This information is current as
of June 18, 2017.
Ig Knock-In Mice Producing
Anti-Carbohydrate Antibodies:
Breakthrough of B Cells Producing Low
Affinity Anti-Self Antibodies
Lorenzo Benatuil, Joel Kaye, Nathalie Cretin, Jonathan G.
Godwin, Annaiah Cariappa, Shiv Pillai and John Iacomini
J Immunol 2008; 180:3839-3848; ;
doi: 10.4049/jimmunol.180.6.3839
http://www.jimmunol.org/content/180/6/3839
Subscription
Permissions
Email Alerts
This article cites 64 articles, 20 of which you can access for free at:
http://www.jimmunol.org/content/180/6/3839.full#ref-list-1
Information about subscribing to The Journal of Immunology is online at:
http://jimmunol.org/subscription
Submit copyright permission requests at:
http://www.aai.org/About/Publications/JI/copyright.html
Receive free email-alerts when new articles cite this article. Sign up at:
http://jimmunol.org/alerts
The Journal of Immunology is published twice each month by
The American Association of Immunologists, Inc.,
1451 Rockville Pike, Suite 650, Rockville, MD 20852
Copyright © 2008 by The American Association of
Immunologists All rights reserved.
Print ISSN: 0022-1767 Online ISSN: 1550-6606.
Downloaded from http://www.jimmunol.org/ by guest on June 18, 2017
References
The Journal of Immunology
Ig Knock-In Mice Producing Anti-Carbohydrate Antibodies:
Breakthrough of B Cells Producing Low Affinity
Anti-Self Antibodies1
Lorenzo Benatuil,* Joel Kaye,* Nathalie Cretin,† Jonathan G. Godwin,* Annaiah Cariappa,‡
Shiv Pillai,‡ and John Iacomini2*
N
atural Abs, those produced without intentional immunization, play a major role in providing protective host
immunity (1–3). However, the study of natural Abs has
been difficult because in many cases their Ag specificity is unknown. The specificity of many natural Abs remains undefined;
however, a significant portion are specific for carbohydrates such
as blood group Ags. In addition to anti-ABO blood group Abs,
natural Abs specific for a carbohydrate Ag Gal␣1–3Gal␤1– 4GlcNAc-R (␣Gal), represent a significant population of natural Abs
(4 –11). The ␣Gal Ag is synthesized by the glucosyltransferase
UDP galactose:␤-D-galactosyl-1,4-N-acetyl-D-glucosaminide
␣(1–3) galactosyltransferase (Enzyme classification E.C.
2.4.1.151), or ␣GT. All placental mammals except humans, apes
and Old World monkeys express a functional ␣GT enzyme and
␣Gal epitopes on most tissues (12), and are consequently tolerant
to ␣Gal because it is recognized as part of self. In contrast, humans, apes and Old World primates carry a nonfunctional ␣GT
gene whose function appears to have been lost ⬃30 million years
*Transplantation Research Center, Brigham and Women’s Hospital, Children’s Hospital Boston and Harvard Medical School, Boston, MA 02115; †Novartis Pharma,
Infectious Diseases, Transplantation & Immunology, Basel, Switzerland; and ‡Massachusetts General Hospital, Center for Cancer Research, Charlestown, MA 02129
Received for publication December 31, 2007. Accepted for publication January
4, 2008.
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 work was supported by Grants R01AI044268-09 and R01 AI050602-06 from
the National Institutes of Health (to J.I.).
2
Address correspondence and reprint requests to Dr. John Iacomini, Transplantation
Research Center, Brigham and Women’s Hospital and Children’s Hospital Boston,
Harvard Medical School, 221 Longwood Avenue, Room LM303, Boston, MA 02115.
E-mail address: [email protected]
Copyright © 2008 by The American Association of Immunologists, Inc. 0022-1767/08/$2.00
www.jimmunol.org
ago (5). Because these species do not recognize ␣Gal structures as
self, they consequently produce ␣Gal-specific Abs.
␣Gal-specific natural Abs are estimated to comprise 1– 8% of
circulating Ig in humans, and ⬃1% of EBV-transformed peripheral
blood B cells make Abs that bind ␣Gal (6, 13). In humans, ␣Galspecific Abs are encoded for by a restricted set of Ig Vh genes from
the VH3 family (14). Production of Abs specific for ␣Gal is believed to be elicited in response to normal bacterial flora that colonize the human gastrointestinal tract (11, 15). The presence of
anti-␣Gal Ab in serum and secretory fluids, such as colostrum and
saliva, suggests that these Abs have evolved to play a protective
role in primate immunity. Viruses produced in ␣GT-expressing
cells that display ␣Gal-modified glycoproteins within their envelop, such as lymphocytic choriomeningitis virus, Newcastle disease virus and vesicular stomatitis virus, as well as C-type retroviruses, have all been shown to be susceptible to inactivation by
serum anti-␣Gal Abs (16, 17). ␣Gal-specific Abs are therefore
believed to play an important role in preventing cross-species infection by pathogens. ␣Gal-specific Abs have also been shown to
play an important role in rejection of xenogeneic tissue when
transplanted into non-human primates (18 –22). Despite the importance of ␣Gal-reactive Abs to host immunity little is known about
how development of B cells producing ␣Gal-reactive or other anticarbohydrate Abs is regulated.
␣Gal-deficient mutant mice lacking a functional ␣GT gene
(GT0/0 mice) generated by gene targeting in embryonic stem cells
lack expression of ␣Gal epitopes and consequently develop ␣Galspecific natural Abs, the majority of which are IgM (23–25). The
serum titer of ␣Gal-specific Abs in GT0/0 mice increases in an
age-dependent fashion. ␣Gal-specific Abs in GT0/0 mice share
many features with human ␣Gal-specific Abs, including usage of
related V genes (26 –31). These mice therefore represent a small
animal model in which ␣Gal-reactive Abs can be studied. However, the frequency of B cells that produce ␣Gal-specific Abs in
Downloaded from http://www.jimmunol.org/ by guest on June 18, 2017
Natural Abs specific for the carbohydrate Ag Gal␣1–3Gal␤1– 4GlcNAc-R (␣Gal) play an important role in providing protective
host immunity to various pathogens; yet little is known about how production of these or other anti-carbohydrate natural Abs is
regulated. In this study, we describe the generation of Ig knock-in mice carrying functionally rearranged H chain and L chain
variable region genes isolated from a B cell hybridoma producing ␣Gal-specific IgM Ab that make it possible to examine the
development of B cells producing anti-carbohydrate natural Abs in the presence or absence of ␣Gal as a self-Ag. Knock-in mice
on a ␣Gal-deficient background spontaneously developed ␣Gal-specific IgM Abs of a sufficiently high titer to mediate rejection of
␣Gal expressing cardiac transplants. In the spleen of these mice, B cells expressing ␣Gal-specific IgM are located in the marginal
zone. In knock-in mice that express ␣Gal, B cells expressing the knocked in BCR undergo negative selection via receptor editing.
Interestingly, production of low affinity ␣Gal-specific Ab was observed in mice that express ␣Gal that carry two copies of the
knocked in H chain. We suggest that in these mice, receptor editing functioned to lower the affinity for self-Ag below a threshold
that would result in overt pathology, while allowing development of low affinity anti-self Abs. The Journal of Immunology, 2008,
180: 3839 –3848.
Ig KNOCK-IN MICE PRODUCING ANTI-␣GAL Abs
3840
Materials and Methods
Mice
C57BL/6 GT0/0 mice and derivation of the colony used in these studies
have been described in (35). C57BL/6 and BALB/c mice were used as
controls and were obtained from The Jackson Laboratory. All mice were
housed in viral Ab-free microisolator conditions. All animal experiments
were conducted in accordance with Institutional guidelines.
Generation of M86VHVL knock-in mice
The functionally rearranged M86 VDJH and VJ␬ regions were isolated
using standard genomic cloning techniques. To construct Ig H chain and L
chain targeting vectors the rearranged M86VH and VL gene segments were
then cloned using standard techniques into the plVhL2Neo (36) and
pVKRNeo (37) H and L chains targeting vectors kindly provided by Dr. K.
Rajewsky (Immune Disease Institute, Harvard Medical School, Boston,
MA). Vector integrity was confirmed by restriction mapping. Gene targeting and generation of chimeric mice was performed essentially as described (38). Briefly, each targeting vector was linearized and transfected
separately into ES-J1 embryonic stem cells. Transfected cells were selected
in presence of G418 (300 mg/ml) and gancyclovir (2 mM). DNA prepared
from double drug-resistant colonies was then screened for the presence of
homologous recombination by Southern blotting. ES clones containing either the targeted M86VH or VL regions were injected into C57BL/6 blastocysts and then transferred into (BALB/c ⫻ C57BL/6) F1 foster mothers.
