SEARCH

SEARCH BY CITATION

Keywords:

  • Allelic exclusion;
  • B cells;
  • DNA repair;
  • Novartis Prize;
  • V(D)J recombination

Abstract

  1. Top of page
  2. Abstract
  3. Introduction: The Novartis Basic Immunology Prize
  4. The early days of V(D)J recombination
  5. Drug resistance and gene amplification
  6. The early days of B cell biology in the Baltimore laboratory
  7. Differential RNA processing, J chain and Marion Koshland
  8. Allelic exclusion
  9. A-MuLV transformants: A surprisingly accurate model for early B cell development
  10. The role of IgH μ chains in signaling B cell developmental events
  11. Accessibility regulation of V(D)J recombination
  12. DNA double-strand breaks and N regions: The V(D)J recombination mechanism
  13. Completion of the V(D)J recombination reaction: Non-homologous end-joining
  14. Concluding statements: Gene amplification, V(D)J recombination, and back
  15. Acknowledgements
  16. Appendix

I have been invited to write a short historical feature in the context of being a co-recipient with Klaus Rajewsky and Fritz Melchers of the 2007 Novartis Prize in Basic Immunology that was given in the general area of the molecular biology of B cells. In this feature, I cover the main points of the short talk that I presented at the Award Ceremony at the International Immunology Congress in Rio de Janeiro, Brazil. This talk focused primarily on the work and people involved early on in generating the models and ideas that have formed the basis for my ongoing efforts in the areas of V(D)J recombination and B cell development.

Abbreviations:
A-MuLV:

Abelson murine leukemia virus

DHFR:

dihydrofolate reductase

DSB:

double-strand break

NHEJ:

non-homologous end-joining

RSs:

recombination signal sequences

Introduction: The Novartis Basic Immunology Prize

  1. Top of page
  2. Abstract
  3. Introduction: The Novartis Basic Immunology Prize
  4. The early days of V(D)J recombination
  5. Drug resistance and gene amplification
  6. The early days of B cell biology in the Baltimore laboratory
  7. Differential RNA processing, J chain and Marion Koshland
  8. Allelic exclusion
  9. A-MuLV transformants: A surprisingly accurate model for early B cell development
  10. The role of IgH μ chains in signaling B cell developmental events
  11. Accessibility regulation of V(D)J recombination
  12. DNA double-strand breaks and N regions: The V(D)J recombination mechanism
  13. Completion of the V(D)J recombination reaction: Non-homologous end-joining
  14. Concluding statements: Gene amplification, V(D)J recombination, and back
  15. Acknowledgements
  16. Appendix

I was extremely flattered to be included with Dr. Rajewsky and Dr. Melchers in this award group. For several decades, these two leading B cell biologists have been great friends, collaborators, and, sometimes, competitors. My interactions with them over the years have certainly helped to shape my own thinking about the field.

I am pleased, at this point in time, of having the opportunity to interact closely with Dr. Rajewsky, as we both have positions in the Immune Disease Institute at Harvard Medical School. We have had many productive collaborations, and have had the opportunity to think and write together about the intriguing, unsolved problem of the mechanisms that enforce allelic exclusion 1, a topic that has captivated us both for decades. Likewise, I have been able to maintain close contact with Dr. Melchers, who is a visiting faculty member of Harvard Medical School and a member of the Immune Disease Institute Scientific Advisory Board. Dr. Rajewsky and I teach the B cell biology portion of the graduate immunology course at Harvard Medical School. Dr. Melchers often participates as a guest lecturer, and we entertain the students with discussions of our somewhat differing points of view of the mechanisms that underlie differential rearrangement and expression of Igκ and Igλ loci.

It is very humbling to receive the Novartis Basic Immunology Prize when one considers all those who have already received this prize and all of those, in many areas of immunology, who could and should receive this distinction. Even within the B cell biology field, there are so many who have done so much, that it is very difficult to distinguish a single set of defining contributions. Max Cooper and his colleagues made huge contributions to establishing the field that I entered. In this context, Max won the original Novartis (Sandoz) prize in 1990.

Another name that must be mentioned is Tasuku Honjo. Dr. Honjo's contributions to B cell biology over the years have been extraordinary, and were highlighted by his crowning achievement of discovering activation-induced cytidine deaminase, the enzyme responsible for initiating the distinct and fundamental processes of IgH class switch recombination, somatic hypermutation, and gene conversion 2, 3. Dr. Honjo was a member of the selection committee for the 2007 Novartis Prize, and, in his comments introducing the awardees, he noted that he was proud to have the award go to the B cell field. If I were on the selection committee, there is absolutely no question that Dr. Honjo and I would have exchanged places, with him being a recipient and I being the proud member of the selection committee. In fact, as I have written in my interview for the 2007 Immunology Congress in Brazil, I consider the major transforming discoveries in the field of adaptive immunology over the past several decades to be the discovery of the RAG genes by the Baltimore group 46 and the discovery of activation-induced cytidine deaminase by the Honjo group 2, 3. These discoveries have enabled many of us to perform new types of studies that have facilitated our ongoing contributions.

