Organization of the Ly-5 gene

Differential lineage-specific regulation of murine CD45 transcription by Oct-1 and PU.1

Although it has been established that CD45 expression is regulated at the transcriptional level, neither the regulatory elements that are responsible for its unique expression pattern nor the relevance of its three distinct transcriptional start sites (P1a, P1b, and P2) has been fully characterized. We studied the contribution of the three start sites to CD45 mRNA production in haematopoietic cell lines and primary haematopoietic cells. In myeloid and lymphoid cells and cell lines most CD45 transcripts originate from P1b with the exception of the thymoma-derived T cell line EL4, in which approximately 90% of CD45 transcripts originate from P1a. The degree of contribution of P1a is highest in lymphoid cells and increases in T cells following mitogen stimulation. In vitro evaluation of sequence upstream of the start sites shows that the P2 start site is sufficient for CD45 expression in lymphoid but not in myeloid cells, confirms the presence of a PU.1-binding site essential for myeloid expression of CD45, and reveals an Octamer-binding site that interacts with both Oct-1 and Oct-2 and activates CD45 transcription in lymphoid and myeloid cells. These findings are the first evidence that Octamer-binding factors are involved in the control of CD45 expression.

Genomic organization of the channel catfish CD45 functional gene and CD45 pseudogenes

CD45 is a transmembrane protein tyrosine phosphatase, which in mammals plays an important role in T and B cell receptor and cytokine signaling. Recently, a catfish cDNA was shown to contain all characteristic CD45 features: an alternatively spliced amino-terminus, a cysteine-rich region, three fibronectin domains, a transmembrane region, and two phosphotyrosine phosphatase domains. However, analyses of CD45 cDNAs from various catfish lymphoid cell lines demonstrated that catfish CD45 is unique in that it contains a large number of alternatively spliced exons. Sequence analyses of cDNAs derived from the catfish clonal B cell line 3B11 indicated that this cell line expresses up to 13 alternatively spliced exons. Furthermore, sequence similarity among the alternatively spliced exons suggested duplication events. To establish the exact number and organization of alternatively spliced exons, a bacterial artificial chromosome library was screened, and the catfish functional CD45 gene plus six CD45 pseudogenes were sequenced. The catfish functional CD45 gene spans 37 kb and contains 49 exons. In comparison, the human and pufferfish CD45 genes consist of 34 and 30 exons, respectively. This difference in the otherwise structurally conserved catfish gene is due to the presence of 18 alternatively spliced exons that were likely derived through several duplication events. In addition, duplication events were also likely involved in generating the six pseudogenes, truncated at the 3' ends. A similarly 3' truncated CD45 pseudogene is also present in the pufferfish genome, suggesting that this specific CD45 gene duplication occurred before catfish and pufferfish diverged (approximately 400 million years ago).

Organization and expression of thirteen alternatively spliced exons in catfish CD45 homologs

CD45, also known as LCA, is a transmembrane protein tyrosine phosphatase encoded by the PTPRC gene. In mammals, it plays an important role in T and B cell receptor and cytokine signaling by maintaining receptor associated kinases in an active state. A prominent CD45 feature is alternative splicing of exons encoding the N-terminus, resulting in the generation of several isoforms. The expression of isoforms is tightly regulated and dependent on the developmental/activation state of the lymphocyte. Nevertheless, the significance of these multiple isoforms in mammals is poorly understood. In this study, the channel catfish CD45 homolog was sequenced and found to be similar to CD45 of other species. However, unlike mammalian CD45, it appears that up to 13 exons are used in producing multiple alternatively spliced CD45 variants in catfish cells. These 13 alternatively spliced exons variably encode for O-linked glycosylation sites. Several of the exons are identical or very similar, suggesting gene duplication of a block of four exons. As demonstrated by RT-PCR, many of the alternatively spliced forms of catfish CD45 are differentially expressed in lymphoid cell lines with B cells expressing larger isoforms than do T cells. Furthermore, immunoprecipitation experiments utilizing anti-catfish CD45 mAbs substantiated that different size CD45 isoforms are expressed at the protein level on catfish T and B cells.

