Summary
Members of the p24 family of type I transmembrane proteins are highly abundant in transport vesicles and are thought to be involved in selective protein transport between the endoplasmic reticulum and the Golgi complex. The p24 proteins have been grouped into four subfamilies (α, β,γ , and δ) and appear to assemble into tetrameric complexes that contain only one representative from each subfamily. Here we molecularly dissected the p24 family in a single cell type, namely in the intermediate pituitary melanotrope cells of the amphibian Xenopus laevis. The biosynthetic activity of these cells for production of their major cargo protein proopiomelanocortin (POMC) can be physiologically manipulated via the process of background adaptation (∼30-fold induction, with highly active cells in black toads and virtually inactive cells in white animals). Extensive cDNA library screening revealed the identity of six p24 proteins expressed in the Xenopus melanotrope cells, namely one member of the p24α (α3), one of the p24β (β1), two of the p24γ (γ2, γ3) and two of the p24δ (δ1, δ2) subfamily. Two other Xenopus p24 proteins, Xp24α2 and -γ1, were not expressed in the melanotrope cells, pointing to cell-type specific p24 expression. Of the six melanotrope p24 proteins, the expression of four (Xp24α3, -β1, -γ3 and -δ2) was 20- to 30-fold induced in active versus inactive melanotropes, whereas that of the other two members (Xp24γ2 and -δ1) had not or only slightly increased. The four proteins were induced only in the intermediate melanotrope cells and not in the anterior pituitary cells, and displayed similar overall tissue distributions that differed from those of Xp24γ1, -γ2 and -δ1. Together, our results reveal that p24 expression can be cell-type specific and selectively induced, and suggest that in Xenopus melanotrope cells anα 3/β1/γ3/δ2 p24 complex is involved in POMC transport through the early stages of the secretory pathway.
Introduction
Proteins synthesized in the endoplasmic reticulum (ER) are transported along the secretory pathway by coated vesicular carriers. Thus far, two types of coats, termed COPI and COPII, have been identified that mediate transport between the ER and the Golgi complex. Cargo proteins exit the ER in COPII-coated vesicles that bud from specialized regions, called ER exit sites ( Aridor et al., 1995). After budding, COPII vesicles quickly shed their coat and fuse to form vesicular-tubular clusters (VTCs) ( Balch et al., 1994; Scales et al., 1997), also referred to as ER-to-Golgi intermediate compartment (ERGIC) ( Schweizer et al., 1990). VTCs are transported as a whole along microtubules to the Golgi complex, where they appear to fuse and form the cis-Golgi network ( Saraste and Svensson, 1991). Retrograde transport of components of the vesicle targeting/fusion machinery and of escaped ER-resident proteins back to the ER are mediated by COPI-coated vesicles ( Aridor et al., 1995; Letourneur et al., 1994; Scales et al., 1997). Furthermore, the COPI coat may be involved in intra-Golgi vesicular transport ( Nickel et al., 1998; Orci et al., 1997).