The early days of V(D)J recombination

  1. Top of page
  2. Abstract
  3. Introduction: The Novartis Basic Immunology Prize
  4. The early days of V(D)J recombination
  5. Drug resistance and gene amplification
  6. The early days of B cell biology in the Baltimore laboratory
  7. Differential RNA processing, J chain and Marion Koshland
  8. Allelic exclusion
  9. A-MuLV transformants: A surprisingly accurate model for early B cell development
  10. The role of IgH μ chains in signaling B cell developmental events
  11. Accessibility regulation of V(D)J recombination
  12. DNA double-strand breaks and N regions: The V(D)J recombination mechanism
  13. Completion of the V(D)J recombination reaction: Non-homologous end-joining
  14. Concluding statements: Gene amplification, V(D)J recombination, and back
  15. Acknowledgements
  16. Appendix

The remainder of this short essay focuses mostly on my recollections about a few of the events and people who have influenced my entry into the field of B cell biology and who helped me to develop some of the models and ideas that captivated my attention for the following decades. When I arrived in David Baltimore's laboratory at Massachusetts Institute of Technology (MIT) to start my postdoctoral training in 1977, Susumu Tonegawa had recently shown that Ig variable region exons are assembled somatically during the differentiation of B cells 7. Ongoing work from Tonegawa and several others groups, notably the Leder and Hood groups, also showed that individual germline V, D, and J segments are flanked by short conserved sequences termed recombination signal sequences (RSs) and led to the notion that these sequences somehow direct V(D)J recombination in the context of the so-called 12/23 rule 8, 9, reviewed in 10, 11.

However, the mechanism of V(D)J recombination, how it was controlled, and how the process was integrated into B cell development was still very unclear and provided a rich area of investigation for many laboratories. This was a wonderful time to enter the field of B cell development. So much had been done already to establish the questions of great interest. So much was still unknown because the methods were not available. The field was ripe for the application of emerging molecular genetic technologies. I was lucky enough to have learned some of this technology in my graduate studies and then to have landed in Baltimore's laboratory, which provided a remarkable environment to apply these and other approaches to the developing field of molecular immunology.

Drug resistance and gene amplification

  1. Top of page
  2. Abstract
  3. Introduction: The Novartis Basic Immunology Prize
  4. The early days of V(D)J recombination
  5. Drug resistance and gene amplification
  6. The early days of B cell biology in the Baltimore laboratory
  7. Differential RNA processing, J chain and Marion Koshland
  8. Allelic exclusion
  9. A-MuLV transformants: A surprisingly accurate model for early B cell development
  10. The role of IgH μ chains in signaling B cell developmental events
  11. Accessibility regulation of V(D)J recombination
  12. DNA double-strand breaks and N regions: The V(D)J recombination mechanism
  13. Completion of the V(D)J recombination reaction: Non-homologous end-joining
  14. Concluding statements: Gene amplification, V(D)J recombination, and back
  15. Acknowledgements
  16. Appendix

I had done my Ph.D. work with Robert Schimke at Stanford, where, along with my close collaborator and friend Rodney Kellems, we studied mechanisms by which cancer cells become resistant to methotrexate (an anti-cancer drug) by producing more of the dihydrofolate reductase (DHFR) target enzyme. Schimke was a dynamic, critical and rigorous scientist. He attracted a diversity of fellows and students, who provided a productive and stimulating environment that was an ideal place to train and to develop independence. Schimke was one of the first to appreciate the importance of protein turnover and he then turned his attention to regulation of gene expression by steroid hormone receptors and development of methods to isolate specific cDNA by the polysome immunoprecipitation methodology. I had convinced him, after hearing some of his pharmacology lectures, to let me work in his laboratory on isolating cultured mammalian cells that harbored mutations in the regulation of gene expression. This was the research goal that led to our studies of methotrexate resistance. The work I did in his laboratory involved nearly 6 years of developing and applying biochemistry and emerging molecular biology and cell culture techniques.

In the fall of 1976, about 5 years into the methotrexate resistance work, we still did not have a precise mechanism, although we began to have strong suspicions of the potential involvement of the then heretical possibility of gene amplification. I had made a commitment to begin postdoctoral work with David Baltimore on transforming retroviruses in the fall of 1977. We needed to have a DHFR cDNA probe to get the final answer, but this was the pre-cloning era. Rod Kellems and I had been trying to apply the polysome immunoprecipitation technique worked out by others in the Schimke laboratory but this was difficult for mRNA of relatively low abundance.

