HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

This Article Abstract Full Text (PDF) Alert me when this article is cited Alert me if a correction is posted Citation Map Services Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Request Permissions Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Kearns-Jonker, M. Articles by Cramer, D. V. Search for Related Content PubMed PubMed Citation Articles by Kearns-Jonker, M. Articles by Cramer, D. V.

The Human Antibody Response to Porcine Xenoantigens Is Encoded by IGHV3-11 and IGHV3-74 IgV_H Germline Progenitors¹

Transplantation Biology Research Laboratory, Department of Cardiothoracic Surgery, University of Southern California School of Medicine, Los Angeles, CA 90033

	Abstract

Top Abstract Introduction Materials and Methods Results Discussion References

 
Preformed and induced Ab responses present a major immunological barrier to the use of pig organs for human xenotransplantation. We generated IgM and IgG gene libraries established from lymphocytes of patients treated with a bioartificial liver (BAL) containing pig hepatocytes and used these libraries to identify IgV_H genes that encode human Ab responses to pig xenoantigens. Genes encoded by the V_H3 family are increased in expression in patients following BAL treatment. cDNA libraries representing the V_H3 gene family were generated, and the relative frequency of expression of genes used to encode the Ab response was determined at days 0, 10, and 21. Ig genes derived from the IGHV3-11 and IGHV3-74 germline progenitors increase in frequency post-BAL. The IGHV3-11 gene encodes 12% of V_H3 cDNA clones expressed as IgM Abs at day 0 and 32.4–39.0% of cDNA clones encoding IgM Abs in two patients at day 10. IGHV3-11 and IGHV3-74 genes encoding IgM Abs in these patients are expressed without evidence of somatic mutation. By day 21, an isotype switch occurs and IGHV3-11 IgV_H progenitors encode IgG Abs that demonstrate somatic mutation. We cloned these genes into a phagemid vector, expressed these clones as single-chain Abs, and demonstrated that the IGHV3-11 gene encodes Abs with the ability to bind to the gal (1,3) gal epitope. Our results demonstrate that the xenoantibody response in humans is encoded by IgV_H genes restricted to IGHV3-11 and IGHV3-74 germline progenitors. IgM Abs are expressed in germline configuration and IgG Abs demonstrate somatic mutations by day 21.

	Introduction

Top Abstract Introduction Materials and Methods Results Discussion References

 
The critical shortage of allogeneic organs available for transplantation into human patients has led to the serious consideration of pigs as an alternative supply of donors for solid organ transplantation. The primary barrier to the use of pigs for human xenotransplantation is a hyperacute rejection of the grafts mediated by natural Abs present in the serum of unimmunized, normal individuals. The binding of xenoreactive Abs to the endothelium of the xenograft triggers the activation of the complement system and the onset of endothelial cell injury, hemorrhage, and thrombosis (1, 2, 3, 4, 5). The importance of natural Abs in xenograft rejection is supported by the observation that large amounts of IgM are deposited on the endothelium of rejecting xenografts and that removal of xenoreactive Abs from the recipient prolongs xenograft survival (5, 6, 7, 8, 9, 10, 11).

In primates, including humans, 80% of the Abs that cause hyperacute rejection of pig grafts are specific for the carbohydrate galactose (1, 3) galactose (gal (1, 3) gal)³ (12, 13). A mutation in the (1, 3) galactosyltransferase (-gal) gene in humans, apes, and Old World monkeys results in an evolutionarily restricted expression of the gal (1, 3) gal carbohydrate epitope (14, 15, 16, 17). The absence of this carbohydrate Ag may be responsible for the appearance in normal humans of high levels of preformed -gal Abs secondary to repeated exposure to infectious agents and enteric bacteria that express the -gal epitope (18). Preexisting peripheral blood levels of Abs directed at the gal (1, 3) gal epitope in adults can be responsible for the rapid rejection of pig xenografts following organ transplantation. In addition to preformed xenoantibodies in normal individuals, exposure of patients to pig xenogeneic tissues stimulates the rapid appearance of new Abs directed at the gal (1, 3) gal epitope (19, 20). We have recently demonstrated that a 2- to 3-fold increase in IgM and IgG xenoreactive Abs can be detected in the peripheral blood of human patients at 10 days following treatment with a bioartificial liver (BAL) containing porcine hepatocytes (21, 22). By day 21, the IgM response begins to decline and is replaced by rising IgG xenoantibody levels. ELISA were used to demonstrate that the xenoreactive Abs bind to pig aortic endothelial cells and that a significant proportion of both IgM and IgG xenoantibodies react with the -gal epitope (21, 22).

The origin of xenoreactive Abs is largely unknown as there has been no direct examination of the structure and/or functional characteristics of the Ig genes responsible for encoding these Abs. We have recently hypothesized that xenoreactive Abs and natural Abs that mediate rapid immune responses to infectious agents are structurally related and are both produced by B cells independent of a requirement for T cell help (23). T cell-independent Ags generally stimulate an Ab response that is encoded by germline IgV_H genes and directed at high m.w. polysaccharide Ags expressing repetitive antigenic epitopes, such as those expressed by enteric bacteria (24, 25, 26). The structural similarity of carbohydrate xenoantigens and epitopes expressed on bacterial cell walls, as described for the -gal epitope, may result in the generation of Abs with cross-reactive specificities and common origin (18, 23, 27). This hypothesis would predict that rapid Ab responses to xenografts, like those directed at bacterial Ags, would be encoded by IgV_H genes expressed in their germline configuration. Our preliminary examination of the nature of the Ab response to xenografts in rodents is consistent with this expectation. Rat mAbs produced from splenic lymphocytes isolated from recipients of hamster heart xenografts have the ability to initiate hyperacute rejection of hamster xenografts in naive recipients following passive transfer (28). The V_H genes encoding these Abs, as well as cDNA clones encoding xenoreactive Abs in vivo for up to 3 wk following graft placement, are restricted to a group of genes expressed in germline configuration (23, 29). Similarly, rat mAbs that specifically react with pig aortic endothelium are encoded by V_H genes that exhibit limited evidence of somatic mutation (30). However, the maturation of the humoral xenograft response at 3 wk posttransplantation is associated with the onset of somatic mutation in a proportion of the IgV_H genes encoding IgG xenoantibodies in rats rejecting hamster heart xenografts.

The nature of the IgV_H genes that initiate the humoral response mounted by patients (or other primate species) to porcine tissues is less well characterized. As described above, preformed natural xenoantibodies are primarily, although not exclusively, directed at -gal epitopes expressed by pig endothelial cells. Isolation and characterization of Abs from EBV-transformed lymphocytes from normal individuals has demonstrated the use of IgV_H genes in both germline and nongermline configurations to encode -gal Abs (27). However, the existence of V_H genes in either configuration in unstimulated individuals provides limited information on the nature of the host responses to xenografts because the route of the exposure to xenoantigens has the potential to influence the specific immune response pathway used by the host. The repertoire of human IgG and IgM anti-gal (1, 3) gal Abs is polymorphic and includes Abs that vary in specificity for -gal presented as di-, tri-, or pentasaccharide conjugates (31). High-affinity anti-gal (1, 3) gal Abs are included in this group (32). Therefore, we have chosen to examine the nature of the host Ab response following the systemic exposure of patients to pig hepatocytes and compare the results of these studies to our data derived from studies conducted in rodents. Our results indicate that, despite widely disparate donor/recipient species combinations, the humoral response of patients to pig tissues shares many important characteristics with the rodent xenograft response, including the use of a small group of closely related V_H genes to encode the Ab response to the graft.

	Materials and Methods

Top Abstract Introduction Materials and Methods Results Discussion References

 
RNA isolation

Total RNA was extracted from the PBL obtained from the same patients in which a strong IgM and IgG xenoantibody response to pig endothelium and to the -gal epitope was detected using ELISA and immunoprecipitation techniques at days 0, 10, and 21 following exposure to pig hepatocytes (21, 22). We used the same time points and the same patient samples to isolate RNA needed to clone the genes encoding this xenoantibody response. The RNA extraction was performed using a solution containing 4 M GuSCN (guanidine thiocyanate), 1.5 M sodium citrate, pH 7.0, and 0.5% sarcosyl (33). RNA was subjected to a phenol/chloroform/isoamyl alcohol extraction, precipitated with isopropanol, then resuspended in 1 mM EDTA, pH 8.0, and precipitated with 0.3 M sodium acetate and ethanol.