Blastocyst injection, and embryo transfer, was performed by the Massachusetts General Hospital Microinjection Core Facility (Boston, MA). Chimeric mice were mated to C57BL/6 mice. DNA prepared from tail biopsy
samples of resulting offspring was analyzed by Southern blotting to confirm germline transmission. Resulting knock-in mice were then crossed
with cre mice (provided by the Massachusetts General Hospital Core Facility) to delete the Neor gene from targeted H chain and L chain loci.
Resulting knock-in mice were then crossed to GT0/0 mice on the C57BL/6
background to generate M86VHGT⫹/0 and M86VLGT⫹/0 mice. These mice
were then bred to generate ␣Gal expressing M86VHVLGT⫹/0 and ␣Galdeficient M86VHVLGT0/0 mice.
Flow cytometry and Abs
Single cell suspensions were prepared from blood or lymphoid tissues and
then stained and analyzed by flow cytometry as described (35). ␣Gal
epitopes were detected using the Gal-specific IB4 lectin (Sigma-Aldrich)
from Bandeiraea simplicifolia (BS-I isolectin B4) (39). The following Abs
used in this study were purchased from BD Pharmingen: RA3-6B2 (antiCD45R/B220), 187.1 (anti-mouse Ig␬), R26-46 (anti-mouse ␭1, ␭2, and
␭3), 11-26c.2a (anti-IgD FITC), R6-60.2 (anti-IgM), 7G6 (anti-CD21), S7
(anti-CD43), and anti-Mac-1 (anti-CD11b). Goat anti-mouse IgM was purchased from Jackson ImmunoResearch Laboratories. RS3.1 (anti-Igh-6a
(40)) and MB86 (anti-Igh-6b (41)) were provided by Dr. H. Wortis (Tufts
University Sackler School of Biomedical Sciences, Boston MA). B cells
capable of binding ␣Gal were detected by staining with FITC- or biotinconjugated Gal-BSA (V-Labs).
ELISA
ELISAs were conducted as previously described (35). Briefly, ELISA
plates (Corning) were coated overnight at 4°C with either ␣Gal conjugated
to BSA (Gal-BSA) or lactosamine conjugated to BSA (Lac-BSA; V-Labs)
in carbonate buffer (pH 9.5), and then washed with PBS containing 0.05%
Tween 20 (PBS-Tween 20). Lac-BSA shares all determinants with GalBSA, except for the terminal galactose structure, and serves as a specificity
control. The wells were blocked with 1% BSA in PBS-Tween for 1 h at
room temperature and then washed. Serum samples were serially diluted in
PBS-Tween 20, added to the plates, and incubated for 1 h at 37°C. The
plates were then washed extensively with PBS-Tween 20, and bound Abs
were detected using HRP-conjugated goat anti-mouse IgM (1/4000; Jackson ImmunoResearch Laboratories). To determine the relative contribution
of transgene-encoded vs endogenously encoded anti-␣Gal, bound Abs
were detected with purified biotinylated RS3.1 or MB86, followed by
HRP-conjugated streptavidin (1/800; Amersham Biosciences). The plates
were incubated for 1 h at 37°C and then washed five times with PBSTween 20. A total of 0.01 mg/ml o-phenylenediamine dihydrochloride
(Sigma-Aldrich) in substrate buffer was then added for 20 min at room
temperature to develop the assays. The reaction was terminated by adding
sulfuric acid to each well, and absorbency was read at 492 nm. Background
values obtained from Lac-BSA-coated plates were subtracted from those
obtained using Gal-BSA-coated plates. Assays were performed in duplicate. In some instances, serum from immunized mice was used. Mice were
immunized i.p. with 107 irradiated (3000 rad) pig PBMC as described (42).
Anti-␣Gal B cell ELISPOT assay
Multiscreen-HA plates (Millipore, Bedford, MA) were coated with 10
␮g/ml of either Gal-BSA or Lac-BSA in PBS at 4°C overnight. The plates
were then washed three times with PBS, allowing the plates to soak for 5
min between each wash. The plates were blocked with IMDM containing
0.4% BSA and penicillin and streptomycin for 2 h at 37°C. The blocking
medium was then removed and 10-fold serial dilutions (starting at 1 ⫻ 106
cells per well) of spleen cells prepared in blocking IMDM were added to
the wells. The cells were incubated at 37°C in 5% CO2 for 24, 48, or 72 h
in the presence or absence of LPS (0.5 ␮g/well). After culture, the plates
were washed three times in PBS, followed by three additional washes in
PBS-Tween 20. HRP-conjugated goat anti-mouse IgM was then added to
each well and incubated for 2 h at 4°C. The plates were washed three more
times with PBS-Tween 20, followed by PBS, at which point the assays
were developed by adding filtered chromogen substrate (3-amino-9-ethylcarbazole) in acetate buffer (pH 5.0). Plates were incubated in the presence
of chromogen substrate at room temperature for 5 min and the reaction
terminated by washing the plate with water. Spots were enumerated using
an automated ELISPOT reader (ImmunoSpot; Cellular Technology). In all
assays, the number of background spots obtained on Lac-BSA-coated
plates was subtracted from the number obtained on corresponding GalBSA-coated plates. All samples were plated in duplicate.
Cell culture
B cell precursors from M86VHVL-GT0/0 mice were grown in vitro as previously described (43, 44). Briefly, bone marrow cells were depleted of
erythrocytes and were cultured in BMB220 medium (IMDM supplemented
with 10% FCS, 2-ME, L-glutamine, penicillin/streptomycin, and 50 –100
U/ml recombinant murine IL-7; R&D Systems) at a concentration of 2 ⫻
106 cells/ml for 5 days. Washed cells were then cultured for 2– 42 h in the
absence of IL-7 in wells with irradiated (2000 rad) confluent primary adherent bone marrow-derived stroma from Ag-negative GT0/0 or Ag-bearing
wild-type C57BL/6 mice. After culturing the cells on the stroma, nonadherant pre-B cells were washed and frozen at ⫺80°C until RNA was extracted. Stromal cultures were initiated by plating erythrocyte-depleted
bone marrow cells to confluence in stromal medium (RPMI 1640, 5% FCS,
2-ME, L-glutamine, penicillin/streptomycin, sodium pyruvate, and nonessential amino acids) for 3 days, then washing away nonadherent cells and
allowing the adherent stroma to grow for at least 2 wk before use. Stromal
cell lines were tested for the expression of ␣Gal by staining with ␣Galspecific IB4 lectin.
Detection of RAG2 expression
RNA was prepared from cells using an RNeasy mini kit (Qiagen). Complementary DNA was prepared from DNaseI-treated (Invitrogen Life
Downloaded from http://www.jimmunol.org/ by guest on June 18, 2017
these mice is low, making direct analysis of B cells producing
␣Gal-specific Abs difficult. We and others have generated Ig transgenic mice to study regulation of B cells producing ␣Gal-specific
Abs (32, 33). However, the use of these mice to address several
aspect of B cell development is limited because the Ig transgenes
are randomly integrated. They are not transcriptionally regulated
in an identical fashion to endogenous Ig genes, and in the case of
Ig L chain transgenes, unlike endogenous self-reactive rearranged
␬ L chain genes, are not subject to deletion during receptor editing.
Therefore, to develop a mouse model in which regulation of ␣Galspecific Ab production could be studied, we used gene targeting in
embryonic stem cells to construct Ig gene knock-in mice. To this
end, the rearranged VH and VL regions encoding specificity for
␣Gal were isolated from M86, a hybridoma (IgM, ␬ L chain),
derived from GT0/0 mice (34). M86VHVL knock-in mice were
generated on either an Ag-deficient GT0/0 or Ag-sufficient GT⫹/⫹
or GT⫹/⫺ background to address fundamental issues in regulation
of anti-carbohydrate natural Abs. Using these mice, we examined
the source of B cells producing ␣Gal-specific Abs and mechanisms
leading to negative selection of B cells producing anticarbohydrate Abs. Our data indicate that restricting the ability of B
cells producing self-reactive anti-carbohydrate Ags to undergo receptor editing significantly affects B cell development and allows
for the production of low affinity anti-self Abs.
The Journal of Immunology
3841
Technologies) RNA with oligo(dT) primers with the Superscript firststrand synthesis kit (Invitrogen Life Technologies). Primer sequences
used are as follows: RAG2 forward primer, 5⬘-CACATCCACAAGCA
GGAAGTACAC-3⬘; RAG2 reverse primer, 5⬘-GGTTCAGGGACATC
TCCTACTAAG-3⬘; ␤-actin forward primer, 5⬘-ACCCCAAGGCCAAC
CGCGAGAAGATGACC-3⬘; ␤-actin reverse primer, 5⬘-GGTGATGA
CCTGGCCGTCAGGCAGCTCGTA-3⬘. PCR were performed in a final
volume of 50 ␮l using 3–5 ␮l of cDNA and 2 U Taq polymerase (Fischer Scientific) on a GeneAmp PCR 2400 thermal cycler
(PerkinElmer). Except for the first cycle, which had a 2 min 94°C denaturation step, each cycle consisted of 1 min at 94°C, 1 min at 60°C,
and 1 min at 72°C, with a 7 min 72°C extension cycle to end the PCR.