Given the short amount of remaining time for my graduate studies, I took a last stab at generating a specific probe by applying and further developing a subtractive hybridization methodology to identify differences in gene expression in methotrexate-resistant and -sensitive mammalian cells. This general approach had at that time been used to compare transforming and non-transforming retroviral viral genomes 12. Luckily, there was a postdoctoral fellow in the Schimke laboratory, Henry Burr, who had trained with Roy Brittin and knew everything about liquid DNA annealing and hybridization methods. He taught me those techniques. With this great technical advice, the subtractive hybridization methods that I designed worked perfectly the first time I tried them and yielded a pure DHFR cDNA, which allowed the use of classical liquid DNA annealing/hybridization methods to discover gene amplification 13. Along with Tonegawa's demonstration of programmed gene rearrangement in developing lymphocytes, the gene amplification discovery helped change the notion that the mammalian genome was static.

The early days of B cell biology in the Baltimore laboratory

  1. Top of page
  2. Abstract
  3. Introduction: The Novartis Basic Immunology Prize
  4. The early days of V(D)J recombination
  5. Drug resistance and gene amplification
  6. The early days of B cell biology in the Baltimore laboratory
  7. Differential RNA processing, J chain and Marion Koshland
  8. Allelic exclusion
  9. A-MuLV transformants: A surprisingly accurate model for early B cell development
  10. The role of IgH μ chains in signaling B cell developmental events
  11. Accessibility regulation of V(D)J recombination
  12. DNA double-strand breaks and N regions: The V(D)J recombination mechanism
  13. Completion of the V(D)J recombination reaction: Non-homologous end-joining
  14. Concluding statements: Gene amplification, V(D)J recombination, and back
  15. Acknowledgements
  16. Appendix

This gene amplification discovery and the methods that we developed set the path for much of my future work in the separate areas of programmed genomic alterations in lymphocytes and genomic instability and cancer. The subtractive hybridization technology, in particular, provided the basis for my entry into the B cell biology field when I came to David Baltimore's laboratory as a postdoctoral fellow. While I carried out work in cancer biology and immunology independently for decades, it was also very satisfying to have them intersect in fundamental ways more recently 14.

Going to the Baltimore laboratory was really a stroke of good fortune for me. David Baltimore attracted a host of exceptional people and the laboratory had a wonderfully intense research atmosphere. There seemed to be one major discovery after another occurring almost every month at one of the benches up and down the laboratory. David had his laboratory meetings each week; everyone talked, and we learned to present our most recent data and get feedback from everyone. David himself had substantial intellectual input into all of the work we did together. Discussing models and ideas with David was an incredibly exciting, demanding, and productive give and take.

In the Baltimore laboratory in 1977, there was no focus on B cell biology or immunology, but David's work on transforming retroviruses had really set the stage. Naomi Rosenberg had established the Abelson murine leukemia virus (A-MuLV) transformation system and shown that the virus transforms early B cells 15. Edward Siden was characterizing IgH and IgL chain gene expression patterns and, along with Max Cooper, John Kearney and their colleagues, had just made the discovery of asynchronous onset of the expression of IgH μ chains and IgL chains 1619. I collaborated with Siden early on to extend some of those findings to fetal liver cells 19. Al Bothwell came to the Baltimore laboratory from the Cold Spring Harbor Laboratory and had brought with him a wealth of emerging recombinant DNA methods including cDNA cloning procedures. Bothwell had a goal of cloning cDNA for the various IgH and IgL chain isotypes.

I had originally come to the laboratory to work on retroviral biology; however, I had been fascinated with the immune system since learning about the enigmatic thymus in a high school summer course at the University of Kansas in 1966. I had written about the possibility of doing research on that organ in my college application essays in 1967. My college and graduate work took me in quite different directions by serendipity, from plant physiology to biochemistry and molecular genetics of drug resistance; it was all great preparation for my work in B cell immunology.

When I arrived at the Baltimore laboratory in 1977, the remarkable opportunities to enter the immunology field and to also continue working on gene rearrangements and genomic stability were compelling. To contribute to this developing effort in the laboratoy, I established a collaboration with another postdoctoral fellow, Vincenzo Enea, with the goal of employing the subtractive hybridization methodologies to isolate cDNA specific for each of the various IgH and IgL isotypes. We worked closely together and applied the subtraction schemes on various hybridomas that expressed different combinations of IgH and IgL chains. Given the high-level expression of Ig transcripts in these cells, the approach worked very well.

We wrote a short paper describing this work for a 1979 Keystone meeting report 20; at that time I had given no name for this methodology and neither had anyone else. David Baltimore noted that the technique needed a name and then proceeded to come up with “subtractive hybridization” and it stuck. We used the subtracted probes to isolate cDNA clones from the libraries generated by Al Bothwell and very early on had a battery of IgH and IgL cDNA reagents that no one else had yet generated. The only problem was that we were all new to the immunology field and, at the time we generated the probes, most of us were not completely sure what to do next.

