One limitation to using pigs for xenotransplantation in humans has been the strong immune response to porcine antigens. The most immunogenic epitopes, generated by alpha 1,3 galactosyl transferase (GGTA1; E.C. 220.127.116.11), are responsible for hyperacute rejection (1, 2). GGTA1-null pigs were developed through the selection of rare GGTA1-null somatic cells generated by homologous recombination followed by complex nuclear transfer methods (3, 4). The removal of the GGTA1-dependent antigens has uncovered other strong antigens that still impose barriers to xenotransplantation (5, 6). The genes responsible for these antigens will need to be modified.
The approach that was successful for GGTA1 will be impractical if multiple, independent selection methods are required. To overcome these limitations we developed a method utilizing a protein called Drosophila recombination-associated protein (DRAP) with homology-dependent strand transferase and topoisomerase activities (7, 8). In a proof-of-concept study, we generated piglets with heritable, modified GGTA1 alleles through pronuclear coinjection of DRAP and DNA oligonucleotides.
Recombinant DRAP (60 ng/ul), with an N-terminal (His6), was expressed in bacteria and purified by Ni-NTA, Phosphocellulose and Superose 12 chromatography (8).
Oligonucleotides with 5′-OH ends (Invitrogen; HPLC-purified) were used for microinjection with DRAP singly, annealed as duplexes, or in non-complementary combinations: Oligonucleotides were suspended in sterile microinjection buffer (10 mM Tris-HCl, 10 mM NaCl, 1.0 mM MgCl2, and 0.1 mM EDTA, pH 7.5) to a final concentration of 75 ng/μl. For microinjection of single oligonucleotide–DRAP complexes into zygotes, 1 μl of 5′C, 3′NC, 5′C2, or 3′NC2 oligonucleotides (75ηg/μl) and 1 μl of DRAP (1.5 mg/ml) were diluted into 98 μl microinjection buffer and incubated on ice for 30 min prior to injection (9).
Ear notch-derived fibroblasts were collected from 100-mm culture dishes maintained at less than 60% confluence. For controls, a porcine fibroblast cell line from a commercial breed (+/+) and SK-N-DZ human neuroblastoma (−/−) cell line were used. Cells were trypsinized for two to three min, transferred to media, pelleted, washed 1× with phosphate-buffered saline solution, pelleted and resuspended in 4.0% paraformaldehyde and incubated at room temperature for 10 min. Fixed cells were pelleted, washed two times with 0.4% BSA, and divided into two aliquots for fluorescein isothiocyanate conjugated isolectin 4 (FITC-IB4) labeling and for unlabeled, autofluorescence measurements. Cells for labeling were resuspended in 0.5 μg/ml FITC-IB4, incubated in the dark at 37°C for 30 min., and washed two times with 0.4% body surface area (BSA). Both labeled and unlabeled cells were re-suspended in 400 μl 0.4% BSA prior to flow cytometry. The raw adjusted geometric mean (AGM) fluorescence was calculated for each cell line by dividing the geometric mean fluorescence of the labeled cells to that of the nonlabeled cells, thus correcting for cell line autofluorescence (10).
The oligonucleotides (31-mers; Table 1 legend) corresponded to the catalytic domain in exon 9 of GGTA1 (11). Each oligonucleotide contained a central point mutation. Preliminary in vitro biochemical studies and exploratory studies in mice suggested a point mutation traps the transferred oligonucleotide and yields altered phenotypes (8).
Semen for in vitro fertilization (IVF) was obtained from three Yorkshire males. Oocytes were collected from ovaries of undefined commercial breeds and matured in vitro. Approximately 50 zygotes 21 to 22 hr after insemination were visually sorted for the presence of two pronuclei whereupon DRAP and DNA were injected into the male pronucleus. Embryos at various stages of development were surgically transferred into uteri of asynchronous recipients. Recipient females (parity 0 or 1) were selected that exhibited first standing estrus 24 hr prior to oocyte insemination. Embryos that were cultured less than 48 hrs (1–2 cell stage) were placed in the ampullar region of the oviduct. Embryos cultured 48 hr or more (≥4 cell stage) were placed in the tip of the uterine horn. Typically, 30–70 injected embryos were placed in the oviduct. The pronuclear injection of DRAP and oligonucleotides showed no untoward effects on the establishment of IVF pregnancies or on litter size (9). From 15 successful pregnancies a total of 96 live-born piglets were delivered (Table 1).