Cargo proteins can leave the ER without prior concentration ( Martínez-Menárguez et al., 1999; Warren and Mellman, 1999), but several studies have demonstrated that the cell has mechanisms for concentration of cargo in ER-derived vesicles and for accelerated transport out of the ER ( Kuehn et al., 1998; Mizuno and Singer, 1993; Nishimura and Balch, 1997), suggesting a selective mechanism of cargo transport, presumably via cargo receptors. Thus far, three evolutionarily conserved families of integral membrane proteins have been proposed to facilitate ER-to-Golgi transport. Representatives of these protein families are fairly abundant and binding sites for COPI and/or COPII coat subunits are found in the cytoplasmic tail of these proteins; these binding sites enable the proteins to cycle constantly within the early secretory pathway. The BAP family seems to regulate trafficking of certain membrane proteins out of the ER ( Adachi et al., 1996; Kim et al., 1994; Terashima et al., 1994). BAP31, a representative of this family, has been shown to bind with high specificity to the endosomal membrane protein cellubrevin and to control its export out of the ER ( Annaert et al., 1997). ERGIC-53/p58, a mannose-specific membrane lectin, belongs to another class of receptors involved in the transport of a number of glycoproteins from the ER to the ERGIC ( Hauri et al., 2000). The third group of putative cargo receptors is a family of structurally related 24-kDa type I transmembrane proteins, collectively termed p24 proteins. Based on their amino acid sequences, these proteins have been classified into four main subfamilies, designated p24α, -β, -γ and -δ ( Dominguez et al., 1998). Members of the various p24 subfamilies exhibit only a low degree of amino acid sequence identity (17-30%) but all p24 proteins have certain structural characteristics in common, such as a relatively large lumenal domain with two conserved cysteine residues forming a disulfide bridge, a C-terminally located transmembrane stretch and a short cytoplasmic tail with sequence motifs known to specify interactions with vesicle coat proteins. Experimental evidence indicates that members of the various subfamilies can interact and tetrameric complexes are formed containing one representative of each subfamily ( Belden and Barlowe, 1996; Füllekrug et al., 1999; Marzioch et al., 1999). Consistent with this view, the stability of other family members is compromised in yeast mutants and knockout mice deficient in the expression of a single p24 member ( Denzel et al., 2000; Marzioch et al., 1999). A function for p24 proteins in cargo transport has been proposed on the basis of the observation that in yeast, deletion of certain p24 members slows ER export of a set of secretory proteins, whereas the export rate of a number of other cargo proteins is normal ( Schimmöller et al., 1995). More recently, it was shown that two of these yeast p24 members, Emp24 (yp24β) or Erv25p (yp24δ), which coexist in a heteromeric complex, can be directly crosslinked to the lumenal cargo protein Gas1p in ER-derived vesicles. Efficient packaging of Gas1p was reduced when vesicles were generated from membranes lacking Emp24p activity ( Muñiz et al., 2000). Furthermore, genetic experiments in yeast and Caenorhabditis elegans indicated that loss of p24 protein activity affects the fidelity of ER sorting ( Elrod-Erickson and Kaiser, 1996; Wen and Greenwald, 1999). Although in yeast, p24 proteins are not essential for vesicular transport ( Springer et al., 2000), deleting a single p24 member (p23) leads to early embryonic death in mice ( Denzel et al., 2000).
In this study, we identified the members of the p24 family that are expressed in the intermediate pituitary of the South African clawed toad Xenopus laevis. The intermediate pituitary consists of a homogenous population of melanotrope cells that are involved in the process of background adaptation of the animal. The central function of the melanotrope cells is the production of proopiomelanocortin (POMC) and in an active cell this prohormone constitutes over 80% of all newly synthesized proteins ( Holthuis et al., 1995a). The processing of POMC yields a number of bioactive peptides of which theα -melanophore stimulating hormone (α-MSH) stimulates the dispersion of the black pigment melanin in skin melanophores, causing darkening of the animal ( Jenks et al., 1977). In the melanotrope cells, the expression levels of POMC can be manipulated in a physiological way simply by changing the background color of the animal. On a black background, the POMC gene is highly active, whereas on a white background the gene is virtually inactive. The high levels of POMC production in black-adapted animals cause an enormous increase in cargo transport in the melanotrope cells, reflected by an extremely well-developed biosynthetic and secretory pathway, and a melanotrope cell size about twice as large as that in white-adapted animals (reviewed by Roubos, 1997). One would thus expect that the p24 proteins which have a presumed function in POMC transport are coordinately expressed with this prohormone. We could demonstrate that the expression of a selective set of p24 proteins is induced in the melanotrope cells of black-adapted animals, whereas others are not or only slightly induced. The coordinate expression of Xp24α3, -β1, -γ3 and -δ2 with POMC suggests that these p24 proteins assemble into a tetrameric complex involved in the ER-to-Golgi transport of the prohormone.
