Gene structure, chromosomal localization, and protein sequence of mouse CD53 (Cd53): evidence that the transmembrane 4 superfamily arose by gene duplication

Tetraspanin CD53: an overlooked regulator of immune cell function

Tetraspanins are membrane organizing proteins that play a role in organizing the cell surface through the formation of subcellular domains consisting of tetraspanins and their partner proteins. These complexes are referred to as tetraspanin enriched microdomains (TEMs) or the tetraspanin web. The formation of TEMs allows for the regulation of a variety of cellular processes such as adhesion, migration, signaling, and cell fusion. Tetraspanin CD53 is a member of the tetraspanin superfamily expressed exclusively within the immune compartment. Amongst others, B cells, CD4+ T cells, CD8+ T cells, dendritic cells, macrophages, and natural killer cells have all been found to express high levels of this protein on their surface. Almost three decades ago it was reported that patients who lacked CD53 suffered from an increased susceptibility to pathogens resulting in the clinical manifestation of recurrent viral, bacterial, and fungal infections. This clearly suggests a vital and non-redundant role for CD53 in immune function. Yet, despite this striking finding, the specific functional roles of CD53 within the immune system have remained elusive. This review aims to provide a concise overview of the published literature concerning CD53 and reflect on the underappreciated role of this protein in immune cell regulation and function.

Down-regulation of CD53 expression in Epinephelus coioides under LPS, poly (I:C), and cytokine stimulation

Tetraspanins are a group of cell surface molecules involved in cell adhesion, motility, metastasis, signal transduction, and immune cell activation. Members of the tetraspanin family include CD9, CD37, CD63, CD53, and others. However, few tetraspanins have been investigated in teleosts. In this study, we obtained the open reading frame of CD53 cDNA from orange spotted grouper (Epinephelus coioices), an economically important fish. The predicted amino acid structure contains four membrane-spanning domains and a conserved CCG motif. The amino acid identity between human and grouper CD53 was only 38%; however, both CD53 proteins share the same structure. Quantitative real-time PCR revealed that mRNA is abundant in immune organs, including the head and trunk kidneys, spleen, thymus, gill, and blood. Immunochemistry and immunofluorescence analyses further revealed that CD53 was majorly expressed in the leukocytes of various organs. Finally, mRNA and protein expression for CD53 was down-regulated in fish treated with immune stimulators, including LPS, Poly (I:C), Vibrio, recombinant grouper IL-6, and CCL4. Our results indicate that the expression of CD53 may play important roles in pathogen invasion and inflammation reaction.

Defining regulatory elements in the human KAI1 (CD 82) metastasis suppressor gene

The human KAI1 metastasis suppressor gene encodes for a 267 amino acid plasma membrane glycoprotein, which has four transmembrane domains and one large and one small extracellular domain. Plasma membrane expression of KAI1 is downregulated during the progression of several cancers to a metastatic state, including prostate, lung, and pancreatic cancers. To elucidate the mechanisms for this downregulation, an understanding of its transcriptional regulation is critical. Therefore, a set of luciferase reporter gene plasmid constructs were generated containing various 5' flanking regions to the transcription initiation site, with or without the first exon and a portion of the first intron of the KAI1 gene. These constructs were transfected into a series of normal and malignant prostate cell lines as well as fibroblasts and the resultant luciferase activities used to define the regulatory elements for expression of the KAI1 gene. Using this approach, the minimal KAI1 promoter was found to be 0.5 kb in size and contains two regions. The first region includes the 197 bp 5' to the transcription initiation site and the second region includes the first exon and a portion of first intron (i.e., +1 to +351 bp). This combined region result in positively regulated strong, but not epithelial specific, luciferase expression. In addition to this minimal promoter element from -197 to +351 bp, there is a second regulatory element further upstream (i.e., -735 to -197 bp), which negatively regulates luciferase expression. Additional studies of Marreiros et al. (Gene 302:155, 2003) have documented that a third regulatory element even further 5' (i.e., -922 to -846 bp) encodes an enhancer for the KAI1 gene.