A CD45 minigene restores regulated isoform expression and immune function in CD45-deficient mice: therapeutic implications for human CD45-null severe combined immunodeficiency

Transgenic mice have been generated that carry a CD45 minigene under control of the human leukocyte function-associated antigen (LFA-1, CD11a) promoter. CD45-null mice carrying the transgene exhibit the lymphocyte lineage-specific isoform expression patterns of wild-type mice. Furthermore, these mice have normal thymocyte development and peripheral T-cell numbers. The proliferative ability of T cells in response to mitogens and antigen also is regained, as is B-cell responsiveness to anti-IgM. The antibody response to antigen is also restored and is similar to that of normal mice. Therefore, introduction of a functional CD45 minigene is sufficient to overcome the principal severe combined immunodeficiency (SCID)-associated defects and represents a potential route to a gene therapy for human CD45-deficent SCID.

PU.1 is a lineage-specific regulator of tyrosine phosphatase CD45

The hematopoietic cell-specific ets family transcription factor PU.1 regulates many lymphoid and myeloid genes. We have determined that PU.1 is critical for lineage-specific expression of the tyrosine phosphatase CD45. CD45 is expressed exclusively in hematopoietic cells at all stages of development, except for mature red cells and platelets. Although CD45 is normally expressed in all leukocyte lineages, it is critically regulated by PU.1 only in myeloid cells. Whereas myeloid cells from PU.1 null mice failed to express CD45, lymphoid cells were CD45(+) by flow cytometry. Additionally, mRNA for CD45 was absent from PU.1-deficient myeloid cells. To understand the molecular basis for these observations, we characterized a transcriptional regulatory region of the murine CD45 gene containing exons 1a, 1b, and 2. Distinct transcriptional initiation sites for CD45 were demonstrated in T and B cells versus myeloid cells. A transcriptional initiation site in exon 1b (P1b) was principally utilized by myeloid cells. A PU.1 binding site was identified upstream of exon 1b by sequence analysis and DNA binding assays. Using this region of the CD45 locus we demonstrated that PU.1 directly transactivated reporter gene expression. Finally, retrovirus-mediated restoration of PU.1 expression to PU.1-deficient myeloid cells resulted in expression of cell surface CD45 and restored phosphatase activity, confirming the role of PU.1 in the positive regulation of this well known signaling molecule. We conclude that CD45 is regulated differentially in myeloid and lymphoid cells and that sequences critical to direct myeloid expression include a PU.1 binding site upstream of the P1b transcriptional initiation site.

Structural and functional analysis of the human CD45 gene (PTPRC) upstream region: evidence for a functional promoter within the first intron of the gene

Expression of the leucocyte common antigen (CD45) in mammals is restricted to the nucleated lineages of haematopoietic cells. It appears in early progenitors in the bone marrow and is expressed at the surface of these cells throughout their differentiation. However, at least in T cells, the pattern of expression switches between different isoforms during the successive stages of differentiation in the thymus and after activation in the periphery. In order to understand the mechanisms controlling the transcription of the human CD45 gene, 2.7 kbp of the 5'-flanking region were sequenced and analysed for their ability to direct expression of a reporter gene. The only region with promoter activity was localized within the first intron of the gene. This promoter shows no tissue specificity but could be enhanced by a heterologous enhancer. Mobility shift assays showed complex but specific protein binding. The sequence in this region lacks similarity with known promoters or initiators but is highly conserved in evolution. No transcription initiation could be detected within or downstream of this region, suggesting that this might be a new type of RNA polymerase II promoter able to drive transcription from an upstream sequence. An additional exon was also found upstream of exon 1. The two exons 1 (1a and 1b) are mutually exclusive and both are spliced to exon 2. This makes the structure of the 5' region of the human CD45 gene identical to its mouse homologue.

