Full genetic rescue of adenosine deaminase-deficient mice through introduction of the human gene
Full genetic rescue of adenosine deaminase-deficient mice through introduction of the human geneAlexandra A. J. Migchielsen, Marco L. Breuer, Michael S. Hershfield1 and Dinko Valerio2,*
Gene Therapy Section, Department of Medical Biochemistry, University of Leiden, Wassenaarseweg 72, P.O. Box 9503, 2300 RA Leiden, The Netherlands, 1Department of Rheumatology and Immunology, Duke University Medical Center, Durham, NC 27710, USA and 2IntroGene BV, Lange Kleiweg 151, 2280 GG Rijswijk, The Netherlands
Received April 16, 1996;Revised and Accepted July 2, 1996
We have shown recently that adenosine deaminase (ADA)-deficient mice die perinatally with severe liver cell degeneration. In addition to enzyme substitution, we report the restoration of viability through introduction of the human ADA gene. The ADA gene is subject to complex developmental and tissue-specific regulation. To include the cis-regulatory elements necessary for correct regulation of the human ADA gene, a large transgenic locus constituting the human ADA gene with 10 kb of 5' and 4 kb of 3' flanking sequences was generated by co-injection of two overlapping DNA fragments into murine zygotes. Probably as a result of extrachromosomal (homologous) recombination between the fragments, one of the two transgenic lines contained a reconstituted, functional human ADA gene. As in man, human ADA expression generally was low in these transgenic mice, but high in the thymus, spleen and gastro-duodenal part of the gut. Apparently, all cis-regulatory elements essential for a human expression pattern were incorporated in the transgene and were functional in the murine background. Similarly to man, the upper alimentary tract of the transgenic mice revealed low human ADA activity in contrast to extremely high levels of murine ADA. The human gene probably lacks the cis-regulatory elements that target high level murine ADA expression to the murine upper alimentary tract. ADA-deficient mice rescued by introduction of the human ADA transgene appeared histologically and immunologically normal. Apparently, human ADA can complement murine ADA in all tissues, even in the epithelium of the upper alimentary tract where human ADA activity is as much as 70-fold lower than murine ADA activity in wild-type mice. Clearly, the lethal phenotype of ADA-deficient mice is due to the absence of ADA.
Deficiency of adenosine deaminase (ADA) causes abnormalities in purine nucleoside metabolism due to the loss of hydrolytic deamination of adenosine and deoxyadenosine into inosine and deoxyinosine, respectively. In humans, this inborn error primarily interferes with lymphocyte maturation. Most patients are severely lymphopenic and lack both cellular (T-cell) and humoral (B-cell) immunity, resulting in severe combined immunodeficiency disease (SCID) (1 ). To test treatment options, such as gene therapy, we set out to generate an ADA-deficient mouse via gene targeting to embryonic stem cells. In mice, however, ADA deficiency results in perinatal death associated with severe liver cell degeneration (2 ,3 ), which suggests a critical role for ADA during murine fetal development. In addition, the thymus of one ADA-deficient mouse, that died in its third postnatal day, showed extensive apoptotic cell death (2 ), a phenomenon also observed in mice after postnatal treatment with the ADA inhibitor 2'-deoxycoformycin (4 ,5 ). This suggests that if the ADA-deficient mouse survives prenatal development, it will develop immune deficiency. Interestingly, liver dysfunction has been described as a complication of ADA deficiency in man (6 ).
To rescue the ADA-deficient mouse, we supplied it with a functional human ADA (hADA) gene by means of crossing the heterozygous ADA-deficient mouse with a hADA-bearing transgenic animal.
Although the ADA gene is ubiquitously expressed in mammals, ADA levels vary widely in various tissues and at different developmental stages. In mice, extremely high levels appear after birth in the epithelial lining of the alimentary mucosa of the tongue, oesophagus, forestomach and proximal small intestine (7 -11 ). Up to 50-fold lower but still relatively elevated levels of ADA are formed in the thymus (10 -14 ) and the placenta (15 -17 ), of both mouse and man, and in the gastro-duodenal part of the gut of man (9 ,18 -20 ).
The complex pattern of ADA expression suggests the involvement of multiple cis- and trans-regulatory elements in achieving developmentally regulated expression in diverse cell types. Because the thymus is the predominantly affected tissue in human ADA-deficient SCID patients, the delineation of the cis-regulatory element directing high level expression to this tissue received special attention. Incorporation of such an element into viral vectors destined for ADA-deficient SCID gene therapy might help to improve transgene expression by genetically corrected thymocytes. Using mice bearing chimeric reporter genes, the element controlling expression in human thymocytes was located to a 2.3 kb locus control region in the first intron of the human gene (9 ,21 -23 ). Similarly, 6.4 kb of 5' flanking sequences of the murine ADA (mADA) gene directed expression of a reporter gene to the fetal placenta and forestomach of transgenic mice (24 ). Insertion of these mADA 5' flanking sequences in front of the mADA cDNA rescued ADA-deficient fetuses from perinatal lethality (25 ). So far the regulatory elements involved in high level expression of mADA in the maternal decidua and proximal small intestine have not been identified.
The human and murine ADA proteins are encoded by single genes consisting of 12 exons (26 -28 ). The human (~32 kb) and murine ADA genes (~26 kb) share a classical mammalian `housekeeping-gene' promoter (GC-rich, lacking obvious GAAC and TTAA boxes), as well as nearly identical exon-intron boundaries. Both genes contain a large first intron, that comprises almost half of their total length. Murine ADA (341 residues) lacks the 11 C-terminal amino acid residues of hADA (29 ), due to an additional stop codon in exon 11. The proteins are 83% identical, and essential residues in the active site have been conserved (30 ).