CD53, a thymocyte selection marker whose induction requires a lower affinity TCR-MHC interaction than CD69, but is up-regulated with slower kinetics

The molecular mechanisms that govern the survival, maturation and export of thymocytes are the subject of intense study, and candidates for involvement in these processes might be identified by their differential expression during thymocyte selection. One such molecule is the tetraspanin CD53, which is not expressed on most CD4(+)CD8(+) double-positive (DP) cells in the normal mouse. We have examined CD53 expression on DP from several class I- and class II-restricted TCR transgenic (Tg) mice, and have found a strong correlation between CD53 expression and positive selection. CD53 expression in DP was formally demonstrated to be dependent upon MHC recognition as evidenced by studying DP from MHC-deficient mice which totally lack expression of this molecule. This link between selection and CD53 expression was reminiscent of CD69, and indeed the majority of selected DP from normal mice that express CD53 also express CD69. We compared CD53 and CD69 induction in vitro using pre-selected thymocytes from TCR-Tg mice that were stimulated either with mAb against TCR or with antigen-presenting cells (APC) pulsed with peptides. The data shows that with either stimulus, CD69 is induced rapidly on the thymocyte surface with expression detected in as little as 2 h. CD53 induction is slower with maximal expression taking up to 20 h. We also stimulated pre-selected thymocytes from the OT-1 TCR-Tg strain with APC pulsed with peptides of varying affinities for the TCR. Here low-affinity peptides which induce CD69 expression poorly were able to induce significant levels of CD53 expression. These data demonstrate that the induction of CD53 and CD69 upon selection is not identical. Thus a combination of the CD69 and CD53 selection markers may be a powerful tool to isolate thymocytes that have either been very recently selected or have arisen from differing MHC--TCR affinity interactions during selection.

The molecular characterisation of mouse tspan-3

The Tetraspanin/Transmembrane-4 Superfamily of cell surface molecules is defined by their four highly conserved transmembrane domains and is found in a wide variety of species and cell types. A common function for these molecules has yet to be discovered, however their broad expression patterns and conservation over evolution suggests that they will have an important general function relevant to many cell lineages. Here we describe the cloning and characterisation of the murine homologue of a recently described member of this superfamily, tspan-3. Murine tspan-3 was remarkably similar to the human molecule showing 88% identity at the nucleic acid level and 98% homology on the amino acid level. Northern blot analyses of mouse tspan-3 show a very broad pattern of expression, with expression readily detected in most organs including neural and bone marrow derived tissues.

The molecular characterisation of a novel tetraspanin protein, TM4-B(1)

TM4-B is a novel member of the Tetraspanin superfamily and displays characteristics typical of the superfamily. It bears significant homology to other superfamily members and is most similar to Tspan-1. This molecule is broadly expressed in most human tissues and cell lines including neural and bone marrow derived tissues. TM4-B was mapped to the q34 region on human chromosome 9.

The Drosophila melanogaster 60A chromosomal division is extremely dense with functional genes: their sequences, genomic organization, and expression

We cloned and sequenced genomic DNA contigs spanning over 45 kb, surrounding the insertion site of the P-element that is responsible for the developmental defects in the ken and barbie (ken) mutant of Drosophila. This region harbors 10 functional transcription units, in addition to the already well-characterized TGFbeta-60A gene. These include the genes, undefined 1 (UD1), UD2, and UD3, each coding for proteins of unknown function, the ken gene encoding a new Krüppel-like putative transcription factor, the fly homologues of the mammalian mitochondrial trifunctional enzyme (thiolase), and the TAR DNA-binding protein-43 (TBPH), the first nonvertebrate member of the transmembrane 4 superfamily (TM4SF) gene, a new homeodomain gene, and a gene coding for a putative nuclear binding protein (PNBP) that is homologous to maleless, and a Copia-like element. UD3 exists in an intron of the maleless homologue, yet is expressed independent of it. The UD1 and TM4SF genes orient in a tail-to-tail manner with their 3' untranslated region sequences overlapping over 44 nucleotides. Thus the partial overlap and intraintronic organization permitted dense packing of the functional genes within a short segment of the genome.

Characterisation of the mouse homologue of CD151 (PETA-3/SFA-1); genomic structure, chromosomal localisation and identification of 2 novel splice forms

CD151 (PETA-3/SFA-1) is a member of the Transmembrane 4 Superfamily (TM4SF) of cell-surface proteins and, like other TM4SF members CD9 and CD63, is expressed by platelets, megakaryocytes and endothelial cells. The precise function of CD151 is unknown however complexes containing CD151 and beta1 integrins have been isolated from a number of cell systems and studies using anti-CD151 monoclonal antibodies have suggested a role in transmembrane signalling and cell adhesion. To further investigate the function of CD151 we have determined the genomic organisation of mouse CD151 (Cd151). Cd151 spans 4 kb and contains six coding region exons. Using 5' RACE and reverse transcriptase-polymerase chain reaction (RT-PCR) we have identified three 5' UTR splice variants which arise through alternate splicing of three exons. Splice variants were detected in a number of mouse tissues by RT-PCR. Analysis of the Cd151 genomic structure reveals a high degree of structural conservation with other TM4SF molecules supporting the theory that family members have arisen from gene duplication of a common ancestral gene. Cd151 maps to chromosome 7, in close linkage to the p gene (OCA2 in humans), and helps define a boundary in the human/mouse homology relationships.