Anchored PCR-ELISA for analysis of the IgV_H gene repertoire

The anchored PCR-ELISA was used to establish the relative distribution of IgV_H gene subgroups in patients exposed to pig hepatocytes. Total RNA isolated from PBL derived from patients before and following exposure to BAL was used to prepare first-strand cDNA using µ-chain-specific primers (Table I), and the cDNA/RNA hybrids were hydrolyzed with NaOH. The first-strand cDNA was poly(dG)-tailed with dGTP and TdT, purified and subjected to an anchored PCR amplification in 10 mM Tris-HCl, pH 8.3, 50 mM KCl, 1.5 mM MgCl_2, 400 µM dNTPs, and 2.5 U of Taq DNA polymerase (Boehringer Mannheim, Indianapolis, IN) (34). The primers used for analysis of the IgV_H genes used by IgM-expressing B cells were Cµ antisense oligonucleotide 1 and a 9:1 mixture of anchor primers 1 and 2 (see Table I for primer sequences). The amplification conditions were one cycle at 94°C for 3 min, then 25 cycles at 94°C for 30 s, 48°C for 30 s, and 72°C for 1 min, followed by a 7-min extension at 72°C. The purified PCR products were used in a nested PCR with an anchor primer and a 5'-biotinylated Cµ antisense primer 2 corresponding to a sequence upstream of the initial Cµ primer (Table I). The nested PCR conditions included 25 cycles of annealing at 55°C. The PCR product obtained from various patient samples was quantitated using the program ZeroDscan, v.1 (Scanalytics, Billerica, MA) and distributed at a concentration of 1 µg/well into ELISA plates for analysis of relative changes in the distribution of IgV_H gene subgroups among IgM-expressing B cells in patients exposed to porcine hepatocytes. Oligonucleotides corresponding to the sense strand of each of the leader sequences from six IgV_H subgroups (Table I) were labeled with digoxigenin. The purified and quantitated PCR product was distributed into streptavidin-coated microtiter plates (Bio-Rad, Hercules, CA) that were preincubated with 1% BSA for 1 h at room temperature. The PCR product was incubated in these plates for 1 h, washed, and 0.1 N NaOH was added to denature the DNA. The plates were washed, and digoxigenin-labeled sense strand oligonucleotide was added to each well and incubated at 65°C for 20 min, then 42°C for 90 min. The plates were washed, incubated with alkaline phosphatase-conjugated anti-digoxigenin Ab (Boehringer Mannheim) in blocking buffer for 30 min at room temperature, then incubated with Attophos substrate (JBL Scientific, San Luis Obispo, CA). Excitation and emission was measured at 450 nm and 580 nm.

A nested PCR amplification was performed to amplify the IgV_H genes encoding IgM and IgG Abs that are members of the V_H3 family. For genes encoding IgM Abs, Ig gene products amplified in the first PCR of the anchored PCR-ELISA were subjected to a nested reaction using Cµ primer 3 and a V_H3 gene family-specific primer containing an SfiI restriction site at the 5' end (see Table I). The PCR was performed in a Perkin-Elmer 9600 GeneAmp PCR system thermocycler (Norwalk, CT) for 1 cycle of 94°C for 5 min, 30 cycles of 94°C for 45 s, 56°C for 60 s, 72°C for 90 s, and one cycle of 72°C for 7 min. IgG cDNA libraries were prepared following amplification using IgG primer 1 (35), and a 9:1 mix of AN and ANC anchor primers for 1 cycle of 94°C for 5 min, 30 cycles of 94°C for 45 s, 48°C for 60 s, 72°C for 90 s, and one cycle of 72°C for 7 min. Two libraries representing smaller groups of genes within the V_H3 family were prepared following a nested PCR amplification using C or Cµ primers and upstream primers V_H3 primer 2 to amplify 20 V_H3 family genes and a V3-74 primer for genes related to the IGHV3-74 germline gene. The PCR conditions for this amplification were denaturation at 94°C for 30 s, annealing at 54°C for 30 s, and extension at 72°C for 60 s for a total of 35 cycles. The PCR products were verified by size on a 1.4% agarose gel and cloned into the pCR 2.1 vector (Invitrogen, San Diego, CA). The cloned DNA was transformed into Escherichia coli, and recombinant colonies were identified for plasmid preparation on the basis of color screening. PCR products were ligated into the pCR 2.1 vector (Invitrogen) or the pT7 blue T-vector (Novagen, Madison, WI) and transformed into INVF' cells or NovaBlue competent cells, respectively, according to the specifications of the manufacturers. We initially selected cDNA clones isolated from each of the cDNA libraries and sequenced 5–10 of these genes before the analysis of the frequency of expression of IgV_H genes encoded by specific germline progenitors using the colony filter hybridization technique in a larger sample size.

Screening of the cDNA library

Recombinant colonies were identified by color screening on Luria-Bertani ampicillin or carbenicillin/tetracycline plates containing x gal. White colonies were replated and transferred onto nylon membranes (Boehringer Mannheim) for colony filter hybridization. Filters were denatured and neutralized before cross-linking the DNA with UV light using a Stratalinker (Stratagene, La Jolla, CA). Filters were treated with 2 mg of proteinase K in solution for 1 h at 37°C before hybridization. Oligonucleotides were labeled with digoxigenin in a solution containing 100 pmol of the oligonucleotide, 200 µM potassium cacodylate, 25 mM Tris-HCl, pH 6.6, 250 µg/ml BSA, 5 mM CaCl₂, 50 µM digoxigenin-dideoxy UTP, and 25 U of TdT (Boehringer Mannheim) for 20 min at 37°C. The oligonucleotides used for screening the cDNA libraries (see Table II) were selected to identify changes in the relative frequency of cDNA clones encoding the following V_H genes: 1) cDNA clones reported to encode Abs reactive with the gal (1, 3) gal epitope in normal individuals (DP58 identified with oligo 583IC (5'-TAGTTATGAAATGAACT-3') (36), V3-7 identified with oligo 543IC (5'-AACATAAAGCAAGATGGA-3') (36), and V3-74 and human IGHV3-74 alleles identified with oligos 193WS (5'-AGTAGCACAAGCTACGCGG-3'), 193GS (5'-AGTAGCACAACGTACGCGG-3'), and 193 (5'-AGTACTACAAACTATGCGG-3'); 2) cDNA clones encoded by the IGHV3-11 germline progenitor were identified with oligo RVH11 (5'-TCACTTTCAGTGACTACTACATGAGCTGGA-3') designed to be specific for all functional alleles of the IGHV3-11 germline gene. The oligonucleotides used to screen for IgV_H genes encoded by the IGHV3-11 and DP58 germline progenitors are specific for the complementarity-determining region (CDR)-1 to distinguish between these two genes, as they are identical in CDR2; 3) the RVH10 oligo (5'-AGAGTACCTGAGTAGTTTGGATGCTTTTGATATCGGCTA-3') is specific for CDR3 of an expressed IGHV3-11 gene; 4) cDNA clones were screened with the oligonucleotide 29IC (5'-TTGGCCGTACTAGAAACAA-3') (36) to determine whether V_H genes with a 1–4 canonical structure demonstrate a relative increase in frequency post-BAL; and 5) the majority of the remaining V_H3 genes not identified using the primers designated above were detected with the oligonucleotide RVH20 (5'-GGATTCACCTTTAGTAGCTAT-3'). The RVH20 oligonucleotide is a primer that will identify relative changes in the expression of 10 of 17 additional V_H3 genes not identified with oligonucleotides 543IC, 583IC, 193WS, RVH11, or 29IC. These primers allowed us to screen for relative changes in the expression of 15 of 22 functional IgV_H genes in the V_H3 family. The sequences of the germline IgV_H progenitors described in this report are named according to the conventional locus nomenclature for each segment (IMGT, the international imMunoGeneTics database, http://imgt.cnusc.fr:8104) (37) and VBase (http://www.mrc-cpe.cam.ac.uk/imt-doc/goldamino.html) (36, 38). Filters were prehybridized for 1 h at room temperature in DIG Easy Hyb buffer before adding the labeled oligonucleotide at a concentration of 7.5 pmol/ml overnight, in the dark, at room temperature with gentle agitation. Filters were washed twice for 5 min in 2x SSC, 0.1% SDS at room temperature, then twice for 15 min in 0.5x SSC, 0.1% SDS at 42°C for the RVH10, 11, 20, and 543IC probes, at 37°C for the 583IC oligonucleotide probe, and at 56°C for the 29IC probe. Optimal concentrations of oligonucleotide probes and optimal wash temperatures were determined on the basis of pilot experiments. Chemiluminescent detection was performed at room temperature according to the specifications of the manufacturer (Boehringer Mannheim). Briefly, filters were blocked and exposed to anti-digoxigenin-AP at a 1:10,000 dilution for 30 min before application of the chemiluminescent substrate for 15 min at 37°C. The filters were exposed to Lumi-Film (Boehringer Mannheim) and developed after 2 h. Positive colonies representing IgV_H genes were counted to determine the relative levels of IgV_H genes encoded by specific germline progenitors in patients exposed to porcine hepatocytes during BAL perfusion. Ig genes that had increased in frequency of expression at days 10 and 21 post-BAL when compared with day 0 using these oligonucleotide primers were selected for sequencing. The oligonucleotides and hybridization conditions were carefully selected to identify both mutated and nonmutated IgV_H genes. The relative frequency of mutations in particular amino acids located in the CDR1 and CDR2 of the IgV_H gene have been published (39). We have used this information and a careful analysis of genes identified under an array of hybridization and wash conditions in the selection of oligonucleotides that identify IgV_H genes encoded by a single germline progenitor and expressed both with and without somatic mutation. The genes isolated using the CDR3-specific probe would not be affected in any way by mutations in the CDR1 and CDR2.

cDNA clones were sequenced using the ALFexpress automated DNA sequencer and the Autoread and Autocycle sequencing kits using T7 and Taq polymerase, respectively (Pharmacia Biotech, Alameda, CA). The cDNA clones were sequenced in both directions using M13 universal and/or -40 primers and reverse primers provided in the sequencing kit.

Analysis of Ig V gene mutations

The number of expected mutations in the IgV_H CDR and framework regions (FR) was calculated using the formula R = n x CDR R_f or FR R_f x CDR_rel or FR_rel in which n is the total number of observed mutations, R_f is the replacement frequency inherent to CDR or FR sequences, and CDR_rel and FR_rel are the relative size of the CDR or FR (40, 41, 42). The determination of whether the excess of R mutations in CDRs or their scarcity in FRs was due to chance alone was calculated using the binomial probability model p = (n!/[k!/n - k!] x q^k x (1 - q^n-k ) in which q is the probability that a R mutation will be located in the CDR or FR (q = CDR _rel x CDR R_f or FR_rel x FR R_f), and k is the number of observed R mutations in the CDR or FR (40, 41, 42).