To rescue the ADA-deficient mouse and to demonstrate additional regulatory elements of the hADA gene, transgenic mice bearing hADA were generated. In order to achieve correctly regulated hADA expression, as many cis-regulatory sequences as possible were to be included in the transgene. Because the cloning of DNA fragments into cosmids is size-limited, two overlapping hADA DNA fragments were co-injected into murine zygotes. Although neither of the individual fragments contained the entire coding region of the hADA gene, extrachromosomal (homologous) recombination between the two fragments apparently resulted in the reconstitution of a functional hADA gene. Together, the two fragments constitute the hADA gene with 10 kb of 5' and 4 kb of 3' flanking sequences.
In this study, we show that the transgenic mouse carrying the hADA gene displays an expression pattern reflecting that of ADA in man. Although there are minor differences between the transgenic and endogenous human ADA activities, all trans-acting factors and cis-regulatory elements necessary for a human-specific ADA expression pattern apparently are present and functional in a murine background. The major difference in expression levels of transgenic hADA and endogenous mADA, observed in the epithelial layer of the upper alimentary tract, reflects the difference in ADA expression between man and mouse, and is possibly associated with structural differences between these species.
Introduction of the hADA transgene into ADA-deficient mice restored viability. Remarkably, although the activities of the rescuing hADA were as much as 70 times lower than the normal mouse levels in the epithelial layer of the tongue, oesophagus and forestomach, the rescued mice appeared normal and showed no histological and immunological abnormalities.
Transgenic mice expressing hADA were generated by co-injection of the two overlapping DNA fragments, 5' hADA and 3' hADA, into murine zygotes (Fig. 1 ). Since none of the individual fragments contains the entire coding region of the hADA gene, extrachromosomal (homologous) recombination of both fragments is required to reconstitute a functional hADA gene. Together, the fragments comprise the genomic hADA gene with 10 kb of 5' and 4 kb of 3' flanking sequences.
Zymogram analysis has the advantage of allowing a semi-quantitative comparison of transgenic hADA activity versus endogenous mADA activity. Only tissues of transgenic origin revealed hADA activity. While in most tissues transgenic hADA and endogenous mADA activity were relatively low, they were high in the thymus, spleen, the duodenum and (glandular) stomach (Fig. 3 A). In most tissues, the ratio of human to murine ADA activity was in the same range of ~1:2 (thymus, spleen, kidney, adrenal gland, bladder, uterus, ovary, lung, trachea, ileum, cecum, colon, bone marrow, femur, blood; see also Fig. 3 A). This is in agreement with the transgenic mice having only one functional hADA gene and two mADA genes. Brain, skin, skeletal and cardiac muscle, however, showed a higher human than murine ADA activity (see also Fig. 3 A).
We recently have generated ADA-deficient mice, which die perinatally (2 ). In order to determine whether hADA could reverse the abnormalities resulting from the absence of ADA, mice heterozygous for the mADA gene (mADA+/-) were crossed with mice heterozygous for the hADA transgene (mADA+/+,hADA+/-). Subsequently, littermates heterozygous for both the murine and human ADA genes (mADA+/-,hADA+/-) were intercrossed to obtain mADA-deficient mice carrying hADA (mADA-/-,hADA+). The progeny were genotyped at weaning. No mADA-deficient mice of the genotype mADA-/-,hADA- were found, in contrast to mADA-/-,hADA+ mice, which were among the survivors (Fig. 1 B). This indicates that the presence of functional hADA in ADA-deficient fetuses protects them from perinatal death. Apparently hADA can complement mADA. In addition, both male and female rescued mice were fertile.
The mADA-deficient mice carrying hADA appeared phenotypically normal. Histological examination showed none of the abnormalities typical of mADA-deficient newborn mice, such as the severe liver cell degeneration, incomplete expansion of the lungs and small intestinal cell death (2 ). Although the ADA activity in mADA-deficient mice carrying hADA is relatively low in the mucosal epithelium lining the upper alimentary tract, where wild-type mice specifically develop extremely high levels of ADA, no pathology was observed in the tongue, oesophagus and forestomach (for liver and oesophagus see Fig. 4 A and B).
Murine ADA-deficient mice carrying hADA express only hADA, so hADA activity could be measured quantitatively without interference from mADA activity. The ADA activities determined in thymus and lung (Table 1 ) confirmed the semi-quantitative zymogram results (Fig. 3 ) that had shown an ~1: 2 ratio for transgenic hADA activity:mADA activity.
Tissue distribution of ADA activity in man and mouse
Tissue
Humana
Mouseb
Mouseb
mADA-
mADA-/-
mADA+/+
hADA+/+
hADA+c
hADA-
Thymus
151-790
180 +- 61 (7)
481 +- 169 (6)
Lung
1-7
22 +- 16 (8)
49 +- 17 (6)
Tongue
Entire tissue
9-19
30 +- 18 (4)
1854 +- 656 (2)
Mucosal layer
ND
383 +- 256 (4)
5268 +- 3023 (4)
Muscle layer
ND
84 +- 48 (4)
1524 +- 907 (4)
Stomach
Entire tissue
86-200
ND
ND
Glandular stomach
-
ND
ND
Forestomach
-
53 +- 17 (3)
2359 +- 750 (2)
Mucosal layer
ND
251 +- 33 (4)
9742 +- 5092 (4)
Muscle layer
ND
216 +- 100 (4)
904 +- 568 (4)
Oesophagus
Entire tissue
9-10
55 +- 32 (4)
3288 +- 194 (2)
Mucosal layer
ND
252 +- 172 (4)
17 828 +- 8492 (3)
Muscle layer
ND
160 +- 72 (4)
1252 +- 1541 (3)
aSpecific activity (nmol/min per mg of protein) given as a range (9,13,14,18,19).bSpecific activity (nmol/min per mg of protein) in mean +- SD of tissues from 4- to 11-month old mice. Numbers in parentheses indicate number of samples examined.cmADA-deficient mice carrying the hADA transgenic locus probably contain one or at most two functional hADA genes per cell. ND, not done.
