Molecular dissection of an antigen-specific immune response

journal of Autoimmunity(1989) 2 (Supplement), 195-201

Molecular Dissection of an Antigen-specific Immune Response

Claudia Berek Institute for Genetics, University of Cologne, Weyertall21,

D-5000 Cologne 41, FRG

Selection defines the repertoire of the primary response, and also which cells difIerentiate and enter the pool of the memory cells. Whereas the primary response is drawn from the germline repertoire, the memory response consists overwhelmingly of cells which have been hypermutated and selected to express receptors of high affinity. The accumulation of somatic mutations over time indicates that the activation of the memory cell by antigen also reactivates the hypermutation mechanism. Since the frequency of somatic mutations correlates with an increase ln antibody affinity, an efficientselectionoperates not only on the primary B cells when first stimulated by antigen, but also on the memory cells after secondaryor tertiary immunization.

Introduction

Immune responses are not only rapid and specific, they display the phenomenon of maturation [ 11.How is all this achieved? The opportunity to examine these problems in detail arose with the introduction of hybridoma techniques [2] and rapid mRNA sequencing [3, 41. The application of these techniques to the murine immune response against the hapten 2-phenyl-oxazolone (phOx) has provided some of the answers. This response is, at least during the early stages, quite homogenous. Considering that genetic mechanisms exist which ensure an enormous diversity in the variable regions of the antibody molecules [5], the restricted repertoire in the phOx specific antibodies is striking. Multiple germline V, D and J gene segments encoding the variable regions of H and the L chains are present in the germline. Combinatorial diversity arises as a result of the use of different combinations of V (D) and J and different pairing of L and H chains. Furthermore there is junctional diversity because of variation in the precise site of joining of the different gene segments. Additional diversity can iesult from insertion of nucleotides (N segments) at the V, to D or D to JH border [6], and also at the V, to Jk border [7]. Not 195 0891%8411/89/038195+07$03.00/0

0 1989 Academic Press Limited

196 C. Berek

surprisingly, therefore, there is an enormous variability in the primary structure of phOx specific antibodies demonstrated by the fact that virtually every one so far sequenced has a unique V(D)J region. Nevertheless, there are certain structural features which recur in the variable regions of most of the high affinity antibodies. This implies that, out of the broad spectrum available in the phOx specific repertoire, only certain B-cell clones will be permitted to proliferate and to further differentiate. These are the clones which will mature into B cells producing antibodies of high affinity.

Selection dejines the antigen-specific response

The diversity in the phOx-specific repertoire becomes most evident in IgM antibodies [8]. IgM-secreting hybridoma lines were obtained at all stages of the immune response, but mainly in experiments in which mice depleted of helper T cells or nude mice were used (Berek, unpublished). The majority of the IgM antibodies had various Vu/V,-gene segment combinations and differed greatly in their D/J region (Figure 1). However, most of these IgM antibodies had only low affinity for the antigen and some even expressed cross-reactive specificities. In contrast to this, antibodies of the y 1 class showed a more restricted repertoire, particularly in the D/J region (Figure 1). When antigen was given to a normal mouse, and 7 d later, spleen cells were fused, 70% of the phOx specific hybridoma lines secreted antibodies expressing one particular Vn- and V,-gene segment, referred to as Vu-Ox1 and V,-0x1 respectively [9]. In all these antibodies, the length of the D/J region was 16 amino acids, with the first residue of the D-segment being Asp and third Gly. Although in the late primary [lo], the secondary [ 1l] and the tertiary response [ 121, the antibody repertoire became more diverse, these late response antibodies did not show the diversity observed in IgM antibodies. For instance, the V,-0x1 L chain was found in combination with H chains which differed in thier Vu-gene segments, but not in the D/J region. Another group of antibodies frequently observed in the secondary and tertiary response had a V,-45.1 L chain with the Vl 1 H chain. In all of these antibodies, the V, segment was joined to Jn3 and a D segment five amino acids long, starting and ending with a Gly. The data suggested that low affinity IgM expressing B-cell clones will neither proliferate to any significant extent, nor switch to the IgG class. Out of the broad spectrum of the phOx specific B-cell repertoire, only those few B-cell clones which express antibody molecules of high affinity are selected to differentiate further and to mature. Hence one factor which is clearly important in determining which members of the naive repertoire are chosen to play a productive role in the (long term) response is antigen selection based on receptor affinity. This is, however, by no means the whole story. Another important factor is certainly the frequency of a particular phOx-specific B-cell clone in the total phOx specific repertoire [8]. Certain V(D)J combinations might be more likely to occur than others which could explain the dominance of the 0x1 antibody in the primary response in BALB/c mice. However, the repertoire shift observed in the memory response is more difficult to understand [ 111. Idiotypic regulation, separate compartimentalization of primary and memory B cells, senescence of early stimulated B-cell clones and other factors might play a role.

