Unusually Similar Patterns of Antibody V Segment Diversity in Distantly Related Marsupials
http://www.100md.com
免疫学杂志 2005年第9期
Abstract
A pattern of coevolution between the V gene segments of Ig H and L chains has been noted previously by several investigators. Species with restricted germline VH diversity tend to have limited germline VL diversity, whereas species with high levels of germline VH diversity have more diverse VL gene segments. Evidence for a limited pool of VH but diverse VL gene segments in a South American opossum, Monodelphis domestica, is consistent with this marsupial being an exception to the pattern. To determine whether M. domestica is unique or the norm for marsupials, the VH and VL of an Australian possum, Trichosurus vulpecula, were characterized. The Ig repertoire in T. vulpecula is also derived from a restricted VH pool but a diverse VL pool. The VL gene segments of T. vulpecula are highly complex and contain lineages that predate the separation of marsupials and placental mammals. Thus, neither marsupial follows a pattern of coevolution of VH and VL gene segments observed in other mammals. Rather, marsupial VH and VL complexity appears to be evolving divergently, retaining diversity in VL perhaps to compensate for limited VH diversity. There is a high degree of similarity between the VH and VL in M. domestica and T. vulpecula, with the majority of VL families being shared between both species. All marsupial VH sequences isolated so far form a common clade of closely related sequences, and in contrast to the VL genes, the VH likely underwent a major loss of diversity early in marsupial evolution.
Introduction
The complexities of the Ig loci make their evolution, genetics, and expression the focus of attention. The typical Ig molecule is composed of two identical H and two identical L chains. Variation in amino acid residues at the N-terminal ends of the H and L chains contributes to Ab diversity and establishes Ab specificity. This variation is initially created through the process of so-called V(D)J recombination where gene segments are recombined somatically to generate an exon encoding the V domain. The H chain V domains are assembled by somatic recombination of three gene segments called VH, DH, and JH, whereas the L chain V domains are assembled from two gene segments, VL and JL. The arrays of V, D, and J segments available for recombination evolve through a process of gene duplication, divergence, and deletion (1). All three living lineages of mammals, the eutherians (placentals), metatherians (marsupials), and prototherians (monotremes), have essentially the same classes of Ig H and L chains. This includes having two IgL loci, designated Ig and Ig (2, 3, 4). In humans and mice, there is a relatively large and diverse set of VH, V, and V gene segments, which are one contribution to the generation of a primary Ig repertoire through V(D)J recombination. The human IgH locus contains 44 VH segments with complete open reading frames, 39 of which are known to be functional, and that comprise seven diverse VH families (5). Human VH can share as little as 47% average interfamily nucleotide identity (6). In contrast, other species (e.g., rabbits, cattle, platypus, and chickens) have limited germline VH diversity. In these cases, the germline VH within a species typically share >80% nucleotide identity, and may be only a single VH family or a set of recently derived VH families (7, 8).
Several investigators have noted a coevolutionary relationship between the H and L chain loci in some species (1, 7, 9, 10). Those species with a limited number of germline VH families (e.g., rabbits and cattle) also have limited numbers of germline VL families. Conversely, those species with multiple, diverse VH families (e.g., humans, sheep, and mice) have similarly complex germline VL families (1, 7, 9, 10). This pattern of similarity seen in VH and VL segments is sufficiently striking to lead some authors to propose that it is not by chance but the result of factors driving the coevolution of the IgH and IgL loci (1, 9, 10). What factor or factors these may be is not clear.
Marsupial Ab diversity has been analyzed for only a single species, the gray short-tailed opossum, Monodelphis domestica. Unlike other mammals and nonmammals examined to date, M. domestica has a contrasting diversity between its germline H and L loci. The M. domestica VH repertoire is derived from recombination of a set of V segments limited both in diversity and number. M. domestica has <15 germline VH segments, all but one of which belong to a single VH family designated VH1 (11). The lone VH segment that is not a VH1, is still closely related, but has been designated to a separate family, VH2 (11). In contrast, M. domestica has a highly diverse pool of VL segments from which to recombine the expressed VL domains. There are at least three V and four V families, and they contain together >60 V gene segments (2, 3). The V and V gene families in M. domestica are diverse and are derived from gene duplications that predate the separation of marsupials and eutherians. Thus, the IgH and IgL loci of M. domestica appear to be evolving divergently with respect to their complexity, rather than converging on similar patterns of germline diversity. To determine whether M. domestica is an exception among species or whether this pattern is common among marsupials, we analyzed the VH and VL repertoire of a second distantly related marsupial, the Australian brushtail possum, Trichosurus vulpecula.
Materials and Methods
PCR amplification of VH, V, and V sequences
A T. vulpecula mesenteric lymph node cDNA library constructed using the ZAPII vector has been described previously (12). The strategy to isolate V region sequences in a random manner from the cDNA libraries is also described elsewhere (2, 3, 11). Briefly, oligonucleotide primers complementary to the C regions of Ig, Ig, and IgH were paired with primers complementary to a region of the vector upstream of the 5' start of the cDNA. T. vulpecula V segments were amplified using a primer specific for marsupial C (5'-TGGTTGGAAGATGAAGGCAG-3') in combination with a primer complementary to the M13 universal sequencing site in ZAPII. V segments from T. vulpecula were amplified using a primer designed specifically for T. vulpecula C (5'-ACATTAACGGTGGGGCTTGC) in combination with a primer complementary to the T7 site in ZAPII. The T. vulpecula cDNA library was used as a target in PCR, and successful amplifications were achieved using 1.5 mM MgCl2 at an annealing temperature of 62°C and 65°C for C and C, respectively, using Taq polymerase (PerkinElmer) for 35 cycles.
T. vulpecula VH fragments were isolated using oligonucleotide primers complementary to T. vulpecula Cμ (5'-CCCACAGCCACAGGGGCATC) and C (5'-AATTCCATGTCACTGTCAC) in combination with the M13 primer. Successful amplifications were achieved using 1.5 mM MgCl2 at 68 and 53°C, respectively, using the T. vulpecula cDNA library as target. PCR products were cloned into the pCR2.1 vector (Invitrogen Life Technologies). Additional T. vulpecula sequences were obtained from phage clones isolated by screening the T. vulpecula cDNA library with an IgG C region probe as described in Belov et al. (13). Positive plaques were used as a target in PCR using a C primer (5'-AGGAGGTTCCCTGTCCACAATTG-3') paired with a primer specific for the T3 site in the ZAPII vector. Successful amplifications were achieved at an annealing temperature of 55°C using Taq polymerase (PerkinElmer) for 35 cycles, and PCR products were cloned into the pGEM-T vector (Promega).
Probes and hybridizations
An Isoodon macrourus spleen cDNA library was constructed using the SMART cDNA construction kit and the TripleX vector (Clontech). Screening of the I. macrourus spleen cDNA library by hybridization was done by plating the phage and preparing plaque lifts on reinforced nitrocellulose. An M. domestica VH probe (MVH356) previously described (11) was used to screen the I. macrourus cDNA library. Probes were constructed from DNA inserts excised from agarose gels and labeled with [32P]dCTP by the Random-Prime-It RmT labeling kit (Stratagene). Plaque lifts were hybridized under conditions of 42°C, 50% formamide, 5x Denhardt’s solution, 5x SSC, 50 mM NaPO4 (pH 6.5), 0.1% SDS, 5 mM EDTA, and 250 mg/ml sheared salmon DNA. Final washes were performed at 65°C in 0.2x SSC, and lifts were autoradiographed at –80°C for 1–4 days.
Sequencing and phylogenetic analysis
Sequencing was performed using the Big Dye sequencing kit (Amersham Biosciences) and analyzed on an automated DNA sequencer (PerkinElmer ABI Prism 377 DNA Sequencer). Minor manual corrections of sequences were made using the Sequencher 4.1 program (Gene Codes Corporation). Comparisons to known sequences in GenBank database were made using the BLAST algorithm (14).
Sequences were aligned using either the Clustal X program (15) with minor manual adjustments or with the Clustal W function bundled within BioEdit (16). For phylogenetic analyses, alignments were made using sequences corresponding to the framework regions (FR) 4 1–3 of the V domains. All phylogenetic tree reconstruction shown in this paper is based on analyses done using nucleotide alignments. Gaps in the nucleotide sequences were determined by first aligning the amino acid translations to establish gap position and then converting the sequence back to nucleotide using the BioEdit program. In this way, nucleotide gaps were established based on codon position. However, analyses done with nucleotide sequence alignments gapped independently of codon position yielded identical results. Based on the nucleotide alignments, phylogenetic trees were constructed by the neighbor joining (NJ) method of Saitou and Nei (17), maximum parsimony (MP), and minimum evolution (ME) using the MEGA2 program (18). NJ and MP analyses using MEGA2 were performed using 1000 bootstrap replicates and the Kimura two-parameter and min-mini heuristic search, respectively.
