| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Department of Biology, University of New Mexico, Albuquerque, NM 87131
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
and 
light chains. From comparative studies of
Ig gene content and expressed V sequences, three distinct mechanisms
for creating V segment diversity have been found to predominate in
different species: 1) germline diversity, 2) gene conversion, and 3)
somatic mutation. Germline diversity refers to the presence of a
significant number of different, functional VH segments in
the germline Igh locus, which V(D)J recombination utilizes
to create a diverse primary Ig repertoire. The diversity of
VH sequences is usually evident by the presence of
divergent families of VH, which have been found in mice and
humans as well as nonmammalian vertebrates including amphibians and
fish (1, 2, 3, 4). A VH family is defined as one in which the
members share >80% nucleotide similarity. In mice, for example,
VH segments have diverged into at least 14 families based
on sequence homology (reviewed in 1 .
In contrast, some species have been shown to lack germline
VH diversity and rely on post-V(D)J rearrangement
mechanisms for diversifying their Ig. The chicken Igh locus,
for example, contains only a single functional VH segment
and several related VH pseudogenes (
VH),
limiting the number of functional germline rearrangements (5).
Variation in the expressed VH region is generated by
modifying the single, functionally rearranged VH through
gene conversion using the VH as donor sequences. A
similar system has been described for the Igh locus in
rabbits, in which the majority of B cells rearrange the most D proximal
VH segment, which is then modified by gene conversion (6).
Alternatively, sheep and swine appear to derive their expressed Ig
repertoire from a few closely related VH that are modified
by somatic mutation before Ag exposure (7, 8). So far, in species using
either gene conversion or somatic mutation, the germline VH
are similar to each other and essentially belong to a single
VH family. In chickens, rabbits, and swine, the
VH family present is most closely related to group III
families found in mice and humans, whereas in sheep the expressed
VH appear to be group II (7, 8, 9). It is interesting that the
three mechanisms of VH diversity (germline diversity, gene
conversion, and somatic mutation) appear not to be linked to phylogeny.
Comparative studies have greatly increased our understanding of the
evolutionary history of these mechanisms and may reveal as yet
undiscovered ways of creating VH diversity.
In the present study, we describe the diversity of VH segments expressed in the Ig repertoire of a metatherian (marsupial) mammal, the South American short-tailed opossum (Monodelphis domestica). Very little is presently known about the molecular development of immunocompetence in metatherians, which diverged from eutherians (placental mammals) at least 130 million years ago during the Cretaceous (10). M. domestica is a member of the didelphid marsupials, which are believed to be among the oldest living mammals and separated early from the other marsupial orders (11, 12). The immunologic studies that have been done with marsupials have yielded interesting differences from placental mammals including the general lack of a MLR, typically poor responses, and unusual patterns of isotype switching (reviewed in Refs. 13 and 14). M. domestica is available as a laboratory-bred marsupial, developed over the last decade as a model didelphid (15) and used in biomedical research for studies of UV-induced melanoma (16). M. domestica has also provided an opportunity to study immunologic development in a species with the general marsupial characteristic of being significantly less developed at birth than a placental mammal (13). Newborn M. domestica are at a developmental point similar to a human or mouse embryo at 8 wk or 13 days of gestation, respectively (13). At birth, there are few detectable lymphocytes, and the thymic epithelium appears undifferentiated (17).
Studies using M. domestica, or any marsupial for that matter, have been restricted by the lack of knowledge concerning the ontogeny of immunocompetence and the generation of immune responses. To develop reagents for such studies, we recently began to characterize the M. domestica homologues of genes important for lymphocyte development, including RAG1 (18) and terminal deoxynucleotidyl transferase (33).5 Here, we report the characterization of the expressed Monodelphis VH repertoire.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
All genomic M. domestica DNAs used were extracted
from liver tissue. Size-selected genomic DNA for library construction
was prepared by electrophoresing EcoRI- or
SpeI-digested DNA through agarose and extracting the DNA in
the selected size range either by melting the agarose and using phenol
and chloroform or using a gel extraction kit following the
manufacturer’s recommended protocol (QIAquick; Qiagen, Chatsworth,
CA). For constructing the EcoRI library, the size range of
genomic DNA fragments isolated and pooled was 1 to 10 kb. The
SpeI library was made specifically to isolate a genomic
VH2 fragment and was constructed using genomic
DNA fragments of a size range of 3 to 5 kb. The extracted DNA was
ligated to ZAPII phage arms, packaged, and plated for screening
following the manufacturer’s recommended protocol (Stratagene, La
Jolla, CA).