Because the major discrepancy between human and murine ADA expression resides in the mucosal epithelium lining the tongue, oesophagus and forestomach (10 ), these tissues were studied in more detail in mADA-deficient mice carrying hADA (see also Fig. 3 B). Transgenic hADA activities in the entire tongue, oesophagus and forestomach were 62-, 59- and 45-fold lower than mADA activities in the respective wild-type tissues (Table 1 ). The tissues were dissected into a mucosal and muscle layer. Transgenic hADA levels appeared much lower than murine levels in the mucosal layer of the oesophagus (71*) and forestomach (52*), whereas the muscle layer showed only an 8- and 4-fold lower transgenic hADA than murine activity; the latter being close to the 2- to 3-fold difference observed for lung and thymus (Table 1 ). In contrast, the ADA activities of the separate layers of the tongue revealed a 14- and 18-fold lower transgenic hADA than murine activity in both the epithelial and muscle layers, respectively. Because separating the mucosal from the muscle layers of the tongue is particularly difficult, the high mADA activity measured for the muscle fraction is most probably due to contamination with mucosal material.
In wild-type mice, mADA activity appeared 4-, 14- and 11-fold higher in the mucosal layer than in the muscle layer of the tongue, oesophagus and forestomach, respectively. In the mADA-deficient mice carrying hADA, transgenic hADA levels of the epithelial and muscle layers appeared about equal in the oesophagus and the forestomach. The mean transgenic hADA activity in the mucosal layer of the tongue was 5-fold higher than in the muscle layer.
Purine nucleoside phosphorylase (PNP) activity was similar in tissues of both genotypes. The PNP activities in tissues of mADA+/+,hADA- mice versus mADA-/-,hADA+ mice, respectively, were (in nmol/min per mg of protein): thymus, 57 +- 11 (n = 5) versus 50 +- 13 (n = 3); lung, 44 +- 16 (n = 5) versus 47 +- 10 (n = 3); tongue (mucosa/muscle), 22 +- 4 (n = 2) versus 24 (n = 1) [11 versus 21/36 versus 25 (n = 1)], oesophagus (mucosa/muscle), 22 +- 4 (n = 2) versus 24 (n = 3) [5 versus 5/29 versus 23 (n = 1)] and forestomach (mucosa/muscle), 22 +- 1 (n = 2) versus 28 (n = 1) [34 versus 12/25 versus 56 (n = 1)].
In addition to the successful genetic approach described here, several other strategies for rescuing the mADA-deficient mouse were attempted.
First, since it has been found that genetic background can influence phenotype (32 ,33 ), heterozygous ADA-deficient 129/Ola mice were backcrossed for up to four generations with wild-type FVB, CBA, C57BL/6 and BALB/c mice. However, upon mating heterozygous ADA-deficient mice, the ADA-deficient newborns died perinatally, no matter what their genetic composition was.
Secondly, mADA-deficient mice were injected with PEG-ADA (bovine ADA modified by attachment of polyethylene glycol; a gift from Enzon, Inc.). In the clinic, intramuscular (bi)weekly injections with PEG-ADA improved the immune function of human ADA-deficient SCID patients (31 ). Three procedures were followed to get the ADA-deficient mice to survive the critical fetal period. First, newborn mice received PEG-ADA (0.25 U) intraperitoneally immediately at birth. This, however, did not protect the ADA-deficient littermates, probably because too much metabolic damage had already been done. Secondly, injecting pregnant females every 2-3 days with PEG-ADA (7.5 U) also had no effect on the viability of ADA-deficient newborns. Testing the seral ADA activity of injected mothers and their litters revealed high (PEG-) ADA levels at the maternal site only, suggesting that PEG-ADA does not cross the placenta. Thirdly, embryos were injected in utero (0.25 U) at day 14 of gestation. Preliminary results indicate that this pharmacological approach, though technically difficult, temporarily rescues the ADA-deficient embryos. So far, we have obtained one ADA-deficient mouse that died only on the ninth day of its life, having received PEG-ADA once at the 14th day of gestation.
We generated transgenic mice expressing a functional hADA gene by co-injecting two overlapping DNA fragments into murine zygotes. Since none of the individual fragments contained the entire coding region of the hADA gene, reconstitution of the gene most probably occurred via (homologous) recombination between the fragments before integration into the genome of the zygote. This hypothesis is sustained by the detection in transgenic tissues of correctly sized hADA mRNA and correctly sized hADA protein. High efficiency extrachromosomal homologous recombination in murine zygotes has been reported previously (34 ). Together, the fragments constitute the 32 kb hADA gene flanked by 10 kb of upstream and 4 kb of downstream sequences.
The expression pattern of the hADA transgene in mice closely matched the expression pattern of hADA in man, ADA activity being generally low, but high in the duodenum, stomach, thymus and spleen (Fig. 3 A). Placental ADA activity, however, high in man, is yet to be determined in hADA transgenic mice. Heterozygous ADA-deficient mice were crossed with transgenic mice bearing the hADA gene to obtain mADA-deficient mice expressing only transgenic hADA. In spite of the considerable variations in hADA activity in man [due to genetic composition and age (9 ,13 ,14 ,18 ,19 )], the transgenic hADA activity in mice approached the activity reported for man (Table 1 ). Together, the results suggest that all cis-regulatory elements essential for a human pattern of ADA expression are contained within the transgene. Moreover, trans-acting factors involved in the correct regulation of ADA gene expression appear conserved between man and mouse. Although previous studies showed that the 230 bp minimal hADA promoter directed expression of a reporter gene to all (except haematopoietic) cells, the expression pattern did not always reflect the tissue distribution of endogenous mADA (35 ). In addition, 4.3 kb of hADA 5' flanking sequences and 12.8 kb of the human first intron revealed correct reporter gene expression in the thymus, but not in the spleen, stomach and duodenum of transgenic mice (9 ). Our present data suggest that the sequences necessary for correct human ADA expression are located within a region starting 10 kb upstream of the translation initiation site extending to 4 kb downstream of the 3' polyadenylation site.
