunization (Fig. 5). 1 SS, sum of squares; df, degrees of freedom; MS, mean square. §Not significant at a=0 1. * Significant at x=0-01; ** significant at a =005. SoBSA
APSo
NPSo v - -4 I* et
Mice no.
(o) I
(*) 3* b0-S
l005
to-?
10-60
10 -8
10?7
(U) 4 (°) 5 (a) 7* (a) 8 (A) 11 ( A) I2 (U) 14
2
E 0 S
( t) 15* 14
290
290
14
14
290
(days) Figure 6. Comparison of serum inhibition values for So-BSA, APSo and NPSo at an early time-point (Day 14) versus a late time-point (Day 290) in individual mice. For some mice (*) it was necessary to use serum collected on Day 17 rather than on Day 14. Time
against soman-protein conjugates only. The lack of improvement in binding free soman hapten analogues might be resolved by using non-immunogenic linkers which might elicit a maturational antibody response to the hapten structure alone. The extreme toxicity of soman precludes its general use for screening
purposes or as an inhibitor in the ELISA, although the toxin has been used in competitive inhibition assays performed by others under special laboratory conditions.2 3 The use of haptenprotein conjugates to detect hapten-specific antibodies is commonly used as a primary screening procedure9l" 13 and soman is
Maturation of the antibody response to the toxin soman
405
Table 3. Fine specificity profiles of secondary and tertiary response anti-So-KLH hybridomas Inhibitors So-BSA
So-KLH
3-8 8-0 1-07 7-73
0-00000008 0-0000005 0-0000002 0-0000003
0-000000003 0 00000006 0 000000008 0-00000004
2-18 1-05 2-15 0-58
1-0 0-9 0-5 > 10
0-0000001 0-0000003 0-00000001 0-000000008
0-00000001 0-00000002 0-000000009 0-00000003
ND
3-29
>10 >10 > 10
> 10
ND > 10
>10 4 03
>10 > 10
0 00000004 0-00000001 0-00000001 0-000000004
ND 0-000000004 0-000000001 0-0000000009
NPPC
NPMPC
NPEP
Secondary response IgG1 hybridomas 0-388 0.009* 0-26 SoGl-3 0-064 0-0003 0-002 SoGl-5 0 053 0-13 0-024 SoGl-9 0-365 0-0003 0-26 SoGI-13
> 10 9-9 > 10 > 10
> 10 > 10 > 10 > 10
> 10 2-98 1-57 3-47
Secondary response IgG2a hybridomas 0-018 0-015 0 001 SoG2a-3 0-014 0-0004 0-02 SoG2a-4 0-025 0-0009 0-026 SoG2a-6 0-129 0-18 0 04 SoG2a-7
1-02 0-58 0-75 3-76
0-63 0-63 0-88 > 10
Tertiary response hybridomas 0-0004 0-002 SoGl-1 0-033 0-033 SoG2a-1 0-0008 0-018 SoG2a-2 0 035 0-02 SoG2b-1
5 64 > 10 >10 > 10
APSo
NPSo
NPDBP
0-151 >10 > 10 > 10
OHSo
* Values represent mm concentration of inhibitor required to inhibit 50°/. binding to So-BSA-coated plates in ELISA. ND, not done.
inactivated as a toxin once it is coupled to the protein carrier. It remains possible, however, that the pattern of fine specificity maturation observed reflects the use of soman-protein as the screening antigen. In terms of fine structure recognition it appears that the phenyl linkage is important in the binding of NPSo-specific IgG serum antibody; this is consistent with what has been found for NPSo-specific mAb3-6 and polyclonal rabbit antisera.3 The generation of antibodies with specificity including recognition of linkage structures has been described for several hapten-protein conjugates.2"-23 The presence of the phenyl structure is also required for recognition by memory Group II antibodies generated against phenylphosphocholine conjugates.7'0 Antibodies produced in response to So-KLH resemble those generated in the memory response to PC-KLH in the diversity of fine specificity and V gene usage. However, in the anti-So-KLH response this molecular heterogeneity is observed as early as the primary response.6 The inability of PC-BSA and its hapten analogue NPPC to inhibit the binding of anti-NPSo antibodies to soman-protein illustrates the specificity of this response for a soman-protein epitope. Since the phenyl-phosphorus linkage is found in both PC-protein and soman-protein, the positively charged choline moiety of PC may play a significant role in preventing its recognition by NPSo-specific antibodies. In addition, we observed partial inhibition of NPSo-specific IgG antibody by NPDBP, which lacks this positive charge. The presence of the negatively charged phosphate in PC-BSA and NPPC may be less significant in preventing cross-reactivity since NPDBP also bears a negatively charged phosphate (Fig. 4, Table 1). Also, NPMPC, a methylphosphonate analogue of NPPC lacking this negative charge, was a very poor inhibitor of NPSo-specific IgG monoclonal antibodies (Table 3, and data not shown). The relative affinity, kinetics and levels of IgM antibody produced after memory boosting were similar to those of the
