Abstract
The leukocyte common (CD45) Ag is essential for normal T lymphocyte function and alternative splicing at the N terminus of the gene is associated with changes in T cell maturation and differentiation. Recently, a statistically significant association was reported in a large series of human thymus samples between phenotypically abnormal CD45 splicing and the presence of the CC chemokine receptor 5 deletion 32 (CCR5del32) allele, which confers resistance to HIV infection in homozygotes. We show here that abnormal splicing in these thymus samples is associated with the presence of the only established cause of CD45 abnormal splicing, a C77G transversion in exon A. In addition we have examined 227 DNA samples from peripheral blood of healthy donors and find no association between the exon A (C77G) and CCR5del32 mutations. Among 135 PBMC samples, tested by flow cytometric analysis, all those exhibiting abnormal splicing of CD45 also showed the exon A C77G transversion. We conclude that the exon A (C77G) mutation is a common cause of abnormal CD45 splicing and that further disease association studies of this mutation are warranted.
The leukocyte common Ag CD45 is an abundant hemopoietic cell-specific transmembrane protein tyrosine phosphatase (1, 2). The phosphatase activity of CD45 Ag is crucial for efficient lymphocyte Ag receptor signal transduction as has been shown in CD45-deficient mice (3, 4) and in humans lacking CD45 expression (5, 6). CD45 knockout mice are severely immunodeficient, with very few T cells but normal B cell numbers, and have a very similar phenotype to the two recently described patients. Both patients presented at 2 mo of age and had SCID. The infants showed low T cells numbers, were unresponsive to mitogen stimulation, and although B cell numbers were normal, Ig production was impaired with low concentrations of IgM and IgA (5, 6). Three different genetic abnormalities in the gene encoding CD45 were shown to be associated with the lack of CD45 expression in the cells of patients and the immune deficiency, providing the first direct evidence for the crucial role of CD45 in immune function in humans.
Eight different isoforms of CD45 Ag may exist, generated by alternative splicing in the N-terminal extracellular domain of the molecule (7, 8). The expression of different CD45 isoforms depends on the state of activation and differentiation of hemopoietic cells, and mAbs to the alternatively spliced exons have been used to separate subsets of T cells with distinct functional properties (9, 10, 11, 12). In humans, postthymic naive T cells express high molecular weight CD45 isoforms, recognized by CD45RA mAbs, but activation of the cells results in a change to expression of low molecular weight isoforms, detected by CD45RB and CD45RO mAbs. These two major subsets of T lymphocytes expressing CD45RA and CD45RO have been termed naive and memory cells. Although these terms are an oversimplification, the two subsets do have very different properties. The majority of CD45RA-positive T cells are small resting cells with long telomeres that divide only rarely, while CD45RO cells show evidence of activation, respond well to recall Ags, have shorter telomeres, and divide relatively frequently (13, 14). Despite this understanding, the mechanisms linking the expression of different isoforms to the function of different T cell subsets remain unclear.
A genetically determined variant pattern of CD45 isoform expression has been recognized in humans (15). Activated or memory lymphocytes in these individuals continue to express both high and low molecular mass CD45 isoforms in contrast to the normal pattern of low molecular mass isoform expression. RT-PCR and immunoprecipitation revealed that only the 205-kDa (AB) isoform of CD45 lacks proper regulation, whereas expression of the 220-kDa (ABC) isoform follows the normal pattern (16). More recently a point mutation in the fourth or A exon of the gene encoding CD45, a C to G transversion at position 77 (C77G), has been shown to prevent the normal splicing of this exon in the affected individuals (17, 18). Family studies revealed that the inheritance of the variant CD45RA pattern is that of a single autosomal dominant gene (19) and although these heterozygous individuals are apparently normal, a recent paper has described an association of exon A (C77G) and abnormal CD45 splicing with the development of multiple sclerosis in some families (20).