Although in transgenic mice the expression of hADA seemed to follow mADA expression with about half the endogenous mADA activity, a discrepancy existed in the upper alimentary tract. In contrast to the extremely high endogenous mADA activity, transgenic hADA activity was relatively low in the epithelial layer of the tongue, oesophagus and forestomach (Fig. 3 B). In mice, ADA is synthesized in extremely large quantities in the keratinized squamous epithelium lining the upper alimentary tract, where it accounts for up to 20% of the total soluble protein (10 ). To study this difference in more detail, ADA activity was measured separately in the mucosal and muscle layer of tongue, oesophagus and forestomach of wild-type mice and of mADA-deficient mice carrying hADA. In the epithelial layer of these tissues, transgenic hADA activity was 14-, 71- and 52-fold lower than mADA activity, respectively (Table 1 ), and about equal in the epithelial and muscle layers in which respect it differs from the situation in wild-type mice. The discrepancy between the low transgenic hADA activity in total tissue compared with the higher activities in the separate layers of mADA-deficient mice bearing hADA (Table 1 ) is not understood, but might result from the slightly different procedures used to process these tissues.
Winston et al. reported that 6.4 kb of 5' flanking sequences of the mADA gene cause elevated expression of a reporter gene in the epithelium of the upper alimentary tract and fetal placenta of transgenic mice, but not in the duodenum and maternal decidua (24 ). Apparently, different mADA gene regulatory elements are active in various murine tissues characterized by high levels of ADA. Interestingly, the mucosal epithelium of the upper alimentary tract is structurally different in man and mouse, having a keratinized squamous epithelium in mice only (11 ). In mice, the appearance of ADA in the forestomach, oesophagus and tongue coincides with maturation of the mucosal epithelium into a keratinized structure (11 ). Our data, showing only low transgenic hADA expression in the keratinized mucosal epithelium of the upper alimentary tract of mice, suggest that this level of expression is defined not by trans-acting factors but by cis-regulatory elements. Because hADA expression in man and in transgenic mice is not up-regulated, the hADA gene most likely lacks the sequences required for high level expression in the upper alimentary tract.
It is noteworthy that zymogram analysis showed that the high activity of transgenic hADA in the stomach (Fig. 3 A) is localized to the glandular stomach and not to the forestomach of transgenic mice (data not shown). In contrast to the situation in mice, the human stomach is not separated into a histologically distinct distal glandular portion and a proximal forestomach. Structurally, the human stomach resembles the glandular stomach of mice in that it has a non-keratinized, mucosal glandular epithelium. Since ADA levels both in man and mouse are relatively high in the (glandular) stomach (Table 1 ) (9 -11 ,18 ,19 ) and different murine cis-regulatory regions drive high level expression in the gastro-duodenal part of the gut versus the forestomach (oesophagus and tongue, 24 ), this structural resemblance might explain why transgenic hADA activity is selectively high in the glandular stomach of transgenic mice. We speculate that man and mouse share cis-regulatory elements targeting expression to the gastro-duodeno-jejunal part of the gut. For comparison, recently a murine homologue of the hADA thymic enhancer was identified in the first intron of the murine gene (36 ). Furthermore, the higher transgenic activity of hADA versus mADA in the skin, brain, skeletal and cardiac muscle of the transgenic mouse reflects the reported differences in organ distribution between these two species (9 -14 ,18 ,19 ).
Mice lacking ADA die around birth. They exhibit liver cell damage associated with purine metabolic disturbances and interference with transmethylation reactions (2 ,3 ). Our rescue strategy of introducing hADA sequences into ADA-deficient mice restored viability and completely corrected the histological abnormalities typical of knock-out mice. Although transgenic hADA activity was <50% of normal in most tissues, this appeared sufficient for a wild-type phenotype. Moreover, despite the fact that in the upper alimentary tract of rescued mice ADA activity is as much as 70 times lower than in the wild-type, this region showed no histological abnormalities. Apparently, high level ADA expression in the upper alimentary tract of wild-type mice is not a prerequisite for a normal physiological condition. In man, ADA levels vary widely among individuals, and levels as low as 7% of normal are sufficient (37 ). In addition, ADA-deficient humans suffer from immune dysfunction, but show no defects in the gastrointestinal tract. When up to 20% of the soluble cellular protein in the mucosal epithelium of the murine upper alimentary tract is ADA, its physiological role might well be structural as well as enzymatic. Nevertheless, maybe under more extreme conditions (housing, feeding, microbiological status), the need for high level expression in the proximal alimentary tract of mice will become apparent. Witte et al. (38 ) have suggested that the high level of degradation of purines might be involved in a defense against parasites that are entirely dependent on exogenous purines derived from the host. Such a hypothesis might now be tested.
The rescue experiment proves that, due to the lack of ADA, ADA-deficient mice die perinatally. Introduction of an ADA gene from a patient possessing a low residual activity might provide us with a mouse not only surviving the critical fetal period, but also presenting the clinical phenotype of human ADA-deficient SCID. Such an animal model would be invaluable to test treatment options such as gene therapy.
Cosmid clones ADA cos 4.3 and ADA cos 1.3, containing hADA genomic sequences, were isolated previously (39 ). The 39.0 kb SalI insert from ADA cos 4.3, called 5' hADA, contains 10 kb of upstream sequences and 29 kb of the hADA transcription unit (Fig. 1 A). The 25.8 kb SalI fragment isolated from ADA cos 1.3, called 3' hADA, starts in the middle of hADA intron I and ends 4 kb downstream of the last exon (Fig. 1 A). Together, 5' and 3' hADA span a 45.3 kb region comprising the total hADA gene with an overlap of 19.5 kb (Fig. 1 A).
