- Email: [email protected]
A New Way to Diversify Antibodies by DNA Transposition
Leading Edge
Previews A New Way to Diversify Antibodies by DNA Transposition Davide F. Robbiani1,* and Michel C. Nussenzweig1,2,* 1Laboratory
of Molecular Immunology Hughes Medical Institute The Rockefeller University, New York, New York 10065, USA *Correspondence: [email protected] (D.F.R.), [email protected] (M.C.N.) http://dx.doi.org/10.1016/j.cell.2016.01.031 2Howard
While searching for new therapeutics against malaria, Lanzavecchia and colleagues discovered that antibodies can be diversified by DNA sequences encoded outside of antibody genes. B lymphocytes are tasked with producing antibodies to neutralize a broad range of pathogens. This challenge is solved by two separate diversification reactions. The initial antibody repertoire is generated during development when B cell precursors randomly assemble V, D, and J gene segments to produce transcription units that encode antibody-heavy and -light chains. This process is catalyzed by the RAG1/2 recombinase, and it produces a vast array of unique B cell clones, each displaying a distinct antibody molecule on the cell surface. Random gene assembly produces a large number of self-reactive B cells that are eliminated by one of several different mechanisms. Non-self-reactive B cells seed peripheral lymphoid tissues, where they can engage incoming viruses, bacteria, and parasites (Jankovic et al., 2004; Schatz and Ji, 2011). The second diversification reaction only occurs after B cells respond to pathogens in germinal centers where they undergo clonal expansion. During the germinal center reaction B cells express activation-induced cytidine deaminase (AID), a mutator enzyme that targets antibody genes. AID induces U:G mismatches in DNA that can be processed by one of a variety of different pathways to create point mutations, or DNA double-strand breaks that are associated with small insertions or deletions. B cell clones that accrue mutations that enhance antibody affinity against the pathogen are selectively expanded and also differentiate into memory and plasma cells (Di Noia and Neuberger, 2007; Victora and Nussenzweig, 2012). Some pathogens, including Plasmodium falciparum, attempt to evade the immune system by diversifying the antigens
they display to the immune system. This strategy can be overcome by the immune system by producing broadly neutralizing antibodies. A broadly neutralizing antibody capable of binding and neutralizing strains of Plasmodium falciparum expressing different protein coats would be beneficial therapeutically and potentially guide improved vaccine design. To find such antibodies, Peter Bull and Antonio Lanzavecchia’s groups screened plasmas from hundreds of malaria patients and identified two with broad binding activity against eight distinct parasite isolates (Tan et al., 2016). Memory B lymphocytes from these donors were immortalized and their antibodies screened for parasite staining. In both cases, they found antibodies that recognize up to eight distinct malaria isolates. The most broadly reactive antibodies bind proteins of the RIFIN family, which are encoded by the parasite and are expressed on the membrane of infected erythrocytes. When the investigators analyzed the genetic make-up of these antibodies, they discovered a new mechanism for antibody gene diversification. Both of the broadly reactive antibodies display large insertions between V and DJ segments (Figure 1). Strikingly, in both donors, the insertion includes sequence originating from the LAIR1 gene, which maps on a different chromosome than the antibodyheavy chain gene. LAIR1 is a collagenbinding receptor. The two broadly reactive antibodies show LAIR1 genomic insertions of 507 and 732 nucleotides, which include the exon coding for the extracellular domain of LAIR1. Moreover, the insertion is biologically significant, since the collagen-binding domain of LAIR1 is required and sufficient for anti-