A premade, commercially available M. domestica spleen cDNA
library, also in ZAPII, was also purchased (Stratagene). A plasmid
clone (P83) containing an unpublished germline VH segment
from another didelphid genus, Didelphis virginiana (North
American opossum) was generously provided by Dr. Roy Riblet (Medical
Biology Institute, La Jolla, CA). All other DNA clones used are
described in this report. All probes used in this study were prepared
as DNA inserts excised from the plasmids and labeled with
[32P]dCTP using the random primer method (Prime-It Kit,
Stratagene). All hybridizations, including library screening and
Southern blots, were done at 42°C in 50% formamide, 5x Denhardt’s
solution, 5x SSC, 50 mM NaPO4 (pH 6.5), 0.1% SDS, 5 mM
EDTA, and 250 mg/ml of sheared salmon DNA. Final wash conditions for
hybridizations were done at 65°C and 0.2x SSC.
PCR screening of cDNA libraries
Oligonucleotide primers complementary to the M.
domestica Cµ and C
regions were paired in separate PCR with
primers for either the M13 forward, M13 reverse, or T3 sites, which
flank the cloning site in pBluescript (Stratagene) to amplify
VH region segments in an unbiased manner (Cµ,
5'-GGAGAAGAGATTTGGTGCAGATGGG; C, 5'-GTCACCAGTTCTAGAGTCACAGAGG).
The spleen cDNA library was used as target in the PCR. The ZAPII
vector is a phagemid that contains the pBluescript sequences flanking
the cloning site. PCR was performed using the conditions of 1.5 mM
MgCl2 and 55°C annealing temperature, using
Taq polymerase (Perkin-Elmer, Foster City, CA) for 35
cycles. PCR products were cloned for sequencing using the pCRII vector
(Invitrogen, Carlsbad, CA) following the manufacturer’s recommended
protocol.
Sequencing and analysis
DNA sequencing of genomic or cDNA clones (in either pUC19 or pBluescript) was performed manually using the Sequenase 2.0 kit (United States Biochemicals, Cleveland, OH) and [35S]dATP labeling or, for most sequences, was performed on an automated DNA sequencer (Perkin-Elmer ABI Prism 377 DNA sequencer) using the DNA cycle sequencing kit (Perkin-Elmer). All DNA sequences reported were derived by sequencing both strands of each clone. Sequences were analyzed using the Sequencher 3.0 program (Gene Codes Corp., Ann Arbor, MI) and aligned using the CLUSTAL W program (19). All VH sequences reported here have been deposited in the GenBank database and are available under accession numbers AF007070 to AF007093 and AF012111 to AF012124. Phylogenetic analyzes were done using the UNIX-based test version of PAUP* (phylogenetic analysis using parsimony), version 4, written by David L. Swofford (Smithsonian Institution). The results are shown here with Dr. Swofford’s permission.
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
The striking nucleotide conservation found in the first framework
region (FR1) of group III VH gene segments from
different species has facilitated using cloned VH from one
species to cross-hybridize VH from another species (20).