VL used in the analyses
Representative opossum M. domestica (Modo) V and V family sequences were as follows: VLI, AF049774; VLII, AF049781; VLIII, AF049756; VKI, AF116930; VKII, AF116924; VKIII, AF116931; and VKIV, AF116935. Human, Homo sapiens (Hosa) V and V were obtained from the VBASE database (6). Mouse, Mus musculus (Mumu) V were as follows: VL1, X82687; VLX, D38129; mouse V were as follows: IGKVR1, X13938; IGKCLM, Z72384; IGKCAM2, M24937; IGVKID, M63611. Rat, Rattus norvegicus (Rano) V was U39609. Hamster, Cricetulus migratorius (Crmi) V was U17165. Horse, Equus caballus (Eqca), VK1 was X75611. Sheep, Ovis aries (Ovar) VK was X54110. Rabbit, Oryctolagus cuniculus (Orcu), were as follows: VL2, M27840; and VL3, M27841. Chicken, Gallus gallus (Gaga) V was M96972. Heterodontus fransciscii (Hefr) V was X15316. Platypus, Ornithorhynchus anatinus (Oran) V were as follows: AF525088, AF525090-AF525092, AF525095-AF525097, AF525100, AF525102-AF525106, AF525108, AF525109, AF525111, AF525115, AF525116, AF525118, AF525119, and AF525123.
VH from other species
Previously published T. vulpecula (Trvu) sequences were as follows: TrvuFA1, AF091140; and TrvuFA2, AF091141. M. domestica VH1 sequences were as follows: MVH10, AF012121; MVH11, AF012122; MVH26, AF012113; and M. domestica VH2 sequence, AF012124. The Didelphis virginiana (Divi) VH is unpublished and was provided by Dr. R. Riblet (Torrey Pines Institute for Molecular Studies, La Jolla, CA). Human, H. sapiens (Hosa), VH sequences were obtained from the VBASE database (6). Mouse, M. musculus (Mumu), VH family representatives were as follows: 7183, U04227; 3660, K01569; 3609N, X55935; DNA4, M20829; J558, Z37145; J606, X03398; Q52, M27021; S107, J00538; SM7, M31285; VH11, Y00743; and X24, X00163. Cow, Bos taurus (Bota) VH was AF015505. Sheep, O. aries (Ovar) VH was Z49180. Pig, Sus scrofa (Susc) VH was U15194. Horned shark, H. fransciscii (Hefr) VH, X13449. Platypus, O. anatinus (Oran) VH were as follows: AF3811324, AF381290-AF381296, AF381298, AF381299, AF381303-AF381308, AF381310-AF381315, AF381317-AF381320, AF381322, AY055778, AY055779, AY055780-AY055782, AY168639, AAL59586 Echidna, Tachyglossus aculeatus (Taac) VH were as follows: AAM61760 AAM60800 AAM60783 AY101439, AY101442, and AY101445.
Results
Limited VH diversity in T. vulpecula
To sample the expressed Ig H chain diversity in the brushtail possum, VH sequences were isolated using a combination of hybridization and PCR amplification. IgH C region sequences were already available for T. vulpecula in GenBank (19). A total of 16 unique T. vulpecula VH sequences were analyzed. Nine were isolated by amplification using anchored PCR and reverse primers complementary to the 5' end of Cμ and C on a T. vulpecula mesenteric lymph node cDNA library. The Cμ and C primers were paired with forward primers in the phage vector so as not to bias the VH sequences amplified. Five clones containing VH were isolated by hybridization using a C fragment as a probe on the same library. Two T. vulpecula VH included in the analysis were available in GenBank (20). Pairwise analysis of these VH sequences revealed that all but two shared >80% nucleotide identity to each other and are by definition of the same VH family. The same result was found when the CDRs were excluded from the analysis, limiting the impact that somatic mutations may have played in family assignment (not shown). The most divergent T. vulpecula VH sequences still shared >78% nucleotide identity to the rest of the VH from that species. The T. vulpecula sequences were also compared with M. domestica VH and found to share >73%, and as much as 82%, interspecies nucleotide identity. VH sequences are available for only two other species of marsupials: a germline VH from an American marsupial, the Virginia opossum, D. virginiana (DiviP83 in Fig. 1) (11), and a VH from an Australian marsupial, the Northern Brown Bandicoot, I. macrourus (Isma5.2 in Fig. 1). The latter was obtained by screening a spleen cDNA library by hybridization (this study). Both the I. macrourus and D. virginiana VH sequences share a high degree of nucleotide identity with M. domestica VH1 sequences (69–83% for I. macrourus and 68–91% for D. virginiana). The most divergent marsupial VH is the M. domestica VH2 sequence, a VH from a family containing a single gene segment (11).
FIGURE 1. Phylogenetic tree based on alignments of VH sequences from T. vulpecula and representatives of other mammals and some nonmammals. The marsupial VH are shown in bold. This tree was constructed using the NJ method (17 ). The numbers at the branch nodes indicate bootstrap values based on 1000 replicates. Roman numerals on the right indicate the three VH groups (21 ). Species names are abbreviated to the first two letters of their scientific names: B. taurus (Bota), D. virginiana (Divi), H. sapiens (Hosa), H. fransciscii (Hefr), I. macrourus (Isma), M. domestica (Modo), M. musculus (Mumu), O. aries (Ovar), O. anatinus (Oran), S. scrofa (Susc), T. aculeatus (Taac), T. vulpecula (Trvu).
Phylogenetic analyses using alignments of the T. vulpecula VH sequences with VH from other marsupials and a variety of eutherian and prototherian mammals were performed (Fig. 1). A VH sequence from the horned shark, H. fransciscii, Hefr, was included in the analyses as outgroup to the mammalian VH. A phylogenetic tree reconstructed from the alignments using the NJ method is shown in Fig. 1. The mammalian VH cluster into three groups (I, II, and III in Fig. 1; Ref. 21), sometimes referred to as clades A, B, and C, respectively (1). All marsupial VH form a single, monophyletic clade within group III. The M. domestica VH2 sequence remains the most divergent marsupial VH in the cluster; M. domestica VH2 is sister to all other marsupial VH. The phylogenetic tree reveals that the T. vulpecula VH cluster as a single family, consistent with the pairwise nucleotide analyses described above. Furthermore, this single T. vulpecula VH clade is sister to the M. domestica VH1 clade. Similar results were obtained when the alignments were analyzed using MP and ME (not shown). In analyses using MP, the M. domestica VH2 sequence was sister to a clade that included all marsupial and platypus VH. Platypus, O. anatinus, Oran, VH form a single clade, and all platypus VH belong to essentially a single VH family that is related to the marsupial VH (Fig. 1). This is not the case for all monotremes, because echidna, T. aculeatus, Taac, VH are highly diverse and are present in all three mammalian VH groups (Fig. 1; Ref. 22). In conclusion, pairwise nucleotide identities and phylogenetic analyses of all marsupial VH sequences currently available reveal that two distantly related species have a closely related set of VH genes.
V and V diversity in T. vulpecula
Given the restricted nature of the T. vulpecula VH repertoire, we would predict that this species might also have a restricted VL repertoire; that is, if the pattern of coevolution described in the introduction occurs in this marsupial. To test this, the expressed diversity of V and V gene segments were sampled in T. vulpecula, using anchored PCR. C and C sequences have been described for T. vulpecula (23, 24). V sequences were amplified from the T. vulpecula mesenteric lymph node cDNA library using anchored PCR and an oligonucleotide primer complementary to the 5' end of C. A total of 40 clones containing partial or full V domains were isolated and sequenced. Each were derived from a unique V-J recombination event, and there were 25 different V gene sequences represented. An additional three V sequences, published previously, were included in the dataset (24). Pairwise analysis of the 19 most complete T. vulpecula V revealed the presence of four distinct V families (V1 to -4), sharing <80% nucleotide identity between families (Fig. 2a). Exclusion of the CDRs from the analysis did not change the assignment of sequences to a particular family (not shown). Most of the clones (14 of 19) contained V that belonged to the V1 family (Fig. 2a). The T. vulpecula V1 family was the most diverse, sharing between 81 and 96% intrafamily nucleotide identity.
FIGURE 2. Range of percent nucleotide identity between 19 T. vulpecula V (a) and 32 T. vulpecula V (b) sequences based on alignments of regions corresponding to FR1 through FR3 inclusive of CDRs. Numbers in parentheses indicate number of clones isolated from each family. Sequences used in analysis were as follows: V1: TrvuVL56, TrvuVL87, TrvuVL126, TrvuVL131, TrvuVL151, TrvuVL157, TrvuVL158, TrvuVL162, TrvuVL182, TrvuVL175, TrvuVL190, TrvuVL204, TrvuVL215, TrvuVL325; V2: TrvuVL1, TrvuVL10, TrvuVL11; V3: TrvuVL12; V4: TrvuVL33; V1: TrvuVk16, TrvuVk46, TrvuVk52, TrvuVk57, TrvuVk58, TrvuVk66, TrvuVk70, TrvuVk84, TrvuVk85, TrvuVk88, TrvuVk113, TrvuVk172, TrvuVk230, TrvuVk231, TrvuVk242; V2: TrvuVk6, TrvuVk30, TrvuVk103, TrvuVk153, TrvuVk228, TrvuVk252, TrvuVk259; V3: TrvuVk94, TrvuVk107, TrvuVk134, TrvuVk141, TrvuVk156, TrvuVk262; V4: TrvuVk68, TrvuVk176, TrvuVk218, TrvuVk236; and V5: TrvuVk158.