Association of transcriptionally silent genes with Ikaros complexes at centromeric heterochromatin

Ikaros proteins are required for normal T, B, and NK cell development and are postulated to activate lymphocyte-specific gene expression. Here we examined Ikaros distribution in the nucleus of B lymphocytes using confocal microscopy and a novel immunofluorescence in situ hybridization (immuno-FISH) approach. Unexpectedly, Ikaros localized to discrete heterochromatin-containing foci in interphase nuclei, which comprise clusters of centromeric DNA as defined by gamma-satellite sequences and the abundance of heterochromatin protein-1 (HP-1). Using locus-specific probes for CD2, CD4, CD8alpha, CD19, CD45, and lambda5 genes, we show that transcriptionally inactive but not transcriptionally active genes associate with Ikaros-heterochromatin foci. These findings support a model of organization of the nucleus in which repressed genes are selectively recruited into centromeric domains.

Alternative splicing generates four different forms of a non-transmembrane protein tyrosine phosphatase mRNA

PTP-S is a widely expressed non-transmembrane protein tyrosine phosphatase (PTPase), which binds to DNA in vitro. The cellular PTP-S gene product is present mainly in the nucleus in association with chromatin. cDNAs related to PTP-S have been described from human and mouse cells. To establish the origin of molecular diversity in these cDNAs, genomic clones of rat PTP-S were isolated that span over 40 kb of the gene and contain 7 axons. The exon-intron splice sites in the catalytic domain are conserved between PTP-S and human PTP1B. Sequences specific to and homologous to human T-cell PTPase (TC-PTP) were found in the genomic clones of PTP-S, which are expressed in rat cells, as determined by using a specific probe and Northern blot analysis. Analysis of RNA from different rat tissues by reverse transcription-polymerase chain reaction (RT-PCR) showed the presence of four different forms of PTP-S mRNA (named PTP-S1, PTP-S2, PTP-S3, and PTP-S4). PTP-S1 is same as PTP-S reported previously by us. PTP-S2, which is the major form, differs from PTP-S1 in having additional 19 amino acids corresponding to exon E1. PTP-S4 is similar to human T-cell phosphatase. PTP-S3 differs from PTP-S4 in having a deletion of 19 amino acids corresponding to exon E1. Our results suggest that four different forms of PTP-S mRNA arise from a single gene by differential splicing. Two of these forms, PTP-S1 and PTP-S3, were not found in human cells, possibly due to the loss of an internal splice acceptor site in one of the exons, suggesting the occurrence of species-specific splicing in this gene.

Expression pattern of CD45 RA/RO isoformic antigens in T-lineage neoplasms

The expression of CD45 RA/RO antigen was investigated in neoplasms including cases expressing CD7 antigen as the sole pan-T antigen (n = 8), T-lineage acute lymphoblastic leukemia (ALL)/lymphoblastic lymphoma (LBL) at various stages of differentiation (n = 32), peripheral stage T-lineage leukemia (n = 10) and adult T-cell leukemia (ATL) (n = 14). The p56lck gene expression was also investigated in selected cases. The expression pattern of CD45 RA/RO antigen was defined as of RA, mixed, or RO type. All but one CD7+ CD5- CD2- case were of the RA type. The CD7+ CD5+ CD2- prothymic stage included seven RA and one mixed type cases. One CD7+ CD5- CD2+ case was of the RA type, but the other was of the RO type. The CD7+ CD5+ CD2+ prothymic stage included three RA and four mixed type cases. All seven CD3- CD4+ CD8+ (double-positive) thymic cases were of the RO type. The CD3+ CD4+ CD8+ (triple-positive) stage included two RO and three mixed-type cases. One CD3+ CD4+ CD8- late thymic case was of the mixed type. The peripheral stage cases included five RA, three RO, and two mixed type cases. All ATL cases were of the RO type. The expression of p56lck gene in the prothymic stage was less marked than that in the thymic stage. On the basis of these results, the following sequence of pattern of the CD45 RA/RO antigen expression along with T-lineage differentiation was reconstructed: prothymic stage [RA and mixed type]-->double-positive thymic stage [RO type]-->triple-positive thymic stage [RO and mixed type]-->peripheral stage [RA, mixed, and RO type]. While one RO-type CD7+ CD5- CD2- and one RO-type CD7+ CD5- CD2+ cases were not in accord with this sequence, the pattern of CD45 RA/RO antigen expression in most of T-lineage neoplasms could be determined by the respective stage of differentiation. The poor expression of the p56lck gene by the prothymic blasts compared with the thymic blasts may be related to the expression pattern of the CD45 RA/RO molecules, which exhibits phosphatase activity. The consistent RO-type expression in the ATL cases may reflect the activated status of the neoplastic T cells due to the presence of the HTLV-I gene. Alternatively, the target cells for HTLV-I-induced neoplastic transformation may possible be of the RO type.