Using a cloned germline VH (P83) from D.
virginiana as a probe, we were able to detect approximately eight
faintly hybridizing bands, ranging in size from 1 to 10 kb, on a
Southern blot of EcoRI-digested M. domestica
genomic DNA (not shown). A phage library was constructed using
size-selected EcoRI fragments from M. domestica
genomic DNA, and the library was screened using P83 as a probe. Three
unique phage clones were identified, isolated, and sequenced and found
to contain partial VH segments with significant homology to
other mammalian group III sequences (sequences 17, 18, and 39 in Fig. 1
A). All three clones
contained the 
sequence in FR1. However, due to a common
EcoRI site in the FR3 region, they lacked approximately 37
nucleotides of 3' coding sequence and, since these clones are likely
germline fragments, the recombination signal sequences also were
missing.
|
Using one of the germline VH clones (clone 17) as
a probe, a spleen cDNA library from M. domestica was
screened, and eight full length clones containing the VH,
DH, JH, and CH regions were
identified. Six of these clones contained CH regions
homologous to other mammalian Cµ region genes, while the other two
were homologous to C (not shown). Not unexpectedly, all
VH sequences present in these clones had a high degree of
sequence homology with the germline VH segments (sequences
10, 11, 22, 23, and 26–28 in Fig. 1
A). The Cµ and
C sequences are available as GenBank accession numbers AF012109 and
AF012110, respectively; a more complete description of the M.
domestica C region genes will be made elsewhere (R. D. Miller and
G. H. Rosenberg manuscript in preparation). To screen for novel
VH sequences without bias, a probe containing the M.
domestica Cµ coding sequence and lacking VH,
DH, and JH was derived from one of the full
length cDNAs and used to screen the spleen cDNA library. A total of
nine clones (sequences 94 and 96 in Fig. 1A are
representative) were identified, and the VH regions were
sequenced. Although identified solely on the basis of the presence of
Cµ, all nine clones contained VH regions that had >85%
nucleotide similarity to the previously identified VH
sequences. All of these clones, therefore, are by definition members of
a common VH family that we have designated
MdoVH1 (M. domestica VH
family 1).
Search for other M. domestica VH families
To expedite and simplify the screening of large numbers of
expressed VH sequences, anchored PCR was performed
using the spleen cDNA library as target. Reverse primers complementary
to the 5' coding sequences of the Cµ and C regions were designed
and paired in PCR with primers specific for cloning vector sequences in
the ZAPII phage used to construct the cDNA library. PAGE of the
amplified products revealed a diffuse band 
600 bp (not shown), which
is close to the predicted product size for amplifying a complete V
region domain, including the CDR3 region and the 5' end of the constant
region. Since the CDR3 region is created by the junction of the
VH, DH, and JH segments and can
contain variable numbers of nucleotides, variable length PCR products
would be expected if amplification were polyclonal. Similar results
were achieved using the Cµ or C primers paired with either the M13
or T3 primers in PCR. A total of 32 new, unique VH
sequences were generated from this strategy (25 using the Cµ primer
and 7 using the C primer). Most of these sequences show >80%
similarity to the known MdoVH1 family (Fig. 1A,
sequences 298, 338, 356, and 364 are representative). However, six
clones were found to differ significantly from the MdoVH1
family (Fig. 1B). This second set of sequences
shared, on average, <75% nucleotide homology with the
MdoVH1 sequences. These differences, by convention, place
the sequences shown in Figure 1B in a second VH
family, which we have designated MdoVH2. Two of the six
MdoVH2 clones were later found to be identical to the
germline MdoVH2 sequence and are not shown. The cloning of
the germline MdoVH2 gene segment is described in the next
section. A comparison between MdoVH1 and MdoVH2
family sequences, found 74% similarity at the nucleotide level and
68% identity at the amino acid level in the coding region (Fig. 2). As shown in Figure 2, most of the
differences are in CDR1 and CDR2, although there are stretches of
nucleotides in both FR1 and FR3 which contribute to the differences
between the V segments. All six MdoVH2 sequences isolated
by PCR were generated using the Cµ primer; none were found using
C. Using clone 340 as a probe, the spleen cDNA library was screened
for additional full length clones containing a VH from the
MdoVH2 family. Screening the equivalent of approximately
8 x 104 phage clones detected only a single clone, an
IgA, containing an MdoVH2 family sequence (sequence 511 in
Fig. 1B).
|
Using a representative VH clone from each of the
two families as probes, a Southern blot containing M.
domestica genomic DNA digested with several different restriction
enzymes was hybridized to analyze the number of VH
sequences present in the genome. As shown in Figure 3, the MdoVH1 (clone 356)
probe hybridized with varying intensities and had 8 to 12 restriction
fragments depending on the enzyme used to cut the DNA. The variation in
intensity between hybridizing bands likely reflects the degree of
homology between the germline fragment and the cDNA used as a probe.