The CC chemokine receptor 5 allele (CCR5del32) confers resistance in homozygotes to infection with HIV isolates that use CCR5 as a coreceptor (21). Recently, linkage of abnormal CD45 splicing with CCR5del32 was reported (22). The cortical and medullary thymocytes in a statistically significant proportion of CCR5del32 carriers failed to down-regulate high molecular mass CD45RA-containing isoforms during thymocyte development and activation, resulting in continued expression of CD45RA. However, the published study did not type the thymus samples for the known CD45 exon A (C77G) mutation.
In this study, we first genotyped the previously studied thymic samples, and, second, we examined the association of CCR5del32 with abnormal CD45 splicing in DNA from PBMC and also tested peripheral blood mononuclear blood (PBMC) samples by flow cytometry to determine whether any other common cause of abnormal CD45 splicing exists. We conclude that abnormal CD45 splicing in the thymus is commonly caused by the exon A (C77G) mutation, that there is no association between CCR5del32 and exon A (C77G) in DNA from PBMC from normal subjects, and that the latter is the only common cause of abnormal splicing of CD45 in peripheral blood T cells in healthy individuals.
Materials and Methods
Materials
Genomic DNA samples were obtained from 10 individuals homozygous and 82 individuals heterozygous for the CCR5del32 mutation (95.6% of the samples were from American Caucasians), as determined in earlier studies (23). Peripheral blood was obtained from 135 healthy donors from the National Blood Transfusion Service (London, U.K.), of whom 19 typed as CCR5del32 heterozygotes. PBMC were isolated by Ficoll-Paque (Amersham Pharmacia Biotech, Piscataway, NJ) density gradient centrifugation and genomic DNA was extracted from 5 × 106 isolated PBMC as previously described (24). Thymic samples were obtained at surgery from children with congenital heart disease as discarded tissue from the Duke University Department of Pathology using an institutional review board-approved research protocol. Ten of the 12 samples, previously shown by Liao et al. (22) to exhibit abnormal splicing of CD45, were available for genotyping along with 10 thymic samples with normal CD45 phenotype as controls. The thymic samples were coded and tested blind before their phenotype was revealed. All samples were genotyped for CD45 exon A (C77G) and CCR5del32 mutations as described below.
CCR5 genotyping
The δ32 portion of the CCR5 gene was amplified by PCR from genomic DNA as previously described (21). Briefly, R5 upstream 5′-CAAAAAGAAGGTCTTCATTACACC-3′ and R5 downstream 5′-CCTGTGCCTCTTCTTCTCATTTCG-3′ primers, flanking the 32-bp deletion, were used to generate wild-type and deleted fragments of 189 and 157 bp, respectively. Each PCR mixture contained 20 pmol of each primer, 0.25 mM dNTP, 2 mM MgCl2, and 0.5 U of Taq polymerase in 1× PCR buffer (Perkin-Elmer, Norwalk, CT). PCR amplification consisted of five cycles of 94°C for 60 s, 55°C for 60 s, 72°C for 90 s, followed by 35 cycles of 94°C for 30 s, 60°C for 30 s, and 72°C for 45 s. PCR products were resolved on a 2% agarose gel.
PCR and restriction analysis for detection of the CD45 exon A (C77G) mutation
Genomic DNA (100 ng) was amplified by PCR using forward (5′-GACTACAGCAAAGATGCCCAGTG-3′) and reverse (5′-GGGATACTTGGGTGGAAGTA-3′) primers on either side of the C77G transversion amplifying a fragment of 155 bp. The PCR conditions included a 4-min incubation at 94°C followed by 30 reaction cycles (1 min at 94°C, 1 min at 55°C, and 4 min at 72°C) and a final 16-min extension at 72°C. The C77G transversion introduces a new restriction site for MspI (New England Biolabs, Beverly, MA), which cleaves the mutant PCR product into two fragments of 72 and 83 bp. The presence of an undigested band of 155 bp indicates the presence of the wild-type allele. The PCR and digestion products were analyzed on a VisiGel Separation Matrix (Stratagene, La Jolla, CA).