Materials and Methods
Animals
Adult South African clawed toads (Xenopus laevis) were obtained from laboratory stock and kept under constant illumination in water of 22°C. Animals were allowed to adapt to the background by placing them in either white or black tanks for at least three weeks.
Library screening and DNA sequencing
The nonredundant GenBank/EMBL/DDBJ database at NCBI was searched for p24 sequences from mouse with the BLAST program. Expressed sequence tags (ESTs) containing the entire ORFs of mouse p24α3, -γ1, -γ2, -γ3 and -γ4 were identified and received from the IMAGE Consortium ( Zehetner and Lehrach, 1994). PCR products, constituting parts of the ORFs, were generated using the mouse EST clones as templates. Degenerated oligonucleotides encoding conserved sequences in vertebrate p24β proteins were used to amplify a 313-bp fragment of Xp24β1 (nucleotides 286-599 in Xp24β1 ORF) from cDNA obtained through standard reverse-transcriptase reactions ( Sambrook et al., 1989) on RNA isolated from Xenopus brain by the Trizol isolation method (Life Technologies-BRL). Fragments were gel purified, labelled with [32P]dATP by random primer extension ( Ausubel et al., 1989), and unincorporated nucleotides were removed using NucTrap Probe purification columns (Stratagene, Cedar Creek, TX). The probes were used to screen an oligo dT-primed cDNA library of neurointermediate lobes of the pituitary gland of black-adapted X. laevis ( Kuiper et al., 2000). In addition, ZapII-cDNA libraries made from whole Xenopus embryos (stage 42; Michael King, Indiana University, Bloomington, IN) or embryo heads (stage 28-30; Richard Harland, University of California, Berkeley, CA) were used. Plaques (density 400/cm2) were replicated on duplicate nylon membrane filters by standard procedures ( Sambrook et al., 1989). Filters were prehybridized for at least 1 hour at 50°C in hybridization solution (10% dextran sulfate, 1% SDS, 1 M NaCl, 0.1% sodiumpyrophosphate, 0.2% bovine serum albumin, 0.2% polyvinylpyrollidone K90, 0.2% Ficoll 400, 50 mM Tris pH 7.5) and hybridized under conditions of low stringency (at 50°C) with a labelled 524 bp probe for mp24α3 (nucleotides 150-674 in accession number AA109932), a 313-bp probe for Xp24β1 (nucleotides 286-599 in Xp24β1 ORF), a 537-bp probe for mp24γ1 (nucleotides 107-644 in accession number W08294), a 462-bp probe for mp24γ2 (nucleotides 248-710 in accession number W58982), a 301 bp probe for mp24γ3 (nucleotides 5-306 in accession number AA020489), and a 453 bp probe for mp24γ4 (nucleotides 274-727 in accession number AA060892). Filters were washed twice for 40 minutes at 50°C in 2×SSPE/0.1% SDS (where 1×SSPE is 0.15 M NaCl, 10 mM sodium phosphate, 1 mM EDTA; pH 7.4) and exposed to X-ray films between two intensifying screens at —70°C. Positive plaques were purified, in vivo excised and analyzed by DNA sequencing using the ABI-PRISM DNA sequencing kit and the ABI-PRISM automatic sequencer (Perkin Elmer-Cetus Applied Biosystems, Foster City, CA). The cDNA libraries were screened under high-stringency hybridization conditions with the 3′-untranslated sequences of isolated cDNA clones encoding the various Xenopus p24 proteins. These sequences were amplified in a PCR reaction yielding for Xp24α3 a 0.43 kb fragment, for Xp24β1∼ 1.8 kb, for Xp24γ2 ∼0.6 kb and for Xp24γ3 ∼1.7 kb. The high-stringency hybridizations were performed at 63°C and filters were washed to a final stringency of 0.1×SSPE/0.1% SDS at 63°C. Standard procedures, such as PCR and single clone in vivo excision of λ-phage, were performed as described ( Ausubel et al., 1989).