Molecular cloning and expression of mouse homologue of SFA-1/PETA-3 (CD151), a member of the transmembrane 4 superfamily

We have isolated mouse SFA-1 (CD151) cDNA from a mouse myelomonocytic leukemia cell line, WEHI-3. Mouse CD151 mRNA comprises approximately 1.8 kb, has 253 amino acid residues with 93% identity to human CD151 and contains four putative transmembrane domains, a number of cysteine residues and one potential N-glycosylation site located at a site corresponding to that in human CD151. Mouse CD151 gene expression was observed in many cell types, but was either absent or present at a low level in brain and lymphoid cells and tissues, including thymus and spleen. The expression patterns of mouse and human CD151 genes are similar.

Anti-CD53 monoclonal antibody induced LFA-1/ICAM-1-dependent and -independent lymphocyte homotypic cell aggregation

CD53 is a pan-leukocyte glycoprotein and belongs to a member of the tetraspan family of cell membrane proteins. The predicted structure and functional characteristics of CD53 suggest that it may play important roles in transmembrane signaling, but its roles in cell adhesion have not been clarified. The present study shows that anti-CD53 monoclonal antibody (mAb), HI29 induced homotypic cell aggregation of lymphoid cell lines including a B cell line from a patient with leukocyte adhesion deficiency syndrome (LAD). The homotypic cell aggregation was blocked by another anti-CD53 mAb, MEM53, in all the examined cell lines and by anti-LFA-1 (CD11a/CD18) or anti-ICAM-1 (CD54) mAbs in the cell lines except for the LAD line, but it was not blocked by anti-CD44 or anti-CD49d mAb. The induced homotypic cell aggregation was energy-dependent. These findings suggest that CD53 relates to LFA-1/ICAM-1-dependent and -independent pathways of homotypic cell aggregation of lymphocytes and that it plays an important role in lymphocyte activation and cell adhesion.

Genomic organization of the human KAI1 metastasis-suppressor gene

Decreased expression of the human KAI1 metastasis-suppressor gene is involved in the progression of human prostatic cancer and possibly lung and breast cancer. To evaluate the frequency of mutation and allelic loss during the progression of human cancer, as well as to determine the regulatory mechanism for the expression of the KAI1 gene in normal and cancerous tissues, we characterized the 5'-promoter region, exon/intron organization, and transcription initiation site of the human KAI1 gene. About 80 kb of DNA was identified as the human KAI1 gene, which contains 8 kb of 5'-region, 10 exons, 9 introns, and 8 kb of DNA following exon 10. The coding region starts in exon 3 and ends in exon 10. The size of intron 1 is 29 kb, which almost equals the sizes of all other introns combined. A CpG island is present in the 5'-promoter region and extends to exon 1 and intron 1. The promoter region has no TATA or CCAAT box but has many putative binding motifs for various transcription factors, including nine Sp1 sites and five AP2 sites. These results suggest a diverse regulatory mechanism for the expression of the KAI1 gene in human tissues. The transcription initiation site of the KAI1 gene is located 181 bp upstream of the first nucleotide of the translation initiation codon. Comparisons of gene structures between KAI1 and seven other members of the transmembrane 4 superfamily revealed that the splicing sites relative to the different structural domains of the predicted proteins are well conserved, suggesting that these genes are evolutionarily related and that they arose through gene duplication and divergent evolution.

Chromosomal localization of the human urothelial "tetraspan" gene, UPK1B, to 3q13.3-q21 and detection of a TaqI polymorphism

The TI1/UPK1b gene codes for a protein of the "tetraspan" family and is expressed as a differentiation product of the mammalian urothelium. A partial genomic clone of the human homologue of the TI1/UPK1b gene was isolated and used as probe to localize the human gene to chromosome 3q13.3-q21 by in situ hybridization. Using the same probe, a TaqI restriction fragment length polymorphism, with 29% heterozygosity, was identified by Southern analysis.