Flow cytometric analysis
PBMC were stimulated for 10 days with 1 μg PHA-P (Sigma, St. Louis, MO) per 106 cells/ml in RPMI 1640/10% FCS. PBMC pre- and post-PHA stimulation were stained with APC-conjugated CD3 mAb (PharMingen, San Diego, CA) along with FITC-conjugated CD45RA (Sigma) and PE-conjugated CD45R0 (Dako, Glostrup, Denmark) mAbs in a single step at 4°C for 20 min and washed with PBS containing 1% BSA. Isotype-matched mAbs were used as controls. Expression of CD45RA and CD45R0 on CD3+ cells was examined. Ten thousand events per sample were collected on a FACScan flow cytometer (Becton Dickinson, Mountain View, CA) and analyzed using CellQuest software.
Statistical analysis
The Fisher exact test was used to analyze the association of the CCR5del32 and CD45 exon A (C77G) mutations.
Results
Determination of the CD45 exon A (C77G) status of thymi showing abnormal CD45 splicing
A C to G transversion at position 77 in exon A in the gene encoding CD45 has been shown to be responsible for the retention of CD45RA on T cells in individuals with variant CD45RA splicing (17, 18). The transversion changes the sequence of exon A from CCCG to CCGG, which can be detected by restriction digestion with MspI that cuts only the mutant allele into two bands of 72 and 83 bp. The presence of an undigested band of 155 bp indicates the presence of the wild-type allele (Fig. 1⇓).
Identification of the CD45 exon A (C77G) mutation. PCR analysis for detection of exon A (C77G) was performed on genomic DNA with primers on either side of the mutation amplifying a fragment of 155 bp. The C77G transition introduces a new restriction site for MspI, which cleaves the mutant PCR product into two fragments of 72 and 83 bp (lane 3). The presence of an undigested band of 155 bp indicates the presence of the wild-type allele.
Among the 10 thymi tested that showed abnormal splicing of CD45, all carried the exon A (C77G) transversion (Table I⇓). Of these samples, five were CCR5del32 heterozygotes, one was CCR5del32 homozygous, and the remaining four had wild-type CCR5. None of the 10 thymi with normal CD45 expression was exon A (C77G) positive, irrespective of their CCR5del32 genotype. Among these thymic samples therefore, the only cause of abnormal CD45 splicing appears to be the C77G transversion.
CD45 exon A (C77G) and CCR5del32 genotype in thymia
Analysis of linkage of CD45 exon A (C77G) mutation with the CCR5del32 mutation in peripheral blood samples
To examine the association between the CD45 exon A(C77G) and CCR5del32 mutations in normal subject peripheral blood samples, we analyzed 10 homozygous, 101 heterozygous, and 116 wild-type CCR5del32 peripheral blood samples for the presence of the CD45 exon A (C77G) transversion.
Using PCR and MspI digestion, we identified 2 individuals with the CD45 exon A (C77G) mutation of the 116 CCR5del32 wild-type samples, 2 of the 101 heterozygous, and 0 of the 10 samples homozygous for the CCR5del32 allele (Table II⇓). The presence of the CD45 exon A (C77G) mutant allele was also confirmed by sequencing of all positive samples.
Frequency of CD45 exon A (C77G) mutation in healthy individual peripheral blood with CCR5del32
The CD45 exon A (C77G) allele was found in 1.8% of those with the CCR5del32 allele, similar to the frequency (1.7%) in the CCR5del32 wild-type controls. Using Fisher’s exact test to analyze the association between the C77G and CCR5del32 mutations, no statistically significant association between the C77G and CCR5del32 mutations was demonstrated (p = 1).
Following the method used to analyze the earlier thymic data (22), we next performed three pairwise comparisons of the homozygous and heterozygous CCR5del32 genotypes with the wild-type CCR5 genotype (Table III⇓). However, in contrast to the finding of Liao et al. (22), the p values for the pairwise comparisons of the homozygous and heterozygous CCR5del32 genotypes with the exon A (C77G) mutation show no statistically significant association. Taken together, our data suggest that there is no association of CD45 exon A (C77G) and CCR5del32 mutations in these DNA samples from PBMC of healthy individuals.