Differential RNA processing, J chain and Marion Koshland

  1. Top of page
  2. Abstract
  3. Introduction: The Novartis Basic Immunology Prize
  4. The early days of V(D)J recombination
  5. Drug resistance and gene amplification
  6. The early days of B cell biology in the Baltimore laboratory
  7. Differential RNA processing, J chain and Marion Koshland
  8. Allelic exclusion
  9. A-MuLV transformants: A surprisingly accurate model for early B cell development
  10. The role of IgH μ chains in signaling B cell developmental events
  11. Accessibility regulation of V(D)J recombination
  12. DNA double-strand breaks and N regions: The V(D)J recombination mechanism
  13. Completion of the V(D)J recombination reaction: Non-homologous end-joining
  14. Concluding statements: Gene amplification, V(D)J recombination, and back
  15. Acknowledgements
  16. Appendix

To help us get started, Bothwell, Enea, Siden and I read a basic immunology book (“Immunology”) by Hood, Weissman, and Wood 21 and vigorously tested ourselves on the problem sets that the book included. Our studies of this book provided our entry into basic (cellular) immunology. However, my true entry into immunology was immensely influenced by my interactions with Marion Koshland of the University of California at Berkeley, who was spending a sabbatical in the Baltimore laboratoy with the goal of isolating a cDNA for the Ig J chain. Marion was simply a wonderful and powerful influence. She was knowledgeable of all of the important questions in immunology and spent countless hours educating me on her views of these questions.

A particularly interesting problem that she raised with me was the nature of how B cells express IgH μ chains first as a membrane-bound protein and then as a secreted effector molecule. Marion believed that there must be a novel genetic basis for this remarkable ability. Based on her persuasion and her acquisition of appropriate cell lines for the studies, I decided to try to solve the problem and applied some of the cDNA manipulations that I developed for other purposes. We were able to show that production of membrane and secreted forms of the IgH chain protein occurs by a differential RNA processing mechanism 22. This discovery, which was perhaps the first example of this generally relevant control mechanism with respect to endogenous mammalian genes, was made simultaneously by Lee Hood and his colleagues 23, 24 and published back to back with our paper.

In return, I was able to help Marion Koshland by designing a subtractive/positive selection hybridization strategy to purify a J chain cDNA probe. Marion and her student Beth Mather carried out the strategy perfectly and obtained the pure J chain cDNA, which they then used to isolate the J chain cDNA clones 25. Marion Koshland was a great inspiration early on in my immunology career with her broad knowledge of B cell biology and immunology in general and her deep understanding of the important problems and determination to solve them, with her own hands if necessary.

Allelic exclusion

  1. Top of page
  2. Abstract
  3. Introduction: The Novartis Basic Immunology Prize
  4. The early days of V(D)J recombination
  5. Drug resistance and gene amplification
  6. The early days of B cell biology in the Baltimore laboratory
  7. Differential RNA processing, J chain and Marion Koshland
  8. Allelic exclusion
  9. A-MuLV transformants: A surprisingly accurate model for early B cell development
  10. The role of IgH μ chains in signaling B cell developmental events
  11. Accessibility regulation of V(D)J recombination
  12. DNA double-strand breaks and N regions: The V(D)J recombination mechanism
  13. Completion of the V(D)J recombination reaction: Non-homologous end-joining
  14. Concluding statements: Gene amplification, V(D)J recombination, and back
  15. Acknowledgements
  16. Appendix

Somewhere along the way I learned about the intriguing phenomenon of allelic exclusion 26, 27 and it totally captivated me. I was given a good background to begin to think about this problem from the knowledge that I had acquired of mouse myelomas (plasmacytomas) and their IgH and IgL expression patterns. Once we had available the cDNA for the various IgH isotypes and for the Igκ and Igλ L chain isotypes, we were able to do genetic, as well as biochemical experiments that allowed us to develop models.

Based on our findings that myelomas often expressed multiple L chain genes at the level of transcripts and cytoplasmic proteins, David Baltimore and I postulated a model in 1980 that suggested expression of a complete Ig could generate a signal to feedback and inhibit expression of the recombination machinery 28. This feedback model carried with it several correlates including the notion that the IgL chain must associate with a pre-existing IgH chain to effect feedback, and also that rearrangement of a given locus must occur one allele at a time to allow time for the feedback signal to be effected. The model was based on a fairly restricted amount of available data at the time, but did provide a view of how the allelic exclusion process is regulated that is still accepted.

Our work and the development of ideas were not done in isolation. We enjoyed extensive interactions and discussions with Martin Weigert, Bob Perry and their colleagues Chris Coleclough and Klaus Karjalainen, who were independently developing similar ideas, with perhaps more of a focus on stochastic aspects of the model 29, 30. While the issues of feedback versus stochastic mechanisms have not been fully resolved to this day 1, 31, there have been a number of important advances in our understanding including the work from Bergman, Cedar and their colleagues who have provided molecular insights into the long-standing question of how cells may choose one allele for rearrangement at a time (reviewed in 1). Also, in those early days, the remarkable process of receptor editing was not known 32. Clearly, receptor editing offers many more challenges than we knew at that time for explaining allelic exclusion and its potential mechanistic basis at the IgL loci.