For PCR analysis, 35 cycles was used. The PCR product was then
fractionated by agarose gel electrophoresis, transferred to HybondN nylon membranes (Amersham Biosciences) and RAG2 amplicons
detected by hybridization to 32P-labeled RAG2 probes as described
(45).
RAG2 expression was also analyzed by quantitative real-time PCR. One
microgram of total RNA was reverse transcribed using the SuperScript III
First-Strand Synthesis SuperMix for quantitative real-time PCR kit (Invitrogen Life Technologies) and used to quantitate relative levels of RAG2
expression using the TaqMan predeveloped assay Mm00501300_m1 for
RAG2, and 4352932E for GADPH (Applied Biosystems). The assays were
performed in triplicates following standard protocols. The values shown
are presented as the difference in cycle threshold (Ct) values normalized to
GADPH for each sample (RQ).
Data analysis and reproducibility
Data throughout the study are representative of experiments containing at
least three age- and sex-matched mice per group. All experiments have
been shown to be reproducible by multiple individuals and have been conducted and confirmed on multiple occasions.
Results
Construction of M86VHVL Ig knock-in mice
The low frequency of B cells that produce ␣Gal-specific Abs in
GT0/0 mice (25, 46, 47) limits the use of these mice to study the
regulation of ␣Gal-specific Ab production. To develop a mouse
model in which B cells producing ␣Gal-specific Abs could be
directly tracked during their development, we generated Ig
knock-in mice that express an ␣Gal-specific BCR. To this end, the
rearranged Ig H chain and L chain variable region segments were
cloned from the hybridoma M86. M86 was derived from GT0 mice
and produces an ␣Gal-specific IgM Ab that uses a ␬ L chain (26,
32, 34). The H chain (rearranged to Jh4) and L chain (rearranged
to J␬1) variable region gene segments were then cloned into the
plVhL2Neo (36) and pVKRNeo (37) H and L chains Ig targeting
vectors. Each targeting vector was then electroporated separately
into 129/Sv embryonic stem cells (ES-J1) as described (38). ES
clones containing either the targeted M86VH or M86VL region
were then injected separately into C57BL/6 blastocysts that were
then transferred into (BALB/c ⫻ C57BL/6) F1 foster mothers to
generate chimeric mice as described (38). Offspring carrying a
knocked in M86VH or M86VL allele were then crossed with Cre
recombinase transgenic mice to delete the neomycin resistance
gene from targeted H chain and L chain loci. Resulting knock-in
mice were then crossed to GT0/0 mice on the C57BL/6 background
to generate M86VHGT⫹/0 and M86VLGT⫹/0 mice (Fig. 1A) that
were then bred to generate ␣Gal-expressing M86VHVLGT⫹/0
mice and ␣Gal-deficient M86VHVLGT0/0 mice.
Downloaded from http://www.jimmunol.org/ by guest on June 18, 2017
FIGURE 1. Characterization of ␣Gal-specific Ab production in M86VHVL knock-in mice. A, Amplification of tail DNA from M86VHVL knock-in mice
and normal controls by PCR. The knocked in H chain was detected using primers that amplify the recombined VHDJ4 segment (left panel). To analyze the
presence of one or two copies, an additional primer was introduced to amplify the intron between JH3 and JH4 (JH3-JH4, top band) present only in wild-type
alleles. A similar approach was used to detect the L chain allelic expression (right panel). B and C, Anti-␣Gal Ab production in naive M86VHVL Ig knock-in
mice. In all cases, binding to Lac-BSA was used as a specificity control. Data shown are representative of multiple experiments containing at least three mice
per group. B, Anti-␣Gal serum titers in naive (marked by x) and immunized M86VHVLGT0/0 (Œ) and M86VHVLGT⫹ mice (). Also shown are anti-␣Gal serum
titers in C57BL/6 (F) and naive GT0/0 (䡺) mice. C, Production of knocked in IgMa () or endogenous IgMb (⽧) anti-␣Gal Abs in naive M86VHVLGT0/0 mice.
D, ELISPOT analysis of ␣Gal-specific IgM production in naive M86VHVLGT0/0, M86VHVLGT⫹, and immunized GT0/0 mice. The frequency of Ab-secreting cells
per 106 total cells in each tissue is shown. E, Survival of GT⫹/⫹ () or GT0/0 (E) hearts transplanted into naive M86VHVLGT0/0 mice.
3842
Ig KNOCK-IN MICE PRODUCING ANTI-␣GAL Abs
Downloaded from http://www.jimmunol.org/ by guest on June 18, 2017
FIGURE 2. Characterization and
early B cell development in M86VHVL
knock-in mice. A, Staining of bone
marrow cells from C57BL/6, M86VH
VLGT⫹, and M86VHVLGT0/0 mice for
expression of B220 and CD43. Shown
are the frequency of B220⫹, CD43⫺ B
cells and B220⫹, CD43⫹ B cells as determined by flow cytometry. B, Analysis of sIgMa and sIgMb expression by
B220⫹ cells in bone marrow by flow
cytometry. C, Analysis of ␣Gal-binding by sIgMa and sIgMb expressing B
cells (B220⫹) in the bone marrow of
M86VHVLGT0 (top row) and M86VH
VLGT⫹ (bottom row) mice. D, Analysis of ␣Gal-binding in B220⫹,
CD43⫺ pre-B cells and B220⫹,
CD43⫹ pro-B cells from M86VHVL
GT0/0 mice. Data shown are representative of multiple experiments
containing at least three mice per
group.
Characterization of ␣Gal-specific Ab production in M86VHVL
knock-in mice
Analysis of natural serum Abs revealed that M86VHVLGT0/0 mice
contain high titers of ␣Gal-specific Ab in their serum (Fig. 1B).
The level of ␣Gal-specific Abs in M86VHVLGT0/0 mice was significantly higher than that observed in GT0/0 controls at all ages
examined (Fig. 1B). Immunization of M86VHVLGT0/0 mice with
␣Gal-expressing pig cells led to an increase in the titer of ␣Galspecific Abs (Fig. 1B). In contrast, we were unable to detect production of ␣Gal-specific Abs in the serum of M86VHVLGT⫹/0 or
M86VHVLGT⫹/⫹ mice even after immunization (Fig. 1B). In
M86VHVL mice the endogenous H chain locus is IgH-6b
(IgMb,␮b), whereas the knocked in H chain locus is IgH-6a
(IgMa,␮a), which can be detected using the anti-allotypic mAbs
MB86 (41) and RS3.1 (40), respectively. ␣Gal-specific Abs in
M86VHVLGT0/0 mice were encoded for by the knocked in IgMa
allotype, rather than the endogenous IgMb allotype (Fig. 1C). B
cells spontaneously producing ␣Gal-specific Abs were detected in
the spleen, lymph node, and bone marrow but not the peritoneum
of M86VHVLGT0/0 mice (Fig. 1D). Immunization with pig cells
did not result in the detection of ␣Gal-producing B cells in the
peritoneum of immunized M86VHVLGT0/0 mice (data not shown).
The frequency of ␣Gal-producing B cells in the spleen, bone marrow, and lymph nodes of M86VHVLGT0/0 mice was at least 10fold higher than in GT0/0 controls (Fig. 1D). We were unable to
detect B cells producing ␣Gal-specific Abs in lymphoid tissues
from M86VHVLGT⫹/0 or M86VHVLGT⫹/⫹ mice (Fig. 1D).
We next examined whether the titer of ␣Gal-specific Abs in
M86VHVLGT0/0 mice was sufficient to induce Ab-mediated
transplant rejection. MHC-matched hearts from littermate
M86VHVLGT⫹ or C57BL6 mice were heterotopically transplanted into the abdomen of M86VHVLGT0/0 mice. Hearts from
M86VHVLGT⫹ and C57BL/6 mice were uniformly rejected within
24 –72 h. In contrast, hearts from M86VHVLGT0/0 mice were accepted long-term (Fig. 1E). These data suggest that ␣Gal-specific
FIGURE 3. Self-antigenic stimulation induces expression of RAG2 in
M86VHVLGT0/0 BM B cells. Analysis of RAG2 up-regulation by bone
marrow pre-B cells upon Ag encounter. M86VHVLGT0/0 bone marrow
pre-B cells were grown in IL-7 for 5 days and then cultured on pre-established primary GT⫹ or GT0/0 bone marrow stromal cells for the times
indicated. A, Real-time PCR analysis by PCR and Southern blot showed
up-regulation of RAG2 mRNA in bone marrow pre-B cells cultured on
GT⫹ stroma. ␤-actin mRNA levels served as an internal control and were
evaluated by ethidium bromide staining. Films were exposed for 2 or 24 h.