The association between CD45 the exon A (C77G) and CCR5del32 mutations in healthy individual peripheral blood
Flow cytometric analysis of variant CD45RA splicing
We next examined the expression of CD45 isoforms in PBMC by flow cytometry to eliminate the possibility that abnormal splicing can be caused by events distinct from the previously described exon A (C77G) transversion. Of the 227 samples genotyped above for CCR5del32 and exon A (C77G) mutations, 135 were available as PBMC. The pattern of CD45 splicing was examined in these samples by flow cytometric analysis pre- and post-T cell activation.
One hundred thirty-three of the 135 samples showed the normal pattern of CD45 isoform expression characterized by the presence of single CD45RA+ as well as single CD45RO+ T lymphocytes (Fig. 2⇓A). After 10 days of stimulation with PHA, >90% of the T cells lose expression of CD45RA and gain the expression of CD45RO associated with the acquisition of the effector/memory function.
Expression of CD45 isoforms in human peripheral T cells pre- and poststimulation. PBMC were stimulated with 1 μg/ml PHA and on days 0 and 10 stained with isoform-specific CD45RO-PE and CD45RA-FITC along with CD3-APC Abs. Analysis was performed on gated CD3+ cells. The normal pattern of CD45 splicing is characterized by loss of CD45RA and gain in expression of CD45RO after activation (A and B). Variant CD45 splicing can be identified by the absence of the CD45RA-negative population and the T cells remain CD45RA/RO double positive after activation (C and D). CCR5del32 heterozygous samples, lacking the exon A (C77G) mutation, exhibit a normal CD45 pattern of CD45 splicing (E and F).
Two of 135 samples showed a variant CD45 splicing pattern. As shown in Fig. 2⇑C, the variant phenotype can be identified by the absence of the single CD45RO+ T cell population and even after 10 days of stimulation the samples with variant CD45 splicing still did not convert to a single CD45RO+ state (Fig. 2⇑D). Both samples identified by flow cytometry with variant CD45 splicing also had the exon A (C77G) mutation and were negative for the CCR5del32 allele. Furthermore, among the 135 samples analyzed, all of the 19 heterozygous CCR5del32 samples showed the characteristic normal pattern of CD45 expression both phenotypically and genotypically (Fig. 2⇑, E and F).
Thus, all of the samples identified phenotypically with abnormal CD45 splicing exhibited the exon A (C77G) mutation, suggesting that the exon A (C77G) transversion is the only common cause for variant CD45 splicing.
Discussion
At present, only one common polymorphism has been described at the CD45 locus, a C to G transversion at position 77 in exon A, which does not change the protein sequence of the CD45RA isoform. We and others have shown previously that this CD45 variant results in the continuous expression of isoform A on activated and memory T cells through the defective splicing of this exon (17, 18).
Recently, a similar abnormal CD45RA splicing pattern in the thymus has been associated with the CCR5del32 mutation (22). Individuals who are homozygous for CCR5del32 are relatively resistant to HIV infection, but heterozygotes are not (25), although progression to AIDS is delayed (26). Furthermore, no difference in susceptibility to HIV infection in vitro was shown between thymus samples with the normal and variant CD45 splicing (22). However, given the similarity between the defect of activation-induced CD45RA down-regulation in thymocytes (22) to that of the previously described lack of CD45RA down-regulation in peripheral T cells (15, 16), we analyzed the association between abnormal CD45 splicing in the thymus and exon A (C77G) mutation. Our data clearly indicate that exon A (C77G) transversion is the cause of abnormal splicing in all of the thymic samples we have examined. It is interesting that there are, nonetheless, subtle differences in the effect of this mutation on splicing in the thymus compared with peripheral blood. Thus, Liao et al. (22) showed a block to down-regulation of both 220- and 200-kDa isoforms whereas in peripheral T cells only the 205-kDa isoform is affected (16).
We have also examined the association between the CD45 exon A (C77G) and CCR5del32 mutations in 227 DNA samples from healthy individual PBMC and find that the frequency of the CD45 exon A (C77G) transversion is no different in the 111 CCR5del32 allele carriers (1.8%) compared with the frequency in the CCR5del32 wild-type controls (1.7%). Therefore, our results suggest that there is no statistically significant association between the CD45 exon A (C77G) and CCR5del32 mutations in these samples from healthy donors.