The 5' and 3' hADA fragments were gel purified and co-injected in equimolar amounts at a final concentration of 10 [mu]g/ml into zygotes from superovulated B6CBA F1 (C57BL/6*CBA) mice. Transgenic mice were generated essentially as described (40 ).
Genomic DNA, isolated from tail tips clipped at weaning, was prepared as described (41 ). Southern blot and slot-blot analyses were performed according to standard procedures (42 ). The probes used were (see also Fig. 1 A): I: a 3.0 kb HindIII genomic mADA fragment recognizing both the 11.0 and 9.5 kb EcoRI fragment diagnostic for the wild-type and the mutated mADA allele, respectively; II: a 1.2 kb KpnI genomic mADA fragment recognizing an 8.3 kb EcoRI mADA fragment; III: a 0.85 kb EcoRI-XhoI genomic hADA fragment hybridizing with a 5.0 kb EcoRI hADA fragment diagnostic for the non-overlapping region of 5' hADA; IV: a 350 bp PstI-BglII hADA cDNA fragment harbouring exons 5-8 derived from pAMG1 (39 ), recognizing the overlapping region of 5' hADA and 3' hADA; and V: a 200 bp PstI-SstII hADA cDNA fragment containing exon 12 isolated from pAMG1, diagnostic for the non-overlapping region of 3' hADA.
The copy numbers of 5' and 3' hADA were determined by slot-blot analyses on calibrated amounts of the transgenic fragments as standards. Probes III, IV and V were used in three consecutive hybridization experiments.
To discriminate between mice heterozygous and homozygous for the transgenic locus, Southern blot analysis of EcoRI-digested DNA was performed with probe II. As an internal standard for the amount of loaded DNA, probe III was used. All quantitations were done with a Phosphor Imager device (Molecular Dynamics). Only mice heterozygous for the transgene were selected for the experiments.
Adult animals were euthanatized with ether. Tissues were removed quickly, rinsed in phosphate-buffered saline and placed on ice. Tissues were stored at -70oC, if not directly processed. Separation of the muscle layer from the epithelial layer of the tongue, oesophagus and forestomach was performed by incubation in 0.02 M EDTA pH 7.5, in RPMI 1640 (Gibco) at 37oC for several hours (11 ).
RNA from transgenic and non-transgenic mouse tissues (stomach, and skeletal muscle) and human cells (Epstein-Barr virus-transformed B-cells, 43 ), was isolated by the LiCl-urea method (44 ). The RNase protection assay was performed as described (45 ). As a probe, an antisense fragment able to protect almost the entire hADA mRNA (1354 of 1498 nt) was used. To this end, the 1402 bp SstII hADA cDNA fragment from pAMG1 (39 ) was inserted into the SstII restriction site of pBluescript II SK (Stratagene). Upon linearization of this vector with EcoRI, the antisense probe was synthesized with T3 polymerase.
Preparation of lysates from tissues was essentially according to Meera Khan (46 ). Tissues were homogenized in 750 [mu]l of lysis buffer (5 mM Na2HPO4/NaH2PO4 pH 6.4, 1 mM EDTA, 1 mM [beta]-mercaptoethanol) on ice. Carbon tetrachloride was used to remove fatty elements. Lysates were used directly, stored at -70oC, or kept at 4oC, in case of spectrophotometric analysis, for up to 1 week. Protein concentrations of tissue lysates were determined by the Bradford method with bovine serum albumin as a standard.Western blot analysis. Tissue lysates (50 [mu]g of protein) were fractionated by 12.5% SDS-PAGE, transferred to nitrocellulose and incubated with F59.1.7, a human ADA-specific monoclonal antibody. Goat anti-mouse IgG-alkaline phosphatase conjugates were used to detect bound antibodies (47 ). Compared with murine ADA, human ADA contains 11 additional amino acid residues at its carboxy-terminus. The hADA-specific monoclonal antibody F59.1.7 was elicited in mice against an 18 amino acid peptide containing 17 amino acids identical to the unique C-terminus of the hADA protein plus a cysteine at the amino-terminus. Enzymatic assays. Zymogram analysis allowed assaying for the presence of both human and murine ADA activity and was performed essentially according to Meera Khan (46 ). Briefly, tissue lysates (10 [mu]g of protein, unless stated otherwise) were loaded on to cellulose acetate gels (Cellogel, Chemetron) which, after electrophoresis, were stained specifically for ADA activity. The spectrophotometric determination of ADA activity was, in principle, as described (48 ). To a reaction mixture of 990 [mu]l of 0.1 mM adenosine (Boehringer) in 0.05 M sodium phosphate pH 7.5, 2 [mu]l of xanthine oxidase (XOD, 0.04 IU, Boehringer), 2 [mu]l of nucleoside phosphorylase (NP, 0.04 IU, Boehringer) and 5-100 [mu]l of tissue lysate were added. The inosine produced by the deamination of adenosine is further converted to uric acid by the added enzymes. The rate of production of uric acid was recorded at 37oC at a wavelength of 293 nm on a Gilford Response spectrophotometer. Similarly, to determine PNP activity spectrophotometrically, 2 [mu]l of xanthine oxidase (XOD, 0.04 IU, Boehringer) and 5-100 [mu]l of tissue lysate were added to a reaction mixture of 990 [mu]l of 0.1 mM inosine (Boehringer) in 0.05 M sodium phosphate pH 7.5.