Genomic structure and promoter analysis of the gene encoding MM3, a member of transmembrane 4 superfamily

We have isolated genomic clones encoding hamster MM3, a member of transmembrane 4 superfamily (TM4SF). Nucleotide (nt) sequence analysis revealed that it is composed of 5 exons spanning about 8 kb. The exon-intron organization of the MM3 gene was quite different from those of other TM4SF members. We also identified its transcription start points (tsp) and the promoter region. Deletion analysis of the promoter revealed that about 160-bp region containing TATA-box, CAAT-box and GC-box was necessary for efficient transcription in cultured cells.

Characterisation of mouse CD37: cDNA and genomic cloning

The leukocyte surface antigen CD37 is a member of the transmembrane 4 superfamily (TM4SF) of glycoproteins which are predicted to span the lipid bilayer four times. The protein sequence and gene structure of mouse CD37 (Cd37) have been deduced through the isolation of cDNA and genomic clones. The Cd37 gene produces a major mRNA transcript of 1.2 kb that is restricted to cells of lymphoid and myeloid origin. Mouse CD37 is a glycoprotein of 281 amino acids in length, encoded by eight exons that span approximately 5.2 kb. CD37 is highly conserved between species, the mouse sequence sharing amino acid identities of 98% and 79% with rat and human, respectively. Cd37 shows a striking similarity in genomic organisation to other members of the TM4SF, which is consistent with the theory that this superfamily has evolved by gene duplication and divergence from a common ancestral gene.

Selective interactions of UPIa and UPIb, two members of the transmembrane 4 superfamily, with distinct single transmembrane-domained proteins in differentiated urothelial cells

The transmembrane 4 (TM4) superfamily contains many important leukocyte differentiation-related surface proteins including CD9, CD37, CD53, and CD81; tumor-associated antigens including CD63/ME491, CO-029, and SAS; and a newly identified metastasis suppressor gene R2. Relatively little is known, however, about the structure and aggregation state of these four transmembrane-domained proteins. The asymmetrical unit membrane (AUM), believed to play a major role in stabilizing the apical surface of mammalian urothelium thus preventing it from rupturing during bladder distention, contains two TM4 members, the uroplakins (UPs) Ia and Ib. In association with two other (single transmembrane-domained) membrane proteins, UPII and UPIII, UPIa and UPIb form 16-nm particles that naturally form two-dimensional crystalline arrays, thus providing unique opportunities for studying membrane structure and function. To better understand how these proteins interact to form the 16-nm particles, we analyzed their nearest neighbor relationship by chemical cross-linking. We show here that UPIa and UPIb, which share 39% of their amino acid sequence, are cross-linked to UPII and UPIII, respectively. We also show that UPIa has a propensity to oligomerize, forming complexes that are stable in SDS, and that UPII can be readily cross-linked to form homodimers. The formation of UPII homodimers is sensitive, however, to octyl glucoside that can solubilize the AUMs. These data suggest that there exist two types of 16-nm AUM particles that contain UPIa/UPII or UPIb/UPIII, and support a model in which the UPIa and UPII occupy the inner and outer domains, respectively, of the UPIa/UPII particle. This model can account for the apparent "redundancy" of the uroplakins, as the structurally related UPIa and UPIb, by interacting with different partners, may play different roles in AUM formation. The model also suggests that AUM plaques with different uroplakin compositions may differ in their assembly, and in their abilities to interact with an underlying cytoskeleton. Our data indicate that two closely related TM4 proteins, UPIa and UPIb, can be present in the same cell, interacting with distinct partners. AUM thus provides an excellent model system for studying the targeting, processing, and assembly of TM4 proteins.

Antibodies against human CD63 activate transfected rat basophilic leukemia (RBL-2H3) cells

CD63 is a widely expressed glycoprotein member of the transmembrane 4 superfamily (TM4SF) that is present on activated platelets, monocytes and macrophages and many non-lymphoid cells. It has been proposed that CD63 and other members of the TM4SF couple to intracellular signal transduction pathways and may have a role in cellular adhesion, proliferation and activation. We have investigated the functions of human CD63 by expression in the rat basophilic leukemia cell line, RBL-2H3, which has previously been reported to respond to antibodies against the rat homolog of CD63. Using a panel of antibodies against human CD63 we have shown that high levels of granular secretion from transfected RBL cells can be stimulated by some, but not all, of the antibodies. The specificity of this response suggests that these activating antibodies may be mimicking a natural ligand for CD63. The secretory response to crosslinking of the high affinity IgE receptor and also that to non-receptor stimuli (phorbol ester and calcium ionophore) is inhibited by an antibody that appears to recognise both human and rat homologs of CD63. These results suggest that stimulus-secretion coupling can occur through human CD63 and that RBL cells transfected with this protein will constitute a valuable tool in elucidating its function.