primary IgM response; this suggests that the majority of IgM produced to repeated challenge with So-KLH represents recruitment of new primary B cells into the ongoing response even after the memory response was well-established, as seen in other systems.24 Whether primary response IgM and secondary response IgG antibodies derive from the same B-cell population or whether they represent distinct B-cell populations is not known; it has recently been suggested that memory B cells may derive from a separate subpopulation.25 IgM-producing B-cell clones are commonly associated with the early response to protein antigens and it has been suggested that they often express , Ittiis therefore possible that the germline V gene sequences. 24.26.27t1 soman-protein-specific IgM antibodies detected here are products of newly stimulated B cells which have not yet been driven to mutate nor subjected to selective pressures, consequently demonstrating poor affinity for the antigen. On the other hand, IgG 1 serum antibodies produced after final boosting are likely to represent products of an antigen-selected memory B-cell population.' 1.13.28 We have found VH gene usage to be heterogeneous among IgM and IgG anti-NPSo hybridomas,6 although the J558 VH family appeared to be over-represented in both populations; members of the 36-60 and 7183 VH families were also present among both IgM and IgG hybridomas. Thus, although the individual V genes have not been identified, the results are consistent with the hypothesis that the IgG population arose selectively from antigen-specific IgM-producing B cells. Our studies show that a fine specificity profile representative of that observed in NPSo-specific hybridomas can be demonstrated among polyclonal serum antibodies produced in vivo. This suggests that in vitro culturing conditions have not altered this repertoire, whereas such a selection has been reported to take place during the in vitro culturing of some T-cell clones.29 The results also suggest that anti-NPSo antibodies produced in vito may be the products of a heterogeneous V gene population, as was found among NPSo-specific hybridomas. Since some of the
406
A. C. Buenafe & M. B. Rittenberg
V gene sequences encoding So-KLH-specific combining sites have recently been characterized,30 it may be possible to use these sequences together with current protein-ligand computer modelling techniques to design high affinity combining sites that would interact with the soman neurotoxin. Finally, the apparent lack of a restricted B-cell response at any point in the So-KLH response may reflect the absence of idiotypic regulation or evolutionary selection for an environmentally common structure such as appears to be the case in the response to PC.3233 Thus, the differences in molecular heterogeneity that characterize the early phases of the antibody responses to these two organophosphorus haptens, soman and PC, may reflect the fact that only one of them, soman, is new to the environment.
ACKNOWLEDGMENTS We wish to thank Drs U. Bruderer, J. Fellman, and D. Lenz for providing the haptens used in this study and M. Fuller for technical assistance. We also thank Dr C. Chen and M. Nutt for helpful discussion and critical review of this manuscript. This work was supported by National Institutes of Health Grants Al 14985 and Al 26827, and by Contract DAMD-17-83-C-3246 from the U.S. Army Medical Research and Development Command.
REFERENCES 1. KOELLE G.B. (1981) Organophosphate poisoning-an overview. Fund. Appl. Toxicology, 1, 129. 2. HUNTER K.W., LENz D.E., BRIMFIELD A.A. & NAYLOR J.A. (1982) Quantification of the organophosphorus nerve agent soman by competitive inhibition enzyme immunoassay using monoclonal antibody. FEBS Lett. 149, 147. 3. LENz D.E., BRIMFIELD A.A., HUNTER K.W., BENSCHOP H.P., DEJONG L.P.A., VAN DIJK C. & CLOW T.R. (1984) Studies using a monoclonal antibody against Soman. Fund. Appl. Toxic. 4, 5156. 4. BRIMFIELD A.A., HUNTER K.W., LENz D.E., BENSCHOP H.P., VAN DIJK C. & DEJONG L.P.A. (1985) Structural and stereochemical specificity of mouse monoclonal antibodies to the organophosphorous cholinesterase inhibitor soman. Molec. Pharmacol. 28, 32. 5. BUENAFE A.C. & RITTENBERG M.B. (1987) Combining site specificity of monoclonal antibodies to the organophosphate hapten Soman. Molec. Immunol. 24, 401. 6. BUENAFE A.C., MAKOWSKI F.F. & RITTENBERG M.B. (1989) Molecular analysis and fine specificity of antibodies against an organophosphorus hapten. J. Immunol. 143, 539. 7. CHANG S.P., BROWN M. & RITTENBERG M.B. (1982) Immunologic memory to phosphorylcholine II. PC-KLH induces two antibody populations that dominate different isotypes. J. Immunol. 128, 702. 8. CHANG S.P., BROWN M. & RITTENBERG M.B. (1982) Immunologic memory to phosphorylcholine III. IgM includes a fine specificity population distinct from TEPC 15. J. Immunol. 129,1559. 9. STENZEL-POORE M.P., BRUDERER U. & RITTENBERG M. B. (1988) The adaptive potential of the memory response: Clonal recruitment and epitope recognition. Immunol. Rev. 105, 113. 10. BRUDERER U., STENzEL-POORE M.P., BACHINGER H.P., FELLMAN J.H. & RITTENBERG M.B. (1989) Antibody combining site heterogeneity within the response to phosphocholine-keyhole limpet hemocyanin. Molec. Immunol. 26, 63. 11. WYSOCKI, L., MANSER T. & GEFTER M. (1986) Somatic evolution of variable region structures during an immune response. PNAS 83, 1847.