However, we cannot rule out the possibility that some more intense
bands represent multiple VH segments comigrating. The
estimate of the total number of MdoVH1 gene segments is
based on observing the number of bands present using a variety of
restriction enzymes. In contrast, the MdoVH2 family probe
(clone 340) hybridizes to only a single restriction fragment in most
lanes, suggesting that the MdoVH2 family comprises only a
single VH. The MdoVH2 family probe also
hybridized a 4.3-kb SpeI fragment from M.
domestica genomic DNA (not shown), which was cloned from a phage
library constructed from SpeI-digested, size-selected
genomic DNA to determine the germline sequence of this single-member
VH family. This sequence is shown as the "germline"
sequence in Figure 1B for comparison with the sequences
generated by PCR amplification.
|
B).
The intron and recombination signal sequence (RSS) sequences for the
germline MdoVH2 gene segment are available from GenBank as
accession number AF012124. Phylogenetic analysis of M. domestica VHsequences
Both the MdoVH1 and MdoVH2 sequences
consistently showed the highest match with human V3 family sequences
(group III) when compared with the GenBank database using the BLAST
algorithms. However, when compared with the Kabat database (21) at the
amino acid level, MdoVH2 was most similar to mouse and
human group II VH (not shown). To investigate these
relationships further, a nucleotide alignment for mammalian
VH was made using representative sequences for each of the
2 opossum, 14 mouse, and 6 human VH families so far
identified, and the single known rabbit, swine, and sheep
VH families available. The human VH7 family was
not included in this analysis as it is closely related to human
VH1. Included in the alignment was the only other known
marsupial VH sequence, from the North American opossum
D. virginiana. Using these alignments, a phylogenetic tree
was generated (Fig. 4) that had similar
topology to those reported by others (1, 9). Interestingly, all of the
marsupial VH sequences, including the North American
opossum, clustered on a common branch, and this branch shared a common
node with some of the human V3 sequences included in the alignment.
|
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
of cattle (22),
it is apparent that mouse or human Ig genetics are not a satisfying
paradigm for all vertebrates. Studies of VH genes in sheep
and pigs have found limited germline diversity in these species as well
(7, 8). We report the first molecular analysis of a marsupial Ig
repertoire and demonstrate that, in at least this one metatherian
species, the available germline VH repertoire also appears
to be limited in number and diversity. It is interesting to speculate
that, as information is collected on more species, rodents and primates
may turn out to be the exceptions among mammals rather than the rule by
maintaining diverse VH families in their genomes.
Although the available germline VH repertoire appears
limited in M. domestica, a significant number of sequence
differences separate the VH into two divergent families,
designated here as MdoVH1 and MdoVH2. By
convention, VH sequences have been grouped into a common
family when they share at least 80% nucleotide homology. Initially,
this cut-off was established as the limit of cross-hybridization
between VH sequences on Southern blots using standard
hybridization stringency (23). The separation of the M.
domestica VH into two distinct families fits these
criteria, with <75% similarity shared at the nucleotide level between
MdoVH2 and any of the MdoVH1 sequences found so
far. It should be cautioned that since many of the sequences presented
in Figure 1 were generated by PCR, some nucleotide differences may be
cloning artifacts. However, the significant sequences differences found
between MdoVH1 and MdoVH2 cannot be accounted
for by cloning artifacts. Based on nucleotide sequence homology and
phylogenetic analysis, both families fall into group III, and most of
the segments contain the sequence (5'-GCTGGTGG) in FR1, a
characteristic found to be conserved across group III families in all
vertebrates (20). The MdoVH1 and VH2 family
sequences are further separated by a two-, and sometimes three-codon
deletion in the CDR2 region of MdoVH1, which would
correspond to amino acid residues 54, 55, and 56 in MdoVH2.
