Commentary

50 million years of chordate evolution: Seeking the origins of adaptive

Diana J. Laird*†, Anthony W. De Tomaso*, Max D. Cooper‡, and Irving L. Weissman*

*Department of Pathology, Stanford University School of Medicine, Stanford, CA 94305-5324; and ‡Howard Hughes Medical Institute, University of Alabama, 378 Wallace Tumor Institute, Birmingham, AL 35294

gnathans, the most primitive chor- ertoire of antigenic shapes the universe sup- to situate this finding from the laboratory of Adates, are poised at a fascinating point plies (2). We would expect that living ag- Jan Klein within the field of evolutionary in evolution. In the 50 million years between nathans, hagfish and lamprey, might retain . agnathan and chondrichthian divergence, the genetic raw materials for these inven- something mysterious, even miraculous oc- tions, but no conclusive evidence of MHC, Rearranging Receptors. Much atten- curred: the adaptive immune system Ig, TCR, or RAG elements exists in meta- tion has focused on the epicenter of rear- evolved. A complex, interdependent system zoans that diverged before chondrichthyes. rangement, RAG1͞2, the - of checks and balances in which fragments Luminaries in the field of evolution and specific proteins that create nicks between of both intracellular and extracellularly de- immunity describe the critical events that germline chromosomal V, D, and J compo- rived foreign molecules () are pre- befell our last common ancestor with nents of Igs and TCR. The curious genomic sented in the clutch of MHC surface sharks between 500 and 450 million years proximity of RAG1 and -2, their absence of molecules to 1016 possible different lympho- ago in cosmic terms. Synthesizing many introns (5), and their detection in sharks but cyte receptors arose seemingly overnight in recent findings, Schluter et al. (3) describe not protochordates first inspired the hypoth- evolutionary time. Moreover, the system the origin of the combinatorial immune esis of their origin by horizontal transfer to that evolved before the divergence of jawed system with the moniker ‘‘Immunological a vertebrate ancestor (6). Similarity between fish was so successful that the basic para- Big Bang.’’ As in astrophysics, the study of the mechanisms of DNA cleavage for RAG digm remained throughout the radiation of immunology must infer the events of the recombination, retroviral integration, and subsequent vertebrate lineages. past from the stuff—whether electromag- transposition corroborated the hypothesis The immune mechanisms shared by all netic or genetic—that remains today. This that RAG1 and RAG2 entered together on vertebrates are collectively termed ‘‘adap- endeavor is complicated by the possibility a retrotransposon (7–11). It is thought that tive immunity,’’ ‘‘combinatorial immunity,’’ that genetic mechanisms and selective a retrotransposon containing RAG1 and -2 or the ‘‘anticipatory response’’ (1) for the forces exerted by parasites differed in the integrated into a large cis promoter ele- unique receptors generated in response to era of early vertebrates—as did the physics ment. Once translated, RAG proteins me- foreign invaders, or antigens, such as mi- of the nascent universe. Although molec- diated transposition of a segment crobes, parasites, and genetically altered ular evolutionists rely on the genomes of flanked by short recombination signal se- cells. Unlike the innate immune compo- extant organisms to peer back in time, this quences (RSS) into a primordial receptor nents found in all multicellular organisms, is commonly misunderstood to be like gene, first splitting the gene into fragments this response is specific, selective, remem- gazing at the light from stars millions of that became V, D, and J through duplication bered, and regulated. T and B , light years away. The genome of living (12). Whether RAG fortuitously planted which bear unique immunoreceptors, circu- sharks is neither ancestral to nor older itself into a uniquely lymphoid promoter or late throughout the body in search of anti- than that of humans, but their comparison acquired specificity later is unclear. gens. B cell receptors, called immunoglobu- may reveal structures that existed in their Non-Rearranging Ig-Like Receptors. lins (Ig), recognize intact macromolecular last common ancestor 450 million years Ig and TCR molecules feature a structural motif complexes on the invaders; the cell progeny ago. comprised of ␤-pleated sheets, known as the of B cells secrete massive amounts of the With the unavailability of either fossilized ‘‘Ig fold,’’ whose archtypical presence in specific Igs, known as , that per- Cambrian nucleic acid or time travel, evo- receptors and adhesion molecules through meate the intercellular milieu. By contrast, lutionary immunologists primarily rely on phylum Chordata and even in invertebrates receptors (TCRs) recognize small comparative tools using extant species. Re- (13, 14) dates it before the entrance of fragments of antigen presented in the cent investigations into the phylogenetic or- RAG. Many investigations have ap- groove of the MHC (major histocompati- igin of this complex system derive from four proached the hypothesized primordial re- bility complex) receptor found on nearly different approaches: (i) the study of the ceptor by searching for relics of Ig-type every cell in the body. TCR and Ig are evolution of rearranging receptors, (ii)the molecules in classes of organisms that di- encoded by interspersed V, D, and J subge- study of the origin and assembly of MHC, verged before agnathans. Sequences iso- neic elements that are combinatorially re- (iii) the study of the origins of self-nonself lated from hagfish (15, 16), tunicate (17), arranged during lymphocyte development; recognition systems such as histocompati- and sponge (14, 18) aligned with canonical rearrangement is initiated by a unique DNA bility, and (iv) the study of lymphocyte phy- Ig domains produce a similarity score that splicing enzyme complex, the recombinase logeny. A paper in this issue of PNAS by falls into the ‘‘twilight zone’’ of questionable activating (RAGs). Each lymphocyte Shintani et al. (4) advances our understand- expresses about 105 identical products of ing of the cellular mediators of adaptive one successful V(D)J rearrangement pair, immunity by tracing the origins of the Spi See companion article on page 7417. and collectively the antigen receptors of T family of lymphocyte-specific transcription †To whom reprint requests should be addressed. E-mail: and B lymphocytes anticipate the full rep- factors in the jawless fish. Here we attempt [email protected].