12. BEREK C. & MILSTEIN C. (1987) Mutation drift and repertoire shift in the maturation of the immune response. Immunol. Rev. 96, 23. 13. SIEKEVITZ M., KOCKS C., RAJEWSKY K. & DILDROP R. (1987) Analysis of mutation and class switching in naive and memory B cells generating adoptive primary and secondary responses. Cell, 48, 757. 14. GEARHART P.J., SIGAL N.H. & KLINMAN N.R. (1975) Heterogeneity of the BALB/c antiphosphorylcholine antibody repsonse at the precursor cell level. J. exp. Med. 141, 56. 15. QUINTANS J. & COZENZA H. (1976) Antibody response to phosphorylcholine in vitro II. Analysis of T-dependent and T-independent responses. Eur. J. Immunol. 6, 399. 16. PERLMUTTER R.M., CREWS S.T., DOUGLAS R., SORENSEN G., JOHNSON N., NIVERA N., GEARHART P.J. & HOOD L. (1984) The generation of diversity in phosphorylcholine-binding antibodies. Adv. Immunol. 35, 1. 17. TODD I., BROWN M. & RITTENBERG M.B. (1985) Immunologic memory to phosphocholine. VI. Heterogeneity in light chain gene expression. Eur. J. Immunol. 15, 177. 18. STENZEL-POORE M.P. & RITTENBERG M.B. (1989) Clonal diversity, somatic mutation, and immune memory to phosphocholine-keyhole limpet hemocyanin. J. Immunol. 143,4123. 19. FISH S. & MANSER T. (1987) Influence ofthe macromolecular form of a B cell epitope on the expression of antibody variable and constant region structure. J. exp. Med. 166, 711. 20. ALLEN D., SIMON T., SABLITZKY F., RAJEWSKY K. & CUMANO A. (1988) Antibody engineering for the analysis of affinity maturation of an anti-hapten response. EMBO J. 7, 1995. 21. RODWELL J.D. & KARUSH F. (1980) Restriction in IgM expression. I. The VH regions of equine anti-lactose antibodies. Molec. Immunol. 17, 1553. 22. HOPPNER W., FISHER K., POSCHMANN A. & PAULSEN H. (1985) Study on the carbohydrate specificity of antibodies formed in rabbits to synthetic glycoproteins with the carbohydrate structure of asialoglycophorin A. Molec. Immunol. 22, 1341. 23. CHOUCHANE L., STROSBERG A.D. & HOEBEKE J. (1988) Stereospecific immuno-recognition of the tetracyclic anti-depressant oxaprotiline. Molec. Immunol. 25, 1299. 24. BEREK C., JARVISJ.M. & MILSTEIN C. (1987) Activation of memory and virgin B cell clones in hyperimmune animals. Eur. J. Immunol. 17, 1121. 25. KLINMAN N.R. & LINTON P.J. (1990) The generation of B cell memory: a working hypothesis. Contemp. Topics Microbiol. Immunol. 159, 19. 26. KARUSH F. (1978) The affinity of antibody: range, variability and the role of multivalence. In: Immunoglobulins (eds G. Litman and R. A. Good), Vol. 5, p. 85 Plenum Press, New York. 27. RODWELL J.D., GEARHART P.J. & KARUSH F. (1983) Restriction in IgM expression. IV. Affinity analysis of monoclonal antiphosphorylcholine antibodies. J. Immunol. 130, 313. 28. EISEN H. & SISKIND G. (1964) Variations in affinities of antibodies during the immune response. Biochemistry, 3, 996. 29. GAMMON G., KLOTZ J., ANDO D. & SERCARZ E. (1990) The T cell repertoire to a multideterminant antigen. Clonal heterogeneity ofthe T cell response, variation between syngeneic individuals, and in vitro selection of T cell specificities. J. Immunol. 144, 1571. 30. BUENAFE A.C. (1989) Specificity of antibodies binding organophosphorus haptens. PhD thesis, Oregon Health Sciences University. 31. PHILLIPS D.S. (1978) Basic Statisticsfor Health Science Students. W. H. Freeman, New York. 32. BOTTOMLY, K. (1984) 1984: All idiotypes are equal, but some are more equal than others. Immunol. Rev. 79, 45. 33. BRILES D.E., FORMAN C., HUDAK S. & CLAFLIN J.L. (1982) Antiphosphorylcholine antibodies of the T15 idiotype are optimally protective against Streptococcuspneumoniae. J. exp. Med. 156, 1177.