B, Analysis of RAG2 expression by quantitative real-time PCR.
M86VHVLGT0/0 bone marrow pre-B cells were grown in IL-7 for 5 days
and then cultured for 2, 4, and 16 h on irradiated (2000 rad) GT⫹ (f) or
GT0/0 (䡺) bone marrow stromal cells. RAG2 levels were then assessed by
quantitative real-time PCR. Data presented are the difference in cycle
threshold values normalized to GAPDH for each sample (RQ) and are the
results of two different experiments. All assays were performed in triplicate. The level of RAG2 transcripts was statistically higher in pre-B cells
cultured in GT⫹ stromal cells (p ⬍ 0.05, at 2, 4 and 16 h, by Student’s t
test, unpaired 95% confidence).
The Journal of Immunology
3843
Abs in M86VHVLGT0/0 mice are functional and of sufficient titer
to induce Ab-mediated rejection.
Characterization of early lymphocyte development in M86VHVL
knock-in mice
M86VHVLGT0/0 and M86VHVLGT0/⫹ mice were sacrificed and
lymphoid tissues analyzed to examine lymphocyte development.
␣␤ T cell development in the thymus was essentially normal when
compared with GT0/0 or GT⫹ controls (data not shown). Similarly,
T cell development in the spleen and lymph nodes was normal
(data not shown). Characterization of B cell development in the
bone marrow of M86VHVLGT0/0 or M86VHVLGT⫹/⫹ knock-in
mice demonstrated that pro-B cells (B220⫹,CD43⫹) and immature
(B220⫹,CD43⫺) B cells, as defined in (48, 49), were present in
similar proportions to those observed in normal mice, although the
frequency of each of these fractions was reduced when compared
with normal controls (Fig. 2A), as observed in other Ig knock-in
mice (32, 33). The frequency of pre-B cells was similar in
M86VHVLGT0/0 or M86VHVLGT⫹/⫹ knock-in mice (data not
shown).
The majority of B220⫹ B cells in the bone marrow of
M86VHVLGT⫹ and M86VHVLGT0/0 mice expressed the knocked
in H chain allele (B220⫹, sIgMa⫹) (40.4 ⫾ 9.3% and 46.6 ⫾
10.1%, respectively) rather than endogenously encoded (B220⫹,
sIgMb⫹) allele reflecting exclusion by the rearranged M86VH gene
(Fig. 2B). Essentially all of the B220⫹, sIgMa⫹ B cells in the bone
marrow of M86VHVLGT0/0 mice were able to bind ␣Gal as determined by staining with fluorescently labeled BSA conjugated to
␣Gal (␣Gal-BSA) (Fig. 2C). B220⫹, sIgMb⫹ B cells able to bind
␣Gal-BSA were not detected in M86VHVLGT⫹ or M86VHVL
GT0/0 mice (Fig. 2C). Staining of M86VHVLGT0/0 bone marrow
cells for expression of B220 and CD43 revealed that B cells capable of binding ␣Gal-BSA were CD43⫺, pre-B or newly formed
B cells (Fig. 2D). B cells able to bind ␣Gal were not detected in the
bone marrow of M86VHVLGT⫹ mice, suggesting that B cells expressing the knocked in transgene are tolerized during their development (Fig. 2C).
Receptor editing prevents the development of B cells producing
self-reactive Abs
Because the frequency of B cells expressing surface IgMa (sIgMa)3
in the bone marrow of M86VHVLGT⫹ is relatively high (Fig. 2B)
even though we were unable to detect B cells that bind ␣Gal in the
bone marrow of M86VHVLGT⫹ mice (Fig. 2C) we reasoned that
tolerance to ␣Gal in M86VHVLGT⫹ mice was not deletional. We
3
Abbreviations used in this paper: sIgM, surface IgM; MZ, marginal zone.
Downloaded from http://www.jimmunol.org/ by guest on June 18, 2017
FIGURE 4. ␣Gal-binding B cells in the periphery of M86VHVLGT0/0 mice. A, Analysis of B220⫹, sIgMa, and sIgMb expressing B cells in the spleen
of M86VHVLGT0/0 and M86VHVLGT⫹ mice. B, Analysis of allelic exclusion in splenic B cells. Spleen cells from M86VHVLGT0/0 and M86VHVLGT⫹
were surface stained with Abs specific for IgMa (RS3.1) and IgMb (MB86) and then analyzed by flow cytometry. Shown is sIgMa and sIgMb expression
within the lymphoid gate. C, Analysis of ␣Gal-binding by sIgMa⫹ and sIgMb⫹ B cells in M86VHVLGT0 (top) and M86VHVLGT⫹ (bottom) mice. D,
Analysis of ␣Gal-binding in the peritoneum of M86VHVLGT0 mice. Shown is staining of peritoneal exudate cells (PeC) for expression of B220 and CD11b
(top left). B220⫹,CD11b⫹ cells were also analyzed for expression of CD5 (bottom left) and ␣Gal-binding by B220⫹, CD11b⫹, CD5⫹ B1a B cells (top right)
and B220⫹, CD11b⫹, CD5⫺ B1b B cells (bottom right) examined. E, ELISPOT analysis of ␣Gal-specific IgM production in naive M86VHVLGT0/0
following stimulation with 0.5 ␮g of LPS mice. The frequency of anti-␣Gal Ab-secreting cells per 106 total cells in each tissue is shown.
3844
Ig KNOCK-IN MICE PRODUCING ANTI-␣GAL Abs
therefore examined whether tolerance was the result of receptor
editing (45, 50 –52). To this end, IL-7 driven pre-B cells from the
bone marrow of M86VHVLGT0/0 mice were cultured on irradiated
bone marrow stromal cells from GT⫹ and GT0/0 mice and the
levels of RAG2 expression assayed by real-time PCR. Expression
of RAG2 was up-regulated in M86VHVLGT0/0 pre-B cells cultured
on stromal cells from GT⫹ cells when compared with pre-B cells
from the same mice cultured on GT0/0 stromal cells (Fig. 3A).
Analysis of RAG2 expression by quantitative real-time PCR confirmed this finding (Fig. 3B). These data suggest that pre-B cells
producing self-reactive anti-carbohydrate Abs undergo tolerance
through a mechanism that involves receptor editing.
Peripheral B cell development in M86VHVL knock-in mice
We next characterized B cell development in the periphery of
M86VHVL mice. In the spleen of M86VHVLGT0/0 and
M86VHVLGT⫹/0 mice the majority of B cells were B220⫹,
sIgMa⫹ (51.8 ⫾ 8.3% and 69.4 ⫾ 7.6%, respectively) with the
frequency of B220⫹, sIgMa⫹ B cells being slightly higher in
M86VHVLGT⫹/0 mice (Fig. 4A). B220⫹, sIgMb⫹ B cells were
detected in both types of mice (Fig. 4A). When compared with
bone marrow (Fig. 2B), the proportion of B220⫹, sIgMb⫹ B cells
observed was higher in the spleen of both M86VHVLGT0/0 and
M86VHVLGT⫹/0 mice (Fig. 4A). B cells expressing both sIgMa
and sIgMb were not detected in either M86VHVLGT0/0 or
M86VHVLGT⫹/0 mice, suggesting allelic exclusion by the
knocked in H chain (Fig. 4B). As observed in the bone marrow, B
cells capable of binding ␣Gal were detected in the spleen of
M86VHVLGT0/0 but not M86VHVLGT⫹/0 mice (Fig. 4C). Although we were unable to detect B cells that spontaneously secrete
␣Gal-specific Abs in the peritoneum of M86VHVLGT0/0 mice
(Fig. 1D), B220⫹, CD11b⫹, CD5⫺ B cells in the peritoneum were
capable of binding ␣Gal in M86VHVLGT0/0 mice (Fig. 4D). In
both the spleen and peritoneum, B cells able to bind ␣Gal were
B220⫹, sIgMa⫹. However, B cells in the peritoneum did not produce ␣Gal-specific Abs even after stimulation with LPS (Fig. 4E).
Consistent with data in Fig. 1E, B cells secreting ␣Gal-specific
Abs were only observed in the spleen, bone marrow, and lymph
nodes.
Immature B cells in the bone marrow first emigrate to the red
pulp in the spleen where they can be detected as sIgMhigh, sIgDlow,
CD21low, CD23⫺ newly formed transitional or T1 cells. This emigration to the spleen is an Ag-regulated process (reviewed in Ref.