Characterization of mouse CD53: epitope mapping, cellular distribution and induction by T cell receptor engagement during repertoire selection

The pan-leukocyte antigen CD53 is a member of the poorly understood transmembrane 4 superfamily (TM4SF) of cell membrane glycoproteins. CD53 is proposed to play a role in thymopoiesis, since rat CD53 is expressed on immature CD4-8-thymocytes and the functionally mature single-positive subset, but is largely absent from the intermediate CD4+8+ cells. We have characterized CD53 in the mouse through the production of two new monoclonal antibodies, MRC OX-79 and OX-80, which were raised against the RAW 264 cell line and screened on recombinant CD53 fusion proteins. The epitopes recognized by both antibodies are dependent on disulfide bonding and map to the major extracellular region of CD53, requiring the presence of a single threonine residue at position 154. Mouse CD53 has a molecular mass of 35-45 kDa and is expressed on virtually all peripheral leukocytes, but not on cells outside the lymphoid or myeloid lineages. CD53 expression distinguishes subpopulations of thymocytes in the mouse and resembles the expression pattern of rat CD53. Amongst the immature CD4-8-thymocytes, mouse CD53 is clearly detectable on the earliest CD44high25- subset, but down-regulated on the later CD44high25+, CD44low25+ and CD44low25- stages. Also, the subsequent transient TcR-/low CD4-8+ cells and most CD4+8+ thymocytes express little or no CD53. This is consistent with the idea that cells which are committed to enter the selectable CD4+8+ compartment switch off CD53. The effect of T cell receptor (TcR) engagement on the re-expression of CD53 on CD4+8+ thymocytes was studied both ex vivo and in vitro using F5 mice, transgenic for the H-2b/influenza nucleoprotein-peptide-specific TcR, back-crossed onto an H-2q or H-2b background of RAG-2-deficient mice. CD4+8+ thymocytes from non-selecting H-2q F5 mice are CD53 negative, but in vitro stimulation through the TcR dramatically induces CD53 expression. In contrast, a fraction of CD4+8+ thymocytes from positively selecting H-2b F5 transgenic mice express CD53. Therefore TcR engagement by selecting major histocompatibility complex peptide complexes, or surrogate ligands, induces CD53 expression on otherwise CD53-negative, non-selected CD4+8+ thymocytes. Whether CD53 itself participates as a signaling molecule in further stages of thymic selection is still a matter of speculation.

Expression of cell interaction molecules by immature rat thymocytes during passage through the CD4+8+ compartment: developmental regulation and induction by T cell receptor engagement of CD2, CD5, CD28, CD11a, CD44 and CD53

Rat thymocytes of the T cell receptorlow (TcRlow) CD4+8+ subset which is the target of repertoire selection are heterogeneous with respect to expression of the cell interaction (CI) molecules CD2, CD5, CD11a/CD18 (LFA-1), CD28 and CD44. We show that this heterogeneity is due to the developmental regulation of these CI molecules during passage through the CD4+8+ compartment, and to up-regulation by TcR engagement. Thus, cohorts of CD4+8+ cells differentiating synchronously in vitro from their direct precursors, the immature CD4-8+ cells, were homogeneous with regard to CI molecule expression. Upon entry into the CD4+8+ compartment, they expressed relatively high levels of CD2 and CD44, and moderate levels of CD5, CD28 and CD11a. CD2, CD28 and CD44 were slightly down-regulated during the following 2 days, whereas CD5 slightly increased and CD11a remained constant. TcR stimulation using immobilized monoclonal antibodies resulted in rapid and dramatic up-regulation of CD2, CD5 and CD28 and, to a lesser extent, of CD11a and CD44. Finally CD53, a triggering structure absent from unstimulated CD4+8+ thymocytes was also rapidly induced by TcR stimulation. Inclusion of interleukin (IL)-2, IL-4, or IL-7 in this in vitro differentiation system did not affect the levels of CI molecules studied. Since the high levels of CI molecules induced by TcR-stimulation correspond to those found in vivo on TcRintermediate thymocytes known to be undergoing repertoire selection, these results suggest that upregulation of CI molecules by TcR engagement provides a mechanism by which thymocytes that have entered the selection process gain preferential access to further interactions with stromal and lymphoid cells in the thymus.