To sample the diversity of V in T. vulpecula, an oligonucleotide primer complementary to marsupial C was used in anchored PCR with the T. vulpecula cDNA library. A total of 43 clones containing partial or full unique V domains were isolated and sequenced, and these contained 31 different V gene sequences. An additional two published T. vulpecula V sequences were also included in this analysis (23). Pairwise analyses of all T. vulpecula V sequences (n = 33) revealed that these sequences shared between 65 and 99% nucleotide identity to each other (Fig. 2b). Five distinct T. vulpecula V families were identified based on the sharing of >80% nucleotide identity within a family (Fig. 2b). Again, analyses excluding the CDRs did not change the assignment of a sequence to a particular family (not shown).
Phylogenetic analyses of V and V were performed using an alignment of the T. vulpecula sequences with V and V from M. domestica, several eutherian species, and a prototherian. Only sequences with complete or nearly complete V regions were included in this analysis. Fig. 3 shows a tree of L chain V regions, where alignments of V and V sequences were combined in a single analysis. A VL region from the horned shark, H. fransciscii, was used as outgroup to the mammalian V and V, because elasmobranch VL genes have been shown to be the most ancient of the vertebrate VL genes (9, 25). The tree shown was reconstructed using the NJ method (Fig. 3); however, identical results were found when MP and ME were used (not shown). The overall topology of the tree is similar to that observed by others for mammalian VL (9). Specifically, there is the presence of two distinct, but weakly supported clusters of V and a single, well-supported V cluster nested within the V. Several observations can be made regarding the T. vulpecula V and V from the tree shown in Fig. 3. The T. vulpecula V sequences fall into the four separate clusters corresponding to the four families described previously (Fig. 2a). Interspersed between the four T. vulpecula V families are eutherian V, consistent with the establishment of these germline VL lineages before the separation of eutherians and marsupials. Additionally, three of the four V families in T. vulpecula (V1, -3, and -4) are each phylogenetically most closely related to a M. domestica V family (Fig. 3). The T. vulpecula V sequences fall into the five separate clusters within the V clade corresponding to the five V families based on nucleotide identity (Fig. 2b). The T. vulpecula V families are also interspersed among V sequences from eutherians. Three of the five T. vulpecula V families (V1, -2, and -4) are also each phylogenetically related to a M. domestica V family. In conclusion, an analysis of the V and V expressed in T. vulpecula reveals a diverse pool of gene segments from which to derive expressed IgL diversity in this species. Furthermore, much of the VL diversity present in T. vulpecula is ancient and overlaps with that of M. domestica.
FIGURE 3. Phylogenetic tree based on alignments of combined V and V sequences from T. vulpecula and representatives of other mammals and some nonmammals. All marsupial sequences are shown in bold in the tree. The position of the distinct T. vulpecula V and V families on the tree are indicated to the right as V1, V1, etc. This tree was constructed using the NJ method (17 ). The numbers at the branch nodes indicate bootstrap values based on 1000 replicates. Species names are as in Fig. 1: B. taurus (Bota), C. migratorius (Crmi), E. caballus (Eqca), G. gallus (Gaga), H. sapiens (Hosa), H. fransciscii (Hefr), M. domestica (Modo), M. musculus (Mumu), O. aries (Ovar), O. anatinus (Oran), O. cuniculus (Orcu), R. norvegicus (Rano), T. vulpecula (Trvu).
Discussion
We analyzed the general level of expressed Ab V diversity in the brushtail possum, T. vulpecula, to infer the germline V segment contribution to the repertoire. The results were compared with what was known from another, distantly related marsupial, M. domestica. M. domestica and T. vulpecula are members of the American Didelphidae family and the Australasian Phalangeridae families, respectively. Marsupials are a monophyletic group that once held wide Gondwanan distribution. Since the breakup of that supercontinent, the marsupial families were fragmented into two groups, the Australasian and the American marsupials (26). These two groups have now undergone a considerable period of independent evolutionary history. The Didelphidae and Phalangeridae have likely been separated for 67–75 million years (26, 27), putting divergence times for M. domestica and T. vulpecula on par with many of the eutherian lineages (27, 28, 29, 30). The overlap in their germline VH and VL repertoires, therefore, is not easily explained by recent common origins.
There are two results from the analysis of marsupial Ig VH and VL diversity that are particularly noteworthy. First, the VH sequences described so far from marsupials are strikingly similar between species. The same could be said for much of the VL diversity as well because many of the M. domestica and T. vulpecula V and V families are closely related. This is not an artifact of sampling because, in the case of M. domestica and T. vulpecula, we used C region sequences of both IgH and IgL to sample the upstream V region diversity. Second, both M. domestica and T. vulpecula share a pattern of diversity where there is limited VH gene segment diversity and extensive VL gene segment diversity. The first observation may have implications for the evolution of Ab genes early in the appearance of marsupials. The VH diversity that is present in many extant eutherians and one monotreme is ancient and predates the appearance of mammals (1). Collectively, mammalian VH fall into three clusters or groups (groups I, II, and III in Fig. 2). VH segments corresponding to all three groups can be also found in the amphibian Xenopus laevis, for example. This is consistent with the appearance of VH groups I, II, and III early in tetrapod evolution at the least (31). The presence of VH groups I, II, and III in some eutherians and one monotreme, the echidna (Fig. 2; Ref. 22), confirms that the last common ancestor of marsupials and eutherians would also have contained VH segments from all three groups. Over time, duplication and deletion (the so-called birth and death process) of VH segments resulted in some lineages retaining this ancient germline diversity, whereas others have lost it (1, 7). Many species have deleted much of the ancestral germline VH diversity and now have a limited set of related VH; in some species, there is only a single VH family remaining (7, 10). An example of such deletion is illustrated by the VH in cattle and swine. These two artiodactyls each have a single VH family; however, they do not have the same or even a closely related VH family. Cattle VH fall into the group I clade, whereas swine VH are group III (7). The VH in groups I and III arose from duplications that occurred likely >350 million years ago, before the appearance of mammals (10). The last common ancestor of cattle and swine must have had at least group I and III VH segments in their germline, and, following speciation, each deleted (or retained) VH from reciprocal lineages.
Contrasting this is the situation in marsupials. The VH sequences isolated from four species of marsupials form a monophyletic family within VH group III. In the case of T. vulpecula and M. domestica, the entire VH repertoires are similar by having a single, related VH family. Because the I. macrourus and D. virginiana VH were obtained by screening with V region probes, rather than the C region probes, there is a bias toward isolating clones having similar V regions, rather than a random sample of VH diversity. We can only conclude that I. macrourus and D. virginiana at least contain VH gene segments related to the M. domestica VH1 family. Whether this is all they have remains to be determined. How could marsupial species on separate continents and separated by at least 65 million years of evolution have unusually similar germline V repertoires? It could be by chance. Alternatively, the similarity of IgV families in M. domestica and T. vulpecula could reflect convergent selection acting on marsupial V segments. This seems unlikely given the time, distance, and habitats separating the American opossums from the Australian possums. Rather, a more parsimonious explanation is that the last common ancestor of M. domestica and T. vulpecula had already undergone a major loss of germline VH diversity resulting in a limited set of available VH segments before the separation of these two species. Given the phylogenetic position of the Didelphidae and Phalangeridae, it is likely that most or all extant marsupial families would have this common, restricted VH pool (26, 32). Such a loss of VH diversity can be treated as a form of a genetic bottleneck at the IgH locus that occurred very early in marsupial evolution, near the base of the extant marsupial family tree. In contrast, both species have maintained ancient VL diversity that must have been present in the earliest marsupials.
The second, but related observation, that M. domestica and T. vulpecula each have limited VH diversity paired with more diverse VL, is in contrast to the pattern of coevolution between VH and VL segments that has been observed in most species examined to date (1, 7, 9, 10). A pattern has emerged in eutherian mammals where species can be divided into two groups with respect to Ig diversity (reviewed in Ref. 7). One group are the species that generate a diverse primary Ig repertoire through the process of V(D)J recombination only, using a diverse pool of available germline V segments (e.g., primates and rodents). The second group are the species that have a limited number of germline V segments, usually of a single family, and rely on mutation, such as gene conversion or somatic mutation, to further diversify their primary Ig repertoire following V(D)J recombination (e.g., rabbits, some artiodactyla, and birds). These mutations occur at lymphoid tissue sites associated with the hindgut (e.g., the appendix in rabbits, the bursa of Fabricius in chickens; Refs. 33 and 34). One hypothesis to explain these restricted VH and VL repertoires is that the mutational mechanisms used to diversify the V domains have placed common constraints on VH and VL region segment evolution, limiting their germline diversity. Evidence from studies of sheep Ab diversity casts doubt on this hypothesis. Sheep have relatively high levels of germline VH and VL gene diversity (35), despite using mutation to further diversify their primary Ig repertoire in the Peyer’s patches (36). The role that mutations and gut associated lymphoid tissue play in generating diversity and development of the primary Ab repertoires in marsupials remains to be determined.