53). T1 cells then colonize the lymphoid follicles present in the
spleen, and up-regulate IgD to become sIgMhigh, sIgDhigh, CD21int
follicular B cell precursors. These cells give rise to sIgMlow,
sIgDhigh, CD21int, CD23⫹ naive follicular B cells, and after additional Ag stimulation, sIgMhigh, sIgDhigh mature follicular B cells.
There is also a subpopulation of B cells that express high levels of
CD21 and CD1d, and are sIgMhigh, sIgDlow, CD21high, CD23⫺,
CD1dhigh marginal zone (MZ) B cells, which exist outside the
follicle (54, 55). Analysis of the spleen of M86VHVLGT0/0 mice
by cell surface staining with Abs specific for IgM, IgD, and CD21
and with FITC-labeled ␣Gal-BSA revealed that within the spleen
B cells capable of binding ␣Gal were sIgMhigh, IgDlow, CD21high
MZ B cells (60 ⫾ 22% MZ B cells are ␣Gal specific, Fig. 5A). We
were unable to detect sIgMlow, IgDhigh, CD21int follicular B cells
capable of binding ␣Gal-BSA (Fig. 5A). Analysis of tissue sections from the spleens of M86VHVLGT0/0 mice revealed that essentially all B cells capable of binding ␣Gal-BSA reside in the MZ
and were excluded from the follicle (Fig. 5B). These data suggest
that in M86VHVLGT0/0 mice ␣Gal-binding B cells are committed
to a MZ fate.
Increasing M86VHVL copy number alters the fate of
␣Gal-specific B cells when self-Ag is encountered
While breeding M86VHVL knock-in mice, we also generated mice
carrying two copies of the knocked in M86VH region and either
Downloaded from http://www.jimmunol.org/ by guest on June 18, 2017
FIGURE 5. Analysis of B cell development in the spleens of
M86VHVLGT0/0 mice. A, Spleen cells
were stained with anti-IgM, -IgD,
-CD21, and ␣Gal-BSA. Based on cell
surface expression levels of IgM and
IgD, B cells within the spleen lymphoid gate were divided (far left
panel) into newly formed (NF) and
MZ fractions (fraction I), follicular
precursors (FP), and MZ precursors
(MZP) (fraction II) and follicular cells
(FO) (fraction III). These fractions
were then analyzed with respect to
CD21 expression levels and the ability to bind ␣Gal. Shown is ␣Gal-binding within the CD21 gate for each B
cell population. B, Staining of spleen
sections from M86VHVLGT⫹ (left)
and M86VHVLGT0/0 (right) mice
with anti-metallophilic macrophage
(MOMA1, blue) and ␣Gal-BSA
(green). All ␣Gal-binding cells are located in the MZ, outside of the marginal sinus. Data shown are representative experiments.
The Journal of Immunology
3845
FIGURE 6. Increasing copy number of
the knocked in VH and VL chains alters the
fate of ␣Gal-specific B cells. Analysis of
B220 expression and ␣Gal-binding by B
cells in peripheral blood of M86VHVL,
M86VH2VL, M86VHVL2, and M86VH2VL2
mice on an ␣Gal-deficient (GT0/0) or positive (GT⫹) background. The frequency of
cells within the lymphoid gates is shown.
The development of B cells producing ␣Gal-specific Ab in
M86VH2VL2GT⫹ mice may reflect lowered affinity for Ag
The ability to detect anti-␣Gal Abs in M86VH2VL2GT⫹ mice was
unexpected. To begin to characterize these Abs, we analyzed the
Production of autoreactive ␣Gal-specific Abs in
M86VH2VL2GT⫹ knock-in mice
Our data indicate that B cells producing ␣Gal-specific Abs were
efficiently tolerized in M86VHVLGT⫹; however, we consistently
observed the presence of ␣Gal-specific serum IgM Abs in
M86VH2VL2GT⫹ mice (Fig. 7A). These Abs did not bind the control Ag Lac-BSA which shares all determinants with ␣Gal-BSA
except for the terminal galactose residue, indicating the binding
observed was specific for ␣Gal. Essentially all ␣Gal-specific Abs
in these mice were of the M86VH-encoded IgMa allotype (data not
shown). The titer of anti-␣Gal Abs in these mice was consistently
lower than that observed in littermate M86VH2VLGT0/0 and
M86VH2VL2GT0/0 mice, but was similar to the titers observed in
GT0/0 control mice (Fig. 7A). Furthermore, the titers of these Abs
could be boosted by pig cell immunization (Fig. 7A). To confirm
these results, spleen cells from pig cell immunized
M86VH2VL2GT⫹/0, C57BL/6 GT⫹ controls and GT0/0 mice were
analyzed for their ability to produce ␣Gal-specific Ab in Ig
ELISPOT assays using ␣Gal-BSA- or Lac-BSA-coated plates. We
detected B cells producing ␣Gal-specific IgM Abs in the spleens of
M86VH2VL2GT⫹ mice, at a frequency similar to that observed in
immunized GT0/0 controls (Fig. 7B). Similar results were observed
in M86VH2VLGT⫹ mice (data not shown). These data suggest that
B cells producing self-reactive anti-␣Gal Abs are able to develop
in M86VH2VL2GT⫹ mice.
FIGURE 7. Production of autoreactive ␣Gal-specific Abs in
M86VH2VL2 mice. A, Production of ␣Gal-specific IgM Abs in naive (f)
and pig cell immunized (Œ) M86VH2VL2-GT⫹/⫺ mice, as well as naive
M86VHVLGT0/0 (), naive GT0/0 (⽧) and C57BL/6 (F) controls. B,
ELISPOT analysis of ␣Gal-specific Ab production in pig cell-immunized
C57BL/6 mice, and GT0/0 and M86VH2VL2GT⫹/0 mice. The frequency of
cells secreting anti-␣Gal Ab in the spleen is shown. Binding to Lac-BSA
was used as a specificity control in all assays. C, ␣Gal-specific Abs in
M86VH2VL2GT⫹ mice bind Ag at a lower affinity. Binding of serum
␣Gal-specific IgM from M86VH2VL2GT⫹ (square symbols) or
M86VHVLGT0/0 (circle symbols) mice to ELISA plates coated with 10
␮g/ml of a 15:1 (open symbols) or 10:1 (closed symbols) molar ratio of
␣Gal to BSA. Data shown are representative.
Downloaded from http://www.jimmunol.org/ by guest on June 18, 2017
one or two copies of knocked in M86VL region (M86VH2VL or
M86VH2VL2 mice) on a GT0/0 or GT⫹ background. Additionally,
we generated mice carrying two copies of the knocked in M86VL
region (M86VHVL2 mice) on a GT0/0 or GT⫹ background. In
blood, the frequency of B cells in M86VH2VL, M86VH2VL2, and
M86VHVL2 mice on the GT0/0 background was similar to the frequency observed in M86VHVLGT0/0 mice (Fig. 6). However, the
frequency of B220⫹ B cells in the periphery of M86VH2VLGT⫹
(18.7 ⫾ 2.3%) mice was significantly reduced when compared
with the frequency in M86VHVLGT⫹ mice (24.7 ⫾ 2.9%; p ⬍
0.05, Student’s t test). M86VHVL2GT⫹ and M86VH2VL2GT⫹
mice exhibited an even greater reduction in B220⫹ cells (10.1 ⫾
0.3% and 6.6 ⫾ 0.6%, p ⬍ 0.05) when compared with the
frequency in M86VHVLGT⫹ mice (Fig. 6). These data suggest that
the presence of multiple knocked in H chain or L chain alleles
affects B cell development. Interestingly, the reduction in B cell
numbers is observed only in mice that express ␣Gal as a self-Ag,
suggesting that the effect observed may be related to alterations in
negative selection.
3846
capacity of ␣Gal-specific Abs from these mice to bind BSA conjugated to different numbers of ␣Gal epitopes. We were readily
able to detect binding of ␣Gal-specific IgM from these mice when
using ELISA plates coated with ␣Gal-BSA conjugates in which
the molar ratio of ␣Gal to BSA molecules was high (15:1) (Fig.
7C). However, the ability to detect binding of ␣Gal in these mice
was significantly reduced when ELISA plates were coated with
␣Gal-BSA conjugates in which the molar ratio of ␣Gal to BSA
was lower (10:1) (Fig. 7C). Importantly, serum IgM from
M86VHVLGT0/0 (Fig. 7C) and M86VH2VL2GT0/0 (data not
shown) was able to bind to both ␣Gal-BSA preparations to a similar degree. These data suggested that ␣Gal-specific Abs produced in M86VH2VL2GT⫹ mice exhibit different binding characteristics than those of littermate mice on the GT0/0
background. We suggest that B cells making self-reactive ␣Galspecific Abs in M86VH2VL2GT⫹ and M86VH2VLGT⫹ mice
(data not shown) may develop because the Ab they produce
exhibits a lowered affinity for ␣Gal.