A-MuLV transformants: A surprisingly accurate model for early B cell development

  1. Top of page
  2. Abstract
  3. Introduction: The Novartis Basic Immunology Prize
  4. The early days of V(D)J recombination
  5. Drug resistance and gene amplification
  6. The early days of B cell biology in the Baltimore laboratory
  7. Differential RNA processing, J chain and Marion Koshland
  8. Allelic exclusion
  9. A-MuLV transformants: A surprisingly accurate model for early B cell development
  10. The role of IgH μ chains in signaling B cell developmental events
  11. Accessibility regulation of V(D)J recombination
  12. DNA double-strand breaks and N regions: The V(D)J recombination mechanism
  13. Completion of the V(D)J recombination reaction: Non-homologous end-joining
  14. Concluding statements: Gene amplification, V(D)J recombination, and back
  15. Acknowledgements
  16. Appendix

The A-MuLV transformation system provided the initial method to address many of the long-standing questions that had captivated me. There were always worries that it was a transformed cell system and that lessons learned might not all apply to normal cells. That was a sound worry but still the system taught us many things that have held up, and new uses of the system are still leading to novel insights into the V(D)J recombination process as illustrated by the recent work of my former fellow Barry Sleckman and his colleagues 33.

Our early analyses of the A-MuLV-transformed cells, along with work from the Perry and Tonegawa groups 3436, showed that the assembly of IgH and IgL chain genes is ordered during early B cell development, with IgH variable region gene assembly preceding that of IgL variable region genes. Studies of the A-MuLV transformants provided the insight that the assembly of D-to-JH rearrangements at the IgH locus occurred first and on both chromosomes before appendage of a VH segment to the pre-existing DJH complex, allowing development of a regulated model for IgH allelic exclusion 35, 37. Studies of A-MuLV transformants, along with studies of normal B lineage cells, also provided the insight that the VH–to-DJH step was the regulated step in the context of allelic exclusion 37.

In 1982, I moved to Columbia University College of Physicians and Surgeons where I continued the work I had started with Baltimore at MIT; shortly thereafter, we were able to articulate a general model for the role of Ig chains in regulating the early stages of B cell development (Fig. 1; from 38). Subsequently, von Boehmer and colleagues, as well as others, discovered that α/β T cells employ a very overlapping set of early developmental strategies 39.

thumbnail image

Figure 1. The regulated model of B cell differentiation and allelic exclusion. The figure is reproduced with permission from 38.

Download figure to PowerPoint

The role of IgH μ chains in signaling B cell developmental events

  1. Top of page
  2. Abstract
  3. Introduction: The Novartis Basic Immunology Prize
  4. The early days of V(D)J recombination
  5. Drug resistance and gene amplification
  6. The early days of B cell biology in the Baltimore laboratory
  7. Differential RNA processing, J chain and Marion Koshland
  8. Allelic exclusion
  9. A-MuLV transformants: A surprisingly accurate model for early B cell development
  10. The role of IgH μ chains in signaling B cell developmental events
  11. Accessibility regulation of V(D)J recombination
  12. DNA double-strand breaks and N regions: The V(D)J recombination mechanism
  13. Completion of the V(D)J recombination reaction: Non-homologous end-joining
  14. Concluding statements: Gene amplification, V(D)J recombination, and back
  15. Acknowledgements
  16. Appendix

I was lucky enough to start my own laboratory with a number of outstanding students and fellows, who contributed greatly to the direction of our research. Of course, I have also been blessed with many other outstanding colleagues who have made major contributions over the years, but the initial days were important for the work I describe here.

My first postdoctoral fellow was Michael Reth, who fortuitously joined my laboratory after obtaining his Ph.D. with Klaus Rajewsky and coming to Boston to work in a different laboratory. Michael, who is now a Professor in Freiburg and a world-renowned immunologist, worked with me for a few weeks in the Baltimore laboratory, after which we packed up my things and moved them in his Volkswagen bus to my new laboratory at Columbia University Medical Center in New York City. In the early days of my laboratory, most people thought he was the “Professor”, perhaps because he looked distinguished and mature. His studies of A-MuLV transformants clearly proved our theory that the protein products of productively assembled IgH genes regulate early B cell differentiation by serving the dual function of feeding back to inhibit further IgH locus rearrangements (effecting allelic exclusion), and by signaling developmental progression to the stage of IgL chain gene assembly (40, Fig. 1).