In marsupials, the VH and VL repertoires may be coevolving to compensate for each other. The lack of germline segment diversity in marsupial IgH is compensated for by higher levels of diversity at the IgL loci. With a severely restricted pool of VH segments to pair with, the L chains are taking up the responsibility of maintaining germline diversity. Crystallographic studies of human and mouse Abs have demonstrated that H chains contribute the greatest contributions to Ag binding (37). Therefore, it would be expected that H chain diversity is critical for generating an effective Ab repertoire. Indeed, one group of mammals, the Camelids, has altogether dispensed with the Ig L chains for some Ab isotypes, without apparent loss of Ab function or specificity. Such discoveries have led some investigators to question the importance of L chains to Ab diversity (reviewed in Ref. 38). In the two marsupial species studied to date, L chains appear to contribute more to the complexity of the Ab repertoire than do H chains. Marsupials as a group, therefore, are important model species to study how L chains can contribute to Ab diversity and to understanding L chain evolution. We would also suggest that the VH and VL segments that exist in an extant species may be shaped more by evolutionary history of the individual loci than by selection for specific patterns of coevolution.
Disclosures
The authors have no financial conflict of interest.
Footnotes
The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
1 This publication was supported by grants from the National Institutes of Health (Institutional Development Award program of the National Center for Research Resources; IP20RR18754) (supporting M.L.B.), the National Science Foundation (MCB-9981960) (to R.D.M.), and the Australian Research Council (to K.B.).
2 The sequences presented in this article have been submitted to EMBL/GenBank under the following accession numbers: for VH sequences, AY074397-AY074406, AY533236-AY533240, and AY586158; for V sequences, AY074407-AY074427, AY074429-AY074434, AY074436, AY074437, and AY074439-AY074446; and for V sequences, AY074448-AY074487.
3 Address correspondence and reprint requests to Dr. Robert D. Miller, Department of Biology, University of New Mexico, Albuquerque, NM 87131. E-mail address: rdmiller@unm.edu
4 Abbreviations used in this paper: FR, framework region; NJ, neighbor joining; MP, maximum parsimony; ME, minimum evolution; Modo, Monodelphis domestica; Hosa, Homo sapiens; Mumu, Mus musculus; Bota, Bos taurus; Crmi, Cricetulus migratorius; Divi, Didelphis virginiana; Eqca, Equus caballus; Gaga, Gallus gallus; Hefr; Heterodontus fransciscii; Isma, Isoodon macrourus; Ovar, Ovis aries; Oran, Ornithorhynchus anatinus; Orcu, Oryctolagus cuniculus; Rano, Rattus norvegicus; Susc, Sus scrofa; Taac, Tachyglossus aculeatus; Trvu, Trichosurus vulpecula.
Received for publication November 10, 2004. Accepted for publication February 22, 2005.
References
Nei, M., X. Gu, T. Sitnikova. 1997. Evolution by the birth-and-death process in multigene families of the vertebrate immune system. Proc. Natl. Acad. Sci. USA 94:7799.
Lucero, J. E., G. H. Rosenberg, R. D. Miller. 1998. Marsupial light chains: complexity and conservation of in the opossum Monodelphis domestica. J. Immunol. 161:6724.
Miller, R. D., E. R. Bergemann, G. H. Rosenberg. 1999. Marsupial light chains: IgK with four V families in the opossum Monodelphis domestica. Immunogenetics 50:329
Nowak, M. A., Z. E. Parra, L. H. Hellman, R. D. Miller. 2004. The complexity of expressed light chains in the egg-laying mammals. Immunogenetics 56:555.
Matsuda, F.. 2004. Human immunoglobulin heavy chain locus. T. Honjo, and F. W. Alt, and M. Neuberger, eds. Molecular Biology of B Cells 1. Elsevier Academic, London.
Tomlinson, I. M., S. C. Williams, O. Ignatovich, S. J. Corbett, G. Winter. 1996. VBASE Sequence Directory MRC Centre for Protein Engineering, Cambridge, U.K.
Butler, J. E.. 1997. Immunoglobulin gene organization and the mechanism of repertoire development. Scand. J. Immunol. 45:455.
Johansson, J., M. Aveskogh, B. Munday, L. Hellman. 2002. Heavy chain V region diversity in the duck-billed platypus (Ornithorhynchus anatinus): long and highly variable complementary determining region 3 compensates for limited germline diversity. J. Immunol. 168:5155.
Sitnikova, T., C. Su. 1998. Coevolution of immunoglobulin heavy- and light-chain variable-region gene families. Mol. Biol. Evol. 15:617.
Ota, T., T. Sitnikova, M. Nei. 2000. Evolution of vertebrate immunoglobulin variable gene segments. L. Du, and Pasquier and, and G. W. Litman, eds. Origin and Evolution of the Vertebrate Immune System 221. Springer, New York.
Miller, R. D., H. Grabe, G. H. Rosenberg. 1998. VH repertoire of a marsupial: Monodelphis domestica. J. Immunol. 160:259.
Belov, K., G. A. Harrison, D. W. Cooper. 1998. Molecular cloning of the cDNA encoding the constant region of the immunoglobulin A heavy chain (C) from a marsupial: Trichosurus vulpecula (common brushtail possum). Immunol. Lett. 60:165.
Belov, K., G. A. Harrison, R. D. Miller, D. W. Cooper. 1999. Isolation and sequence of a cDNA coding for the heavy chain constant region of IgG from the Australian brushtail possum, Trichosurus vulpecula. Mol. Immunol. 36:535.
Altschul, S. F., W. Gish, W. Miller, E. W. Myers, D. J. Lipman. 1990. Basic local alignment search tool. J. Mol. Biol. 215:403.
Thompson, J. D., D. G. Higgins, T. J. Gibson. 1994. CLUSTAL W: improving the sensitivity of the progressive multiple sequence alignment through sequence weighting, position specific gap penalties, and weight matrix choice. Nucleic Acids Res. 22:4673.
Hall, T. A.. 1999. BioEdit: a user-friendly biological sequence alignment editor and analysis program for Windows 95/98/NT. Nucleic Acids Symp. Ser. 41:95.
Saitou, N., M. Nei. 1987. The neighbor joining method: a new method for reconstructing phylogenetic trees. Mol. Biol. Evol. 4:406.
Kumar, S., K. Tamura, M. Nei. 1994. MEGA: molecular evolutionary genetics analysis software for microcomputers. Comput. Appl. Biosci. 10:189
Miller, R. D., K. Belov. 2000. Immunoglobulin genetics of marsupials. Dev. Comp. Immunol. 24:485.
Adamski, F. M., J. Demmer. 1999. Two stages of increased IgA transfer during lactation in the marsupial, Trichosurus vulpecula (Brushtail possum). J. Immunol. 162:6009.
Tutter, A., R. Riblet. 1989. Conservation of an immunoglobulin variable-region gene family indicates a specific, noncoding function. Proc. Natl. Acad. Sci. USA 86:7460.
Belov, K., L. Hellman. 2003. Immunoglobulin genetics of Ornithorhynchus anatinus (platypus) and Tachyglossus aculeatus (short-beaked echidna). Comp. Biochem. Physiol. A Mol. Integr. Physiol. 136:811.
Belov, K., G. A. Harrison, R. D. Miller, D. W. Cooper. 2001. Characterisation of the light chain of the brushtail possum (Trichosurus vulpecula). Vet. Immunol. Immunopathol. 78:317.
Belov, K., G. A. Harrison, R. D. Miller, D. W. Cooper. 2002. Molecular cloning of four light chain cDNAs from the Australian brushtail possum (Trichosurus vulpecula). Eur. J. Immunogenet. 29:95.
Rast, J. P., M. K. Anderson, T. Ota, R. Litman, M. Margittai, M. J. Shamblott, G. W. Litman. 1994. Immunoglobulin light chain class multiplicity and alternative organizational forms in early vertebrate phylogeny. Immunogenetics 40:83.
Kirsch, J. A. W., F. Lapointe, M. S. Springer. 1997. DNA-hybridisation studies of marsupials and their implications for metatherian classification. Aust. J. Zool. 45:211.
Janke, A., X. Xu, U. Arnason. 1997. The complete mitochondrial genome of the wallaroo (Macropus robustus) and the phylogenetic relationship among Monotremata, Marsupialia, and Eutheria. Proc. Natl. Acad. Sci. USA 94:1276.
Kumar, S., S. B. Hedges. 1998. A molecular timescale for vertebrate evolution. Nature 392:917.