The MdoVH1 sequences shown in Figure 1A were
chosen to represent the diversity of sequences found in this family.
Based on FR, CDR, and leader sequence differences, 6 would be a
conservative estimate of the number of independent VH1
segments representative of the 10 to 12 germline segments seen by
Southern blot analysis. A comparison of sequence 11 with sequences 338
(codon deletion in CDR2), 364 (codon insertion in FR3), 10, and 57
(distinct leader sequences) suggest that they are derived from five
unique VH segments. Sequence 26 has a significantly
different FR1 sequence and likely represents a sixth germline
VH segment. All of the sequences shown have significant
nucleotide differences and probably represent other distinct germline
VH1 members, but at this time, without all of the germline
MdoVH1 having been sequenced, cloning artifacts and somatic
mutations in rearranged VH segments cannot be ruled out.
Some of the germline MdoVH1 segments are likely to be
pseudogenes that are not expressed. The recent completion of the entire
human Igh locus found that 43 of 87 VH segments
were pseudogenes (24, 25). How many of the MdoVH1 members
are functional will be determined when all germline members are
sequenced, which will also provide the opportunity to analyze the role
of gene conversion in VH diversity in this species. Most of
the expressed VH repertoire is derived from the
MdoVH1 family members, which may simply reflect
MdoVH1 having multiple members while MdoVH2 has
only a single member. VH families with one or only a few
members have been described in the Igh locus of other
species as well. The mouse VH12 and VH3609N
families (26, 27) and the human VH6 (28) family are
examples. Any duplications, functional or not, which might have
occurred during the evolution of MdoVH2 in the marsupial
Igh have not been maintained in the Monodelphis
genome. Since additional copies of MdoVH2 family members
appear not to exist in Monodelphis, it seems unlikely that
gene conversion would contribute to diversification of expressed
MdoVH2 sequences. However, it is possible that, as in sheep
(7), somatic mutation before Ag selection may contribute to the
diversity of a primary MdoVH2 repertoire. Further analysis
of rearrangements using MdoVH2 during early B cell ontogeny
in this species is needed to address this question. Whether
MdoVH2 homologues exist in other marsupial species also
remains to be determined.
Evolution of marsupial VH genes
All Monodelphis VH found so far are most similar to the group III-related sequences from other species; however, the degree of nucleotide differences between MdoVH1 and MdoVH2 suggests that they are not recent duplications. Phylogenetic analysis of MdoVH1 and MdoVH2 places these two VH families on a common branch within the group III VH families, most closely to some of the human V3 gene segments. This is not unusual given that the divergence times of some of the VH families predate mammalian radiation (9, 20, 29). The only other known marsupial VH is from another didelphid, D. virginiana, which is also a group III member and is most closely related to MdoVH1. From the phylogenetic tree, it is impossible to determine whether the separation of MdoVH1 and MdoVH2 occurred before or after the divergence of marsupials from placental mammals 130 million years ago, but it probably did occur before the radiation of the didelphid marsupials (i.e., Monodelphis from Didelphis), which has been estimated to have occurred anywhere between 15 to 50 million years ago (11).
It is curious that MdoVH1 and MdoVH2 are so
closely related phylogenetically given that they share <75%
nucleotide similarity on average. In contrast, all the human V3
sequences share >75% nucleotide similarity, yet V3–15 is separated
phylogenetically from the other V3 members in the alignment shown in
Figure 4. The sequence dissimilarity between MdoVH1 and
MdoVH2 is apparent in the phylogenetic tree by the long
branch length for MdoVH2. Phylogenetic relatedness with a
low overall sequence similarity can be explained by the pattern of
nucleotide sequence distribution and distinctly different codon usage
in the CDR1 and CDR2 regions of MdoVH2 as compared with
MdoVH1. As described below, this may reflect a very
different pattern of somatic mutation in MdoVH2 compared
with MdoVH1.