Expression of the IGHV3-11 cDNA clone in a phagemid vector

One of the genes (IGHV3-11) displaying an increased frequency of expression in two patients exposed to pig cells was cloned into a phagemid vector (pHEN2; Center for Protein Engineering, Medical Research Council Center, Cambridge, U.K.) using an overlap extension PCR technique. The phagemid vector was used to express this cDNA clone as a single-chain Ab to determine whether the Ab encoded by this gene has the ability to react with the -gal epitope. The IGHV3-11 cDNA clones were amplified in a PCR using the sense primer (5'-GTCCTCGCAACTGCGGCCCAGCCGGCCATGGCCCAGGTGCAGCTGGTGGAGTCTGG-3') and the antisense primer (5'-TCGACCTCGAGTTGAAGAGACGGTGACCATTGTCCCTTGGCCCCAGATATCAAAAGCATCCAAACTACTCAGGTACTCTCGC-3'), gel purified and ligated into the pHEN2 vector, which had been restricted with the enzymes SfiI and XhoI. The ligation was transformed into competent bacteria and screened using a V_H3 primer and J_H3 primer to check for the insert. Phagemid containing the cloned genes were grown in 2XTY containing 100 µg/ml ampicillin and 1% glucose and were infected with VCSM13 helper phage (Strategene) at a ratio of 1:20 (number of bacterial cells:helper phage). The infected cells were then grown in 2XTY containing 100 µg/ml ampicillin and 25 µg/ml kanamycin overnight at 30°C. The phage were precipitated with polyethylene glycol 6000/NaCl (20% polyethylene glycol, 2.5 M NaCl), titered, and resuspended in PBS to 10¹³ transducing U/ml. Nucleic acid sequencing was used to confirm that no nucleic acid substitutions were introduced into the cDNA clones during the PCR that was used to modify the ends of the clones for compatibility with the phagemid vector. Then, 10⁹–10¹² phage were used in ELISA to assess binding of the single-chain Ab encoded by this cDNA clone to bovine thyroglobulin (Sigma, St. Louis, MO) and mouse laminin (expressing 50–70 -gal epitopes per molecule) (Sigma) (43). Previous experiments have demonstrated that all serum Ig molecules that bind to mouse laminin in ELISA display specificity for the -gal epitope (43, 44).

Screening phage particles by ELISA

The binding of phage particles expressing the IGHV3-11 germline gene as a single-chain Ab to mouse laminin and bovine thyroglobulin was addressed using Falcon Microtest III flexible assay plates (Becton Dickinson, San Jose, CA) coated with 50 µl/well of protein Ag at a concentration of 20 µg/well in PBS. The plates were coated overnight at room temperature, rinsed, blocked with 1% BSA for 2 h at 37°C, and incubated with 10⁹–10¹² phage particles for 90 min at room temperature, washed and incubated with HRP-anti-M13 monoclonal conjugate (Pharmacia Biotech, Piscataway, NJ) at a concentration of 1:5000 for 30 min at room temperature. The reaction was developed with substrate solution (100 µl/well tetramethylbenzidine (Kirkegaard & Perry Laboratories, Gaithersburg, MD)) in 100 mM sodium acetate, pH 6.0. Then, 50 µl of 0.18 M sulfuric acid was used to stop the reaction, and the OD was read at 650 nm and 450 nm using a Molecular Devices microplate reader (Menlo Park, CA).

	Results

Top Abstract Introduction Materials and Methods Results Discussion References

 
Patients exposed to porcine hepatocytes use genes in the V_H3 family to respond to Ag exposure

The 100 germline IgV_H genes in the human haploid genome can be divided into seven IgV_H gene subgroups based on similarities in their primary structure (45, 46, 47, 48). Analysis of the IgV_H genes expressed in the normal human adult B cell repertoire indicate that IgV_H gene subgroups are not used randomly. Approximately 50% of the IgV_H genes expressed in normal adults are members of the V_H3 subgroup, 17% are IgV_H4, 15% are IgV_H1, and 7% are IgV_H5 (34, 49, 50). The least well-represented groups are the IgV_H 2, 6, and 7 families, where each family constitutes <2% of the expressed V_H repertoire in normal adults. We used the anchored PCR-ELISA to determine whether an increase in Ig gene expression associated with a particular V_H gene family could be demonstrated following exposure of the patients to porcine xenoantigens. This technique has been used previously to analyze the expressed IgV_H repertoire of polyclonal and monoclonal B cell populations in patients with leukemia (51), and an analysis of the expressed repertoire of IgV_H genes in patients exposed to porcine hepatocytes offers an opportunity to examine the relative frequency of expression of individual IgV_H gene subgroups.

We initially evaluated the distribution of IgV_H gene subgroups in IgM-expressing B cells in pooled normal human peripheral blood samples. The distribution of IgV_H gene frequencies in pooled samples of normal individuals was comparable to the data reported in the literature using both conventional techniques and the anchored PCR-ELISA (34, 49). The relative levels of Ig gene expression associated with IgV_H families 1, 3, and 4 at days 0 and 10 post-BAL were then examined in B cells isolated from the peripheral blood of three patients. Two patients (patient 1 and patient 3) demonstrated a substantial increase in the production of IgM and IgG Abs that bound to pig endothelial cells and thyroglobulin (indicating reactivity with the gal (1, 3) gal epitope) in their peripheral blood at days 0, 10, and 21 following BAL treatment (21, 22). The results of the anchored PCR-ELISA analysis indicated that there is an increase in the relative distribution of Ig V_H genes associated with the V_H3 subgroup in cDNA libraries prepared from lymphocytes from these two patients (Fig. 1). The IgV_H genes expressed were amplified independent of any genetic polymorphisms or somatic mutation because of the design of the primers. The digoxinin-labeled oligonucleotide probes corresponding to each IgV_H subgroup allowed for detection of each group using an anti-digoxinin Ab labeled with alkaline phosphatase in an ELISA. Therefore, the method does not employ nonlinear detection methods, allowing for a more direct comparative analysis of expressed IgV_H genes. A similar analysis was conducted in a patient (patient 8) that did not mount an Ab response to a single BAL procedure. The anchored PCR-ELISA analysis of IgV_H gene expression in this patient did not display an increased frequency of use of any of the V_H gene families.

View larger version (26K): [in this window] [in a new window]   FIGURE 1. PCR-ELISA demonstrating relative levels of Ig gene expression associated with the VH1, VH3, and VH4 families in pooled normal human lymphocyte samples (A) and one patient (patient 3) that mounted an immune response following exposure to pig cells (B). The relative level of Ig gene expression associated with the VH1, VH3, and VH4 families in this normal human pooled serum sample remained relatively stable in two independent experiments whereas the expression of Ig genes related to the VH3 family rises following exposure of this patient to porcine cells. The relative levels of IgVH gene expression are indicated in fluorescent units.

As described above, the anchored PCR-ELISA indicated that there was an expansion in the use of V_H3 family genes in the two patients producing elevated levels of xenoantibodies following BAL perfusion. We prepared cDNA libraries to clone V_H3 family genes from these two patients at days 0 (before BAL treatment) and days 10 and 21 (following BAL exposure) using linker-mediated PCR. The primers used to clone IgV_H genes expressed in these patients were specifically selected to include all V_H3 genes previously reported to encode -gal Abs in the circulation of normal humans, as well as 20 of 22 functional genes in the V_H3 family (27, 36). Our cDNA libraries were produced from PBL of the same patients that demonstrated high levels of IgM and IgG xenoantibodies at days 10 and 21, as described in two previously published manuscripts from our laboratory (21, 22). We used primers specific for the constant region of either IgM or IgG genes on the same days (10 and 21) sampled for xenoantibody analysis.

The cDNA clones obtained from patient samples at days 0, 10, and 21 were screened by colony filter hybridization using oligonucleotide probes specific for individual IgV_H genes to determine the relative percentage of specific IgV_H genes before and following exposure to pig cells. A sample size of at least 100 cDNA clones per group was examined. We initially tested pooled samples from normal individuals to establish the optimal experimental conditions for each oligonucleotide probe used in the colony filter hybridization analysis. The frequency of expression associated with each IgV_H gene in normal individuals was comparable to established frequencies reported in the literature, as were samples from each patient at the day 0 time point (51, 52, 53). cDNA clones from selected individual colonies that were positive in the colony filter analysis were sequenced to confirm the specificity of the hybridization results. The data obtained from the patient samples indicated that Ig genes encoded by the IGHV3-11 and IGHV3-74 (COS 6) germline progenitors were increased in frequency in cDNA libraries representing IgV_H genes encoding IgM Abs at day 10 in these cDNA libraries. cDNA clones related to the IGHV3-11 germline gene represented 12–14% of V_H3 gene expression at day 0 and rose to 32.4–39.0% at day 10 in two patients following exposure to pig cells. cDNA clones related to the IGHV3-74 germline progenitor represented 2–4% of clones isolated at day 0 and rose to 40–49.4% at day 10. The relative frequency of Ig genes previously reported to encode -gal Abs in normal individuals (27), the IGHV3-7 and DP58 germline progenitors, and a somatically mutated IgV_H gene encoded by the IGHV3-74 germline progenitor (identified with oligonucleotide probe 193), remained unchanged (Table III). The frequency of expression associated with the IGHV3-72 germline gene (a gene representing the unrelated 1–4 canonical binding group) also remained unchanged.