RNA isolation and Northern blot analysis
Total RNA was isolated with RNAgents isolation system according to the instructions of the manufacturer (Promega, Madison, WI) and quantified by spectrophotometry. Aliquots of 5 μg per lane were separated by electrophoresis on 2.2 M formaldehyde-containing 1.2% agarose gels in MOPS buffer ( Sambrook et al., 1989) and blotted onto nylon membranes using downward capillary transfer. Hybridizations were performed for 18 hours at 42°C in ULTRAhyb hybridization solution (Ambion, Austin, TX) with probes comprising the entire ORFs of Xp24α3, -γ1, -γ2, and -γ3, a 364-bp fragment of Xp24β1 (nucleotides 77 to 441 in Xp24β1 ORF) or a 230 bp fragment of XGAPDH (nucleotides 266 to 496 in XGAPDH ORF), as a control for RNA loading and integrity. Blots were washed at 60°C to a final stringency of 0.1×SSPE/0.1% SDS (twice for 30 minutes) and exposures were taken using a PhosphorImager (BioRad Personal FX).
Antibodies
For antibody production, a region in the lumen of Xp24α3 (residues 31-128) was cloned into the expression vector pQE30 (Qiagen, Chatsworth, CA), the recombinant protein was produced in E. coli and purified by Ni2+-NTA agarose affinity chromatography. The lumenal domains of Xp24γ1 (residues 18-194) and Xp24γ2 (residues 32-191) were cloned as GST fusion proteins into the bacterial expression vector pGEX-2T (Pharmacia Biotech Benelux, NL). The expressed GST fusions were largely insoluble. Therefore, the aggregated fusion proteins were isolated as inclusion bodies from E. coli ( Nagai and Thogersen, 1987). A synthetic peptide against the cytoplasmic tail sequence of Xp24γ3 (C-FSDKRTTTTRVGS) was coupled to keyhole limpet hemocyanin (Pierce, Rockford, IL). All antigens described were injected into rabbits and the immunization was done as described ( Kuiper et al., 2000). Polyclonal antibodies against amino acid sequences in the lumenal part of Xp24δ1 (C-FDSKLPAGAGRVP; anti-δ1) and -δ2 (residues 72-150; anti-δ2), as well as the C-terminally directed p24δ antibody (anti-δC) have been described previously ( Kuiper et al., 2000). The p24β1 and -γ3 peptide antibodies (anti-β1L and anti-γ3) recognize orthologs in human and Xenopus, and were kindly provided by T. Nilsson (EMBL, Heidelberg, Germany) ( Dominguez et al., 1998). Affinity purifications of antisera using antigen-sepharose 4B columns were performed according to standard protocols ( Harlow and Lane, 1988).
Immunological characterization
Extraction of proteins from different tissues of X. laevis, SDS-PAGE gel electrophoresis, Western blotting, antibody detection and immunocytochemistry were performed as described ( Kuiper et al., 2000).
Results
Cloning of cDNAs encoding the p24 proteins in the Xenopus intermediate pituitary
The melanotrope cells in the intermediate pituitary of the amphibian Xenopus laevis are primarily dedicated to the production of POMC and the expression levels of this prohormone can be readily manipulated by changing the background color of the animal. On a black background, the POMC gene is actively transcribed and the prohormone represents ∼80% of all newly synthesized proteins, whereas on a white background the gene is nearly inactive. Therefore, the Xenopus melanotrope cell is an attractive model system to study the role of p24 proteins in POMC transport. The degree of amino acid sequence identity between vertebrate p24 orthologs is usually above 68%, allowing us to search for Xenopus p24 proteins expressed in the melanotrope cells by low-stringency hybridization of a neurointermediate pituitary (NIL) cDNA library of black-adapted toads with [32P]-labeled mouse p24 cDNA fragments. We identified the Xenopus members of the four p24 subfamilies p24α, -β, -γ, and -δ.