Antigen-specific immune response

h “I.4 0 0+.J

OII Or.1

ox, VW

Or Qrli

VW

45.1

0x1

“or

“II

“or

Alp x Gly 16 08

Sly xxx

197

“0, “0,

my

21aa

“0,

v(lr

Figure 1. Comparison of the diversity in IgG and IgM phOx specific antibodies. Total number of phOx specific hybridoma lines analysed are given n : IgM; E3: IgG. Antibodies having a characteristic D-J region of 16 or 21 amino acids (aa) or various (var) lengths are shown. V,- and V,- gene segments utilized are indicated (data are taken from [8]).

Selection refines the antigen-specific repertoire Since primary response antibodies have no significant number of somatic mutations, it is assumed that the hypermutation mechanism is only activated after induction of B cells by antigen [ 131. This mechanism operates over the rearranged V region and flanking sequences introducing single nucleotide exchanges [ 141. No mutations could be detected in the constant regions of the H and L chains even in hypermutated antibodies [ 131. An analysis of mutations in phOx specific antibodies showed that the somatic mutations were not randomly distributed over the V region 1131. A few examples may be taken from Table 1, to highlight this point. Antibodies NQlO/ 12.4.6 andNQ1 l/8.1 from the secondary responseandantibodies NQ22/10.17,18.7, 16.4 and 61.1 from the tertiary response all showed an exchange at residue 34 of the V,-0x1 L chain from His to Asp or Gln. In most of these antibodies, residue 36 Tyr is changed as well. Chain recombination experiments demonstrated that the substitution His to Asp or Gln increases the affinity by ten- or eight-fold respectively [ 131. The Tyr to Phe exchange by itself does not increase the aflinity for the antigen phOx [ 131. However, it has not yet been excluded that the Tyr+Phe exchange has an effect on the antibody affinity when it occurs at the same time as the His+Asn or +Gln substitution. In most tertiary response antibodies, residue 50, the first residue of the CDR2 of V,-0x1, was mutated as well (antibodies NQ22116.4, 16.1,10.17 and 18.7 in Table 1). Further experiments are necessary to show whether the substitution of the negatively charged Asp to the neutral Gly or Ala causes the further increase in the aflinity. In the H chains of these antibodies, recurring substitutions were also found: for instance, in V,-0x1 and ahoin Vu-M21, residue 31, the first residue of CDRl,