Eizirik, E., W. J. Murphy, S. J. O’Brien. 2001. Molecular dating and biogeography of the early placental mammal radiation. J. Hered. 92:212.(Michelle L. Baker, Kather)
A pattern of coevolution between the V gene segments of Ig H and L chains has been noted previously by several investigators. Species with restricted germline VH diversity tend to have limited germline VL diversity, whereas species with high levels of germline VH diversity have more diverse VL gene segments. Evidence for a limited pool of VH but diverse VL gene segments in a South American opossum, Monodelphis domestica, is consistent with this marsupial being an exception to the pattern. To determine whether M. domestica is unique or the norm for marsupials, the VH and VL of an Australian possum, Trichosurus vulpecula, were characterized. The Ig repertoire in T. vulpecula is also derived from a restricted VH pool but a diverse VL pool. The VL gene segments of T. vulpecula are highly complex and contain lineages that predate the separation of marsupials and placental mammals. Thus, neither marsupial follows a pattern of coevolution of VH and VL gene segments observed in other mammals. Rather, marsupial VH and VL complexity appears to be evolving divergently, retaining diversity in VL perhaps to compensate for limited VH diversity. There is a high degree of similarity between the VH and VL in M. domestica and T. vulpecula, with the majority of VL families being shared between both species. All marsupial VH sequences isolated so far form a common clade of closely related sequences, and in contrast to the VL genes, the VH likely underwent a major loss of diversity early in marsupial evolution.
Introduction
The complexities of the Ig loci make their evolution, genetics, and expression the focus of attention. The typical Ig molecule is composed of two identical H and two identical L chains. Variation in amino acid residues at the N-terminal ends of the H and L chains contributes to Ab diversity and establishes Ab specificity. This variation is initially created through the process of so-called V(D)J recombination where gene segments are recombined somatically to generate an exon encoding the V domain. The H chain V domains are assembled by somatic recombination of three gene segments called VH, DH, and JH, whereas the L chain V domains are assembled from two gene segments, VL and JL. The arrays of V, D, and J segments available for recombination evolve through a process of gene duplication, divergence, and deletion (1). All three living lineages of mammals, the eutherians (placentals), metatherians (marsupials), and prototherians (monotremes), have essentially the same classes of Ig H and L chains. This includes having two IgL loci, designated Ig and Ig (2, 3, 4). In humans and mice, there is a relatively large and diverse set of VH, V, and V gene segments, which are one contribution to the generation of a primary Ig repertoire through V(D)J recombination. The human IgH locus contains 44 VH segments with complete open reading frames, 39 of which are known to be functional, and that comprise seven diverse VH families (5). Human VH can share as little as 47% average interfamily nucleotide identity (6). In contrast, other species (e.g., rabbits, cattle, platypus, and chickens) have limited germline VH diversity. In these cases, the germline VH within a species typically share >80% nucleotide identity, and may be only a single VH family or a set of recently derived VH families (7, 8).
Several investigators have noted a coevolutionary relationship between the H and L chain loci in some species (1, 7, 9, 10). Those species with a limited number of germline VH families (e.g., rabbits and cattle) also have limited numbers of germline VL families. Conversely, those species with multiple, diverse VH families (e.g., humans, sheep, and mice) have similarly complex germline VL families (1, 7, 9, 10). This pattern of similarity seen in VH and VL segments is sufficiently striking to lead some authors to propose that it is not by chance but the result of factors driving the coevolution of the IgH and IgL loci (1, 9, 10). What factor or factors these may be is not clear.
Marsupial Ab diversity has been analyzed for only a single species, the gray short-tailed opossum, Monodelphis domestica. Unlike other mammals and nonmammals examined to date, M. domestica has a contrasting diversity between its germline H and L loci. The M. domestica VH repertoire is derived from recombination of a set of V segments limited both in diversity and number. M. domestica has <15 germline VH segments, all but one of which belong to a single VH family designated VH1 (11). The lone VH segment that is not a VH1, is still closely related, but has been designated to a separate family, VH2 (11). In contrast, M. domestica has a highly diverse pool of VL segments from which to recombine the expressed VL domains. There are at least three V and four V families, and they contain together >60 V gene segments (2, 3). The V and V gene families in M. domestica are diverse and are derived from gene duplications that predate the separation of marsupials and eutherians. Thus, the IgH and IgL loci of M. domestica appear to be evolving divergently with respect to their complexity, rather than converging on similar patterns of germline diversity. To determine whether M. domestica is an exception among species or whether this pattern is common among marsupials, we analyzed the VH and VL repertoire of a second distantly related marsupial, the Australian brushtail possum, Trichosurus vulpecula.
Materials and Methods
PCR amplification of VH, V, and V sequences
A T. vulpecula mesenteric lymph node cDNA library constructed using the ZAPII vector has been described previously (12). The strategy to isolate V region sequences in a random manner from the cDNA libraries is also described elsewhere (2, 3, 11). Briefly, oligonucleotide primers complementary to the C regions of Ig, Ig, and IgH were paired with primers complementary to a region of the vector upstream of the 5' start of the cDNA. T. vulpecula V segments were amplified using a primer specific for marsupial C (5'-TGGTTGGAAGATGAAGGCAG-3') in combination with a primer complementary to the M13 universal sequencing site in ZAPII. V segments from T. vulpecula were amplified using a primer designed specifically for T. vulpecula C (5'-ACATTAACGGTGGGGCTTGC) in combination with a primer complementary to the T7 site in ZAPII. The T. vulpecula cDNA library was used as a target in PCR, and successful amplifications were achieved using 1.5 mM MgCl2 at an annealing temperature of 62°C and 65°C for C and C, respectively, using Taq polymerase (PerkinElmer) for 35 cycles.
T. vulpecula VH fragments were isolated using oligonucleotide primers complementary to T. vulpecula Cμ (5'-CCCACAGCCACAGGGGCATC) and C (5'-AATTCCATGTCACTGTCAC) in combination with the M13 primer. Successful amplifications were achieved using 1.5 mM MgCl2 at 68 and 53°C, respectively, using the T. vulpecula cDNA library as target. PCR products were cloned into the pCR2.1 vector (Invitrogen Life Technologies). Additional T. vulpecula sequences were obtained from phage clones isolated by screening the T. vulpecula cDNA library with an IgG C region probe as described in Belov et al. (13). Positive plaques were used as a target in PCR using a C primer (5'-AGGAGGTTCCCTGTCCACAATTG-3') paired with a primer specific for the T3 site in the ZAPII vector. Successful amplifications were achieved at an annealing temperature of 55°C using Taq polymerase (PerkinElmer) for 35 cycles, and PCR products were cloned into the pGEM-T vector (Promega).
Probes and hybridizations
An Isoodon macrourus spleen cDNA library was constructed using the SMART cDNA construction kit and the TripleX vector (Clontech). Screening of the I. macrourus spleen cDNA library by hybridization was done by plating the phage and preparing plaque lifts on reinforced nitrocellulose. An M. domestica VH probe (MVH356) previously described (11) was used to screen the I. macrourus cDNA library. Probes were constructed from DNA inserts excised from agarose gels and labeled with [32P]dCTP by the Random-Prime-It RmT labeling kit (Stratagene). Plaque lifts were hybridized under conditions of 42°C, 50% formamide, 5x Denhardt’s solution, 5x SSC, 50 mM NaPO4 (pH 6.5), 0.1% SDS, 5 mM EDTA, and 250 mg/ml sheared salmon DNA. Final washes were performed at 65°C in 0.2x SSC, and lifts were autoradiographed at –80°C for 1–4 days.
Sequencing and phylogenetic analysis
Sequencing was performed using the Big Dye sequencing kit (Amersham Biosciences) and analyzed on an automated DNA sequencer (PerkinElmer ABI Prism 377 DNA Sequencer). Minor manual corrections of sequences were made using the Sequencher 4.1 program (Gene Codes Corporation). Comparisons to known sequences in GenBank database were made using the BLAST algorithm (14).
Sequences were aligned using either the Clustal X program (15) with minor manual adjustments or with the Clustal W function bundled within BioEdit (16). For phylogenetic analyses, alignments were made using sequences corresponding to the framework regions (FR) 4 1–3 of the V domains. All phylogenetic tree reconstruction shown in this paper is based on analyses done using nucleotide alignments. Gaps in the nucleotide sequences were determined by first aligning the amino acid translations to establish gap position and then converting the sequence back to nucleotide using the BioEdit program. In this way, nucleotide gaps were established based on codon position. However, analyses done with nucleotide sequence alignments gapped independently of codon position yielded identical results. Based on the nucleotide alignments, phylogenetic trees were constructed by the neighbor joining (NJ) method of Saitou and Nei (17), maximum parsimony (MP), and minimum evolution (ME) using the MEGA2 program (18). NJ and MP analyses using MEGA2 were performed using 1000 bootstrap replicates and the Kimura two-parameter and min-mini heuristic search, respectively.