Fibroblasts were obtained for culture from collagen-digested ear notches of each offspring. The amount of alpha 1,3 galactosyl sugar residues on the fibroblast cell surface was determined by staining with FITC-labeled isolectin B4 (FITC-IB4) and quantified by flow cytometry of 10,000 cells. The AGM for each experimental cell line was normalized to the AGM obtained for the positive control line to correct for any run-to-run variability during each of the 15 separate flow studies carried out over a period of 18 months (Table 1).
The bimodal distribution of the normalized AGMs for all of the experimental cell lines (Fig. 1) spanned a broad range relative to replicates of the positive control cell line. Each oligonucleotide produced animals yielding fibroblast lines exhibiting reduced FITC-IB4 staining. The AGM values were comparable for both males and females. Mean AGM values for six out of seven oligonucleotides tested were significantly reduced (Table 1; Student’s t test; P<0.02 or lower). None of the experimental cell lines had a reduction in FITC-IB4 staining comparable to the human null cell line.
Six F1 animals, developed from injections of DRAP and duplex oligonucleotide possessing 5′ single stranded extensions and having among the lowest normalized AGMs, were bred. Three independent litters yielded a total of 22 healthy F2 piglets in one breeding cycle. Cultured ear notch fibroblasts of the F2 piglets were used to quantify FITC-IB4 staining of alpha 1,3 galactosyl epitopes. Furthermore, new fibroblast cultures were established from each of the F1 pigs to examine the stability of the reduced FITC-IB4 staining phenotype observed in the neonatal period. The reduction in FITC-IB4 staining is stable over the course of more than one year (not shown) and the reduced expression of alpha 1,3 galactosyl epitopes is heritable (Fig. 1).
A majority of the F1 piglet cell lines had a normalized AGM less than half of the mean value for the positive control line. That is, less than what might be expected from the complete ablation of one active allele. This might occur if the animals were compound heterozygotes with the activity of the gene products from each allele being diminished relative to the activity of the wild-type allele. Heterozygosity was confirmed by sequencing.
A 1.2 Kbp genomic fragment, corresponding to the GGTA1 coding portion of exon 9 (694 bp) along with several hundred bp of flanking sequence, was amplified by high- fidelity polymerase chain reaction (PCR) and cloned from 25 cell lines having the most reduced FITC-IB4 staining as well as the reference control cell line. This analysis identified several single nucleotide polymorphisms (cSNPs) in the GGTA1 sequence as well as frequent and consistent nucleotide substitutions in the experimental lines.
Mutations occurred downstream from the 3′ end(s) of the mutagenic oligonucleotide(s). The different oligonucleotides produced very similar mutations within the gene and there appear to be preferred sequences for mutagenesis. The most prevalent mutation occurred at AGA/TCT followed by GTG/CAC triplets—although mutations did not occur at every such triplet—and other trinucleotides were modified (Fig. 1).
Biallelic mutations occur. This is illustrated in the sequence data from pig 676—the animal with the most reduced FITC-IB4 staining—generated from an injection of DRAP and a duplex oligonucleotide. Each allele shows unique substitutions as well as three common, defined, substitutions and a fourth, indeterminant nucleotide (n). The latter is likely to be the common C→T cSNP that was identified. There are several silent mutations as well as mutations leading to non-conservative amino acid substitutions (Fig. 2).
DRAP is encoded by a variant (T312G; ins603G) of a cDNA identified subsequently during the Drosophila genome project (NM_168011) having both strand transferase and topoisomerase activity. DRAP contains phosphotyrosine binding and pleckstrin homology domains associated with protein-protein interactions and/or signal transduction (12). Neither domain predicts a role in homology-dependent strand transfer. However, DRAP has residues, appropriately spaced, consistent with the Asp-Asp-Glu (DDE) catalytic domain found in various site-specific recombinases and viral integrases (13).
We hypothesized that if DRAP and a mutant oligonucleotide were introduced into a cell nucleus, homology-dependent strand transfer would occur (7, 14). The DNA intermediate would need to be resolved, possibly by the topoisomerase activity inherent in DRAP (8), and result in a recombinogenic DNA strand break (15). Subsequent modifications in the gene could be brought about by the actions of endogenous DNA repair enzymes. No evidence of general genotoxicity such as altered fertility, cancer or anatomic abnormalities in F1 or F2 animals was seen.
The predominant mutations induced by DRAP and oligonucleotides in pigs are G/C→A/T transitions and T/A→G/C transversions. This bears a striking similarity to the single nucleotide mutations that arise during the somatic hypermutation phases of immunogobulin gene maturation and cellular defense against retroviral infection (16, 17).