body binding to RIFINs. Hence, under pressure to increase antibody diversity during Plasmodium falciparum infection, the heavy chain gene acquired DNA from a distinct chromosome (Tan et al., 2016). But how did this happen? The precise location of the LAIR1 insert between V and DJ segments and the presence of N nucleotides at the junctions suggest RAG1/2 involvement in this process. Consistent with this idea, cryptic recombination signal sequences have been identified at the insertion boundaries, implying that aberrant cleavage by RAG1/2 at the LAIR1 locus may release a DNA fragment that can be inserted into a V(D)J recombination intermediate in trans. RAG1/2-mediated DNA transposition has been reported using plasmid substrates in vitro (Agrawal et al., 1998; Hiom et al., 1998), but there is little supporting evidence for the existence of this reaction in vivo. More importantly, a RAG1/2 mediated event would require excision of the LAIR1 DNA, and the endogenous LAIR1 locus is intact, indicating that the insertions are generated by a ‘‘copy-andpaste’’ rather than a ‘‘cut-and-paste’’ mechanism that is characteristic for RAG1/2. A second possibility is that the substrates for insertion were created by AID. This possibility is supported by the finding of somatic mutations in the V-LAIR1-DJ antibodies. AID preferentially targets DNA at complementarity determining regions (CDRs) such as CDR3, which is situated precisely at the V-to-DJ junction, and it produces DNA double-strand breaks in a significant fraction of antibody genes (Di Noia and Neuberger, 2007; Wilson et al., 1998). According to this model, a double-stranded DNA break catalyzed
Cell 164, February 11, 2016 ª2016 Elsevier Inc. 601
precedent for this phenomenon was recently reported (Onozawa et al., 2014). In addition to understanding the precise mechanism that mediates the LAIR1 insertion, an important question that requires further investigation relates to the frequency of this phenomenon. How much do insertions contribute to the normal antibody repertoire? Are they exacerbated by chronic infection, such as malaria? Beyond LAIR1, do other loci contribute to antibody diversification and if so which ones? Like other groundbreaking discoveries, the observation that antibodies can be diversified by large insertions raises a number of interesting and important biological questions. ACKNOWLEDGMENTS D.F.R. is supported by NIH grant AI112602 and M.C.N. by AI037526 and the HHMI.
REFERENCES Agrawal, A., Eastman, Q.M., and Schatz, D.G. (1998). Nature 394, 744–751. Boehm, T., McCurley, N., Sutoh, Y., Schorpp, M., Kasahara, M., and Cooper, M.D. (2012). Annu. Rev. Immunol. 30, 203–220. Di Noia, J.M., and Neuberger, M.S. (2007). Annu. Rev. Biochem. 76, 1–22. Hiom, K., Melek, M., and Gellert, M. (1998). Cell 94, 463–470.
Figure 1. Antibody Diversification by DNA Transposition (A) Diagram of classical antibody proteins and antibody proteins containing a portion of LAIR1. Not to scale. (B) Candidate mechanisms to produce the LAIR1 fragment. Cryptic Recombination Signal Sequences (RSS) are represented by triangles in 1. DNA breaks are represented by arrows in 2. Not to scale.
by AID in CDR3 could provide the landing pad for a DNA insertion. The genesis of the LAIR1 DNA fragment is unclear, but if this reaction were to occur in germinal centers where B cells undergo rapid DNA replication, it could occur by any of a number of different mechanisms. For example, in chickens and some fish, AID triggers antigen-receptor diversification by gene conversion, an intra-chromosomal reaction that requires homology
between donor and acceptor sequences (Boehm et al., 2012; Di Noia and Neuberger, 2007). By analogy, AID might also initiate the LAIR1 insertion reaction. Or, as proposed by the authors, the LAIR1 fragments could first be duplicated by reverse transcription of pre-mRNA followed by insertion at a DNA break generated by either RAG1/2 or AID. Similar to retro-transposons, this would solve the copy-and-paste issue, and a
602 Cell 164, February 11, 2016 ª2016 Elsevier Inc.
Jankovic, M., Casellas, R., Yannoutsos, N., Wardemann, H., and Nussenzweig, M.C. (2004). Annu. Rev. Immunol. 22, 485–501. Onozawa, M., Zhang, Z., Kim, Y.J., Goldberg, L., Varga, T., Bergsagel, P.L., Kuehl, W.M., and Aplan, P.D. (2014). Proc. Natl. Acad. Sci. USA 111, 7729– 7734. Schatz, D.G., and Ji, Y. (2011). Nat. Rev. Immunol. 11, 251–263. Tan, J., Pieper, K., Piccoli, L., Abdi, A., Foglierini, M., Geiger, R., Tully, C.M., Jarrossay, D., Ndungu, F.M., Wambua, J., et al. (2016). Nature 529, 105–109. Victora, G.D., and Nussenzweig, M.C. (2012). Annu. Rev. Immunol. 30, 429–457. Wilson, P.C., de Bouteiller, O., Liu, Y.J., Potter, K., Banchereau, J., Capra, J.D., and Pascual, V. (1998). J. Exp. Med. 187, 59–70.