To eliminate the possibility that the xenoantibody response could be a polyclonal activation of genes encoded by several germline progenitors, we screened the Ig V_H3 gene libraries at days 0 and 10 with an oligonucleotide probe designed to hybridize with 10 of 17 additional germline genes within the V_H3 library that would not be detected with individual oligonucleotide probes as used in these experiments (RVH20, see Materials and Methods). This primer displayed at least 95% identity with 320 additional rearranged V_H genes present in the human database. No increase in the frequency of V_H gene expression, other than the IGHV3-11 and IGHV3-74 genes, was detected following exposure to pig cells. Additionally, a primer that identifies a germline gene encoding Abs with a different canonical structure (oligo 29IC, canonical structure 1–4; IGHV3-72 germline progenitor) was used for the same colony filter hybridization. This gene also displayed no increase in expression post-BAL (0.9% at day 10). These results indicate that the xenoantibody response is specific and that it is restricted to two IgV_H genes (IGHV3-11 and IGHV3-74), both exhibiting a 1–3 canonical structural group configuration.

V_H genes encoding IgM xenoantibodies are expressed in germline configuration

A comparison of the V_H gene nucleic acid sequences identified in cDNA clones expressed at day 10 with their closest identifiable germline counterpart was conducted to establish whether these Ig genes exhibit any evidence of somatic mutation. We have previously shown that V_H genes encoding rat anti-hamster and rat anti-pig aortic endothelial cell xenograft responses are expressed in their original germline configuration (29, 30, 54). A comparison of the nucleic acid sequence of the V3-11 and V3-74 genes with their germline counterparts indicates that the IgM response to pig xenoantigens in patients is encoded by IgV_H genes that are 98.1–99.3% and 97.6–100% identical with their germline progenitors, respectively (Table IV). The base pair substitutions observed following comparison of the cDNA clones for the V3-11 genes with their corresponding germline progenitors did not include more than one replacement and/or silent change(s) in the CDR and 1–2 replacement substitutions in the V_H FR (Table IV, Fig. 2). A total of 63 cDNA clones were sequenced in these experiments. Most cDNA clones derived from the IGHV3-74 germline gene and isolated at day 10 (20 clones sequenced) bore neither replacement nor silent substitutions in the CDR and/or FR (Fig. 3). The high level of nucleic acid similarity for the rearranged V_H genes and their germline progenitors and the absence of preferential accumulation of mutations within the CDR is consistent with the use of these genes in their original germline configuration. The ratio of replacement over silent changes (R/S) in the V_H gene are considered to indicate Ag-driven selection when the ratio is >2.9 (55, 56). The R/S ratios for all V_H genes sequenced in patients exposed to pig cells were consistently <2.9 in both CDR and FR (see Table IV). The probability that replacement mutations in the CDR of the V_H gene occurred randomly was calculated using the binomial distribution model of Shlomchik et al. (57). In the absence of a positive or negative selection pressure on a gene product, a random distribution of R and S mutations would be expected to occur throughout the protein sequence. If the number of R mutations is lower than expected by chance only, it is likely that selective pressure to maintain the structure of the Ab occurred. If the number of R mutations is higher than expected by chance alone, a positive pressure to mutate is likely to have occurred. Because the predominant role in Ag binding is believed to be played by the V_H segment of the Ab, the probability that R mutations in the V_H gene arose by chance was calculated using the binomial distribution model p = [n!/k! (n - k)!] q^k (1 - q)^n-k where q = R_f x CDR_f is the probability that a R mutation will occur in the CDR (q = 0.22 x 0.75) and k = number of observed mutations in the CDR. Statistically, the limited number of mutations in the CDR of cDNA clones isolated from patients at day 10 were consistent with their occurrence by chance alone. The results of this analysis, presented in Table IV, provide support for the concept that cDNA clones encoding IgM xenoantibodies are expressed without evidence of the accumulation of selective mutations within the CDR.

View larger version (42K): [in this window] [in a new window]   FIGURE 2. A, Nucleotide sequences of VH genes related to the IGHV3-11(DP35) germline progenitor in patients at day 0 and day 10 following exposure to porcine hepatocytes. Stars indicate identities in nucleic acid sequence. Nomenclature of the germline genes, alignments, and numbering is based on the list in the Human IGHV germline gene table at http://www.mrc-cpe.cam.ac.uk/imt-doc/goldamino.html. B, Translated amino acid sequences of VH genes isolated from patients at days 0 (clone 642) and day 10 (clones 103, 793, 802) following exposure to porcine hepatocytes. Shown for comparison are the translated sequences of the IGHV3-11 germline gene. C, CDR3 of the clonally expanded IGHV3-11 gene. Data are listed with the GenBank database under accession numbers AF119805, AF119800, AF119801, AF119802, and AF119803.

View larger version (46K): [in this window] [in a new window]   FIGURE 3. Nucleotide sequences of VH genes related to the IGHV3-74 (COS 6) germline progenitor in patients at day 10 (clones 1385, 1401, 1362) following exposure to porcine hepatocytes (A). Predicted amino acid sequences for the cDNA clones isolated from patients responding to porcine cells are indicated (B). Data are listed with the GenBank database under accession numbers AF119799, AF119798, and AF119797.

Colony filter hybridization using a CDR3-specific probe followed by nucleic acid sequencing was used to identify the clonal expansion of a specific V_H gene at day 10 post-BAL (1% at day 0 to 25.6% at day 10). A comparison of the nucleotide sequence of 19 Ig genes encoded by the IGHV3-11 germline progenitor indicates that the majority of clones expanded at day 10 were entirely identical for their VDJ sequences to those identified on day 0. The two amino acid substitutions identified in the sequence of these genes (located in positions 58 and 67) at day 10 post-BAL could be detected before BAL exposure. These amino acid changes may have influenced the expansion of this Ab if they induce a higher-affinity interaction with porcine xenoantigens. One of these substitutions occurs in the CDR (Fig. 2) and may therefore play a role in the preferential expansion of individual cDNA clones in response to exposure to pig cells. We believe that the bias toward cDNA clones with a unique VDJ gene configuration at day 10 was not due to artifact for the following reasons: 1) these clones were identified from independently amplified libraries; 2) they were uniquely expressed at day 10 and not present in high levels at day 0; 3) the genes were sequenced in forward and reverse directions using both Taq and T7 polymerase to minimize the possibility of Taq-associated errors and the individual sequences were identical (Fig. 3); 4) an identical VDJ gene configuration was identified in libraries prepared using different primers (to isolate IgG clones) from these patients at day 21 post-BAL; and 5) these results were consistent when two different V_H3 family-specific sense primers were used to generate V_H3 family libraries. The nucleic acid sequences of 20 cDNA clones related to the IGHV3-74 germline gene, in contrast, demonstrated that cDNA clones with identical CDR3 could be identified at day 10 but that the majority of cDNA clones using the IGHV3-74 germline progenitor were expanded as a population of independent Abs, including many cDNA clones that displayed differences in their CDR3 (see Table V).

We then extended our analysis of the human response to pig xenoantigens to include the identification of V_H genes that encode IgG xenoantibodies in the same patients. Our previous studies have demonstrated that patients mounting an immune response following exposure to at least two BAL treatments display strong IgG responses to pig endothelium at days 10 and 21 when measured using the ELISA (21, 22). To identify the genes that encode xenoantibodies at day 21, we prepared cDNA libraries that are specific for IgG Abs encoded by the V_H3 family. The libraries were prepared from lymphocytes isolated from the peripheral blood at days 0 (before BAL treatment) and day 21 (following BAL exposure) using IgG-specific primers in a nested PCR amplification. The cDNA clones obtained from patient samples at days 0 and 21 were screened by colony filter hybridization to determine the relative percentage of specific IgV_H genes identified before and following exposure to pig cells. A sample size of at least 100 cDNA clones per group was examined. The oligonucleotide probes selected to identify IgV_H genes expressed in these patients at day 21 were identical with the probes used to analyze Ig gene expression in IgM libraries at day 10 (see Table II). Included in the analysis of the day 21 IgG cDNA clones were oligonucleotide probes designed to identify all V_H3 genes reported to encode -gal Abs in the circulation of normal humans, as well as 20 of 22 functional genes in the V_H3 family (27, 36). The controls for these experiments were pooled samples from normal individuals. The IgV_H gene expression in normal individuals and in patients was comparable at day 0 (see Fig. 4). The data obtained from the patient samples indicated that IgG Ig genes encoded by the IGHV3-11 germline progenitor were specifically increased in frequency at day 21 (Fig. 4B). cDNA clones related to the IGHV3-11 germline gene represented 2.9% of V_H3 gene expression at day 0 and rose to 20% at day 21 following exposure to pig cells. The cDNA clones related to the IGHV3-74 germline progenitor represented 0.8% of clones isolated at day 0 and 5.5% at day 21. The IGHV3-7 germline V_H gene encodes the H chain in three of nine of the IgV_H genes identified in a group of nine anti-gal-producing Abs cloned from EBV-transformed lymphocytes isolated from the peripheral blood of normal individuals (27). Therefore, we screened the cDNA libraries isolated from these patients with an oligonucleotide probe (543IC) designed to identify both germline and somatically mutated IgV_H genes derived from this specific germline progenitor. Our results demonstrated no increase in the expression of IgV_H genes encoded by the IGHV3-7 germline progenitor in these patients (0 at day 0 and 2.0% at day 21, Fig. 4). The relative frequency of genes encoded by the IGHV3-72 germline progenitor (a structurally unrelated gene in the 1–4 canonical structural group) also demonstrated no significant increase post-BAL (0- 2.0% at day 21).