Codon usage and the likelihood of somatic mutation
Targeting somatic mutation to the CDRs has been associated with
nonrandom distribution of nucleotide sequences in V region gene
segments. Rogozin and Kolchanov (30) reported finding that two sites,
RGY and TAA, are more likely to show a mutation than other sequences,
with the mutation most likely occurring at the underlined position. The
most notable mutational "hot spot" is the serine codon AGY, and an
unusual pattern of codon bias for serines is found in variable regions,
with the nonmutable serine codon, TCN, predominating in FRs and the
mutable AGY in CDRs (31, 32). All VH families in mammals
demonstrate this pattern, and the MdoVH1 family sequences
are no exception, further demonstrating the conserved nature of this
mechanism for directing mutation to the CDRs. However, while the CDRs
of most MdoVH1 sequences are serine rich, the CDR1 and CDR2
regions of MdoVH2 are strikingly devoid of serines, as can
be seen in the alignment shown in Figure 2. This suggests that the
MdoVH2 gene segment may be under very different selection
pressure(s) than the MdoVH1 family members. For example, in
the MdoVH1 germline sequence from clone 17 (Fig. 1A), there are 13 serines, 8 in the FRs and 5 in the
CDRs. Of the FR serines, 6 of 8 use the TCN codon and the other 2 use
AGY. Four of the 5 serines present in the CDRs use the AGY codon and
the other uses TCN. It is likely then that the members of the
MdoVH1 family are good targets for somatic mutation
directed toward the CDRs by maintaining bias toward AGY in the CDRs. In
contrast, the sequence "AGY" is noticeably absent from the
MdoVH2 CDR1 and CDR2 regions in any reading frame, although
other potential mutation hot spots are still present. The codon
"GGT" encoding glycine satisfies the RGY motif and TAA is present
in a CDR as GTA ATA, although because of the reading frame, changing
the first A in the sequence would result in a silent mutation. Whether
MdoVH2 demonstrates an unusual pattern of somatic mutation
or is devoid of somatic mutations remains to be seen through further
analysis. All of the Ig cDNA clones that contain MdoVH2
have been, so far, IgM clones; an immediate goal will be to search for
the use of this VH family in other isotypes to analyze
patterns of somatic mutation presumably following Ag activation and
affinity maturation.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
2 All sequences reported, as well as additional Monodelphis VH sequences not described, have been deposited in the GenBank/EMBL database and assigned accession numbers AF007070 to AF007093, AF012109 to AF012124, AF020794, and AF020795.
3 Address correspondence and reprint requests to Dr. R. D. Miller, Department of Biology, University of New Mexico, Albuquerque, NM 87131-0001.
4 Abbreviations used in this paper: Igh, immunoglobulin heavy chain gene; CDR, complementarity-determining region; FR, framework region.
Received for publication July 14, 1997. Accepted for publication September 22, 1997.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
This article has been cited by other articles:
|
|
 |
|
  M. L. Baker, K. Belov, and R. D. Miller Unusually Similar Patterns of Antibody V Segment Diversity in Distantly Related Marsupials J. Immunol., May 1, 2005; 174(9): 5665 - 5671. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. Johansson, M. Aveskogh, B. Munday, and L. Hellman Heavy Chain V Region Diversity in the Duck-Billed Platypus (Ornithorhynchus anatinus): Long and Highly Variable Complementarity-Determining Region 3 Compensates for Limited Germline Diversity J. Immunol., May 15, 2002; 168(10): 5155 - 5162. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
F. M. Adamski and J. Demmer Two Stages of Increased IgA Transfer During Lactation in the Marsupial, Trichosurus vulpecula (Brushtail Possum) J. Immunol., May 15, 1999; 162(10): 6009 - 6015. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. E. Lucero, G. H. Rosenberg, and R. D. Miller Marsupial Light Chains: Complexity and Conservation of {lambda} in the Opossum Monodelphis domestica J. Immunol., December 15, 1998; 161(12): 6724 - 6732. [Abstract] [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