The tongue, oesophagus, forestomach, duodenum, jejunum, ileum, thymus, spleen, lung, liver, cardiac and skeletal muscle were isolated from 5- to 9-month old mice. Tissues were fixed in 10% phosphate-buffered formalin. After processing to paraffin blocks, semi-serial sections stained with haematoxylin-phloxine-saffron (HPS) were produced. In total, four mADA-/-,hADA+ mice and three mADA+/+,hADA- controls were examined histologically.
Thymuses and spleens were isolated, and single-cell suspensions were prepared. Non-erythroid cells were counted and subjected to two-colour immunofluorescence analysis. The antibodies used, either biotinylated or conjugated with fluorescein isothiocyanate (FITC) were: H129-19 (anti-CD4), 53-6.7 (anti-CD8), 53-2.1 (anti-Thy1.2) and RA3-6B2 (anti-CD45R). Antibodies were purchased from Pharmingen (53-2.1), or were a gift from F. Ossendorp (Immunohematology and Bloodbank, Academical Hospital Leiden, The Netherlands). Cells incubated with biotinylated antibodies were stained subsequently by incubation with periodinin chlorophyll protein (PerCP)-conjugated streptavidin. Flow cytometric analysis was performed on a FACscan (Becton Dickinson); 2*104 cells were analysed in each sample.
The mitogenic stimulation assays were performed in 96-well round-bottom microtitre plates (Greiner, Nurtingen). Thymocytes and splenocytes (105) were cultured in 200 [mu]l of RPMI 1640 tissue culture medium supplemented with 10% heat-inactivated fetal calf serum (Seralab), 2 mM L-glutamine, 1*10-4 M [beta]-mercaptoethanol, 100 U/ml penicillin and 100 [mu]g/ml streptomycin. All cultures were performed in triplicate. The cells were stimulated with LPS (15 [mu]g/ml; Difco), PHA [1% v/v (+-90 [mu]g/ml); Wellcome] and ConA (6 [mu]g/ml; Pharmacia). In the case of thymocyte stimulation, recombinant human interleukin 1[beta] (rhIL-1[beta], 25 U/ml ) was added. The cells were cultured in an atmosphere of 5% CO2 at 37oC for 72 h, the last 4-5 h of which were in the presence of [3H]thymidine (9.2 KBq/ well, sp. act. 74 Gbq/mmol, Amersham). The cells were harvested on glass fibre filters and the incorporated radioactivity was determined by liquid scintillation counting (TriCarb, Packard).
We are grateful to Harry Brok, Jan van de Brugge and Ko de Vast for animal care. We also thank Yolanda Berkhoudt-Vankan, Henny Lemmink, Chris Zurcher, Professor Meera Khan, Juul Wijnen, Ada Struijk, Shosh Knaan and Stephan Verlinden for practical and theoretical assistance, and Hans van Ormondt and Rob Hoeben for critically reading the manuscript. Finally, we are greatly indebted to Enzon, Inc. for the gratis supply of PEG-ADA.
This work was supported by the Netherlands Organization for Scientific Research (NWO) and the Praeventiefonds.
1 Giblett, E.R., Anderson, J.E., Cohen, F., Pollara, B. and Meuwissen, H.J. (1972) Adenosine deaminase deficiency in two patients with severely impaired cellular immunity. Lancet, ii, 1067-1069.MEDLINE Abstract
2 Migchielsen, A.A.J.,Breuer, M.L., van Roon, M., te Riele, H., Zurcher, C., Ossendorp, F., Toutain, S., Hershfield, M.S., Berns, A. and Valerio, D. (1995) Adenosine-deaminase-deficient mice die perinatally and exhibit liver-cell degeneration, atelectasis and small intestinal cell death. Nature Genet., 10, 279-287.
3 Wakamiya, M., Blackburn, M.R., Jurecic, R., McArthur, M.J., Geske, R.S., Cartwright J., Jr, Mitani, K., Vaishnav, S., Belmont, J.W., Kellems, R.E., Finegold, M.J., Montgomery A.C. Jr, Bradley, A. and Caskey, C.T. (1995) Disruption of the adenosine deaminase gene causes hepatocellular impairment and perinatal lethality in mice. Proc. Natl Acad. Sci. USA, 92, 3673-3677.MEDLINE Abstract
4 Tedde, A., Balis, M.E., Ikehara, S., Pahwa, R., Good, R.A. and Trotta, P.P. (1980) Animal model for immune dysfunction associated with adenosine deaminase deficiency. Proc. Natl Acad. Sci. USA, 77, 4899-4903.MEDLINE Abstract
5 Ratech, H., Hirschhorn, R. and Thorbecke, G.J. (1985) Effects of deoxycoformycin in mice III. A murine model reproducing the multi-system pathology of human adenosine deaminase deficiency. Am. J. Pathol., 119, 65-72.MEDLINE Abstract
6 Bollinger, M.E., Arredondo-Vega, F.X., Santisteban, I., Schwarz, K., Hershfield, M.S. and Lederman, H.M. (1996) Brief report: hepatic dysfunction as a complication of adenosine deaminase deficiency. N. Engl. J. Med., 334, 1367-1371.
7 Brady, T.G. and O'Donovan, C. (1965) A study of the tissue distribution of adenosine deaminase in six mammal species. Comp. Biochem. Physiol., 14, 101-120.
8 Lee, P.C. (1973) Developmental changes of adenosine deaminase, xanthine oxidase, and uricase in mouse tissues. Dev. Biol., 31, 227-233.MEDLINE Abstract
9 Aronow, B.J., Lattier, D., Silbiger, R., Dusing, M., Hutton, J., Jones, G., Stock, J. McNeish, J., Potter, S., Witte, D. and Wiginton, D. (1989) Evidence for a complex regulatory array in the first intron of the human adenosine deaminase gene. Genes Dev., 3, 1384-1400.MEDLINE Abstract
10 Chinsky, J.M., Ramamurthy, V., Knudsen, T.B., Higley, H.R., Fanslow, W.C., Trentin, J.J. and Kellems, R.E. (1989) Developmental and tissue specific regulation of adenosine deaminase in mice. In Beaudet, A.L., Mulligan, R. and Verma, I.N. (eds), Gene Transfer and Gene Therapy. Alan R. Liss, Inc., New York, pp.255-268.