VL used in the analyses
Representative opossum M. domestica (Modo) V and V family sequences were as follows: VLI, AF049774; VLII, AF049781; VLIII, AF049756; VKI, AF116930; VKII, AF116924; VKIII, AF116931; and VKIV, AF116935. Human, Homo sapiens (Hosa) V and V were obtained from the VBASE database (6). Mouse, Mus musculus (Mumu) V were as follows: VL1, X82687; VLX, D38129; mouse V were as follows: IGKVR1, X13938; IGKCLM, Z72384; IGKCAM2, M24937; IGVKID, M63611. Rat, Rattus norvegicus (Rano) V was U39609. Hamster, Cricetulus migratorius (Crmi) V was U17165. Horse, Equus caballus (Eqca), VK1 was X75611. Sheep, Ovis aries (Ovar) VK was X54110. Rabbit, Oryctolagus cuniculus (Orcu), were as follows: VL2, M27840; and VL3, M27841. Chicken, Gallus gallus (Gaga) V was M96972. Heterodontus fransciscii (Hefr) V was X15316. Platypus, Ornithorhynchus anatinus (Oran) V were as follows: AF525088, AF525090-AF525092, AF525095-AF525097, AF525100, AF525102-AF525106, AF525108, AF525109, AF525111, AF525115, AF525116, AF525118, AF525119, and AF525123.
VH from other species
Previously published T. vulpecula (Trvu) sequences were as follows: TrvuFA1, AF091140; and TrvuFA2, AF091141. M. domestica VH1 sequences were as follows: MVH10, AF012121; MVH11, AF012122; MVH26, AF012113; and M. domestica VH2 sequence, AF012124. The Didelphis virginiana (Divi) VH is unpublished and was provided by Dr. R. Riblet (Torrey Pines Institute for Molecular Studies, La Jolla, CA). Human, H. sapiens (Hosa), VH sequences were obtained from the VBASE database (6). Mouse, M. musculus (Mumu), VH family representatives were as follows: 7183, U04227; 3660, K01569; 3609N, X55935; DNA4, M20829; J558, Z37145; J606, X03398; Q52, M27021; S107, J00538; SM7, M31285; VH11, Y00743; and X24, X00163. Cow, Bos taurus (Bota) VH was AF015505. Sheep, O. aries (Ovar) VH was Z49180. Pig, Sus scrofa (Susc) VH was U15194. Horned shark, H. fransciscii (Hefr) VH, X13449. Platypus, O. anatinus (Oran) VH were as follows: AF3811324, AF381290-AF381296, AF381298, AF381299, AF381303-AF381308, AF381310-AF381315, AF381317-AF381320, AF381322, AY055778, AY055779, AY055780-AY055782, AY168639, AAL59586 Echidna, Tachyglossus aculeatus (Taac) VH were as follows: AAM61760 AAM60800 AAM60783 AY101439, AY101442, and AY101445.
Results
Limited VH diversity in T. vulpecula
To sample the expressed Ig H chain diversity in the brushtail possum, VH sequences were isolated using a combination of hybridization and PCR amplification. IgH C region sequences were already available for T. vulpecula in GenBank (19). A total of 16 unique T. vulpecula VH sequences were analyzed. Nine were isolated by amplification using anchored PCR and reverse primers complementary to the 5' end of Cμ and C on a T. vulpecula mesenteric lymph node cDNA library. The Cμ and C primers were paired with forward primers in the phage vector so as not to bias the VH sequences amplified. Five clones containing VH were isolated by hybridization using a C fragment as a probe on the same library. Two T. vulpecula VH included in the analysis were available in GenBank (20). Pairwise analysis of these VH sequences revealed that all but two shared >80% nucleotide identity to each other and are by definition of the same VH family. The same result was found when the CDRs were excluded from the analysis, limiting the impact that somatic mutations may have played in family assignment (not shown). The most divergent T. vulpecula VH sequences still shared >78% nucleotide identity to the rest of the VH from that species. The T. vulpecula sequences were also compared with M. domestica VH and found to share >73%, and as much as 82%, interspecies nucleotide identity. VH sequences are available for only two other species of marsupials: a germline VH from an American marsupial, the Virginia opossum, D. virginiana (DiviP83 in Fig. 1) (11), and a VH from an Australian marsupial, the Northern Brown Bandicoot, I. macrourus (Isma5.2 in Fig. 1). The latter was obtained by screening a spleen cDNA library by hybridization (this study). Both the I. macrourus and D. virginiana VH sequences share a high degree of nucleotide identity with M. domestica VH1 sequences (69–83% for I. macrourus and 68–91% for D. virginiana). The most divergent marsupial VH is the M. domestica VH2 sequence, a VH from a family containing a single gene segment (11).
FIGURE 1. Phylogenetic tree based on alignments of VH sequences from T. vulpecula and representatives of other mammals and some nonmammals. The marsupial VH are shown in bold. This tree was constructed using the NJ method (17 ). The numbers at the branch nodes indicate bootstrap values based on 1000 replicates. Roman numerals on the right indicate the three VH groups (21 ). Species names are abbreviated to the first two letters of their scientific names: B. taurus (Bota), D. virginiana (Divi), H. sapiens (Hosa), H. fransciscii (Hefr), I. macrourus (Isma), M. domestica (Modo), M. musculus (Mumu), O. aries (Ovar), O. anatinus (Oran), S. scrofa (Susc), T. aculeatus (Taac), T. vulpecula (Trvu).
Phylogenetic analyses using alignments of the T. vulpecula VH sequences with VH from other marsupials and a variety of eutherian and prototherian mammals were performed (Fig. 1). A VH sequence from the horned shark, H. fransciscii, Hefr, was included in the analyses as outgroup to the mammalian VH. A phylogenetic tree reconstructed from the alignments using the NJ method is shown in Fig. 1. The mammalian VH cluster into three groups (I, II, and III in Fig. 1; Ref. 21), sometimes referred to as clades A, B, and C, respectively (1). All marsupial VH form a single, monophyletic clade within group III. The M. domestica VH2 sequence remains the most divergent marsupial VH in the cluster; M. domestica VH2 is sister to all other marsupial VH. The phylogenetic tree reveals that the T. vulpecula VH cluster as a single family, consistent with the pairwise nucleotide analyses described above. Furthermore, this single T. vulpecula VH clade is sister to the M. domestica VH1 clade. Similar results were obtained when the alignments were analyzed using MP and ME (not shown). In analyses using MP, the M. domestica VH2 sequence was sister to a clade that included all marsupial and platypus VH. Platypus, O. anatinus, Oran, VH form a single clade, and all platypus VH belong to essentially a single VH family that is related to the marsupial VH (Fig. 1). This is not the case for all monotremes, because echidna, T. aculeatus, Taac, VH are highly diverse and are present in all three mammalian VH groups (Fig. 1; Ref. 22). In conclusion, pairwise nucleotide identities and phylogenetic analyses of all marsupial VH sequences currently available reveal that two distantly related species have a closely related set of VH genes.
V and V diversity in T. vulpecula
Given the restricted nature of the T. vulpecula VH repertoire, we would predict that this species might also have a restricted VL repertoire; that is, if the pattern of coevolution described in the introduction occurs in this marsupial. To test this, the expressed diversity of V and V gene segments were sampled in T. vulpecula, using anchored PCR. C and C sequences have been described for T. vulpecula (23, 24). V sequences were amplified from the T. vulpecula mesenteric lymph node cDNA library using anchored PCR and an oligonucleotide primer complementary to the 5' end of C. A total of 40 clones containing partial or full V domains were isolated and sequenced. Each were derived from a unique V-J recombination event, and there were 25 different V gene sequences represented. An additional three V sequences, published previously, were included in the dataset (24). Pairwise analysis of the 19 most complete T. vulpecula V revealed the presence of four distinct V families (V1 to -4), sharing <80% nucleotide identity between families (Fig. 2a). Exclusion of the CDRs from the analysis did not change the assignment of sequences to a particular family (not shown). Most of the clones (14 of 19) contained V that belonged to the V1 family (Fig. 2a). The T. vulpecula V1 family was the most diverse, sharing between 81 and 96% intrafamily nucleotide identity.
FIGURE 2. Range of percent nucleotide identity between 19 T. vulpecula V (a) and 32 T. vulpecula V (b) sequences based on alignments of regions corresponding to FR1 through FR3 inclusive of CDRs. Numbers in parentheses indicate number of clones isolated from each family. Sequences used in analysis were as follows: V1: TrvuVL56, TrvuVL87, TrvuVL126, TrvuVL131, TrvuVL151, TrvuVL157, TrvuVL158, TrvuVL162, TrvuVL182, TrvuVL175, TrvuVL190, TrvuVL204, TrvuVL215, TrvuVL325; V2: TrvuVL1, TrvuVL10, TrvuVL11; V3: TrvuVL12; V4: TrvuVL33; V1: TrvuVk16, TrvuVk46, TrvuVk52, TrvuVk57, TrvuVk58, TrvuVk66, TrvuVk70, TrvuVk84, TrvuVk85, TrvuVk88, TrvuVk113, TrvuVk172, TrvuVk230, TrvuVk231, TrvuVk242; V2: TrvuVk6, TrvuVk30, TrvuVk103, TrvuVk153, TrvuVk228, TrvuVk252, TrvuVk259; V3: TrvuVk94, TrvuVk107, TrvuVk134, TrvuVk141, TrvuVk156, TrvuVk262; V4: TrvuVk68, TrvuVk176, TrvuVk218, TrvuVk236; and V5: TrvuVk158.