Reth was one of the first to think of membrane-bound IgM in the context of a signaling complex. While many options were considered at the time for how an isolated μ H chain could provide its signal, he was convinced that it involved incorporation into the membrane and tried many experiments early on including tunicamycin treatment of A-MuLV-transformed cells to test the idea. Ultimately, transgenic models of others provided the first proof of the role of the membrane-bound form of the μ H chain in signaling allelic exclusion, with the work of Nussenzweig and Leder being particularly definitive 41, 42. At the same time, Reth and I were also able to show the role of the membrane-bound form of the μ protein in signaling developmental progression to the stage of IgL rearrangement via our studies of A-MuLV transformants 43.

Rajewsky and his colleagues 44 ultimately proved the mechanism in vivovia targeted mutation of the membrane-binding exon of the Cμ gene, and Melchers and his colleagues 45 offered a molecular mechanism for how the μ chain actually signaled in the absence of IgL chain with their discovery of the surrogate L chains and the pre-B receptor. Overall, our early work, which focused mainly on the A-MuLV transformation system, established the model that sequentially rearranged Ig genes are linked to the progression of developing B lymphocytes through a series of checkpoints which test for the functionality of their antigen receptor chains (Fig. 1).

Accessibility regulation of V(D)J recombination

  1. Top of page
  2. Abstract
  3. Introduction: The Novartis Basic Immunology Prize
  4. The early days of V(D)J recombination
  5. Drug resistance and gene amplification
  6. The early days of B cell biology in the Baltimore laboratory
  7. Differential RNA processing, J chain and Marion Koshland
  8. Allelic exclusion
  9. A-MuLV transformants: A surprisingly accurate model for early B cell development
  10. The role of IgH μ chains in signaling B cell developmental events
  11. Accessibility regulation of V(D)J recombination
  12. DNA double-strand breaks and N regions: The V(D)J recombination mechanism
  13. Completion of the V(D)J recombination reaction: Non-homologous end-joining
  14. Concluding statements: Gene amplification, V(D)J recombination, and back
  15. Acknowledgements
  16. Appendix

I have benefited greatly from my graduate students over the years, and two of my first students, Keith Blackwell and George Yancopoulos, made extremely important contributions to our early work on the regulation of V(D)J recombination. Blackwell, who is now working on Caenorhabditis elegans and is a faculty member at Harvard Medical School, led efforts in our laboratory to develop novel methods to assay the assembly of exogenous antigen receptor genes after their introduction into A-MuLV-transformed B lineage lines [46]. Yancopoulos, who is now a member of the US National Academy of Sciences and the President of Regeneron Pharmaceuticals Research Laboratories, made the discovery that proximal VH genes are highly preferentially rearranged in developing mouse B cells compared to distal VH genes [47].

The initial basis for the latter discovery came when Yancopoulos and I were going over a sequence of a V(D)J rearrangement that I had cloned from an A-MuLV transformant and had given him to sequence as a learning project. As we looked through the sequencing gel which he produced, I indicated to him that the sequence looked familiar, like one I had recently sequenced from the 18-81 A-MuLV-transformed cell line and had named VH81X 48. Indeed it looked identical and, for a moment, we had concern that I might have provided him with the wrong plasmid preparation. But, when we examined the junction region closely, it was clearly different from VH 81X. It was the same VH segment rearranged to a different D and JH in a different cell line. This led us to look at other rearrangements in more cell lines and we quickly found many more examples of proximal VH gene segment rearrangements 47.

Proximal VH gene rearrangement was an exciting finding and many others went on to study the phenomenon. However, preferential rearrangement of proximal VH genes does not appear to occur in humans and its overall significance in mice remains unclear; although some of the recent locus contraction work may provide some mechanistic insights into its basis 49, 50.

Yancopoulos, through studies of A-MuLV transformants, contributed to a number of major insights during his graduate work in my laboratory. His finding of transcription of germ-line VH genes before they are rearranged allowed us to formally articulate the accessibility model for regulation of V(D)J recombination 51, an idea that I had considered since my days in the Baltimore laboratory but did not have evidence in support of it until then. This model postulated that lineage, stage, and allelic excluded assembly of Ig and TCR genes is controlled by modulating relative accessibility of substrate gene segments to a common V(D)J recombinase. Subsequently, Yancopoulos exploited the recombination substrate transfection approach developed by Blackwell, utilizing newly identified germ-line TCRβ D and J segments that I had obtained from Leroy Hood in one of our several collaborations, to prove the accessibility model and demonstrate DNAse sensitivity as one of the first of many correlates of V(D)J recombinational accessibility 52.

Over the years, others have added many more correlates but we still do not know the cause and effect. However, we do know, through efforts of Pierre Ferrier, who was one of my earlier postdoctoral fellows, and now is a leader in the V(D)J recombination regulation field in Luminy (France), that transcriptional enhancer elements act to make a linked antigen receptor gene segment accessible to V(D)J recombinase 53, a role that subsequent studies demonstrated for most endogenous Ig and TCR locus enhancer elements via gene-targeted mutation (reviewed in 54).