View larger version (25K): [in this window] [in a new window]   FIGURE 4. IgVH gene expression in patients exposed to pig cells. The relative levels of Ig genes derived from the IGHV3-11, IGHV3-7, and IGHV3-72 germline progenitors are depicted in (A) IgM cDNA libraries prepared from normal individuals and one patient (patient 1) following exposure to pig cells in a bioartificial liver and (B) IgG cDNA libraries prepared from normal control individuals and the same patient pre- and post-BAL (days 0 and day 21). C, A clonally expanded VDJ gene configuration (identified with the CDR3-specific probe RVH10) accounts for 60–63% of the increase in expression of IgM clones sequenced at day 10 and IgG cDNA clones derived from IGHV3-11 at day 21.

To exclude the possibility that IgG xenoantibody responses may be encoded by additional germline genes, we hybridized the filters with the oligonucleotide probe (RVH20) designed to identify the majority of additional V_H3 genes (see Materials and Methods). No increase in the frequency of V_H gene expression, other than the IGHV3-11, was detected following exposure to pig cells. These results indicate that the xenoantibody response is restricted to IgV_H genes encoded by the IGHV3-11 germline progenitor.

The clonally expanded V_H gene encoded by the IGHV3-11 germline progenitor encodes xenoantibodies that react with the -gal epitope

The IgV_H gene (IGHV3-11) displaying a clonal expansion in patients exposed to pig cells was cloned into an expression vector to determine whether an Ab encoded by this IgV_H gene demonstrates reactivity with the -gal epitope, the major target of xenoreactive Abs expressed in humans that mount a xenoantibody response to pig cells. We cloned the VDJ gene (D103; Fig. 2) isolated from these patients and the V L chain DPK9 gene (expressed in germline configuration) into a phagemid vector (pHEN2). Phagemid expressing the IGHV3-11 genes were purified and screened for binding to the -gal epitope in an ELISA using bovine thyroglobulin and mouse laminin as antigenic targets. We constructed two clones producing single-chain Abs encoded by IGHV3-11 germline progenitors and sequenced these clones to demonstrate that no mutations had occurred in the process of cloning these genes into the phagemid vector. Phagemid-expressing IgV_H genes encoded by IGHV3-11 demonstrated strong reactivity for the -gal epitope expressed on both bovine thyroglobulin and mouse laminin (Fig. 5). This data demonstrates that this clonally expanded IgV_H gene encodes xenoreactive Abs with specificity for the -gal epitope.

View larger version (18K): [in this window] [in a new window]   FIGURE 5. ELISA assay demonstrating the binding of phagemid clones 71A and 12 2B (in which the IGHV3-11 gene is expressed as a single-chain Ab) to mouse laminin and bovine thyroglobulin.

A comparison of the V_H gene nucleic acid sequences identified in cDNA clones encoding IgG Abs expressed at day 21 with their closest identifiable germline counterpart was then conducted to establish whether these Ig genes exhibit any evidence of somatic mutation by 3 wk post-BAL exposure. Rat anti-hamster, rat anti-pig aortic endothelial cell, and human IgM xenoantibody responses are encoded by genes expressed in their original germline configuration. However, the switch in rodents from IgM to IgG xenoantibody responses includes genes that demonstrate evidence of somatic mutation (58). A comparison of the nucleic acid sequence of six V3-11 genes isolated from cDNA libraries specific for IgG Ig genes with their germline counterparts indicates that the IgG response to pig xenoantigens in patients at day 21 is encoded by IgV_H genes that are 87–96% similar to their closest germline progenitor (Table VI). The base pair substitutions observed following comparison of the cDNA clones for the V3-11 genes with their germline progenitor included multiple R and/or S change(s) predominantly localized to the CDR2 and V_H FR3 (Table VI, Figs. 6 and 7). The ratio of R/S in the V_H gene are considered to indicate Ag-driven selection when the ratio is >2.9 (56, 57). The R/S ratios for 66% of the V_H genes sequenced in this patient following exposure to pig cells was >2.9 in the CDR but was <2.9 for all the clones in the FR (see Table VI). The probability that R mutations in the CDR of the V_H gene occurred randomly was calculated using the binomial distribution model of Shlomichik et al. (57). Statistically, mutations in the CDR of highly mutated cDNA clones isolated from this patient at day 21 do not occur by chance alone. The results of this analysis, presented in Table VI for the six genes whose nucleic acid sequence and amino acid sequences are presented in Figs. 6 and 7, provide support for the concept that cDNA clones encoding IgG xenoantibodies reflect an Ag-driven response by 21 days following exposure to pig cells.

View larger version (55K): [in this window] [in a new window]   FIGURE 6. Nucleotide sequences of IgG VH genes related to the IGHV3-11 (DP35) germline progenitor at day 21 following exposure to porcine hepatocytes. Stars indicate identities in nucleic acid sequence. Nomenclature of the germline genes is based on the list in the human IGHV germline gene table at http://imgt.cnusc.fr.8104/textes/tables/IGH.html. Data are listed with the GenBank database under accession numbers AF151699, AF151700, AF151701, AF151702, AF151703, and AF151704.

View larger version (31K): [in this window] [in a new window]   FIGURE 7. Translated amino acid sequences of VH genes isolated from patients at day 21 following exposure to porcine hepatocytes. Shown for comparison are the translated sequences of the IGHV3-11 germline gene (A) and a comparison of IgG clone 33 to its nearest germline progenitor, DP47 (B).

	Discussion

Top Abstract Introduction Materials and Methods Results Discussion References

The types of Ags that stimulate Ab production without the need for T cell help are generally LPSs that activate B cells in a polyclonal fashion (TI-1) and repetitive polysaccharides that depend on noncognate signal 2 (TI-2). TI-2 responses occur in B cells activated by exposure to bacteria and to viruses, such as vesicular stomatitis virus, whose envelopes express Ags with a rigid structure (24). A correlation between Ag repetitiveness and the degree to which B cell activation is dependent on T cells has been clearly demonstrated (24). B cells (B1a/B1b cells) that mediate T cell-independent responses express IgV_H genes in their original, germline configuration (24, 25, 26, 59, 60). Alternatively, T cell-dependent B cell Ab responses to Ags are characterized by T cell-driven proliferation and expansion of B-2 B cells, a process that induces somatic mutations in the Ig H and L chain V regions responsible for Ag binding. These mutations allow for gradual increases in binding affinity for the Ab and improved specificity for Ag recognition. Viruses with repetitive antigenic structures initiate T cell-independent Ab responses in the early stages of the immune response and T cell-dependent responses in the later stages of the immune response (24). Therefore, Abs that initiate and maintain an immune defense may include a contribution from both T cell-independent and T cell-dependent B cell pathways.

Our laboratory has conducted an extensive series of experiments examining the structural and functional characteristics of the Ig genes encoding rodent Ab responses to xenografts (29, 54). The experiments we conducted to examine which pathway of Ab production predominates in the hamster-to-rat model involved the placement of a hamster heart graft in rats to stimulate anti-donor xenoantibody production and the use of splenic lymphocytes from graft recipients to produce rat anti-hamster mAbs. We have isolated and sequenced Ig genes that encode mAbs capable of causing the immediate rejection of hamster grafts following passive transfer to naive recipients (28, 54). Analysis of the Ig gene sequences of six independently derived Abs indicated that five of these Abs use the same germline V_H gene (V_H HAR.1). In vivo, 1% of the total IgV_H genes in newborn and naive animals are closely related to the V_H HAR gene family. A sharp increase in the expression of these genes (14% at day 10) is seen in vivo at graft rejection. The majority of the V_HHAR family genes (75%) encoding IgM xenoantibodies are expressed in a germline configuration, whereas the IgG genes encoding xenoantibodies at day 21 are expressed in both germline and nongermline configurations (29, 58). We have concluded from this work that the rejection of hamster xenografts in rats is initially the result of a T cell-independent pathway of Ab production. The early Ab response is primarily an IgM response in which a closely related group of V_H genes are used to encode Abs to a small group of target Ags expressed by the graft. As the humoral response matures with time, a switch to IgG Ab production occurs with the continued use of the V_HHAR family to encode Ab production. The IgG isotype Abs that appear at 3 wk posttransplantation express higher levels of nucleic acid sequence variation in their V_H genes, either due to somatically induced mutations or the use of as yet unidentified germline V_H progenitors. In this rodent model, the stimulus for Ab production is the hamster heart xenograft, which is left in situ following rejection. The persistence of the xenograft may lead to the simultaneous stimulation of T cell-dependent and T cell-independent pathways of Ab production.

In this report, we have compared our rodent data to Ig gene usage in patients exposed to a BAL containing pig hepatocytes and demonstrated that humans exhibit a very similar pattern of IgV_H gene usage in response to xenogeneic tissues. Patients exposed to porcine hepatocytes following BAL treatment use genes in the IgV_H3 family to mediate a humoral immune response to Ag exposure. Ig genes encoded by IGHV3-11 and IGHV3-74 germline progenitors were increased in frequency at day 10, while the relative frequency of expression for other Ig genes within the V_H3 family remained unchanged. Colony filter hybridization and nucleic acid sequencing demonstrated that the increase in the expression of genes encoded by the IGHV3-11 germline progenitors at day 10 include genes exhibiting a specific VDJ (CDR3) configuration. Somatic mutation, normally activated in the first week after antigenic stimulation, does not occur in genes encoding IgM xenoantibodies in these patients (61, 62). Single-chain Abs encoded by cDNA clones isolated from these patients bind to the -gal epitope, indicating that these IgV_H genes encode Abs with functional activity for the appropriate human anti-pig xenogeneic target Ag(s).