To sample the diversity of V in T. vulpecula, an oligonucleotide primer complementary to marsupial C was used in anchored PCR with the T. vulpecula cDNA library. A total of 43 clones containing partial or full unique V domains were isolated and sequenced, and these contained 31 different V gene sequences. An additional two published T. vulpecula V sequences were also included in this analysis (23). Pairwise analyses of all T. vulpecula V sequences (n = 33) revealed that these sequences shared between 65 and 99% nucleotide identity to each other (Fig. 2b). Five distinct T. vulpecula V families were identified based on the sharing of >80% nucleotide identity within a family (Fig. 2b). Again, analyses excluding the CDRs did not change the assignment of a sequence to a particular family (not shown).
Phylogenetic analyses of V and V were performed using an alignment of the T. vulpecula sequences with V and V from M. domestica, several eutherian species, and a prototherian. Only sequences with complete or nearly complete V regions were included in this analysis. Fig. 3 shows a tree of L chain V regions, where alignments of V and V sequences were combined in a single analysis. A VL region from the horned shark, H. fransciscii, was used as outgroup to the mammalian V and V, because elasmobranch VL genes have been shown to be the most ancient of the vertebrate VL genes (9, 25). The tree shown was reconstructed using the NJ method (Fig. 3); however, identical results were found when MP and ME were used (not shown). The overall topology of the tree is similar to that observed by others for mammalian VL (9). Specifically, there is the presence of two distinct, but weakly supported clusters of V and a single, well-supported V cluster nested within the V. Several observations can be made regarding the T. vulpecula V and V from the tree shown in Fig. 3. The T. vulpecula V sequences fall into the four separate clusters corresponding to the four families described previously (Fig. 2a). Interspersed between the four T. vulpecula V families are eutherian V, consistent with the establishment of these germline VL lineages before the separation of eutherians and marsupials. Additionally, three of the four V families in T. vulpecula (V1, -3, and -4) are each phylogenetically most closely related to a M. domestica V family (Fig. 3). The T. vulpecula V sequences fall into the five separate clusters within the V clade corresponding to the five V families based on nucleotide identity (Fig. 2b). The T. vulpecula V families are also interspersed among V sequences from eutherians. Three of the five T. vulpecula V families (V1, -2, and -4) are also each phylogenetically related to a M. domestica V family. In conclusion, an analysis of the V and V expressed in T. vulpecula reveals a diverse pool of gene segments from which to derive expressed IgL diversity in this species. Furthermore, much of the VL diversity present in T. vulpecula is ancient and overlaps with that of M. domestica.
FIGURE 3. Phylogenetic tree based on alignments of combined V and V sequences from T. vulpecula and representatives of other mammals and some nonmammals. All marsupial sequences are shown in bold in the tree. The position of the distinct T. vulpecula V and V families on the tree are indicated to the right as V1, V1, etc. This tree was constructed using the NJ method (17 ). The numbers at the branch nodes indicate bootstrap values based on 1000 replicates. Species names are as in Fig. 1: B. taurus (Bota), C. migratorius (Crmi), E. caballus (Eqca), G. gallus (Gaga), H. sapiens (Hosa), H. fransciscii (Hefr), M. domestica (Modo), M. musculus (Mumu), O. aries (Ovar), O. anatinus (Oran), O. cuniculus (Orcu), R. norvegicus (Rano), T. vulpecula (Trvu).
Discussion
We analyzed the general level of expressed Ab V diversity in the brushtail possum, T. vulpecula, to infer the germline V segment contribution to the repertoire. The results were compared with what was known from another, distantly related marsupial, M. domestica. M. domestica and T. vulpecula are members of the American Didelphidae family and the Australasian Phalangeridae families, respectively. Marsupials are a monophyletic group that once held wide Gondwanan distribution. Since the breakup of that supercontinent, the marsupial families were fragmented into two groups, the Australasian and the American marsupials (26). These two groups have now undergone a considerable period of independent evolutionary history. The Didelphidae and Phalangeridae have likely been separated for 67–75 million years (26, 27), putting divergence times for M. domestica and T. vulpecula on par with many of the eutherian lineages (27, 28, 29, 30). The overlap in their germline VH and VL repertoires, therefore, is not easily explained by recent common origins.
There are two results from the analysis of marsupial Ig VH and VL diversity that are particularly noteworthy. First, the VH sequences described so far from marsupials are strikingly similar between species. The same could be said for much of the VL diversity as well because many of the M. domestica and T. vulpecula V and V families are closely related. This is not an artifact of sampling because, in the case of M. domestica and T. vulpecula, we used C region sequences of both IgH and IgL to sample the upstream V region diversity. Second, both M. domestica and T. vulpecula share a pattern of diversity where there is limited VH gene segment diversity and extensive VL gene segment diversity. The first observation may have implications for the evolution of Ab genes early in the appearance of marsupials. The VH diversity that is present in many extant eutherians and one monotreme is ancient and predates the appearance of mammals (1). Collectively, mammalian VH fall into three clusters or groups (groups I, II, and III in Fig. 2). VH segments corresponding to all three groups can be also found in the amphibian Xenopus laevis, for example. This is consistent with the appearance of VH groups I, II, and III early in tetrapod evolution at the least (31). The presence of VH groups I, II, and III in some eutherians and one monotreme, the echidna (Fig. 2; Ref. 22), confirms that the last common ancestor of marsupials and eutherians would also have contained VH segments from all three groups. Over time, duplication and deletion (the so-called birth and death process) of VH segments resulted in some lineages retaining this ancient germline diversity, whereas others have lost it (1, 7). Many species have deleted much of the ancestral germline VH diversity and now have a limited set of related VH; in some species, there is only a single VH family remaining (7, 10). An example of such deletion is illustrated by the VH in cattle and swine. These two artiodactyls each have a single VH family; however, they do not have the same or even a closely related VH family. Cattle VH fall into the group I clade, whereas swine VH are group III (7). The VH in groups I and III arose from duplications that occurred likely >350 million years ago, before the appearance of mammals (10). The last common ancestor of cattle and swine must have had at least group I and III VH segments in their germline, and, following speciation, each deleted (or retained) VH from reciprocal lineages.
Contrasting this is the situation in marsupials. The VH sequences isolated from four species of marsupials form a monophyletic family within VH group III. In the case of T. vulpecula and M. domestica, the entire VH repertoires are similar by having a single, related VH family. Because the I. macrourus and D. virginiana VH were obtained by screening with V region probes, rather than the C region probes, there is a bias toward isolating clones having similar V regions, rather than a random sample of VH diversity. We can only conclude that I. macrourus and D. virginiana at least contain VH gene segments related to the M. domestica VH1 family. Whether this is all they have remains to be determined. How could marsupial species on separate continents and separated by at least 65 million years of evolution have unusually similar germline V repertoires? It could be by chance. Alternatively, the similarity of IgV families in M. domestica and T. vulpecula could reflect convergent selection acting on marsupial V segments. This seems unlikely given the time, distance, and habitats separating the American opossums from the Australian possums. Rather, a more parsimonious explanation is that the last common ancestor of M. domestica and T. vulpecula had already undergone a major loss of germline VH diversity resulting in a limited set of available VH segments before the separation of these two species. Given the phylogenetic position of the Didelphidae and Phalangeridae, it is likely that most or all extant marsupial families would have this common, restricted VH pool (26, 32). Such a loss of VH diversity can be treated as a form of a genetic bottleneck at the IgH locus that occurred very early in marsupial evolution, near the base of the extant marsupial family tree. In contrast, both species have maintained ancient VL diversity that must have been present in the earliest marsupials.
The second, but related observation, that M. domestica and T. vulpecula each have limited VH diversity paired with more diverse VL, is in contrast to the pattern of coevolution between VH and VL segments that has been observed in most species examined to date (1, 7, 9, 10). A pattern has emerged in eutherian mammals where species can be divided into two groups with respect to Ig diversity (reviewed in Ref. 7). One group are the species that generate a diverse primary Ig repertoire through the process of V(D)J recombination only, using a diverse pool of available germline V segments (e.g., primates and rodents). The second group are the species that have a limited number of germline V segments, usually of a single family, and rely on mutation, such as gene conversion or somatic mutation, to further diversify their primary Ig repertoire following V(D)J recombination (e.g., rabbits, some artiodactyla, and birds). These mutations occur at lymphoid tissue sites associated with the hindgut (e.g., the appendix in rabbits, the bursa of Fabricius in chickens; Refs. 33 and 34). One hypothesis to explain these restricted VH and VL repertoires is that the mutational mechanisms used to diversify the V domains have placed common constraints on VH and VL region segment evolution, limiting their germline diversity. Evidence from studies of sheep Ab diversity casts doubt on this hypothesis. Sheep have relatively high levels of germline VH and VL gene diversity (35), despite using mutation to further diversify their primary Ig repertoire in the Peyer’s patches (36). The role that mutations and gut associated lymphoid tissue play in generating diversity and development of the primary Ab repertoires in marsupials remains to be determined.