There is a discrepancy between our analysis of PBMC samples and the earlier data from 394 thymus samples, which clearly showed an association between abnormal CD45 splicing, now shown to be due to the C77G transversion and the CCR5del32 mutation. We are not able at this time to account for this difference but one likely contributing cause is that whereas the peripheral blood samples were from healthy individuals, the thymic samples were not. Ninety-four and one-half percent of the thymus samples and all those exhibiting C77G were from patients with congenital heart disease, though the link among CD45 exon A (C77G), CCR5del32, and heart disease remains to be determined. However, it is interesting that an association of immunodeficiency with cardiac defects has been reported (27). It is also interesting that in the thymic material no association of abnormal splicing with myasthenia gravis (5.4% of all thymic samples) was observed, although exon A (C77G) and abnormal splicing have been shown to be associated with multiple sclerosis (20). In addition, a mouse model has been proposed in which the inhibition of dimerization of CD45 caused lymphoproliferation and autoimmunity (28).
Extensive studies of the frequency of the exon A (C77G) transversion have not been performed so far. It has been previously reported that 8% of the individuals examined by flow cytometric analysis in the region of Hannover, Germany, exhibited the variant CD45RA expression pattern (15). Another study from Germany has shown a much lower frequency (<1%) (20). Using PCR and MspI digestion analysis, we have so far genotyped 227 samples and found the frequency of CD45 exon A (C77G) transversion to be 1.76%. Differences between frequencies of abnormal splicing seen in different populations could be due to the presence of another (non-C77G), but as yet unidentified mutation responsible for variant CD45RA splicing. However, this seems unlikely, since in total we have examined by flow cytometric analysis and genotyping for exon A (C77G) 200 samples (65 of these were from a separate study to the one detailed in Tables II⇑ and III⇑; E. Tchilian, D. Wallace, R. Dawes, N. Imamiti, C. Burton, F. Gotch, and P. Beverley, unpublished data). All of the samples showing abnormal CD45 splicing identified by flow cytometry exhibited the exon A (C77G) transversion. However, it is noteworthy that in the study of multiple sclerosis and abnormal CD45 splicing, one family was indeed found to exhibit abnormal CD45 splicing, but not the exon A(C77G) transversion (20). Interestingly, also in the study of multiple sclerosis, differences in the CD45 exon A (C77G)-multiple sclerosis association were observed in North American compared with German study populations. Obviously ethnic differences could contribute to the differences we have found between North American thymus samples and the blood samples we have assayed. More extensive studies of the frequency and distribution of the CD45 exon A(C77G) mutation in normal populations will provide a crucial comparison for studies of disease association.
There are no published data indicating that the individuals with variant CD45 splicing have abnormal immunological functions but studies of transgenic mice support the idea that expression of only a high molecular mass isoform compromises immune function (29). The association of exon A (C77G) with multiple sclerosis and exon A (C77G) and CCR5del32 with congenital heart disease strongly suggest however that this mutation may be associated with a higher risk of some diseases. Despite the finding of one family with abnormal CD45 splicing lacking the C77G transversion (20), our studies indicate that this mutation is a common cause of abnormal splicing of CD45 and that the disease associations of this mutation warrant further study.
Acknowledgments
We thank Drs. Derek Macallan and Anna Vyakarnam for help with donor and patient samples and Ritu Dawes for expert technical assistance.
Footnotes
-
↵1 This work was supported by Wellcome Trust Grant 058700 (to N.I.). This is paper number 31 of the Edward Jenner Institute for Vaccine Research.
-
↵2 Address correspondence and reprint requests to Dr. Tchilian, Edward Jenner Institute for Vaccine Research, Compton, Berkshire RG20 7NN, U.K. E-mail address: elma.tchilian{at}jenner.ac.uk
- Received February 2, 2001.
- Accepted March 6, 2001.
- Copyright © 2001 by The American Association of Immunologists
References
In this issue