Anti-carbohydrate natural Abs are thought to play a key role in
providing protective host immunity and have been well described
as being important in mediating transplantation rejection across
blood group disparities as well as discordant species. Yet little is
known about B cells that produce anti-carbohydrate Abs, in part
due to a lack of small animal models in which the development and
regulation of B cells producing anti-carbohydrate Abs can be studied in a physiologically relevant manner. The development of
GT0/0 mice has made it possible to begin to address some of these
issues. However, the frequency of B cells producing ␣Gal specific
Abs in these mice is too low to allow for their direct analysis. We
and others have previously attempted to overcome this difficulty by
generating Ig transgenic mice that express transgenes encoding
␣Gal-specific Abs (32, 33). However, these models are limited
because the expressed transgenes were not subject to regulation by
elements within the endogenous Ig loci, making it difficult to assess physiological relevance. To overcome these issues we used
gene targeting in embryonic stem cells to construct Ig knock-in
mice that carry rearranged H chain and L chain V regions that
encode specificity for ␣Gal that are under regulatory control of the
endogenous Ig H chain and ␬ L chain loci. Breeding M86VHVL
mice to GT0/0 mice offered us the opportunity to directly examine
the development of B cells producing ␣Gal-specific Abs in the
presence or absence of ␣Gal as a self-Ag.
M86VHVLGT0 mice show high serum levels of ␣Gal-specific
Abs produced by the knocked in allele. The titer of these Abs was
sufficient to mediate rejection of H-2-matched heart transplants
from GT⫹ donors. The titer of ␣Gal-specific Abs in M86VHVL
GT0 mice could be increased by immunization. ␣Gal-specific Abs
were not detected in the serum of M86VHVLGT⫹ mice, indicating
that expression of ␣Gal in these mice prevented development of
␣Gal-specific Abs. However, we were still able to detect sIgMa⫹
B cells in the spleens of M86VHVLGT⫹ mice, suggesting that B
cells expressing the knocked in M86VH region were not deleted
during their development. Because the frequency of B cells expressing surface IgMa in the bone marrow of M86VHVLGT⫹ and
M86VHVLGT0/0 knock-in mice was similar even though we were
unable to detect B cells that bind ␣Gal in the bone marrow of
M86VHVLGT⫹ mice, we reasoned that tolerance to ␣Gal in
M86VHVLGT⫹ mice was most likely the result of receptor editing
rather than deletion. Analysis of IL-7 driven pre-B cells from the
bone marrow of M86VHVLGT0/0 mice revealed that exposure to
␣Gal as a self-Ag led to an up-regulation of RAG2 expression in
M86VHVLGT0/0 pre-B cells. These data are consistent with the
idea that pre-B cells producing self-reactive anti-carbohydrate Abs
undergo tolerance induction via receptor editing in the bone marrow. Interestingly, although receptor editing has been suggested to
lead to an increase in the frequency of immature B cells expressing
␭ light chains (37, 45, 56), we did not detect an increase in the
frequency of B cells expressing a ␭ L chain in the bone marrow or
periphery of M86VHVLGT⫹ mice (data not shown). Therefore, in
this model, receptor editing most likely occurs preferentially on the
␬ L chain. At least three distinct mechanisms that shape the naive
B cell repertoire have been described, including clonal deletion
(57– 61); anergy (62, 63); and receptor editing (45, 50, 51, 64). The
mechanism by which B cells producing Abs to self-carbohydrate
Ags are tolerized has been studied in relatively little detail. Although receptor editing has been described as a major mechanism of B cell tolerance, to our knowledge this is the first description suggesting that receptor editing is the mechanism of
tolerance for B cells producing anti-carbohydrate Abs.
Our analysis of B cell development in M86 knock-in mice
shows that anti-␣Gal B cells undergo editing at the H chain locus
even in the absence of ␣-Gal, resulting in the development of
sIgMb⫹ B cells. When comparing the frequency of sIgMb⫹ B cells
in bone marrow and spleen, it is apparent that the relative frequency of such cells is increased in spleen (see Figs. 2 and 4),
suggesting a selection of B cells expressing sIgMb⫹ in the periphery. One possibility is that the knocked in H chain does not exclude
well and that a selection process in GT⫹ mice ensures that only B
cells expressing the endogenous H chain differentiate into peripheral mature lymphocytes. However, based on the analysis shown in
Fig. 4B in which we were unable to detect B cells producing both
sIgMa and IgMb, we would suggest that apparently allelic exclusion in this system is not very leaky. Rather, there seems to be a
selection process by which B cells expressing endogenous sIgMb
are selected for in the periphery. It seems reasonable to suggest
that this selection would be important to allow for an increase in
the diversity of the Ig repertoire.
There is an unresolved issue regarding which B cell subsets
produce anti-carbohydrate Abs. Using Ig transgenic mice, it has
been suggested previously that B cells producing ␣Gal-specific
Abs are skewed to a MZ B cell phenotype (33). However, the use
of Ig transgenic mice has made it difficult to determine whether
this observation is of physiological relevance. Analysis of spontaneous ␣Gal Ab production in lymphoid tissue of M86VHVLGT0/0
mice revealed the presence ␣Gal-specific Ab production in the
spleen, bone marrow, and lymph nodes, but not the peritoneum. In
the spleen of M86VHVLGT0/0 mice, B cells capable of binding
␣Gal were sIgMhigh, IgDlow, CD21high MZ B cells. We were unable to detect sIgMlow, IgDhigh, CD21int follicular B cells capable
of binding ␣Gal-BSA. These data together with analysis of ␣Galbinding B cells in tissue sections suggest that B cells producing
␣Gal anti-carbohydrate Abs are committed to a MZ B cell fate.
Analysis of mice carrying two copies of the knocked in M86VH
region and either one or two copies of the knocked in M86VL
region revealed that copy number can significantly impact B cell
development in mice expressing ␣Gal as a self-Ag. Although the
frequency of B cells in M86VH2VL2, M86VH2VL, and M86VH
VL2 mice on the GT0/0 background was similar to the frequency in
M86VHVLGT0 mice, the frequency of B cells in M86VH2VLGT⫹
mice was significantly reduced. M86VHVL2GT⫹ and M86VH
2VL2GT⫹ mice exhibited an even greater reduction in B cells
when compared with M86VHVL animals. Interestingly, the greatest affect on B cell numbers was observed in M86VHVL2GT⫹ and
M86VH2VL2GT⫹ mice that carry two copies of the knocked in L
chain allele. Because the effect of copy number on B cell number
was seen only in mice expressing ␣Gal as a self-Ag, we suggest
Downloaded from http://www.jimmunol.org/ by guest on June 18, 2017
Discussion
Ig KNOCK-IN MICE PRODUCING ANTI-␣GAL Abs
The Journal of Immunology
Acknowledgments
We thank Uri Galili, (University of Massachusetts) for providing the M86
hybridoma, members of the Dr. Iacomini laboratory for helpful discussions, and John Kearney (University of Alabama, Birmingham) for instruction in the staining of spleen sections.
Disclosures
The authors have no financial conflict of interest.
References
1. Briles, D. E., M. Nahm, K. Schroer, J. Davie, P. Baker, J. Kearney, and
R. Barletta. 1981. Antiphosphocholine antibodies found in normal mouse serum
are protective against intravenous infection with type 3 Streptococcus pneumoniae. J. Exp. Med. 153: 694 –705.
2. Casali, P., and E. W. Schettino. 1996. Structure and function of natural antibodies. Curr. Top. Microbiol. Immunol. 210: 167–179.
3. Haury, M., A. Sundblad, A. Grandien, C. Barreau, A. Coutinho, and A. Nobrega.
1997. The repertoire of serum IgM in normal mice is largely independent of
external antigenic contact. Eur. J. Immunol. 27: 1557–1563.
4. Cooper, D. K., E. Koren, and R. Oriol. 1994. Oligosaccharides and discordant
xenotransplantation. Immunol. Rev. 141: 31–58.
5. Galili, U. 1993. Evolution and pathophysiology of the human natural anti-␣galactosyl IgG (anti-Gal) antibody. Springer Semin. Immunopathol. 15: 155–171.
6. Galili, U., F. Anaraki, A. Thall, C. Hill-Black, and M. Radic. 1993. One percent
of human circulating B lymphocytes are capable of producing the natural anti-Gal
antibody. Blood 82: 2485–2493.
7. Good, A. H., D. K. Cooper, A. J. Malcolm, R. M. Ippolito, E. Koren,
F. A. Neethling, Y. Ye, N. Zuhdi, and L. R. Lamontagne. 1992. Identification of
carbohydrate structures that bind human antiporcine antibodies: implications for
discordant xenografting in humans. Transplant. Proc. 24: 559 –562.