v,-0x1 NQ10/12.4.6 22/10.17 18.7

V,-M2 1 NQl l/8.1 22116.4 61.1

Am -

NQ10112.4.6 22/10.17 18.7

Ala Ala

GlY

26

FRWl

2 Ile Val

v,-0x1 NQ10/8.1 22116.4 61.1

Asn -

5 Thr Ile -

FRWl

CDRl

31 Ser Thr Asn Thr

31 Ser hg Asn -

-

22 Thr Ala Ile -

-

-

Ala Ala

Asp 53 Ala Ser -

-

Aa -

53

Phe Phe Phe

Phe Phe

36 Tyr -

GlY

Gln Asn Asn

34 His Am Asn Gln

44 Gly

-

FRW2

33 Met -

32 Tyr Phe -

50 Val Leu

-

31 Ser Asn -

33 Gly -

Ala

33 Gly -

-

24 Ser Asn -

CDRl

Asn -

54 Ser -

CDR2

Phe

55 Ser Gly -

Val

-

-

Phe

-

49 Tyr

48 Ile

---

Leu

47 Trp ---

FRW2

Thr -

56 Ser -

56 Thr Ile -

-Ala Ala

Gly Tyr

50 Asp -_

60 Asn Gly -

60 Ala Ser Thr

-

Thr

55 Ala

CDR2

Thr -

-

60 Ala -_

68 Ser A%

-

Leu -

78 Met

FRW3

89 Met Ile

Val -

Ile

89 Met -

Gly -

93 Ser -

82 Met -

FRW3

-

92 Ser Asn -

CDR3

Ile -

110 Thr -

FRW4

Ser _

100 Ala -

FRW4

Table 1.Recurring amino acid exchanges inphOx speci$c antibodies. Sequences are compared with thegermline sequences of VK-0x1, V,0x1 or V,-M21. Substitutions were observed in framework (FR W) and in complementary determining regions (CDR) . The positions of the exchanges are indicated (data are takenfrom [lo], [11] and [12])

Antigen-specific immune response

199

Figure 2. Accumulation of somatic mutations in 0x1 antibodies correlated with the increase in antibody atlinity. Data are shown for early primary ( q), late primary (H), secondary (0) and tertiary (A) response antibodies. V,/V,-0x1 sequences differed by the number of mutations indicated. They may also differ by the J,, segment or one single amino acid in the D segment (mutation or N-region diversity) (data are taken from [13]).

was frequently substituted. However, it seems difficult to imagine that the conserved exchange from a Ser to either Thr or Asn has a significant effect on the antibody affinity. Another possibility could be that these recurring substitutions are an indication of hot spots for somatic mutations. For instance, the frequent occurrence of silent mutations at certain positions in the V regions suggested that there must be specific sequences or structures in the DNA which are prone to somatic mutation [ 131. Figure 2 shows that the number of somatic mutations in 0x1 antibodies continuously increases with time after immunization. Whereas in the early primary response antibodies, only one single nucleotide exchange in 9 H and L chains analysed was found, one week later, several substitutions in each H and L chain could be detected. The left part of Figure 2 shows the al%nity of these antibodies. Antibodies of the early primary response had a dissociation constant of > 20 x 10e8. Only two of these antibodies had a slightly higher al%nity because of junctional diversity in the D-J region. With the accumulation of somatic mutations the afIinity of the antibodies increased. This is best demonstrated by 0x1 antibodies which use identical V(D)J gene segments but differ in the number of somatic mutations and athnity (Table 2). One good example is an early primary response antibody NQ2/12.4 which has no mutations and a dissociation constant of 28 x 10w8, in comparison with the tertiary response antibody NQ22/ 15.18 which differs by 17 somatic mutations, of which 12 were expressed. This antibody had a dissociation constant of ~0.2 x lo-*. This means that by somatic mutation and selection a more than lOO-fold increase in affinity has been achieved. The picture which thus emerges is of naive B cells undergoing antigenic stimulation which permits the initiation of the somatic hypermutation mechanism. Cells carrying mutations which lead to an increase in af5nity will then be favoured, both in terms of their making a contribution to the response, and of their selection for inclusion within the memory pool. The accumulation of somatic mutations from the

200 C. Berek Table 2. Ajinity

of antibodies with identical V(D)J

rearrangements Diss constant

Line

Vu

D

Ju

V,

JK

Expressed mutations

1” 2”