Transcriptional activity of the CD45 gene promoter in retroviral vector constructs

Both specific and housekeeping transcriptional control elements have been successfully used to generate retroviral vectors that express in a stable tissue specific manner. However, no such vectors have been developed specifically for expression in haematopoietic cells. This paper describes the construction of retroviral vectors incorporating sequences from the promoter region of the murine CD45 (leucocyte common antigen) gene and using the human alpha-L-iduronidase cDNA as a reporter gene. These vectors were tested for expression of the alpha-L-iduronidase gene in skin fibroblasts and a lymphoblastoid cell line which is permissive for CD45 expression, derived from patients with mucopolysaccharidosis type I (MPS I, alpha-L-iduronidase deficiency). The results obtained demonstrate that, in the context of a retroviral construct, the activity of the CD45 gene promoter in lymphocytes and fibroblasts parallels the activity of the retroviral long terminal repeat (LTR). These results suggest that additional transcriptional control elements from the CD45 gene, or other haematopoietic specific genes, will be necessary for the construction of a haematopoietic specific vector.

Identification and typing of members of the protein-tyrosine phosphatase gene family expressed in mouse brain

Protein-tyrosine phosphatases (PTPases) form a novel and important class of cell regulatory proteins. We evaluated the expression of PTPases in mouse brain by polymerase chain amplification of cDNA segments that encode the catalytic domains of these enzymes. Degenerate primer pairs devised on the basis of conserved protein motifs were used to generate a series of distinct PCR-derived clones. In this way, murine homologues of the human PTPases LRP, PTP beta, PTP delta, PTP epsilon and LAR were obtained. Corresponding regions in their catalytic domains were used to reveal the evolutionary relationships between all currently known mammalian PTPase protein family members. Phylogenetic reconstruction displayed considerable differences in mutation rates for closely related PTPases.

Characterization and expression of the human leukocyte-common antigen (CD45) gene contained in yeast artificial chromosomes

The leukocyte-common antigen (CD45) is a transmembrane protein tyrosine phosphatase expressed uniquely by cells of hematopoietic origin. There are multiple isoforms of CD45 that are generated by the variable use of three exons (exons 4-6). The use of the variable exons results in changes near the amino-terminus of the mature glycoprotein. The gene is located on chromosome 1 for both human and mouse in a region that is homologous between these two species. This conserved linkage group contains a number of genes of immunological interest, such as the genes for complement regulatory proteins and the FCG2 receptor. Yeast artificial chromosomes provide a vector system in which large fragments of foreign DNA can be isolated and are suited to long-range physical mapping. To this end, three yeast artificial chromosomes containing the human CD45 gene have been isolated and characterized. They overlap to span 475 kb, establishing the largest physical map for DNA within the conserved linkage group. The CD45 gene is entirely encoded within one yeast artificial chromosome clone as determined by mapping with cDNA probes. A mouse B cell line transfected with this YAC clone expressed the low-molecular-weight isoform of the protein into the cell surface. The size of the human CD45 gene was determined to be approximately 120 +/- 10 kb.

Analysis of Ly5 chromosome 1 position using allelic differences and recombinant inbred mice

Differences between the mouse Ly5a and Ly5b alleles can be distinguished on the basis of polymerase chain reaction (PCR)-restriction enzyme analysis and differential monoclonal antibody reactivities. To more precisely map the Ly5 gene on the mouse chromosome 1, analytical DNA and protein tests were performed on recombinant inbred strains of mice prepared from SJL/J (Ly5a) and BALB/cke (Ly5b) progenitor strains. Each recombinant inbred strain was characterized to determine whether it carried the Ly5a or Ly5b allele. Both assays, DNA-PCR and protein-immunofluorescence, yielded identical results for each strain examined. Placement of the Ly5 gene with respect to other characterized markers of mouse chromosome 1 for these recombinant inbred mouse strains shows a gene order of Idh-1:Ity:Pep3:[Ly5, Cfh].