In the later stages of the immune response to xenografts in humans, IgG xenoantibodies are encoded by the same family of V_H3 genes as IgM xenoantibodies. However, the IgG Abs are expressed in somatically mutated configurations. These somatic mutations are localized to the CDR2 and FR3 of the gene. Secondary responses that demonstrate R/S ratios that are >2.9 in the CDR and <2.9 in the FR are generally associated with positive antigenic selection (56, 57). Four of the six IgG genes sequenced at day 21 exhibit R/S ratios that are >2.9 in the CDR, a pattern consistent with Ag-driven selection. Chang and Casali have suggested that calculation of the inherent susceptibility to amino acid replacement (R_f) for the germline progenitor sequence is prerequisite to the assessment of whether somatic point mutations are Ag selected in an IgV gene (40, 41). This number is then used to calculate the theoretically expected number of mutations in the CDR and FR. The number of expected CDR and FR R mutations in the IgG genes cloned at day 21 were calculated and used to determine, based on a binomial distribution model, the probability that excess or scarcity of R mutations in the CDR occurred by chance alone. The results of these calculations indicated that all six induced IgG Abs were subjected to positive pressure for mutation in the CDR. Our results indicate that as the immune response to pig xenoantigens matures, a class switch associated with the onset of somatic mutations occurs in a pattern consistent with positive selection.

IgV_H genes encoding -gal-reactive Abs isolated from the peripheral blood of normal individuals are expressed in both germline and somatically mutated configurations (27). A comparison of the structure and sequence of the IgV_H genes encoding anti--gal Abs passively isolated from normal individuals and those encoding -gal-reactive Abs in human patients actively exposed to pig cells indicates that these genes have a common structural configuration that we believe may be characteristic of Abs that react with the carbohydrate Ags expressed by pig xenografts. By definition, a canonical structural group is one of seven main chain conformations that characterize hypervariable regions of IgV_H genes (38). These structural groups have been formulated based on an analysis of 83 human V_H genes with open reading frames (36). Within a specified canonical group, amino acids located in key sites are hypothesized to contribute to the antigenic specificity and structural conformation of the Ab molecule. The shape of the Ag binding site has been hypothesized to be the first step in a two-stage process in which recognition of a particular structural determinant occurs by a "docking process" mediated by Abs with a unique canonical structure (63). The second stage involves specific recognition, where responding Abs are further selected based on the sequence of the hypervariable regions (63). Natural -gal Abs and -gal Abs induced by placement of a xenograft share a single conserved conformational structure (1–3 canonical structure), which may be relevant in the recognition of carbohydrate structural determinants. We have demonstrated that the conformational structure of these Abs is not altered as a result of somatic mutation, suggesting that the structural configuration of the Ab is maintained, perhaps due to selective pressures. Our data supports the concept that somatic mutation may alter the affinity of the interaction of xenoantibodies with their xenoantigen targets without altering a basic conserved Ab configuration characteristic of several species and humans (29, 30, 38, 58). A more extensive analysis, including the definition of the structure of the third hypervariable loop (64) and the L chain conformation associated with Abs that react with the -gal epitope, is needed to determine whether xenoantibodies are encoded by Ig genes in a unique structural class and the role of other sites within the V_H region on xenoantibody binding and affinity.

Despite the similarities in our experimental observations, Galilli and colleagues conclude, based on the sequences of nine anti-gal-producing cDNA clones isolated from normal individuals, that the use of the V_H3 family of genes in the synthesis of Abs that recognize the gal (1, 3) gal epitope in humans represents a more traditional T cell-dependent response to antigenic challenge (27). Our data provides information that is unique and distinct from the findings reported by this group. We conclude, based on Ig sequencing of 69 IgV_H genes cloned from IgM and IgG cDNA libraries isolated from human patients exposed to pig cells, that IgM xenoantibodies are expressed in germline configuration, suggesting that xenoantibodies are initially produced without the requirement for T cell help. Two specific IgV_H germline progenitors, IGHV3-11 and IGHV3-74, encode IgM and IgG xenoantibody responses. The IGHV3-11 germline progenitor was not identified in the sequences reported by Wang et al. (27). Our data demonstrates that this gene encodes a clonally expanded group of xenoantibodies of both IgM and IgG isotypes in human patients undergoing an active immune response to pig cells. The IGHV3-7 germline gene that encodes three of nine of the anti-gal Abs passively isolated from normal individuals by Wang et al. is not increased in expression in these patients at any time. We have used PCGENE to confirm that somatic mutation does not interfere with the ability of our oligonucleotide primers to identify these genes. In addition, we directly screened the patient samples using oligonucleotide probes that include the mutations identified in this manuscript and found no increase in the expression of these genes (for example, oligonucleotide 193 is 100% identical with the mutated sequence identified in the V3-74 gene, but this specific gene demonstrated no increase in expression in patients exposed to pig cells; see Materials and Methods). Our studies demonstrate that the V_H3-11 genes encoding IgM Abs expressed in germline configuration undergo clonal expansion and an isotype switch 3 wk post-BAL exposure. Somatic mutation does not contribute to the xenoantibody response until day 21. However, whether or not the xenoantibody response is T cell-independent or T cell-dependent cannot be established on the basis of nucleic acid sequencing alone. It has recently been demonstrated that human B1-a cells encoding natural Abs and autoantibodies can demonstrate somatic mutation, Ag-driven selection, somatic diversification, and affinity maturation (65, 66, 67, 68, 69). Depletion of CD4⁺ cells and an analysis of the IgV_H response in the absence of Th cells is necessary to address this question. While we expect that both germline and nongermline genes encoding anti-gal Abs may be present in resting individuals in the absence of immune stimulation, the expansion of a small group of IgV_H genes in response to Ag exposure has allowed us to identify the specific genes responsible for mediating the humoral response of patients to pig tissues. The use of only two IgV_H germline genes to mediate xenoantibody responses to porcine hepatocytes, despite the potential for other IgV_H germline progenitors to encode anti-gal Abs, may be due to the route of exposure of these patients to pig xenoantigens. The configuration of the gal (1, 3) gal epitope on the surface of pig cells may determine the pathway and selection of IgV_H genes used by the patient to respond to the xenograft. The diverse repertoire of anti-pig natural Abs is well documented and includes Abs with high affinity for the -gal epitope (32, 70, 71), and Abs with differential binding affinities for -gal expressed as di-, tri-, or pentasaccharides (31). Our data suggests that a relatively small subset of this group bind to the -gal epitope expressed on pig cells.

In summary, the defining characteristics of the human humoral response to xenografts includes the following: 1) a restriction in the use of IgV_H genes to encode IgM and IgG Ab responses to the expression of two specific V_H3 genes, IGHV3-11 and IGHV3-74; 2) the use of V_H3 genes encoding IgM xenoantibodies in the early phases of the response expressed in a germline configuration; 3) an isotype switch and the use of the same IgV_H genes to encode IgG xenoantibodies that display evidence of somatic mutation as the response matures; and 4) the use of IgV_H genes that exhibit a conserved 1–3 canonical structure to encode xenoantibodies. The ability to control the human anti-pig Ab response by manipulation of this small group of IgV_H genes may ultimately contribute to the development of therapeutic strategies to prolong the survival of xenografted organs in humans.

	Acknowledgments

 
Peripheral blood samples obtained from patients exposed to a BAL support device were provided by Dr. Achilles Demetriou, Cedars-Sinai Medical Center. We thank Laura Rassenti and Dr. T. J. Kipps at the University of California, San Diego for providing us with advice and the control plasmid used to identify specific human V_H gene families in the PCR-ELISA.

	Footnotes

 
¹ This work was supported in part by National Institutes of Health Grants RR12186 and AI45484.

² Address correspondence and reprint requests to Dr. Mary Kearns-Jonker, Transplantation Biology Research Laboratory, Division of Cardiothoracic Surgery, Childrens’ Hospital of Los Angeles, 4650 Sunset Boulevard, Mail Stop 66, Los Angeles, CA 90027; E-mail address:

³ Abbreviations used in this paper: gal (1,3) gal, galactose (1, 3) galactose; BAL, bioartificial liver; CDR, complementarity-determining region; FR, framework region; R, replacement; S, silent.

Received for publication June 4, 1999. Accepted for publication August 3, 1999.