8. Oriol, R., Y. Ye, E. Koren, and D. K. Cooper. 1993. Carbohydrate antigens of pig
tissues reacting with human natural antibodies as potential targets for hyperacute
vascular rejection in pig-to-man organ xenotransplantation. Transplantation 56:
1433–1442.
9. Sandrin, M. S., and I. F. McKenzie. 1994. Gal␣(1,3)Gal, the major xenoantigen(s) recognised in pigs by human natural antibodies. Immunol. Rev. 141:
169 –190.
10. Sandrin, M. S., H. A. Vaughan, P. L. Dabkowski, and I. F. McKenzie. 1993.
Anti-pig IgM antibodies in human serum react predominantly with Gal(␣1–3)Gal
epitopes. Proc. Natl. Acad. Sci. USA 90: 11391–11395.
11. Galili, U., R. E. Mandrell, R. M. Hamadeh, S. B. Shohet, and J. M. Griffiss. 1988.
Interaction between human natural anti-␣-galactosyl immunoglobulin G and bacteria of the human flora. Infect. Immun. 56: 1730 –1737.
12. Galili, U., S. B. Shohet, E. Kobrin, C. L. Stults, and B. A. Macher. 1988. Man,
apes, and Old World monkeys differ from other mammals in the expression of
␣-galactosyl epitopes on nucleated cells. J. Biol. Chem. 263: 17755–17762.
13. McMorrow, I. M., C. A. Comrack, P. P. Nazarey, D. H. Sachs, and
H. DerSimonian. 1997. Relationship between ABO blood group and levels of
Gal␣,3galactose-reactive human immunoglobulin G. Transplantation 64:
546 –549.
14. Wang, L., M. Z. Radic, and U. Galili. 1995. Human anti-Gal heavy chain genes:
preferential use of VH3 and the presence of somatic mutations. J. Immunol. 155:
1276 –1285.
15. Hamadeh, R. M., G. A. Jarvis, U. Galili, R. E. Mandrell, P. Zhou, and
J. M. Griffiss. 1992. Human natural anti-Gal IgG regulates alternative complement pathway activation on bacterial surfaces. J. Clin. Invest. 89: 1223–1235.
16. Takeuchi, Y., C. D. Porter, K. M. Strahan, A. F. Preece, K. Gustafsson,
F. L. Cosset, R. A. Weiss, and M. K. Collins. 1996. Sensitization of cells and
retroviruses to human serum by (␣1–3) galactosyltransferase. Nature 379: 85– 88.
17. Welsh, R. M., C. L. O’Donnell, D. J. Reed, and R. P. Rother. 1998. Evaluation
of the Gal␣1–3Gal epitope as a host modification factor eliciting natural humoral
immunity to enveloped viruses. J. Virol. 72: 4650 – 4656.
18. Bach, F. H., S. C. Robson, H. Winkler, C. Ferran, K. M. Stuhlmeier,
C. J. Wrighton, and W. W. Hancock. 1995. Barriers to xenotransplantation. Nat.
Med. 1: 869 – 873.
19. Galili, U., D. C. LaTemple, A. W. Walgenbach, and K. R. Stone. 1997. Porcine
and bovine cartilage transplants in cynomolgus monkey: II. Changes in anti-Gal
response during chronic rejection. Transplantation 63: 646 – 651.
20. Lin, S. S., D. L. Kooyman, L. J. Daniels, C. W. Daggett, W. Parker, J. H. Lawson,
C. W. Hoopes, C. Gullotto, L. Li, P. Birch, et al. 1997. The role of natural
anti-Gal␣1–3Gal antibodies in hyperacute rejection of pig-to-baboon cardiac xenotransplants. Transpl. Immunol. 5: 212–218.
21. Platt, J. L. 1997. Approaching the clinical application of xenotransplantation.
Am. J. Med. Sci. 313: 315–321.
22. Stone, K. R., A. W. Walgenbach, J. T. Abrams, J. Nelson, N. Gillett, and
U. Galili. 1997. Porcine and bovine cartilage transplants in cynomolgus monkey:
I. A model for chronic xenograft rejection. Transplantation 63: 640 – 645.
23. Tearle, R. G., M. J. Tange, Z. L. Zannettino, M. Katerelos, T. A. Shinkel,
B. J. Van Denderen, A. J. Lonie, I. Lyons, M. B. Nottle, T. Cox, et al. 1996. The
␣-1,3-galactosyltransferase knockout mouse: implications for xenotransplantation. Transplantation 61: 13–19.
24. Thall, A. D., P. Maly, and J. B. Lowe. 1995. Oocyte Gal␣1,3Gal epitopes implicated in sperm adhesion to the zona pellucida glycoprotein ZP3 are not required for fertilization in the mouse. J. Biol. Chem. 270: 21437–21440.
25. Thall, A. D., H. S. Murphy, and J. B. Lowe. 1996. ␣1,3-Galactosyltransferasedeficient mice produce naturally occurring cytotoxic anti-Gal antibodies. Transplant. Proc. 28: 556 –557.
26. Chen, Z. C., and U. Galili. 2000. Genes coding for anti-Gal in knock-out mice for
the ␣1,3galactosyltransferase gene: analysis by hybridomas. Transplant. Proc.
32: 846 – 847.
27. Chen, Z. C., M. Z. Radic, and U. Galili. 2000. Genes coding evolutionary novel
anti-carbohydrate antibodies: studies on anti-Gal production in ␣1,3galactosyltransferase knock out mice. Mol. Immunol. 37: 455– 466.
28. McKenzie, I. F., Y. Q. Li, K. Patton, A. D. Thall, and M. S. Sandrin. 1998. A
murine model of antibody-mediated hyperacute rejection by galactose-␣(1,3)galactose antibodies in Gal0/0 mice. Transplantation 66: 754 –763.
29. Nozawa, S., P. X. Xing, G. D. Wu, E. Gochi, M. Kearns-Jonker, J. Swensson,
V. A. Starnes, M. S. Sandrin, I. F. McKenzie, and D. V. Cramer. 2001. Characteristics of immunoglobulin gene usage of the xenoantibody binding to gal␣(1,3)gal target antigens in the gal knockout mouse. Transplantation 72:
147–155.
30. Pearse, M. J., E. Witort, P. Mottram, W. Han, L. Murray-Segal, M. Romanella,
E. Salvaris, T. A. Shinkel, D. J. Goodman, and A. J. d’Apice. 1998. Anti-Gal
antibody-mediated allograft rejection in ␣1,3-galactosyltransferase gene knockout mice: a model of delayed xenograft rejection. Transplantation 66: 748 –754.
Downloaded from http://www.jimmunol.org/ by guest on June 18, 2017
that the effect observed is related to alterations in negative selection and that increasing the copy number of knocked in alleles may
alter mechanisms of negative selection by restricting receptor editing, thereby skewing tolerance toward a deletional mechanism.
This would support the idea that B cells primarily use editing to
eliminate self-reactive B cells, and that negative selection via deletion is secondary to receptor editing.
Although B cells producing ␣Gal-specific Abs were efficiently
tolerized in M86VHVL mice on the GT⫹ background, we consistently observed the presence of ␣Gal-specific serum IgM Abs in
M86VH2VL2GT⫹ mice. This unanticipated finding prompted us to
examine the binding characteristics of ␣Gal specific Abs in these
animals. Interestingly, although we were readily able to detect
binding of ␣Gal-specific IgM from these mice when using ELISA
plates coated with ␣Gal-BSA conjugates in which the molar ratio
of ␣Gal to BSA molecules was high, the ability to detect binding
of ␣Gal in these mice was significantly reduced when ELISA
plates were coated with ␣Gal-BSA conjugates in which the molar
ratio of ␣Gal to BSA was relatively low. These data suggest that
␣Gal-specific Abs in M86VH2VL2GT⫹ mice have a lower affinity
for ␣Gal than that observed in GT0 mice. Such low affinity Abs
appear to require an increase in binding avidity to be detected,
which can be achieved by using BSA molecules that are highly
substituted with ␣Gal epitopes. Why then would such Abs develop
in mice that express ␣Gal? Although this issue is under study, it is
important to point out that production of ␣Gal-specific Abs are
produced only in mice containing two copies of the knocked in H
chain allele and either one or two copies of the knocked in L chain
allele. In M86 mice, the knocked in H chain allele is rearranged to
Jh4, which prevents receptor editing but not V gene replacement.
We suggest that in the context of multiple copies of the knocked in
H chain, V gene replacement via homologous recombination may
select for VH genes of similar sequences to the M86VH region,
which can encode H chain variable regions that bind ␣Gal with an
affinity that is sufficiently low to allow for their development in
GT⫹ mice. We are currently evaluating this hypothesis.