6924–6926 ͉ PNAS ͉ June 20, 2000 ͉ vol. 97 ͉ no. 13 Downloaded by guest on October 9, 2021 homology (19). Another line of investiga- MHC. It is difficult to imagine a need for lead to identification of these histocompat- tion begins with the analysis of vertebrate peptide-presenting MHC receptors before ibility genes. non-rearranging Ig-like receptors in hopes the existence of TCRs that simultaneously of gleaning information about the structure bind MHC and peptide. To date, MHC Lymphocytes. Another approach to the of the ancestral pre-RAG receptor. One receptors have not been isolated in species problem of immune evolution is to inquire such group of related receptors includes the that lack RAG or TCR. However, each type into the phylogenetic origins of the cellular paired Ig-like receptors (PIR) in mice (20, of MHC receptor resides in its own genetic bearers of recombined receptors. One 21), the Ig-like transcripts (ILT) in humans locus spanning 2,000–3,000 kb and contain- would like to know the characteristics of [also called leukocyte inhibitory receptors ing many linked immune and some non- lymphocytes in that ancestral recipient of (LIR) or monocyte inhibitory receptors immune genes, including complement and the first RAG transposon; whether separate (MIR) (22)], and the killer inhibitory recep- inflammatory genes. Because many of the T and B cell lineages existed before TCR tors (KIR) of humans (23). The PIR, ILT, genes associated with vertebrate MHC loci and Ig; and what mechanisms of lymphocyte are phylogenetically conserved and perform fate determination were in place. The paper and KIR multigene families are located in functions common to all metazoans, it is by Shintani et al. (4) in this issue makes an syntenic regions of the mouse and human feasible to test linkage between homologs of important foray into this question with the genomes. Members of these receptor fami- MHC-related genes in nonvertebrate spe- identification and characterization of the lies vary slightly in their extracellular Ig cies (2). In the earliest known chordates, the ancestor to a family of lymphocyte transcrip- domain, and they come in two functionally protochordates, close linkage is not ob- tion factors called Spi. In mammals, three distinct forms. They either possess intracel- served between two HSP70 genes, which are known family members, Spi-1 (also called lular tyrosine-based inhibitory motifs contained in the MHC region of verte- PU.1), Spi-B, and Spi-C, are present at (ITIM) to serve as inhibitory receptors, or brates, or to the self-nonself histocompati- various points of lymphocyte lineage and they have transmembrane region polarity bility system called Fu͞HC (refs. 25 and 26; interact via an Ets DNA-binding domain to allowing them to associate with another see below). It remains to be tested whether turn on many genes important in lympho- transmembrane chain containing a ty- homologs of other MHC-associated genes cyte development. Jan Klein’s group has rosine-based activation motif (ITAM) to such as TNF, LMP, TAP, complement, and isolated a single homolog of the Spi genes in form a cell activating receptor complex. The collagen lie in genetic proximity in ag- the jawless fish, as verified by both amino ligands so far identified for these pairs of nathans or protochordates. The present acid similarity and exon structure. Data not activating and inhibitory receptors include data suggest that the MHC assembled very presented in the paper confirm that no other MHC class I and class I-like molecules. The quickly during the 50 million years of ‘‘big family members with similar Ets domains sequence diversity, polymorphic nature, and bang’’ in question. exist in the hagfish, which lends credence to counterfunctional capability of these recep- the hypothesis that this gene is derived from tors make them reasonable candidates for a Other Histocompatibility Systems. Although a single common ancestor to all three ver- primitive allorecognition system. Although the MHC and adaptive immunity appar- tebrate forms. the KIR family of receptors in humans ently first coexist in the chondrichthians, The present study raises many new ques- histocompatibility is a broader phenomenon tions about the ancestral lymphocytes that