The p24α subfamily
Within the p24α subfamily two branches of p24 proteins can be distinguished, namely, the α1-branch and theα 2/α3-branch ( Dominguez et al., 1998) ( Fig. 1B). Thus far, the only isolated representative of the α1-branch is from dog pancreatic microsomes ( Wada et al., 1991). Two closely related representatives of theα 2/α3-branch are expressed in mouse, with mouse p24α2 (mp24α2; GMP25/mp25) being 80.8% identical to mp24α3 (GMP25iso) ( Dominguez et al., 1998). In our screening for p24α subfamily members, we used a 524 bp fragment of a mouse EST p24α3 clone as a probe, yielding 12 hybridization-positive plaques from 2×105 plaques of the Xenopus NIL cDNA library screened. Nucleotide sequence analysis revealed that the positive clones all contained overlapping cDNAs. The insert size of two full-length clones was approximately 1.2 kb with a 687-bp open reading frame (ORF). Since the deduced Xenopus p24α protein (excluding the signal peptide region) was more closely related to mp24α3 (92.4% identity) than to mp24α2 (81.3% identity), the protein was named Xp24α3 ( Fig. 1). The low degree of amino acid sequence conservation between the α1- and theα 2/α3-branch (identity <60%) does not allow the isolation of a p24α1-related protein using mouse Xenopus p24α3 cDNA as a probe. However, the degree of homology within the α2/α3-branch (identity >80%) should be high enough for the identification of a p24α2 ortholog in Xenopus. We therefore performed a low-stringency hybridization using the coding region of Xp24α3 cDNA as a probe. From a total number of 4.2×105 plaques, 51 hybridization-positive plaques were obtained. All clones positive in this screening were also recognized by the 0.43-kb 3′-untranslated region of Xp24α3 cDNA on a duplicate filter under stringent hybridization conditions, indicating that only p24α3, and not p24α2, is expressed in the Xenopus intermediate pituitary. Nevertheless, p24α2 does exist in X. laevis, because recently performed database searches revealed two Xenopus EST clones isolated from embryo and liver cDNA libraries, of which the deduced amino acid sequences were more similar to mp24α2 than to mp24α3 ( Fig. 1).
Xenopus p24 proteins and their relationship with p24 proteins from other species. (A) Alignment of members of the p24 protein family in X. laevis. Aligned are the amino acid sequences deduced from cDNA clones, the EST database entry BF611875 (representing Xp24α2) and the two p24δ subfamily members identified previously ( Kuiper et al., 2000). The cDNAs of the Xenopus p24 proteins have been isolated from a neurointermediate pituitary cDNA library, except for those of Xp24α2 and -γ1 (isolated from an embryo library). Amino acids that are conserved in at least five sequences are in black boxes. The putative signal peptidase cleavage sites of the N-terminal signal sequences (incomplete for Xp24α2) are indicated by an arrow. Asterisks indicate the two conserved cysteine residues present in the lumenal domains of all p24 proteins. The predicted transmembrane region (TM) is underlined. (B) Phylogenetic tree of the p24 proteins from mouse (m), Xenopus (X), Drosophila melanogaster (d), Caenorhabditis elegans (c) and p24α1 from dog, and the classification in the four proposed p24 subfamilies. (C) Amino acid sequence identity (%) between p24 proteins of mouse and X. laevis. For sequence comparisons and phylogenetic tree construction, the p24 proteins without their signal sequences were aligned by the Clustal W algorithm using default parameters (AlignX program in Vector NTI Suite 6; InforMax, North Bethesda, MD). The mp24α1 protein lacks part of the N-terminal region (indicated with an asterisk). Sequences are mostly compilations of several data base entries and accession numbers are available upon request.