11 Chinsky, J.M., Ramamurthy, V., Fanslow, W.C., Ingolia, D.E., Blackburn, M.R., Shaffer, K.T., Higley, H.R., Trentin, J.J., Rudolph, F.B., Knudsen, T.B. and Kellems, R.E. (1990) Developmental expression of adenosine deaminase in the upper alimentary tract of mice. Differentiation, 42, 172-183.MEDLINE Abstract
13 Adams, A. and Harkness, R.A. (1976) Adenosine deaminase activity in thymus and other human tissues. Clin. Exp. Immunol., 26, 647-649.MEDLINE Abstract
14 Hirschhorn, R., Martiniuk, F. and Rosen, F.S. (1978) Adenosine deaminase activity in normal tissues from a child with severe combined immunodeficiency and adenosine deaminase deficiency. Clin. Immunol. Immunopathol., 9, 287-292.MEDLINE Abstract
15 Knudsen, T.B., Green, J.D., Airhart, M.J., Higley, H.R., Chinsky, J.M. and Kellems, R.E. (1988) Developmental expression of adenosine deaminase in placental tissues of the early postimplantation mouse embryo and uterine stroma. Biol. Reprod., 39, 937-951.MEDLINE Abstract
16 Knudsen, T.B., Blackburn, M.R., Chinsky, J.M., Airhart, M.J. and Kellems, R.E. (1991) Ontogeny of adenosine deaminase in the mouse decidua and placenta: immunolocalization and embryo transfer studies. Biol. Reprod., 44, 171-184.MEDLINE Abstract
17 Dooley, T., Fairbanks, L.D., Simmonds, H.A., Rodeck, C.H., Nicolaidies, K.H., Soothill, P.W., Stewart, P., Morgan, G. and Levinsky, R.J. (1987) First trimester diagnosis of adenosine deaminase deficiency. Prenatal Diagnosis, 7, 561-565. MEDLINE Abstract
18 van der Weyden, M.B. and Kelley, W.N. (1976) Human adenosine deaminase: distribution and properties. J. Biol. Chem., 251, 5448-5456.MEDLINE Abstract
19 ten Kate, J. (1985) Adenosine deaminase complexing protein as a marker in cancer studies. Ph.D. Thesis. Leiden University, The Netherlands.
20 Hirschhorn, R. (1990) In Rosen, F.S. and Seligmann, M. (eds), Immunodeficiency Reviews. Harwood Academic, New York. pp. 175-198.MEDLINE Abstract
21 Aronow, B.J., Silbiger, R.N., Dusing, M.R., Stock, J.L., Yager, K.L., Potter, S.S., Hutton, J.J. and Wiginton, D.A. (1992) Functional analysis of the human adenosine deaminase gene thymic regulatory region and its ability to generate position-independent transgene expression. Mol. Cell. Biol., 12, 4170-4185.MEDLINE Abstract
22 Aronow, B.J., Ebert, C.A., Valerius, M.T., Potter, S.S., Wiginton, D.A., Witte, D.P. and Hutton, J.J. (1995) Dissecting a locus control region: facilitation of enhancer function by extended enhancer-flanking sequences. Mol. Cell. Biol., 15, 1123-1135.MEDLINE Abstract
23 Ess, K.C., Whitaker, T.L., Cost, G.J., Witte, D.P., Hutton, J.J. and Aronow, B.J. (1995) A central role for a single c-myb binding site in a thymic locus control region. Mol. Cell. Biol., 15, 5707-5715.MEDLINE Abstract
24 Winston, J.H., Hanten, G.R., Overbeek, P.A. and Kellems, R.E. (1992) 5' Flanking sequences of the murine adenosine deaminase gene direct expression of a reporter gene to specific prenatal and postnatal tissues in transgenic mice. J. Biol. Chem., 267, 13472-13479.MEDLINE Abstract
25 Blackburn, M.R., Wakamiya, M., Caskey, C.T. and Kellems, R.E. (1995) Tissue-specific rescue suggests that placental adenosine deaminase is important for fetal development in mice. J. Biol. Chem., 270, 23891-23894.MEDLINE Abstract
26 Valerio, D., McIvor, S., Williams, S.R., Duyvesteyn, M.G.C., van Ormondt, H., van der Eb, A.J. and Martin D.W., Jr (1984) Cloning of human adenosine deaminase cDNA and expression in mouse cells. Gene, 31, 147-153.MEDLINE Abstract
27 Wiginton, D.A., Kaplan, D.J., States, J.C., Akeson, A.L., Perme, C.M., Bilyk, I.J., Vaughn, A.J., Lattier, D.L. and Hutton, J.J. (1986) Complete sequence and structure of the gene for human adenosine deaminase. Biochemistry, 25, 8234-8244.MEDLINE Abstract
28 Yeung, C.-Y., Ingolia, D.E., Roth, D.R., Shoemaker, C., Al-Ubaidi, M.R., Yen, J.-Y., Ching, C. Bobonis, C., Kaufman, R.J. and Kellems, R.E. (1985) Identification of functional murine adenosine deaminase cDNA clones by complementation in Escherichiacoli.J. Biol. Chem., 260, 10299-10307.MEDLINE Abstract
29 Al-Ubaidi, M.R., Ramamurthy, V., Maa, M.-C., Ingolia, D.E., Chinsky, J.M., Martin, B.D. and Kellems, R.E. (1990) Structural and functional analysis of the murine adenosine deaminase gene. Genomics, 7, 476-485.MEDLINE Abstract
30 Wilson, D.K., Rudolph, F.B. and Quiocho, F.A. (1991) Atomic structure of adenosine deaminase complexed with a transition-state analog: understanding catalysis and immunodeficiency mutations. Science, 252, 1278-1284.MEDLINE Abstract
31 Hershfield, M.S. and Mitchell, B.S. (1995) In Scriver, C.R., Beaudet, A.L., Sly, W.S. and Valle, D. (eds), The Molecular and Metabolic Basis of Inherited Disease. McGraw-Hill, New York. pp. 1725-1768.