NQ 2/12.4 NQ22/15.18

0x1 0x1

Asp Arg Gly Asp Arg Gly

3 3

0x1 0x1

5 5

0 12

28 lOma <0.2 10-s

2” 3” 3”

NQ10/12.4.6 NQ22/10.17 18.7

0x1 0x1 0x1

Asp Trp Asp Trp Asp Trp

Gly Gly Gly

4 4 4

0x1 0x1 0x1

5 5 5

8 13 10

0.2 10-S <0.2 10-S <0.2 10-8

2” 3” 3”

NQll/ 8.1 NQ22/16.4 61.1

M21 M21 M21

Asp Gly Gly Asp Gly Gly Asp Gly Gly

2 2 2

0x1 0x1 0x1

5 5 5

5 14 10

10 10-E 1.5 10-8 2 10-E

Response

primary to the secondary and tertiary response suggests that the memory cells follow a differentiation pathway comparable to that of the primary B cells. Upon further immunization and hence activation of the memory B cells, the somatic hypermutation mechanism seems to be reactivated and again cells expressing high affinity receptors will be selected to differentiate into effector cells. Acknowledgements

I would like to thank Cesar Milstein for stimulating discussion throughout the course of this work. References 1. Siskind, G. D. and B. Benaceraff. 1969. Cell selection by antigen in the immune response. Adv. Immunol. 10: 1 2. Kohler, G. and C. Milstein. 1975. Continuous culture of fused cells secreting antibody of predelined specificity. Nature 256: 495-497 3. Griffiths, G. M. and C. Milstein. 1985. The analysis of structural diversity in the antibody response. In Hybridoma Technology in the Biosciences and Medicine. T. Springer, ed. Plenum Publishing, New York. pp. 103-l 15 4. Hamlyn, P. H., G. G. Brownlee, C. C. Cheng, M. J. Gait, and C. Milstein. 1978. Complete sequence of constant and 3’ non-coding regions of an immunoglobulin mRNA using the dideoxy nucleotide method of RNA sequencing. Cell 15: 1067-1075 5. Tonegawa, S. 1983. Somatic generation of antibody diversity. Nature 302: 575-582 6. Alt, F. W. and D. Baltimore. 1982. Joining of immunoglobulin heavy chain gene segments: implications from a chromosome with evidence of three D(J)H fusions. Proc. Natl.

Acad. Sci. USA 78: 4118-4120 7. Heller, M., J. D. Owens, J. F. Mushinski, and S. Rudikoff. 1987. Amino acids at the site of Vx-J, recombination not encoded by germline sequences. 3. Exp. Med. 166: 637-646 8. Berek, C. and C. Milstein. 1988. The dynamic nature of the antibody repertoire. Zmmunol. Reet. 105: l-26 9. Kaartinen, M., G. M. Gritliths, A. F. Markham, and C. Milstein. 1983. mRNA sequences define an unusually restricted IgG response to 2-phenyloxazolone and its early diversification. Nature 304: 32&324

Antigen-specific immune response 201 10. Griffiths, G. M., C. Berek, M. Kaartinen, and C. Milstein. 1984. Somatic mutation and the maturation of the immune response to Z-phenyl-oxazolone. Nature 312: 271-275 11. Berek, C., G. M. Griffiths, and C. Milstein. 1985. Molecular events during the maturation of the immune response to oxazolone. Nature 316: 412-418 12. Berek, C., J. M. Jarvis, and C. Milstein. 1987. Activation of memory and virgin B cell clones in hyperimmune animals. Eur. J. Zmmunoll7: 1121-l 129 13. Berek, C. and C. Milstein. 1987. Mutation drift and repertoire shift in the maturation of the immune response. Zmmunol. Rev. 96: 23-41 14. Gearhart, P. J. and D. F. Bogenhagen. 1983. Clusters of point mutations are found exclusively around rearranged antibody variable regions. Proc. Nat. Acad. Sci. USA 80: 3439