DNA double-strand breaks and N regions: The V(D)J recombination mechanism

  1. Top of page
  2. Abstract
  3. Introduction: The Novartis Basic Immunology Prize
  4. The early days of V(D)J recombination
  5. Drug resistance and gene amplification
  6. The early days of B cell biology in the Baltimore laboratory
  7. Differential RNA processing, J chain and Marion Koshland
  8. Allelic exclusion
  9. A-MuLV transformants: A surprisingly accurate model for early B cell development
  10. The role of IgH μ chains in signaling B cell developmental events
  11. Accessibility regulation of V(D)J recombination
  12. DNA double-strand breaks and N regions: The V(D)J recombination mechanism
  13. Completion of the V(D)J recombination reaction: Non-homologous end-joining
  14. Concluding statements: Gene amplification, V(D)J recombination, and back
  15. Acknowledgements
  16. Appendix

Despite the discovery of the 12/23 rule and RSs, the mechanism of V(D)J recombination was unclear in 1980 and many were still thinking of RSs in the terms of stem loops and some type of homology-directed recombination. While in the Baltimore laboratory, I was fortunate to molecularly clone an IgH locus rearrangement (from an A-MuLV transformant) that was very unusual in that it resulted from several rearrangements including an inverted joining of a D to a JH segment. I was sitting in my office one night looking at the sequence with David Raulet (a friend who is now a Professor at the University of California, Berkeley, CA), while he was waiting for me to go out for a late happy hour at a local Mexican restaurant, when I realized that it contained a lot of information. It showed that RSs of the involved segments were precisely joined and that coding sequences were not, and that the coding sequences contained extra nucleotides that were not in the germ-line sequences – something we could be sure of since all of the sequence between the D and JH was known. I worked out a primitive set of ideas for a V(D)J recombination model based on this small bit of data and went to David Baltimore's house for dinner a few days later where we refined the model and wrote a paper on it in the same evening 55.

The model proposed that V(D)J recombination is initiated by a specific and precise DNA double-strand break (DSB) between the V, D, or J coding sequences and flanking RSs and that coding and RSs ends are differentially processed (coding ends being modified with loss or addition of nucleotides with RSs ends being precisely joined and not modified) (Fig. 2, 10). Another key aspect of the model was our proposal that non-templated nucleotides could be added at junctions of antigen receptor-coding gene segments. While our own studies were based on a single join, Yoshikazu Kurosawa and Susumu Tonegawa had cloned several V(D)J junctions and the presence of non-templated nucleotides appeared likely 56. Yoshi was one of my best friends at the time; so we clearly benefited by being able to discuss our ideas with him. We also noted from Yoshi's data that many, but not all, junctions had “micro-homologies”, leading us to suggest that overlaps on two broken ends might be used to prime specific joins 55.

thumbnail image

Figure 2. The “Alt–Baltimore” model for V(D)J recombination: The figure is reproduced with permission from 10.

Download figure to PowerPoint

The final question to be answered was the origin of the non-germ-line nucleotides in the junctions. I had already worked in molecular cloning long enough to know a likely candidate and that the candidate was an enzyme that caught David Baltimore's attention for a long time. During the dinner with David to develop the model, I asked, with expectation, what he thought was a good candidate for the activity that added the extra nucleotides. David's eyes lit up and he said what I expected: “TdT” (terminal deoxynucleotidyl transferase). Much later, studies from Toshi Komori (now a Professor in Japan) in my laboratory and from the Diane Mathis and Christophe Benoist laboratory proved our model that terminal deoxynucleotidyl transferase evolved specifically to add diversity at V(D)J junctions 57, 58.

Once again, David was great with giving a name that stuck. When we were writing the paper, I was puzzling through more cumbersome names for the region, such as non-germ-line-encoded nucleotides and David said that we would simply call them the “N” regions (with “N” denoting nucleotides). N regions are now known to represent a major source of antibody and TCR diversity. It is pleasing that this early model for V(D)J recombination has stood the test of time.

Completion of the V(D)J recombination reaction: Non-homologous end-joining

  1. Top of page
  2. Abstract
  3. Introduction: The Novartis Basic Immunology Prize
  4. The early days of V(D)J recombination
  5. Drug resistance and gene amplification
  6. The early days of B cell biology in the Baltimore laboratory
  7. Differential RNA processing, J chain and Marion Koshland
  8. Allelic exclusion
  9. A-MuLV transformants: A surprisingly accurate model for early B cell development
  10. The role of IgH μ chains in signaling B cell developmental events
  11. Accessibility regulation of V(D)J recombination
  12. DNA double-strand breaks and N regions: The V(D)J recombination mechanism
  13. Completion of the V(D)J recombination reaction: Non-homologous end-joining
  14. Concluding statements: Gene amplification, V(D)J recombination, and back
  15. Acknowledgements
  16. Appendix

The V(D)J recombination model that we generated predicted that the reaction would be initiated by a specific endonuclease that makes a break between coding sequences and RSs, and also suggested a possible role for general DNA repair components in the reaction. David Baltimore's later studies showed that V(D)J recombination depends on the RAG gene proteins (RAG-1 and RAG-2) 46, which were ultimately shown through the work of Martin Gellert and colleagues and others to be the endonuclease that introduces the DSB between V, D, and J segments and flanking RSs 59. Our group focused on the joining mechanism. Barbara Malynn, another early postdoctoral fellow in my laboratory, generated evidence that the protein affected by the mouse SCID mutation was involved in a downstream step of the joining phase of V(D)J recombination 60. Work from Bob Philips and colleagues and from David Weaver showed that the mouse SCID mutation conferred general radio-sensitivity, suggesting that the affected activity might be involved in general DNA repair 61, 62.