Previews A New Way to Diversify Antibodies by DNA Transposition Davide F. Robbiani1,* and Michel C. Nussenzweig1,2,* 1Laboratory
of Molecular Immunology Hughes Medical Institute The Rockefeller University, New York, New York 10065, USA *Correspondence: [email protected] (D.F.R.), [email protected] (M.C.N.) http://dx.doi.org/10.1016/j.cell.2016.01.031 2Howard
While searching for new therapeutics against malaria, Lanzavecchia and colleagues discovered that antibodies can be diversified by DNA sequences encoded outside of antibody genes. B lymphocytes are tasked with producing antibodies to neutralize a broad range of pathogens. This challenge is solved by two separate diversification reactions. The initial antibody repertoire is generated during development when B cell precursors randomly assemble V, D, and J gene segments to produce transcription units that encode antibody-heavy and -light chains. This process is catalyzed by the RAG1/2 recombinase, and it produces a vast array of unique B cell clones, each displaying a distinct antibody molecule on the cell surface. Random gene assembly produces a large number of self-reactive B cells that are eliminated by one of several different mechanisms. Non-self-reactive B cells seed peripheral lymphoid tissues, where they can engage incoming viruses, bacteria, and parasites (Jankovic et al., 2004; Schatz and Ji, 2011). The second diversification reaction only occurs after B cells respond to pathogens in germinal centers where they undergo clonal expansion. During the germinal center reaction B cells express activation-induced cytidine deaminase (AID), a mutator enzyme that targets antibody genes. AID induces U:G mismatches in DNA that can be processed by one of a variety of different pathways to create point mutations, or DNA double-strand breaks that are associated with small insertions or deletions. B cell clones that accrue mutations that enhance antibody affinity against the pathogen are selectively expanded and also differentiate into memory and plasma cells (Di Noia and Neuberger, 2007; Victora and Nussenzweig, 2012). Some pathogens, including Plasmodium falciparum, attempt to evade the immune system by diversifying the antigens
they display to the immune system. This strategy can be overcome by the immune system by producing broadly neutralizing antibodies. A broadly neutralizing antibody capable of binding and neutralizing strains of Plasmodium falciparum expressing different protein coats would be beneficial therapeutically and potentially guide improved vaccine design. To find such antibodies, Peter Bull and Antonio Lanzavecchia’s groups screened plasmas from hundreds of malaria patients and identified two with broad binding activity against eight distinct parasite isolates (Tan et al., 2016). Memory B lymphocytes from these donors were immortalized and their antibodies screened for parasite staining. In both cases, they found antibodies that recognize up to eight distinct malaria isolates. The most broadly reactive antibodies bind proteins of the RIFIN family, which are encoded by the parasite and are expressed on the membrane of infected erythrocytes. When the investigators analyzed the genetic make-up of these antibodies, they discovered a new mechanism for antibody gene diversification. Both of the broadly reactive antibodies display large insertions between V and DJ segments (Figure 1). Strikingly, in both donors, the insertion includes sequence originating from the LAIR1 gene, which maps on a different chromosome than the antibodyheavy chain gene. LAIR1 is a collagenbinding receptor. The two broadly reactive antibodies show LAIR1 genomic insertions of 507 and 732 nucleotides, which include the exon coding for the extracellular domain of LAIR1. Moreover, the insertion is biologically significant, since the collagen-binding domain of LAIR1 is required and sufficient for anti-