	References

Top Abstract Introduction Materials and Methods Results Discussion References

 

1. Platt, J. L., F. H. Bach. 1991. Mechanism of tissue injury in hyperacute xenograft rejection. ed. Xenotransplantation: The Transplantation of Organs and Tissues Between Species 1st Ed.69. Springer-Verlag, Berlin.
2. Cooper, D. K. C., Y. Ye, L. L. J. Rolf, N. Zuhdi. 1991. The pig as potential organ donor for man. ed. The Transplantation of Organs and Tissues Between Species 481. Springer-Verlag, Berlin.
3. Lu, C. Y., T. A. Khair-El-Din, I. A. Dawidson, T. M. Butler, K. M. Brasky, M. A. Vazquez, S. C. Sicher. 1994. Xenotransplantation. FASEB J. 8:1122.[Abstract]
4. Bach, F. H.. 1991. Revisiting a challenge in transplantation: discordant xenografting. Hum. Immunol. 30:262.[Medline]
5. Platt, J. L., R. J. Fischel, A. J. Matas, S. A. Reif, R. M. Bolman, F. H. Bach. 1991. Immunopathology of hyperacute xenograft rejection in a swine-to-primate model. Transplantation 52:214.[Medline]
6. Dalmasso, A., G. M. Vercellotti, R. J. Fischel, R. M. Bolman, F. H. Bach, J. L. Platt. 1992. Mechanism of complement activation in the hyperacute rejection of porcine organs transplanted into primate recipients. Am. J. Path. 140:1157.[Abstract]
7. Geller, R. L., F. H. Bach, M. A. Turman, P. Casali, J. L. Platt. 1993. Evidence that polyreactive antibodies are deposited in rejected discordant xenografts. Transplantation 55:168.[Medline]
8. Kaplon, R. J., R. E. Michler, H. Xu, P. A. Kwiatkowski, N. M. Edwards, J. L. Platt. 1995. Absence of hyperacute rejection in newborn pig-to-baboon cardiac xenografts. Transplantation 59:1.[Medline]
9. Kroshus, T. J., III R. M. Bolman, A. P. Dalmasso. 1996. Selective IgM depletion prolongs organ survival in an ex vivo model of pig-to-human xenotransplantation. Transplantation 62:5.[Medline]
10. Fischel, R. J., III R. M. Bolman, J. L. Platt, J. S. Najarian, F. H. Bach, A. J. Matas. 1990. Removal of IgM anti-endothelial antibodies results in prolonged cardiac xenograft survival. Transplant. Proc. 22:1077.[Medline]
11. Cooper, D. K. C., P. A. Human, G. Lexer, A. G. Rose, J. Rees, M. Keraan, E. Du Toit. 1988. Effects of cyclosporine and antibody adsorption on pig cardiac xenograft survival in the baboon. J. Heart Transplant. 7:238.[Medline]
12. Parker, W., D. Bruno, Z. E. Holzknecht, J. L. Platt. 1994. Characterization and affinity isolation of xenoreactive human natural antibodies. J. Immunol. 153:3791.[Abstract]
13. Sandrin, M. S., H. A. Vaughan, P. L. Dabkowski, I. F. C. McKenzie. 1993. Anti-pig IgM antibodies in human serum react predominantly with Gal(1–3)Gal epitopes. Proc. Natl. Acad. Sci. USA 90:11391.[Abstract/Free Full Text]
14. Galili, U., S. B. Shohet, E. Kobrin, C. L. M. Stults, 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.[Abstract/Free Full Text]
15. Galili, U., and K. Swanson. 1991. Gene sequences suggest inactivation of 1–3 galactosyltransferase in catarrhines after the divergence of apes from monkeys. Proc. Natl. Acad. Sci. USA 7401.
16. Larsen, R. D., C. A. Rivera-Marrero, L. K. Ernst, R. D. Cummings, J. D. Lowe. 1990. Frameshift and nonsense mutations in a human genomic sequence homologous to a murine UDP-Gal:ß-D-Gal(1, 4)-D-GlcNAc (1–3) galactosyltransferase cDNA. J. Biol. Chem. 265:7055.[Abstract/Free Full Text]
17. Joziasse, D. H., J. H. Shaper, D. H. Van den Eijnden, A. H. Van Tunen, N. L. Shaper. 1989. Bovine 1–3 galactosyltransferase: isolation and characterization of a cDNA clone-identification of homologous sequences in human genomic DNA. J. Biol. Chem. 264:290.
18. Galili, U., R. E. Mandrell, R. M. Hamadeh, S. B. Shohet, J. M. Griffiss. 1988. Interaction between human natural anti--galactosyl immunoglobulin G and bacteria of the human flora. Infect. Immun. 56:1730.[Abstract/Free Full Text]
19. Makowka, L., D. V. Cramer, A. Hoffman, M. Breda, L. Sher, G. Eiras-Hreha, P. J. Tuso, C. Yasunaga, C. A. Cosenza, G. D. Wu, F. A. Chapman, L. Podesta. 1995. The use of a pig liver xenograft for temporary support of a patient with fulminant hepatic failure. Transplantation 59:1654.[Medline]
20. Cotterell, A. H., B. H. Collins, W. Parker, R. C. Harland, J. L. Platt. 1995. The humoral immune response in humans following cross-perfusion of porcine organs. Transplantation 60:861.[Medline]
21. Baquerizo, A., A. Mhoyan, H. Shirwan, J. Swensson, R. W. Busuttil, A. A. Demetriou, D. V. Cramer. 1997. Xenoantibody response of patients with severe acute liver failure exposed to porcine antigens following treatment with a bioartificial liver. Transplant. Proc. 29:964.[Medline]
22. Baquerizo, A., A. Mhoyan, M. Kearns-Jonker, W. S. Arnaout, C. Shackleton, R. W. Busuttil, A. A. Demetriou, D. V. Cramer. 1999. Characterization of human xenoreactive antibodies in liver failure patients exposed to pig hepatocytes after bioartificial liver treatment. Transplantation 67:5.[Medline]
23. Cramer, D. V.. 1996. Natural Antibodies. ed. Transplantation Biology: Cellular and Molecular Aspects 473. Raven Press, New York.
24. Fehr, T., H. Y. Naim, M. F. Bachmann, A. F. Ochsenbein, P. Spielhofer, E. Bucher, H. Hengartner, M. A. Billeter, R. M. Zinkernagel. 1998. T-cell independent IgM and enduring protective IgG antibodies induced by chimeric measles viruses. Nat. Med. 4:945.[Medline]
25. Mond, J. J., A. Lees, C. M. Snapper. 1995. T cell-independent antigens type 2. Annu. Rev. Immunol. 13:655.[Medline]
26. Thomas, J. W., P. M. Kralick, U. K. Ewulonu. 1997. T cell-independent response to Brucella-insulin identifies a preimmune repertoire for insulin. J. Immunol. 159:2334.[Abstract/Free Full Text]
27. Wang, L., M. Z. Radic, U. Galili. 1995. Human anti-Gal heavy chain genes: preferential use of V_H3 and the presence of somatic mutations. J. Immunol 155:1276.[Abstract]
28. Wu, G. D., D. V. Cramer, F. A. Chapman, R. Oriol, L. Makowka. 1995. Genetic control of the humoral immune response to xenografts. I. Functional characterization of rat monoclonal antibodies to hamster heart xenografts. Transplantation 60:1497.[Medline]
29. Cramer, D. V., G. D. Wu, M. Kearns-Jonker, E. Gochi, S. Wakiyama, H. Shirwan, D. Borie. 1998. The humoral response to xenografts is controlled by a restricted repertoire of immunoglobulin V_H genes. Transplantation 66:1375.[Medline]
30. Kearns-Jonker, M., M. Fraiman, W. Chu, E. Gochi, J. Michel, G. D. Wu, D. V. Cramer. 1998. Xenoantibodies to pig endothelium are expressed in germline configuration and share a conserved immunoglobulin V_H gene structure with antibodies to common infectious agents. Transplantation 65:1515.[Medline]
31. McKane, W., J. Lee, R. Preston, A. Hacking, P. Simpson, S. Lynds, L. Goldberg, T. Cairns, D. Taube. 1998. Polymorphism in the human anti-pig natural antibody repertoire. Transplantation 66:626.[Medline]
32. Lee, J., T. Cairns, W. McKane, M. Rashid, A. J. T. George, D. Taube. 1998. Demonstration of IgM antibodies of high affinity within the anti-gal 1–3 gal antibody repertoire. Transplantation 66:1117.[Medline]
33. Chomczynsky, P., N. Sacchi. 1987. Single-step method of RNA isolation by acid guanidium-thiocyanate-phenol-chloroform extraction. Anal. Biochem. 162:156.[Medline]
34. Rassenti, L. Z., H. Kohsaka, T. J. Kipps. 1995. Analysis of immunoglobulin V_H gene repertoire by an anchored PCR-Elisa. Ann. NY Acad. Sci. 764:463.[Medline]
35. Huang, C., B. D. Stollar. 1991. Construction of representative immunoglobulin variable region cDNA libraries from human peripheral blood lymphocytes without in vitro stimulation. J. Immunol. Methods 141:227.[Medline]
36. Tomlinson, I. M., G. Walter, J. D. Marks, M. B. Llewelyn, G. Winter. 1992. The repertoire of human germline V_H sequences reveals about fifty groups of V_H segments with different hypervariable loops. J. Mol. Biol. 227:776.[Medline]
37. Giudicelli, V., D. Chaume, J. Bodmer, W. Muller, C. Busin, S. Marsh, R. Bontrop, L. Marc, A. Malik, M. P. Lefranc. 1997. IMGT, the international ImMunoGene Tics database. Nucleic Acids Res. 25:206.[Abstract/Free Full Text]