In marsupials, the VH and VL repertoires may be coevolving to compensate for each other. The lack of germline segment diversity in marsupial IgH is compensated for by higher levels of diversity at the IgL loci. With a severely restricted pool of VH segments to pair with, the L chains are taking up the responsibility of maintaining germline diversity. Crystallographic studies of human and mouse Abs have demonstrated that H chains contribute the greatest contributions to Ag binding (37). Therefore, it would be expected that H chain diversity is critical for generating an effective Ab repertoire. Indeed, one group of mammals, the Camelids, has altogether dispensed with the Ig L chains for some Ab isotypes, without apparent loss of Ab function or specificity. Such discoveries have led some investigators to question the importance of L chains to Ab diversity (reviewed in Ref. 38). In the two marsupial species studied to date, L chains appear to contribute more to the complexity of the Ab repertoire than do H chains. Marsupials as a group, therefore, are important model species to study how L chains can contribute to Ab diversity and to understanding L chain evolution. We would also suggest that the VH and VL segments that exist in an extant species may be shaped more by evolutionary history of the individual loci than by selection for specific patterns of coevolution.
Disclosures
The authors have no financial conflict of interest.
Footnotes
The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
1 This publication was supported by grants from the National Institutes of Health (Institutional Development Award program of the National Center for Research Resources; IP20RR18754) (supporting M.L.B.), the National Science Foundation (MCB-9981960) (to R.D.M.), and the Australian Research Council (to K.B.).
2 The sequences presented in this article have been submitted to EMBL/GenBank under the following accession numbers: for VH sequences, AY074397-AY074406, AY533236-AY533240, and AY586158; for V sequences, AY074407-AY074427, AY074429-AY074434, AY074436, AY074437, and AY074439-AY074446; and for V sequences, AY074448-AY074487.
3 Address correspondence and reprint requests to Dr. Robert D. Miller, Department of Biology, University of New Mexico, Albuquerque, NM 87131. E-mail address: rdmiller@unm.edu
4 Abbreviations used in this paper: FR, framework region; NJ, neighbor joining; MP, maximum parsimony; ME, minimum evolution; Modo, Monodelphis domestica; Hosa, Homo sapiens; Mumu, Mus musculus; Bota, Bos taurus; Crmi, Cricetulus migratorius; Divi, Didelphis virginiana; Eqca, Equus caballus; Gaga, Gallus gallus; Hefr; Heterodontus fransciscii; Isma, Isoodon macrourus; Ovar, Ovis aries; Oran, Ornithorhynchus anatinus; Orcu, Oryctolagus cuniculus; Rano, Rattus norvegicus; Susc, Sus scrofa; Taac, Tachyglossus aculeatus; Trvu, Trichosurus vulpecula.
Received for publication November 10, 2004. Accepted for publication February 22, 2005.
References
Nei, M., X. Gu, T. Sitnikova. 1997. Evolution by the birth-and-death process in multigene families of the vertebrate immune system. Proc. Natl. Acad. Sci. USA 94:7799.
Lucero, J. E., G. H. Rosenberg, R. D. Miller. 1998. Marsupial light chains: complexity and conservation of in the opossum Monodelphis domestica. J. Immunol. 161:6724.
Miller, R. D., E. R. Bergemann, G. H. Rosenberg. 1999. Marsupial light chains: IgK with four V families in the opossum Monodelphis domestica. Immunogenetics 50:329
Nowak, M. A., Z. E. Parra, L. H. Hellman, R. D. Miller. 2004. The complexity of expressed light chains in the egg-laying mammals. Immunogenetics 56:555.
Matsuda, F.. 2004. Human immunoglobulin heavy chain locus. T. Honjo, and F. W. Alt, and M. Neuberger, eds. Molecular Biology of B Cells 1. Elsevier Academic, London.
Tomlinson, I. M., S. C. Williams, O. Ignatovich, S. J. Corbett, G. Winter. 1996. VBASE Sequence Directory MRC Centre for Protein Engineering, Cambridge, U.K.
Butler, J. E.. 1997. Immunoglobulin gene organization and the mechanism of repertoire development. Scand. J. Immunol. 45:455.
Johansson, J., M. Aveskogh, B. Munday, L. Hellman. 2002. Heavy chain V region diversity in the duck-billed platypus (Ornithorhynchus anatinus): long and highly variable complementary determining region 3 compensates for limited germline diversity. J. Immunol. 168:5155.
Sitnikova, T., C. Su. 1998. Coevolution of immunoglobulin heavy- and light-chain variable-region gene families. Mol. Biol. Evol. 15:617.
Ota, T., T. Sitnikova, M. Nei. 2000. Evolution of vertebrate immunoglobulin variable gene segments. L. Du, and Pasquier and, and G. W. Litman, eds. Origin and Evolution of the Vertebrate Immune System 221. Springer, New York.
Miller, R. D., H. Grabe, G. H. Rosenberg. 1998. VH repertoire of a marsupial: Monodelphis domestica. J. Immunol. 160:259.
Belov, K., G. A. Harrison, D. W. Cooper. 1998. Molecular cloning of the cDNA encoding the constant region of the immunoglobulin A heavy chain (C) from a marsupial: Trichosurus vulpecula (common brushtail possum). Immunol. Lett. 60:165.
Belov, K., G. A. Harrison, R. D. Miller, D. W. Cooper. 1999. Isolation and sequence of a cDNA coding for the heavy chain constant region of IgG from the Australian brushtail possum, Trichosurus vulpecula. Mol. Immunol. 36:535.
Altschul, S. F., W. Gish, W. Miller, E. W. Myers, D. J. Lipman. 1990. Basic local alignment search tool. J. Mol. Biol. 215:403.
Thompson, J. D., D. G. Higgins, T. J. Gibson. 1994. CLUSTAL W: improving the sensitivity of the progressive multiple sequence alignment through sequence weighting, position specific gap penalties, and weight matrix choice. Nucleic Acids Res. 22:4673.
Hall, T. A.. 1999. BioEdit: a user-friendly biological sequence alignment editor and analysis program for Windows 95/98/NT. Nucleic Acids Symp. Ser. 41:95.
Saitou, N., M. Nei. 1987. The neighbor joining method: a new method for reconstructing phylogenetic trees. Mol. Biol. Evol. 4:406.
Kumar, S., K. Tamura, M. Nei. 1994. MEGA: molecular evolutionary genetics analysis software for microcomputers. Comput. Appl. Biosci. 10:189
Miller, R. D., K. Belov. 2000. Immunoglobulin genetics of marsupials. Dev. Comp. Immunol. 24:485.
Adamski, F. M., J. Demmer. 1999. Two stages of increased IgA transfer during lactation in the marsupial, Trichosurus vulpecula (Brushtail possum). J. Immunol. 162:6009.
Tutter, A., R. Riblet. 1989. Conservation of an immunoglobulin variable-region gene family indicates a specific, noncoding function. Proc. Natl. Acad. Sci. USA 86:7460.
Belov, K., L. Hellman. 2003. Immunoglobulin genetics of Ornithorhynchus anatinus (platypus) and Tachyglossus aculeatus (short-beaked echidna). Comp. Biochem. Physiol. A Mol. Integr. Physiol. 136:811.
Belov, K., G. A. Harrison, R. D. Miller, D. W. Cooper. 2001. Characterisation of the light chain of the brushtail possum (Trichosurus vulpecula). Vet. Immunol. Immunopathol. 78:317.
Belov, K., G. A. Harrison, R. D. Miller, D. W. Cooper. 2002. Molecular cloning of four light chain cDNAs from the Australian brushtail possum (Trichosurus vulpecula). Eur. J. Immunogenet. 29:95.
Rast, J. P., M. K. Anderson, T. Ota, R. Litman, M. Margittai, M. J. Shamblott, G. W. Litman. 1994. Immunoglobulin light chain class multiplicity and alternative organizational forms in early vertebrate phylogeny. Immunogenetics 40:83.
Kirsch, J. A. W., F. Lapointe, M. S. Springer. 1997. DNA-hybridisation studies of marsupials and their implications for metatherian classification. Aust. J. Zool. 45:211.
Janke, A., X. Xu, U. Arnason. 1997. The complete mitochondrial genome of the wallaroo (Macropus robustus) and the phylogenetic relationship among Monotremata, Marsupialia, and Eutheria. Proc. Natl. Acad. Sci. USA 94:1276.
Kumar, S., S. B. Hedges. 1998. A molecular timescale for vertebrate evolution. Nature 392:917.
Eizirik, E., W. J. Murphy, S. J. O’Brien. 2001. Molecular dating and biogeography of the early placental mammal radiation. J. Hered. 92:212.(Michelle L. Baker, Kather)