probably arose as recently as 40 million COMMENTARY years ago, the identification of PIR͞ILT in multicellular organisms. Natural tissue became T and B cells through transforma- relatives in birds, the chicken paired Ig-like compatibility systems are well documented tion with the RAG genes. Very little is receptors (CHIR),§ suggests a much earlier in organisms ranging from slime molds, known about the functional properties of which spend the majority of their lives as prevertebrate lymphocytes. Data from te- origin for this family of Ig-like activating and single cells and only aggregate to sexually leost and shark suggest that ancient lympho- inhibitory receptors. reproduce (27), to sponges, the simplest cytes had a spontaneous or NK-like cyto- Another large multigene family of novel metazoans (28), and more complex inverte- toxicity against parasites (30), but neither immune-type receptors (NITR) has been brate phyla (cnidarians) to the preverte- the mechanism of recognition nor the mode identified in the pufferfish (24). These non- brates, or protochordates (e.g., ascidians). of killing are known. Parsimony argues that rearranging Ig-like genes encode patterned However, with little exception, nothing is the ontogeny of lymphocytes is conserved sequence variability in their V-like domains. known about the molecular mechanisms across evolution, but how did the present Many members of this family possess ITIM that underlie allorecognition in phyla that system of lineage determination evolve? In sequences in their cytoplasmic region predated vertebrates. mouse, one can isolate a pure population of whereas others may have polar transmem- Protochordates, which occupy a strategic common lymphoid progenitors that use spe- brane sequences that would allow them to phylogenetic position at the cusp of chor- cific cytokine receptors (e.g., IL-7, Flt-3, associate with ITAM-bearing partners. date evolution, provide an excellent model SLF) to interact with bone marrow stroma Apart from shared general features that to study the possible origins of vertebrate (31) and that express distinct signal trans- support their candidacy as primitive al- adaptive immunity. Colonial protochor- duction molecules and transcription factors lorecognition receptors, the NITR and PIR- dates are endowed with a genetically regu- (TFs) (32). Do similar microenvironments types of receptor genes are not obviously lated histocompatibility system, termed Fu͞ in the agnathan tissues support similar lym- related, and data acquired to date suggest it HC, in which a single Mendelian locus with phoid progenitors? Furthermore, is the sub- is unlikely they are located in syntenic chro- hundreds of codominantly expressed alleles sequent differentiation of those cells di- mosomal regions (G. W. Litman, personal controls fusion or rejection of allogeneic rected by homologous TFs? Detective work communication). In the ongoing search for individuals (25). Although early attempts using TFs may be complicated. Ikaros, for non-rearranging Ig-like allorecognition re- failed to identify the protochordate histo- example, is a family of TFs with many ceptors that may predate the TCR and Ig compatibility genes by homology to verte- known mammalian isoforms that interact receptors, it is interesting to note the up- brate MHC receptors, modern genomics with one another as well as many promoter places us on the brink of identifying the key elements. Although early investigations regulation of two Ig-like receptor molecules ͞ at the interface of sponge autografts (14). players via positional cloning. The Fu HC identified Ikaros as a lymphocyte-specific locus in the protochordate Botryllus schlos- TF (33), the Ikaros gene has a number of seri has been mapped within a 1-cM region zinc finger encoding domains, and virtually §Dennis, G., Kubagawa, H. & Cooper, M. D. (2000) FASEB J. of the genome (29, 2), and the analysis of all bloodforming stem and progenitor cells, 14, 1020 (abstr.). expressed sequences in this region should as well as lymphocytes, express different