body binding to RIFINs. Hence, under pressure to increase antibody diversity during Plasmodium falciparum infection, the heavy chain gene acquired DNA from a distinct chromosome (Tan et al., 2016). But how did this happen? The precise location of the LAIR1 insert between V and DJ segments and the presence of N nucleotides at the junctions suggest RAG1/2 involvement in this process. Consistent with this idea, cryptic recombination signal sequences have been identified at the insertion boundaries, implying that aberrant cleavage by RAG1/2 at the LAIR1 locus may release a DNA fragment that can be inserted into a V(D)J recombination intermediate in trans. RAG1/2-mediated DNA transposition has been reported using plasmid substrates in vitro (Agrawal et al., 1998; Hiom et al., 1998), but there is little supporting evidence for the existence of this reaction in vivo. More importantly, a RAG1/2 mediated event would require excision of the LAIR1 DNA, and the endogenous LAIR1 locus is intact, indicating that the insertions are generated by a ‘‘copy-andpaste’’ rather than a ‘‘cut-and-paste’’ mechanism that is characteristic for RAG1/2. A second possibility is that the substrates for insertion were created by AID. This possibility is supported by the finding of somatic mutations in the V-LAIR1-DJ antibodies. AID preferentially targets DNA at complementarity determining regions (CDRs) such as CDR3, which is situated precisely at the V-to-DJ junction, and it produces DNA double-strand breaks in a significant fraction of antibody genes (Di Noia and Neuberger, 2007; Wilson et al., 1998). According to this model, a double-stranded DNA break catalyzed
Cell 164, February 11, 2016 ª2016 Elsevier Inc. 601
precedent for this phenomenon was recently reported (Onozawa et al., 2014). In addition to understanding the precise mechanism that mediates the LAIR1 insertion, an important question that requires further investigation relates to the frequency of this phenomenon. How much do insertions contribute to the normal antibody repertoire? Are they exacerbated by chronic infection, such as malaria? Beyond LAIR1, do other loci contribute to antibody diversification and if so which ones? Like other groundbreaking discoveries, the observation that antibodies can be diversified by large insertions raises a number of interesting and important biological questions. ACKNOWLEDGMENTS D.F.R. is supported by NIH grant AI112602 and M.C.N. by AI037526 and the HHMI.
REFERENCES Agrawal, A., Eastman, Q.M., and Schatz, D.G. (1998). Nature 394, 744–751. Boehm, T., McCurley, N., Sutoh, Y., Schorpp, M., Kasahara, M., and Cooper, M.D. (2012). Annu. Rev. Immunol. 30, 203–220. Di Noia, J.M., and Neuberger, M.S. (2007). Annu. Rev. Biochem. 76, 1–22. Hiom, K., Melek, M., and Gellert, M. (1998). Cell 94, 463–470.
Figure 1. Antibody Diversification by DNA Transposition (A) Diagram of classical antibody proteins and antibody proteins containing a portion of LAIR1. Not to scale. (B) Candidate mechanisms to produce the LAIR1 fragment. Cryptic Recombination Signal Sequences (RSS) are represented by triangles in 1. DNA breaks are represented by arrows in 2. Not to scale.
by AID in CDR3 could provide the landing pad for a DNA insertion. The genesis of the LAIR1 DNA fragment is unclear, but if this reaction were to occur in germinal centers where B cells undergo rapid DNA replication, it could occur by any of a number of different mechanisms. For example, in chickens and some fish, AID triggers antigen-receptor diversification by gene conversion, an intra-chromosomal reaction that requires homology
between donor and acceptor sequences (Boehm et al., 2012; Di Noia and Neuberger, 2007). By analogy, AID might also initiate the LAIR1 insertion reaction. Or, as proposed by the authors, the LAIR1 fragments could first be duplicated by reverse transcription of pre-mRNA followed by insertion at a DNA break generated by either RAG1/2 or AID. Similar to retro-transposons, this would solve the copy-and-paste issue, and a
602 Cell 164, February 11, 2016 ª2016 Elsevier Inc.
Jankovic, M., Casellas, R., Yannoutsos, N., Wardemann, H., and Nussenzweig, M.C. (2004). Annu. Rev. Immunol. 22, 485–501. Onozawa, M., Zhang, Z., Kim, Y.J., Goldberg, L., Varga, T., Bergsagel, P.L., Kuehl, W.M., and Aplan, P.D. (2014). Proc. Natl. Acad. Sci. USA 111, 7729– 7734. Schatz, D.G., and Ji, Y. (2011). Nat. Rev. Immunol. 11, 251–263. Tan, J., Pieper, K., Piccoli, L., Abdi, A., Foglierini, M., Geiger, R., Tully, C.M., Jarrossay, D., Ndungu, F.M., Wambua, J., et al. (2016). Nature 529, 105–109. Victora, G.D., and Nussenzweig, M.C. (2012). Annu. Rev. Immunol. 30, 429–457. Wilson, P.C., de Bouteiller, O., Liu, Y.J., Potter, K., Banchereau, J., Capra, J.D., and Pascual, V. (1998). J. Exp. Med. 187, 59–70.