Genomic structure of SAS, a member of the transmembrane 4 superfamily amplified in human sarcomas

SAS is a recently identified member of the transmembrane 4 superfamily (TM4SF) that is frequently amplified in human sarcomas. To further its characterization and to confirm its classification, the genomic structure of the SAS gene was determined. The SAS gene covers approximately 3.2 kb of DNA. It contains six exons within its translated region, three of which are highly conserved in the TM4SF. 5' to the translation start site are two putative transcription start sites, two CCAAT consensus sequences, and potential binding sites for both Sp1 and ATF transcription factors. Comparison of SAS organization to human ME491, CD9, and CD53 and murine CD53 and TAPA-1 confirms that SAS is a member of this family of genes and is consistent with the theory that these genes arose through duplication and divergent evolution.

The ins and outs of the transmembrane 4 superfamily

The recently discovered transmembrane 4 superfamily comprises a group of cell-surface proteins that are characterized by the presence of four hydrophobic domains, which are presumed to be membrane spanning. At least seven of these molecules are expressed on leukocytes, and it seems likely that they mediate signal transduction events that play a role in the regulation of cell development, activation, growth and motility.

CD9 antigen is an accessory subunit of the VLA integrin complexes

The CD9 antigen is a cell surface glycoprotein of unknown function which belongs to the tetraspans family. We demonstrate here, by precipitation, Western blotting and co-capping experiments, that this molecule is associated with a large fraction of beta 1 integrins in two cell lines, the pre-B cell line NALM-6 and the megakaryocytic cell line HEL. In HEL cells, CD9 antigen is only associated with VLA-4. In contrast, in NALM-6 cells, CD9 antigen is associated with both VLA-4 and VLA-5. On the other hand, only the beta 1 chain is co-precipitated with the CD9 antigen in transfected L cells. These data show that the CD9 antigen is associated with the beta 1 chain rather than with a particular integrin. CD9 monoclonal antibodies (mAb) did not modify the binding of HEL and NALM-6 cells to fibronectin, laminin or collagen. The association of CD9 antigen to VLA integrins is strengthened by the fact that both CD9 and anti-VLA mAb induce aggregation of the two cell lines and inhibit their migration in Transwell chambers. Because the aggregating effect, but not the inhibition of migration, is observed in CEM or CD9-transfected CEM cells, these two effects are likely to be mediated by different mechanisms.

Uroplakins Ia and Ib, two major differentiation products of bladder epithelium, belong to a family of four transmembrane domain (4TM) proteins

The mammalian bladder epithelium elaborates, as a terminal differentiation product, a specialized plasma membrane called asymmetric unit membrane (AUM) which is believed to play a role in strengthening and stabilizing the urothelial apical surface through its interactions with an underlying cytoskeleton. Previous studies indicate that the outer leaflet of AUM is composed of crystalline patches of 12-nm protein particles, and that bovine AUMs contain three major proteins: the 27- to 28-kD uroplakin I, the 15-kD uroplakin II and the 47-kD uroplakin III. As a step towards elucidating the AUM structure and function, we have cloned the cDNAs of bovine uroplakin I (UPI). Our results established the existence of two isoforms of bovine uroplakin I: a 27-kD uroplakin Ia and a 28-kD uroplakin Ib. These two glycoproteins are closely related with 39% identity in their amino acid sequences. Hydropathy plot revealed that both have four potential transmembrane domains (TMDs) with connecting loops of similar length. Proteolytic digestion of UPIa inserted in vitro into microsomal vesicles suggested that its two main hydrophilic loops are exposed to the luminal space, possibly involved in interacting with the luminal domains of other uroplakins to form the 12-nm protein particles. The larger loop connecting TMD3 and TMD4 of both UPIa and UPIb contains six highly conserved cysteine residues; at least one centrally located cysteine doublet in UPIa is involved in forming intramolecular disulfide bridges. The sequences of UPIa and UPIb (the latter is almost identical to a hypothetical, TGF beta-inducible, TI-1 protein of mink lung epithelial cells) are homologous to members of a recently described family all possessing four transmembrane domains (the "4TM family"); members of this family include many important leukocyte differentiation markers such as CD9, CD37, CD53, and CD63. The tissue-specific and differentiation-dependent expression as well as the naturally occurring crystalline state of uroplakin I molecules make them uniquely suitable, as prototype members of the 4TM family, for studying the structure and function of these integral membrane proteins.