Further data on the selective expression of Ly-5 isoforms

Ly-5 (CD45) glycoproteins of the mouse, expressed by all or most hematopoietic cell lineages and specified by a single Ly-5 gene, range in size from isoform T200 of T cells (the smallest), in which exons 4, 5, and 6 are not represented, to isoform B220 of B cells (the largest), in which all three of these optional exons are represented. The main purpose of the present study, utilizing the polymerase chain reaction (PCR), was to ascertain whether known isoforms of intermediate size are generated by single or dual usage of optional exons 4, 5, and 6. Transcripts representing all eight isoforms predictable from varied use of three exons were observed among a diverse panel of nine B-cell tumors in culture, but there was no evident concordance with known contrasting differential features that distinguish members of the B-cell tumor panel. No two B tumors exhibited the same variety of transcripts and the relative quantities of transcripts expressed varied greatly from tumor to tumor. Cloning of B-cell tumors did not alter their distinctive transcript patterns. Separation methods (sodium dodecyl sulfate polyacrylamide gel electrophoresis; SDS-PAGE) did not suffice to segregate all corresponding expressed isoforms but did establish that transcripts representing usage of a single optional exon and of two optional exons were actually translated, which supports a provisional inference that all eight isoforms exist. The considerable diversity of B-cell transcript phenotypes was not seen among seven T-cell leukemias, two cytolytic T-cell lines, and three Th 1 helper T-cell lines, all of which displayed a uniform phenotype comprising major expression of the T200 transcript (no optional exon) and minor expression of a transcript employing exon 5. However, a panel of five cloned Th2 T-cell lines, which represent a second and functionally different branch of the helper/inducer T-cell category, exhibited a characteristic transcript pattern which distinguished them from a panel of three Th1 T-cell lines. The major transcript in the Th2 lines was also T200, but the Th2 lines showed higher representation of transcripts containing optional exons. A single Th2 clone expressed an unusual transcript suggesting a potential isoform not compounded simply by varied inclusion of the three identified optional exons. After activation of the helper T-cell lines with concanavalin A (Con A), expression of transcripts containing optional exons appeared to decrease.

Structural features of Ly-5 glycoproteins of the mouse and counterparts in other mammals

The Ly-5 system of the mouse defines a set of transmembrane glycoprotein isoforms (T200, B220, etc) that hallmark various lineages and stages of hematopoietic differentiation. These isoforms are the products of a single Ly-5 gene comprising 34 exons, 32 of them (Exs-3-34) protein-coding and three (Exs-5-7) selectively represented in different isoforms (e.g., all three in isoform B220 but none in isoform T200). Probable structural features of Ly-5 glycoproteins, largely inferred from Ly-5 gene composition, are presented and compared with the rat L-CA and human LCA/T200 systems, which are phylogenetic counterparts of Ly-5 as an index of the extent and nature of structural conservation. The outer (N-terminal) region of the Ly-5 T200 isoform comprises three broadly similar domains (Exs-4, 8, 9) with salient features that jointly favor free interaction with the aqueous environment and are shared by the L-CA and human LCA/T200 systems despite an overall interspecies protein sequence similarity in this region of only about 50%. In the larger B220 isoform this region includes epitopes dictated by the selective exons Exs-5, 6, 7, these being more conserved than the shared exons Exs-4, 8, 9 and no doubt sustaining the differential functions of the respective isoforms. Comparison of the genomic sequences of Ex-5 in the Ly-5 and human systems suggests that a shift in splice donor site accounts for an extra 23 amino acids in the human Ex-5-coding domain, which is the only salient structural difference between the mouse Ly-5 and human systems. The inner extracellular region (Exs-10-16) includes subregions of high variability, but again there are shared salient interspecies similarities such as sites and numbers of Cys residues that imply a conserved, tightly-folded conformation, in contrast to the more open conformation predicted for the outer extracellular region. The transmembrane region (Ex-17) is highly conserved, as is the very large cytoplasmic region (Exs-17-34) which may interact with the plasma membrane but probably does not traverse it.