The development of M86 Ig knock-in mice provides us with a
unique model to study the development of B cells producing anticarbohydrate Abs. The initial description of these mice has provided a novel insight into mechanisms of tolerance for B cell producing anti-carbohydrate Abs. Importantly, unanticipated findings
resulting from this work may aid our understanding of self-nonself recognition. Although M86VH2VL2GT⫹ mice producing
␣Gal-specific Abs appear to be grossly healthy, the presence of
␣Gal-specific Abs in these mice may provide us with a model to
examine how autoimmunity can be precipitated and lead to
pathology.
3847
3848
47. Ohdan, H., Y. G. Yang, K. G. Swenson, A. D. Thall, and M. Sykes. 2000. In vivo
T-cell depletion enhances production of anti-GAL␣1,3GAL natural antibodies in
␣1,3-galactosyltransferase-deficient mice. Transplantation 69: 910 –913.
48. Hardy, R. R., C. E. Carmack, S. A. Shinton, J. D. Kemp, and K. Hayakawa. 1991.
Resolution and characterization of pro-B and pre-pro-B cell stages in normal
mouse bone marrow. J. Exp. Med. 173: 1213–1225.
49. Pillai, S. 1999. The chosen few? Positive selection and the generation of naive B
lymphocytes. Immunity 10: 493–502.
50. Chen, C., E. L. Prak, and M. Weigert. 1997. Editing disease-associated autoantibodies. Immunity 6: 97–105.
51. Gay, D., T. Saunders, S. Camper, and M. Weigert. 1993. Receptor editing: an
approach by autoreactive B cells to escape tolerance. J. Exp. Med. 177:
999 –1008.
52. Radic, M. Z., J. Erikson, S. Litwin, and M. Weigert. 1993. B lymphocytes may
escape tolerance by revising their antigen receptors. J. Exp. Med. 177:
1165–1173.
53. Cariappa, A., and S. Pillai. 2002. Antigen-dependent B-cell development. Curr.
Opin. Immunol. 14: 241–249.
54. Amano, M., N. Baumgarth, M. D. Dick, L. Brossay, M. Kronenberg,
L. A. Herzenberg, and S. Strober. 1998. CD1 expression defines subsets of follicular and marginal zone B cells in the spleen: ␤2-microglobulin-dependent and
independent forms. J. Immunol. 161: 1710 –1717.
55. Roark, J. H., S. H. Park, J. Jayawardena, U. Kavita, M. Shannon, and
A. Bendelac. 1998. CD1.1 expression by mouse antigen-presenting cells and
marginal zone B cells. J. Immunol. 160: 3121–3127.
56. Nussenzweig, M. C. 1998. Immune receptor editing: revise and select. Cell 95:
875– 878.
57. Chen, C., Z. Nagy, E. L. Prak, and M. Weigert. 1995. Immunoglobulin heavy
chain gene replacement: a mechanism of receptor editing. Immunity 3: 747–755.
58. Hartley, S. B., J. Crosbie, R. Brink, A. B. Kantor, A. Basten, and C. C. Goodnow.
1991. Elimination from peripheral lymphoid tissues of self-reactive B lymphocytes recognizing membrane-bound antigens. Nature 353: 765–769.
59. Nemazee, D., and K. Buerki. 1989. Clonal deletion of autoreactive B lymphocytes in bone marrow chimeras. Proc. Natl. Acad. Sci. USA 86: 8039 – 8043.
60. Nemazee, D. A., and K. Burki. 1989. Clonal deletion of B lymphocytes in a
transgenic mouse bearing anti-MHC class I antibody genes. Nature 337:
562–566.
61. Russell, D. M., Z. Dembic, G. Morahan, J. F. Miller, K. Burki, and D. Nemazee.
1991. Peripheral deletion of self-reactive B cells. Nature 354: 308 –311.
62. Fulcher, D. A., and A. Basten. 1994. Whither the anergic B-cell? Autoimmunity
19: 135–140.
63. Goodnow, C. C., J. Crosbie, S. Adelstein, T. B. Lavoie, S. J. Smith-Gill,
R. A. Brink, H. Pritchard-Briscoe, J. S. Wotherspoon, R. H. Loblay, K. Raphael,
et al. 1988. Altered immunoglobulin expression and functional silencing of selfreactive B lymphocytes in transgenic mice. Nature 334: 676 – 682.
64. Lang, J., M. Jackson, L. Teyton, A. Brunmark, K. Kane, and D. Nemazee. 1996.
B cells are exquisitely sensitive to central tolerance and receptor editing induced
by ultralow affinity, membrane-bound antigen. J. Exp. Med. 184: 1685–1697.
Downloaded from http://www.jimmunol.org/ by guest on June 18, 2017
31. Xu, H., A. Sharma, L. Chen, C. Harrison, Y. Wei, A. S. Chong, J. S. Logan, and
G. W. Byrne. 2001. The structure of anti-Gal immunoglobulin genes in naive and
stimulated Gal knockout mice. Transplantation 72: 1817–1825.
32. Cretin, N., and J. Iacomini. 2002. Immunoglobulin heavy chain transgenic mice
expressing Gal␣(1,3)Gal-reactive antibodies. Transplantation 73: 1558 –1564.
33. Xu, H., A. Sharma, Y. Lei, J. Okabe, H. Wan, A. S. Chong, J. S. Logan, and
G. W. Byrne. 2002. Development and characterization of anti-Gal B cell receptor
transgenic Gal⫺/⫺ mice. Transplantation 73: 1549 –1557.
34. Galili, U., D. C. LaTemple, and M. Z. Radic. 1998. A sensitive assay for measuring ␣-Gal epitope expression on cells by a monoclonal anti-Gal antibody.
Transplantation 65: 1129 –1132.
35. Bracy, J. L., and J. Iacomini. 2000. Induction of B-cell tolerance by retroviral
gene therapy. Blood 96: 3008 –3015.
36. Sonoda, E., Y. Pewzner-Jung, S. Schwers, S. Taki, S. Jung, D. Eilat, and
K. Rajewsky. 1997. B cell development under the condition of allelic inclusion.
Immunity 6: 225–233.
37. Pelanda, R., S. Schaal, R. M. Torres, and K. Rajewsky. 1996. A prematurely
expressed Ig␬ transgene, but not V␬J␬ gene segment targeted into the Ig␬ locus,
can rescue B cell development in ␭5-deficient mice. Immunity 5: 229 –239.
38. Mombaerts, P., A. R. Clarke, M. A. Rudnicki, J. Iacomini, S. Itohara,
J. J. Lafaille, L. Wang, Y. Ichikawa, R. Jaenisch, M. L. Hooper, et al. 1992.
Mutations in T-cell antigen receptor genes ␣ and ␤ block thymocyte development
at different stages. Nature 360: 225–231.
39. Wood, C., E. A. Kabat, L. A. Murphy, and I. J. Goldstein. 1979. Immunochemical studies of the combining sites of the two isolectins, A4 and B4, isolated from
Bandeiraea simplicifolia. Arch. Biochem. Biophys. 198: 1–11.
40. Schuppel, R., J. Wilke, and E. Weiler. 1987. Monoclonal anti-allotype antibody
towards BALB/c IgM. Analysis of specificity and site of a V-C crossover in
recombinant strain BALB-Igh-Va/Igh-Cb. Eur. J. Immunol. 17: 739 –741.
41. Nishikawa, S., Y. Sasaki, T. Kina, T. Amagai, and Y. Katsura. 1986. A monoclonal antibody against Igh6 – 4 determinant. Immunogenetics 23: 137–139.
42. Cretin, N., J. Bracy, K. Hanson, and J. Iacomini. 2002. The role of T cell help in
the production of antibodies specific for Gal␣1–3Gal. J. Immunol. 168:
1479 –1483.
43. Melamed, D., R. J. Benschop, J. C. Cambier, and D. Nemazee. 1998. Developmental regulation of B lymphocyte immune tolerance compartmentalizes clonal
selection from receptor selection. Cell 92: 173–182.
44. Melamed, D., and D. Nemazee. 1997. Self-antigen does not accelerate immature
B cell apoptosis, but stimulates receptor editing as a consequence of developmental arrest. Proc. Natl. Acad. Sci. USA 94: 9267–9272.
45. Tiegs, S. L., D. M. Russell, and D. Nemazee. 1993. Receptor editing in selfreactive bone marrow B cells. J. Exp. Med. 177: 1009 –1020.
46. Ohdan, H., K. G. Swenson, H. S. Kruger Gray, Y. G. Yang, Y. Xu, A. D. Thall,
and M. Sykes. 2000. Mac-1-negative B-1b phenotype of natural antibody-producing cells, including those responding to Gal␣1,3Gal epitopes in ␣1,3-galactosyltransferase-deficient mice. J. Immunol. 165: 5518 –5529.
Ig KNOCK-IN MICE PRODUCING ANTI-␣GAL Abs