32 Harvey, M., McArthur, M.J., Montgomery, C.A., Bradley, A. and Donehower, L.A. (1993) Genetic background alters the spectrum of tumors that develop in p53-deficient mice. FASEB J., 7, 938-943.MEDLINE Abstract
33 Ramirez-Solis, R., Zheng, H., Whiting, J., Krumlauf, R. and Bradley, A. (1993) Hoxb-4 (Hox-2.6) mutant mice show homeotic transformation of a cervical vertebra and defects in the closure of the sternal rudiments. Cell, 73, 279-294.MEDLINE Abstract
34 Pieper, F.R., de Wit, I.C.M., Pronk, A.C.J., Kooiman, P.M., Strijker, R., Krimpenfort, P.J.A., Nuygens, J.H. and de Boer, H.A. (1992) Efficient generation of functional transgenes by recombination in murine zygotes. Nucleic Acids Res., 20, 1259-1264.MEDLINE Abstract
35 Valerio, D., van der Putten, H., Botteri, F.M. and Hoogerbrugge, P.M. (1988) Activity of the adenosine deaminase promoter in transgenic mice. Nucleic Acids Res., 16, 10083-10097.MEDLINE Abstract
36 Brickner, A.G., Gossage, D.L., Dusing, M.R. and Wiginton, D.A. (1995) Identification of a murine homolog of the human adenosine deaminase thymic enhancer. Gene, 167, 261-266.MEDLINE Abstract
37 Hirschhorn, R., Martiniuk, F., Roegner-Maniscalco, V., Ellenbogen, A. Perignon, J.-L. and Jenkins, T. (1983) Genetic heterogeneity in partial adenosine deaminase deficiency. J. Clin. Invest., 71, 1887-1892.MEDLINE Abstract
38 Witte, D.P., Wiginton, D.A., Hutton, J.J. and Aronow, B.J. (1991) Coordinate developmental regulation of purine catabolic enzyme expression in gastro- intestinal and postimplantation reproductive tracts. J. Cell. Biol., 115, 179-190.MEDLINE Abstract
39 Valerio, D., Duyvesteyn, M.G.C., Dekker, B.M.M., Weeda, G., Berkvens, T.M., van der Voorn, L., van Ormondt, H. and van der Eb, A.J. (1985) Adenosine deaminase: characterization and expression of a gene with a remarkable promoter. EMBO J., 4, 437-443.MEDLINE Abstract
40 Hogan, B., Constantinin, F. and Lacy, E. (1986) Manipulating the Mouse Embyo: A Laboratory Manual. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY.
41 Laird, P.W., Zijderveld, A., Linders, K., Rudnicki, M.A., Jaenisch, R. and Berns, A. (1991) Simplified mammalian DNA isolation procedure. Nucleic Acids Res., 19, 4293-4295.MEDLINE Abstract
42 Maniatis, T., Fritsch, E.F. and Sambrook, J. (1982) Molecular Cloning: A Laboratory Manual. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY.
43 van de Griend, R.J., van Krimpen, B.A., Bol, S.J.L., Thompson, A. and Bolhuis, R.L.H. (1984) Rapid expansion of human cytotoxic T cell clones: growth promotion by a heat-labile serum component and by various types of feeder cells. J Immunol. Methods, 66, 285-298.MEDLINE Abstract
44 Auffray, C. and Rougeon, F. (1980) Purification of mouse immunoglobulin heavy-chain messenger RNAs from total myeloma tumor RNA. Eur. J. Biochem., 107, 303-314.MEDLINE Abstract
45 Zinn, K., DiMaio, D. and Maniatis, T. (1983) Identification of two distinct regulatory regions adjacent to the human [beta]-interferon gene. Cell, 34, 865-879.MEDLINE Abstract
46 Meera Khan, P. (1971) Enzyme electrophoresis on cellulose acetate gel: zymogram patterns in man-mouse and man-Chinese hamster somatic cell hybrids. Arch. Biochem. Biophys., 145, 470-483.MEDLINE Abstract
47 Harlow, E. and Lane, D. (1988) Antibodies: A Laboratory Manual. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY.
48 Herbschleb-Voogt, E., Pearson, P.L., Vossen, J.M. and Meera Khan, P. (1981) Basic defect in the expression of ADA in ADA-SCID disease investigated through the cells of an obligate heterozygote. Hum. Genet., 56, 379-386.MEDLINE Abstract
49 Valerio, D., Duyvesteyn, P., Meera Kahn, P., Geurts van Kessel, A., de Waard, A. and van der Eb, A.J. (1983) Isolation of cDNA clones for human adenosine deaminase. Gene, 25, 321-240.
50 van Beusechem, V.W., Kukler, A., Einerhand, M.P.W., Bakx, T.A., van der Eb, A.J., van Bekkum, D.W. and Valerio, D. (1990) Expression of human adenosine deaminase in mice transplanted with hemopoietic stem cells transduced with amphotropic retroviruses. J. Exp. Med., 172, 329-336.
*To whom correspondence should be addressed
This page is maintained by OUP admin. Last updated Thu Oct 31 15:27:49 GMT 1996. Part of the OUP Journals World Wide Web service.Copyright Oxford University Press, 1996