38. Chothia, C., A. M. Lesk, E. Gherardi, I. M. Tomlinson, G. Walter, J. D. Marks, M. B. Llewelyn, G. Winter. 1992. Structural repertoire of the human V_H segments. J. Mol. Biol. 227:799.[Medline]
39. Walter, G., I. M. Tomlinson, P. H. Dear, E. L. L. Sonnhammer, G. P. Cook, G. Winter. 1995. Comparison of the human germline and rearranged V_H repertoire reveals complementarity between germline variability and somatic mutation. Ann. NY Acad. Sci. 764:180.[Medline]
40. Chang, B., P. Casali. 1994. The CDR1 sequences of a major proportion of human germline Ig V_H genes are inherently susceptible to amino acid replacement. Immunol. Today 15:367.[Medline]
41. Chang, B., P. Casali. 1995. A sequence analysis of human germline Ig V_H and V_L genes: the CDR1 sequences of a major proportion of human germline V_H but not V_L genes display a high inherent susceptibility to amino acid replacement. Ann. NY Acad. Sci. 764:171.
42. Schettino, E. W., A. Cerutti, N. Chiorazzi, P. Casali. 1998. Lack of intraclonal diversification in Ig heavy and light chain V region genes expressed by CD5⁺IgM⁺ chronic lymphocytic leukemia B cells: a multiple time point analysis. J. Immunol. 160:820.[Abstract/Free Full Text]
43. Galili, U., A. Tibell, B. Samuelsson, L. Rydberg, C. G. Groth. 1995. Increased anti-Gal activity in diabetic patients transplanted with fetal porcine islet cell clusters. Transplantation 59:1549.[Medline]
44. Gabrielli, A., M. Candel, A. M. Ricciatti, M. L. Caniglia, J. Wieslander. 1991. Antibodies to mouse laminin in patients with systemic sclerosis (scleroderma) recognize galactose (1–3)galactose epitopes. Clin. Exp. Immunol. 86:367.[Medline]
45. Pascual, V., J.D. Capra. 1991. Human immunoglobulin heavy-chain variable region genes: organization, polymophism, and expression. Adv. Immunol. 49:1.[Medline]
46. Kodaira, M., T. Kinashi, I. Umemura, F. Matsuda, T. Noma, Y. Ono, T. Honjo. 1986. Organization and evolution of variable region genes of the human immunoglobulin heavy chain. J. Mol. Biol. 190:529.[Medline]
47. van Dijk, K. W., F. Mortari, P. M. Kirkham, Jr H. W. Schroeder, E. C. B. Milner. 1993. The human immunoglobulin V_H7 gene family consists of a small, polymorphic group of six to eight gene segments dispersed throughout the V_H locus. Eur. J. Immunol. 23:832.[Medline]
48. Rechavi, G., D. Ram, L. Glazer, R. Zakut, D. Givot. 1983. Evolutionary aspects of immunoglobulin heavy chain variable region (V_H) gene subgroups. Proc. Natl. Acad. Sci. USA 80:855.[Abstract/Free Full Text]
49. Huang, C., A. K. Stewart, R. S. Schwartz, B. D. Stollar. 1992. Immunoglobulin heavy chain gene expression in peripheral blood B lymphocytes. J. Clin. Invest. 89:1331.
50. Huang, C., B. D. Stollar. 1993. A majority of Ig H chain cDNA of normal human adult blood lymphocytes resembles cDNA for fetal Ig and natural autoantibodies. J. Immunol. 151:5290.[Abstract]
51. Johnson, T. A., L. Z. Rassenti, T. J. Kipps. 1997. Ig V_H1 genes expressed in B cell chronic lymphocytic leukemia exhibit distinctive molecular features. J. Immunol. 158:235.[Abstract]
52. Milner, E. C. B., W. O. Hufnagle, A. M. Glas, I. Suzuki, C. Alexander. 1995. Polymorphism and utilization of human V_H genes. Ann. NY Acad. Sci. 764:50.[Medline]
53. Tonnelle, C., A. M. Cuisinier, L. Gauthier, V. Guelpa-Fonlupt, M. Milili, C. Schiff, M. Fougereau. 1995. Fetal versus adult preB or B cells: the human V_H repertoire. Ann. NY Acad. Sci. 764:231.[Medline]
54. Borie, D. C., D. V. Cramer, H. Shirwan, G. D. Wu, O. Rodriguez, F. A. Chapman, H. K. Wang, L. Makowka. 1995. Genetic control of the humoral immune response to xenografts. II. Monoclonal antibodies that cause rejection of heart grafts are encoded by germline immunoglobulin genes. Transplantation 60:1504.[Medline]
55. Andris, J. S., J. D. Capra. 1995. The molecular structure of human antibodies specific for the human immunodeficiency virus. J. Clin. Immunol. 15:17.[Medline]
56. Shlomchik, M. J., A. Marshak-Rothstein, C. B. Wolfowicz, T. L. Rothstein, M. G. Weigert. 1987. The role of clonal selection and somatic mutation in autoimmunity. Nature 328:805.[Medline]
57. Shlomchik, M. J., A. H. Aucoin, D. S. Pisetsky, M. G. Weigert. 1987. Structure and function of anti-DNA autoantibodies derived from a single autoimmune mouse. Proc. Natl. Acad. Sci. USA 84:9150.[Abstract/Free Full Text]
58. Gochi, E., G. D. Wu, S. Wakiyama, M. Kearns-Jonker, J. Swensson, D. V. Cramer. 1999. Genetic control of the humoral responses to xenografts. III. Identification of the immunoglobulin V_H genes responsible for encoding rat immunoglobulin G xenoantibodies to hamster heart grafts. Transplantation 68:15.[Medline]
59. Kantor, A. B., L. A. Herzenberg. 1993. Origin of murine B cell lineages. Annu. Rev. Immunol. 11:501.[Medline]
60. Haughton, G., L. W. Arnold, A. C. Whitmore, S. H. Clarke. 1993. B-1 cells are made, not born. Immunol. Today 14:84.[Medline]
61. Weiss, U., R. Zoebelein, K. Rajewsky. 1992. Accumulation of somatic mutants in the B cell compartment after primary immunization with a T cell-dependent antigen. Eur. J. Immunol. 22:511.[Medline]
62. McHeyzer-Williams, M. G., M. J. McLean, P. A. Lalor, G. J. V. Nossal. 1993. Antigen-driven B cell differentiation in vivo. J. Exp. Med. 178:295.[Abstract/Free Full Text]
63. Lara-Ochoa, F., J. C. Almagro, E. Vargas-Madrazo, M. Conrad. 1996. Antibody-antigen recognition: a canonical structure paradigm. J. Mol. Evol. 43:678.[Medline]
64. Morea, V., A. Tramontano, M. Rustici, C. Chothia, A. M. Lesk. 1998. Conformations of the third hypervariable region in the V_H domain of immunoglobulins. J. Mol. Biol. 275:269.[Medline]
65. Harindranath, N., I. S. Goldfarb, H. Ikematsu, S. E. Burastero, R. L. Wilder, A. L. Notkins, P. Casali. 1991. Complete sequences of the genes encoding the V_H and V_L regions of low- and high-affinity monoclonal IgM and IgA1 rheumatoid factors produced by CD5⁺ B cells from a rheumatoid arthritis patient. Int. Immunol. 3:865.[Abstract/Free Full Text]
66. Mantovani, L., R. L. Wilder, P. Casali. 1993. Human rheumatoid B-1a (CD5⁺ B) cells make somatically hypermutated high affinity IgM rheumatoid factors. J. Immunol 151:488.
67. Kasaian, M. T., H. Ikematsu, J. E. Balow, P. Casali. 1994. Structure of the V_H and V_L segments of monoreactive and polyreactive IgA autoantibodies to DNA in patients with systemic lupus erythematosus. J. Immunol. 152:3137.[Abstract]
68. Kasaian, M. T., P. Casali. 1995. B-1 cellular origin and V_H segment structure of IgG, IgA, and IgM anti-DNA autoantibodies in patients with systemic lupus erythematosus. Ann. NY Acad. Sci. 764:410.[Medline]
69. Schettino, E. W., S. K. Chai, M. T. Kasaian, Jr H. W. Schroeder, P. Casali. 1997. V_HDJ_H gene sequences and antigen reactivity of monoclonal antibodies produced by human B-1 cells. Evidence for somatic selection. J. Immunol. 158:2477.[Abstract]
70. Vanhove, B., F. H. Bach. 1993. Human xenoreactive natural antibodies—avidity and targets on porcine endothelial cells. Transplantation 56:1251.[Medline]
71. Galili, U., O. P. Minanov, R. E. Michler, K. R. Stone. 1997. High-affinity anti-Gal immunoglobulin G in chronic rejection of xenografts. Xenotransplantation 4:127.

This article has been cited by other articles:

				H. Xu, D. Yin, B. Naziruddin, L. Chen, A. Stark, Y. Wei, Y. Lei, J. Shen, J. S. Logan, G. W. Byrne, et al. The In Vitro and In Vivo Effects of Anti-Galactose Antibodies on Endothelial Cell Activation and Xenograft Rejection J. Immunol., February 1, 2003; 170(3): 1531 - 1539. [Abstract] [Full Text] [PDF]

				D. Yin, L. Ma, A. Varghese, J. Shen, and A. S.-F. Chong Intact Active Bone Transplantation Synergizes with Anti-CD40 Ligand Therapy to Induce B Cell Tolerance J. Immunol., May 15, 2002; 168(10): 5352 - 5358. [Abstract] [Full Text] [PDF]

This Article Abstract Full Text (PDF) Alert me when this article is cited Alert me if a correction is posted Citation Map Services Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Request Permissions Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Kearns-Jonker, M. Articles by Cramer, D. V. Search for Related Content PubMed PubMed Citation Articles by Kearns-Jonker, M. Articles by Cramer, D. V.