Laird et al. PNAS ͉ June 20, 2000 ͉ vol. 97 ͉ no. 13 ͉ 6925 Downloaded by guest on October 9, 2021 metazoan parasites, etc.), many complex invertebrate metazoans have an equally pressing danger to their genomic integrity— stem cells from other members of their species (38). Many protochordate and other invertebrate metazoans can fuse their vas- cular systems with other individuals, open- ing their bodies to foreign cells, including germline stem cells (39, 40). As might be expected, selection for more predatory germline stem cells occurs, and the only barriers the species have from this spread- ing, species-homogenizing protoneoplasm is a highly genetically polymorphic histocom- patibility barrier, such as Fu͞HC (25), that limits genome sharing to siblings (40). Thus, the kind of histocompatibility immune sys- tem that allows sibling cell lineage transfer (by a shared Fu͞HC allele) but prevents Fig. 1. Dendrogram showing evolutionary relationships between selected animal phyla and classes invasion by cells with no common allele according to the current model (http:͞͞phylogeny.arizona.edu͞tree͞phylogeny.html). Speculative origin might have been the primordial alloimmune times of adaptive immune structures are indicated, as are documented histocompatibility systems. state in the protochordate ancestor that Divergence times are not to scale unless indicated. bordered the emerging chordate phyla (Fig. 1). Finally, it should be pointed out that collections of splice-variant isoforms (34). tebrate cytokines to ask whether candidate there are serious limits to morphological Perhaps an easier route is found by exam- lymphocytes respond to proliferative factors descriptions of lymphoid cells in both the ining the downstream genes regulated by from other species. contexts of ontogeny and phylogeny. Nei- the transcription factor in question. In mam- Perhaps another pathway to discovery of ther microscopic (histological or immuno- mals, Spi activates transcription of several B the phylogenetic origins of adaptive immu- histochemical) nor molecular (collections cell genes CD72, LSP1, CD20, BTK, mb-1, nity will come from a reconsideration of of transcription factors, etc) morphologies ␮ HC, ␬ and ␭ LCs, and the J chain (35), the immune system functions critical to the sur- can substitute for the isolation of the cells last of which may be identifiable in lamprey vival and the genomic integrity of each in question, the identification of their by hybridization to known homologs in examined species. Although the prevailing ontogenetic precursors and progeny, and chicken and segmented worm (36, 37). The paradigm of vertebrate immune and MHC the direct demonstration between species study of phylogenetically ancient lympho- system function is the protection from par- of their morphological and functional cytes might exploit the conservation of ver- asites with distinct genomes (microbes, similarity (41).