Mutations in AGA or GTG triplets define a repertoire of 36 possible codon changes that can potentially induce 17 silent, 3 conservative, and 16 nonconservative substitutions involving numerous amino acids. The extent to which the changes affect total gene product activity in a cell depend upon the specific modifications in each allele and the degree to which each is expressed. We anticipate that ongoing studies will lead to a more complete understanding of this mutagenic mechanism and refinements in the method, allowing for more precise mutagenesis or homologous recombination.
At present, the DRAP method offers two advantages over approaches producing a null allele—that of speed and the ability to create gene products with variable activity. Furthermore, as there are no selection steps it should be possible to modify several genes simultaneously or in succession. This will be important in the production of pigs for xenotransplantation.
We wish to thank Drs. B. L. Linder and A. Griffith, for their interest, support, insightful comments and editorial assistance throughout this project. We would also like to acknowledge Robin Stephens for technical assistance in sequencing.
1. Sandrin MS, Vaughan HA, Dabkowski PL, McKenzie IF. Anti-pig IgM antibodies in human serum react predominantly with Gal(alpha 1–3)Gal epitopes. Proc Natl Acad Sci U S A
1993; 90(23): 11391.
2. Sandrin MS, McKenzie IF. Gal alpha (1,3)Gal, the major xenoantigen(s) recognised in pigs by human natural antibodies. Immunol Rev
1994; 141: 169.
3. Dai Y, Vaught TD, Boone J, et al. Targeted disruption of the alpha1,3-galactosyltransferase gene in cloned pigs. Nat Biotechnol
2002; 20(3): 251. PMID: 11875425.
4. Lai L, Kolber-Simonds D, Park KW, et al. Production of alpha-1,3-galactosyltransferase knockout pigs by nuclear transfer. Science
2002; 295(5557): 1089.
5. Tseng YL, Kuwaki K, Dor FJ, et al. alpha1,3-Galactosyltransferase gene-knockout pig heart transplantation in baboons with survival approaching 6 months. Transplantation
2005; 80(10): 1493.
6. Kuwaki K, Tseng YL, Dor FJ, et al. Heart transplantation in baboons using alpha1,3-galactosyltransferase gene-knockout pigs as donors: initial experience. Nat Med
2005; 11: 29.
7. Eisen A, Camerini-Otero RD. A recombinase from Drosophila melanogaster embryos. Proc Natl Acad Sci U S A
1988; 85: 7481.
8. Eisen A. 48 U.S. Patent:6,534,643 B1. New York: Albert Einstein College of Medicine of Yeshiva University; 2003.
9. Betthauser J, Forsberg E, Augenstein M, et al. Production of cloned pigs from in vitro systems. Nat Biotechnol
2000; 18(10): 1055.
10. McKenzie IF, Xing PX, Vaughan HA, et al. Distribution of the major xenoantigen (gal (alpha 1–3)gal) for pig to human xenografts. Transpl Immunol
1994; 2(2): 81.
11. Boix E, Swaminathan GJ, Zhang Y, et al. Structure of UDP complex of UDP-galactose:beta-galactoside-alpha-1,3-galactosyltransferase at 1.53-A resolution reveals a conformational change in the catalytically important C terminus. J Biol Chem
2001; 276: 48608.
12. Sonnhammer EL, Eddy SR, Durbin R. Pfam: a comprehensive database of protein domain families based on seed alignments. Proteins
1997; 28(3): 405.
13. Polard P, Chandler M. Bacterial transposases and retroviral integrases. Mol Microbiol
1995; 15(1): 13.
14. Zhang Y, Muyrers JP, Rientjes J, Stewart AF. Phage annealing proteins promote oligonucleotide-directed mutagenesis in Escherichia coli and mouse ES cells. BMC Mol Biol
2003; 4(1): 1.
15. Johnson RD, Jasin M. Double-strand-break-induced homologous recombination in mammalian cells. Biochem Soc Trans
2001; 29(Pt 2): 196.
16. Durandy A. Activation-induced cytidine deaminase: a dual role in class-switch recombination and somatic hypermutation. Eur J Immunol
2003; 33(8): 2069.
17. Zhang H, Yang B, Pomerantz RJ. The cytidine deaminase CEM15 induces hypermutation in newly synthesized HIV-1 DNA. Nature
2003; 424(6944): 94.