We moved our laboratories to the Children's Hospital and Center for Blood Research (now the Immune Disease Institute) at Harvard Medical School in 1991 (where we have remained). There, Guillermo Taccioli, a postdoctoral fellow in my laboratory, completed long and very insightful effort aimed at identifying DNA repair pathways that might be involved in V(D)J recombination. His effort paid off when he showed that a set of ionizing radiation-sensitive cell lines, harboring different (complementing) mutations, also had defects in their ability to perform the V(D)J recombination reaction 6365. These findings clearly demonstrated that a general DSB repair pathway is used to fuse cleaved V, D, and J segments during V(D)J recombination.

Subsequently, along with our long-time collaborators (Penny Jeggo, Steve Jackson, and Thomas Stamato) we, and others, identified the mutant gene products (including the gene mutated in SCID mice), which were the first characterized non-homologous end-joining (NHEJ) factors 6671. Later, we (and others) were able to generate mice deficient for each of the various NHEJ factors, which, in turn, allowed proof of their role in V(D)J recombination and elucidation of their functions in other cellular processes including suppression of genomic instability, lymphomas, and other cancers 7274.

Concluding statements: Gene amplification, V(D)J recombination, and back

  1. Top of page
  2. Abstract
  3. Introduction: The Novartis Basic Immunology Prize
  4. The early days of V(D)J recombination
  5. Drug resistance and gene amplification
  6. The early days of B cell biology in the Baltimore laboratory
  7. Differential RNA processing, J chain and Marion Koshland
  8. Allelic exclusion
  9. A-MuLV transformants: A surprisingly accurate model for early B cell development
  10. The role of IgH μ chains in signaling B cell developmental events
  11. Accessibility regulation of V(D)J recombination
  12. DNA double-strand breaks and N regions: The V(D)J recombination mechanism
  13. Completion of the V(D)J recombination reaction: Non-homologous end-joining
  14. Concluding statements: Gene amplification, V(D)J recombination, and back
  15. Acknowledgements
  16. Appendix

In the above paragraphs, I have outlined how I got started in my studies of B cell development and molecular immunology. This feature necessarily focused on very early work. I did not have the opportunity to discuss the very large amount of downstream work by my many other outstanding students and fellows, work that centered, among other things, on B and T development, regulation of V(D)J recombination, characterization of the NHEJ pathway of DSB repair, IgH class switch recombination, and on more general aspects of DNA repair and the DSB response in the maintenance of genomic stability and suppression of cancer.

During our early work at Columbia University, we were fortunate in being able to perform, in addition to our B cell and V(D)J recombination work, further studies of gene amplification and genomic instability. In 1978, shortly after joining the Baltimore laboratory, I had initiated with Cliff Tabin, who was a graduate student with Bob Weinberg and now is chairman of the Department of Genetics at Harvard Medical School, a subtractive hybridization strategy aimed at isolating a gene that we thought might be commonly amplified in human neuroblastomas. This effort was side-tracked by technical difficulties with reagents and also by the distracting success of some of the immunology experiments. Shortly after arriving at Columbia, the third of my first set of graduate students, Nancy Kohl (who now is a leader of oncology at Merck), finished that project (which was also done contemporaneously by Michel Bishop, Jeff Trent and their colleagues) and isolated the N-Myc gene, which is indeed commonly amplified in advanced-stage neuroblastoma 75, 76.

The isolation of N-Myc led another of my early fellows, Ron DePinho (now a Professor at Dana-Farber Cancer Institute and Havard Medical School), to perform studies that helped to define a Myc family of cellular oncogenes 77. Our more recent work showed that V(D)J recombination in NHEJ-deficient progenitor lymphocytes routinely leads to progenitor B cell lymphomas that have amplified Myc family genes as a result of RAG-dependent IgH translocations and helped to elucidate a mechanism for oncogene amplification 78, 79. The work also led us to more in depth studies of the DNA DSB response. I feel remarkably fortunate to have the two different arms of my research efforts, V(D)J recombination and B cell development, and genomic instability and cancer, come back to intersect so intimately over the past decade 14, thereby providing many exciting new models and ideas for studying these inter-related processes in the future.

  • 1

    WILEY-VCH

  • 2

    WILEY-VCH