1. Klein, J. (1990) Immunology (Blackwell Scientific, Natl. Acad. Sci. USA. 88, 1746–1750. 28. Humphreys, T. & Reinherz, E. L. (1994) Immunol. Boston.). 16. Raison, R. L. & Hildemann, W. H. (1984) Dev. Today 15, 316–320. 2. Magor, B. G., De Tomaso, A., Rinkevich, B. & Comp. Immunol. 8, 99–108. 29. De Tomaso, A. W., Saito, Y., Ishizuka, K. J., Weissman, I. L. (1999) Immunol. Rev. 167, 69–79. 17. Pancer, Z., Diehl-Seifert, B., Rinkevich, B. & Palmeri, K. J. & Weissman, I. L. (1998) 3. Schluter, S. F., Bernstein, R. M., Bernstein, H. & Muller, W. E. (1997) DNA Cell Biol. 16, 801–806. 149, 277–287. Marchalonis, J. J. (1999) Dev. Comp. Immunol. 23, 18. Pancer, Z., Skorokhod, A., Blumbach, B. & Muller 30. Haynes, L. & McKinney, E. C. (1991) Dev. Comp 107–111. W. E. (1998) Gene 207, 227–233. Immunol. 15, 123–134. 31. Kondo, M., Weissman, I. L. & Akashi, K. (1997) 4. Shintani, S., Terzic, J., Sato, A., Saraga-Babic, M., 19. Doolittle, R. F. (1987) URFs and ORFs: A Primer Cell 91, 661–672. O’hUigin, C., Tichy, H. & Klein, J. (2000) Proc. on How to Analyze Derived Amino Acid Sequences 32. Akashi, K., Traver, D., Miyamoto, T. & Weissman, Natl. Acad. Sci. USA 97, 7417–7422. (University Science Books, Mill Valley, CA). I. L. (2000) Nature (London) 404, 193–197. 5. Oettinger, M. A., Schatz, D. G., Gorka, C. & 20. Kubagawa, H., Burrows, P. D. & Cooper, M. D. 33. Georgopoulos, K., Bigby, M., Wang, J. H., Molnar, Baltimore, D. (1990) Science 248, 1517–1522. (1997) Proc. Natl. Acad. Sci. USA 94, 5261–5266. A., Wu, P., Winandy, S. & Sharpe, A. (1994) Cell 6. Bartl, S., Baltimore, D. & Weissman, I. L. (1994) 21. Hayami, K., Fukuta, D., Nishikawa, Y., Ya- 79, 143–156. Proc. Natl. Acad. Sci. USA 91, 10769–10070. mashita, Y., Inui, M., Ohyama, Y., Hikida, M., 34. Klug, C. A., Morrison, S. J., Masek, M., Hahm, K., 7. Schatz, D. G. (1999) Immunol. Res. 19, 169–182. Ohmori, H. & Takai, T. (1997) J. Biol. Chem. 272, Smale, S. T. & Weissman, I. L. (1998) Proc. Natl. 8. Lewis, S. M. & Wu, G. E. (1997) Cell 88, 159–162. 7320–7327. Acad. Sci. USA 95, 657–662. 9. Thompson, C. B. (1995) Immunity 3, 531–539. 22. Colonna, M., Nakajima, H. & Cella, M. (2000) 35. Rao, S., Garrett-Sinha, L. A., Yoon, J. & Simon, 10. Hiom, I., Melek, M. & Gellert, M. (1998) Cell 94, Semin. Immunol. 12, 121–127. M. C. (1999) J. Biol. Chem. 274, 34245–34252. 463–470. 23. Long, E. O. (1999) Annu. Rev. Immunol. 17, 36. Takahashi, T., Iwase, T., Takenouchi, N., Saito, M., 11. Agrawal, A., Eastman, Q. M. & Schatz, D. G. 875–904. Kobayashi, K., Moldoveanu, Z., Mestecky, J. & (1998) Nature (London) 394, 744–751. 24. Strong, S. J., Mueller, M. G., Litman, R. T., Moro, I. (1996) Proc. Natl. Acad. Sci. USA 93, 12. Yu, W., Nagaoka, H., Jankovic, M., Misulovin, Z., Hawke, N. A., Haire, R. N., Miracle, A. L., Rast, 1886–1891. 37. Takahashi, T., Takashi, I., Tachibana, T., Suh, H., Rolink, A., Melchers, F., Meffre, E. & J. P., Amemiya, C. T. & Litman, G. W. (1999) Komiyama, K., Kobayashi, K., Chen, C., Mestecky, Nussenzweig, M. C. (1999) Nature (London) 400, Proc. Natl. Acad. Sci. USA 96, 15080–15085. J. & Moro, I. (2000) Immunogenetics 51, 85–91. 682–687. 25. Scofield, V., Schlumpberger, J. M., West, L. A. & 38. Buss, L. W. (1987) The Evolution of Individuality 13. Williams, A. F. & Barclay, A. N. (1988) Annu. Rev. Weissman, I. L. (1982) Nature (London) 295, (Princeton Univ. Press, Princeton). Immunol. 6, 381–405. 499–502. 39. Stoner, D. S. & Weissman, I. L. (1996) Proc. Natl. 14. Blumbach, B., Diehl-Seifert, B., Seack, J., Steffen, 26. Fagan, M. B. & Weissman, I. L. (1998) Immuno- Acad. Sci. USA 93, 15254–15259. R., Muller, I. M. & Muller, W. E. G. (1999) genetics 47, 468–476. 40. Stoner, D. S., Rinkevich, B. & Weissman, I. L. Immunogenetics 49, 751–776. 27. Buss, L. W. (1982) Proc. Natl. Acad. Sci. USA 79, (1999) Proc. Natl. Acad. Sci. USA 96, 9148–9153. 15. Varner, J., Neame, P., Litman, G. W. (1991) Proc. 5337–5341. 41. Weissman, I. L. (1994) Immunity 1, 529–531.

6926 ͉ www.pnas.org Laird et al. Downloaded by guest on October 9, 2021