The p24β subfamily
Only one member of the p24β subfamily exists in higher vertebrates and its high degree of sequence conservation enabled us to amplify a 313-bp fragment of Xp24β1 in a PCR reaction using Xenopus brain-derived cDNA and degenerate primers corresponding to two conserved domains in vertebrate p24β1 proteins. This fragment was used to screen the Xenopus NIL cDNA library, and two full-length clones of Xp24β1 with an insert size of 2.4 kb were isolated. A comparison of the deduced primary sequence of Xp24β1 with mouse p24β1 revealed that, when the signal peptide sequence is excluded, the two proteins share a sequence identity of 99.4% ( Fig. 1). To assess whether other representatives of the p24β subfamily exist and are expressed in the library, duplicate filters were prepared, of which one filter was hybridized under low-stringency conditions with a probe comprising the entire ORF of Xp24β1, whereas the second filter was hybridized under high stringency conditions with a probe directed against the ∼1.8 kb 3′-untranslated sequence of Xp24β1. All 345 hybridization-positive plaques from 6.7×105 plaques screened were recognized by both probes, indicating that Xenopus melanotrope cells express only one member of the p24β subfamily.
The p24γ subfamily
The p24γ subfamily is more diverse than the other p24 subfamilies, and database searches revealed four members in mouse. An evolutionary tree showed that two groups of p24γ proteins can be distinguished, one represented by p24γ1 and -γ2, and the other by p24γ3 and -γ4 ( Fig. 1B). Low-stringency hybridization of the Xenopus NIL cDNA library (3.4×105 plaques) with a 462-bp mouse p24γ2 fragment allowed the identification of 9 hybridization-positive cDNA clones, which carried an insert that coded for a p24γ2 ortholog in X. laevis. The deduced protein sequence of Xp24γ2 displays 72.4% identity with mouse p24γ2 and 52.9% with mouse p24γ1 (excluding signal peptides) ( Fig. 1). Out of 7.3×105 NIL cDNA plaques screened under stringent conditions, 30 plaques were recognized by a ∼0.6 kb probe corresponding to the 3′-untranslated region of Xp24γ2. A subsequent low-stringency screening of the same library filters with the coding sequence of Xp24γ2 yielded no additional hybridizing plaques. Thus, under the conditions used, no cross hybridization with other Xp24γ sequences was found.
A low-stringency hybridization of the NIL cDNA library detected no positive clones when a mouse p24γ1 fragment was used. However, screening of two Xenopus embryo cDNA libraries with the mouse p24γ1 probe resulted in the identification of several overlapping cDNA clones, encoding the Xp24γ1 protein ( Fig. 1). Therefore, Xp24γ1 does not appear to be expressed in the Xenopus intermediate pituitary, but is expressed in other Xenopus tissues. In its mature form (without signal peptide), the Xp24γ1 protein shares an overall sequence identity of 68.5% with mouse p24γ1, 56.4% with mouse p24γ2 and 56.2% with Xp24γ2. Moreover, nucleotide sequence analysis revealed that 2 out of the 17 Xp24γ1 clones isolated from the embryo head library showed an insertion of 30 nucleotides at position +442 in Xp24γ1 cDNA. Consequently, the protein encoded by the two cDNAs contained an in-frame insertion of 10 amino acid residues following amino acid 148, with the amino acid sequence VRFCPLTFEE (single-letter code). This finding suggests that in a low incident of cases the transcripts of Xp24γ1 are subject to alternative splicing, giving rise to two structurally distinct Xp24γ1 proteins that differ in size by ∼1.1 kDa. The in-frame insertion was not found in other known p24γ1 proteins and so far alternative splicing has not been described for other p24 proteins. In conclusion, since under identical hybridization conditions Xp24γ1 could be isolated from an embryo cDNA library but not from a NIL cDNA library, only Xp24γ2 but not -γ1 is expressed in the melanotrope cells of the Xenopus intermediate pituitary.
For the isolation of Xenopus orthologs of p24γ3 and p24γ4, the NIL cDNA library was screened with a 453 bp fragment derived from a mouse p24γ4 cDNA. Out of the 4×105 plaques screened, 25 were positive on a duplicate set of filters. Nucleotide sequence analysis of 10 of these clones revealed a single ORF and the corresponding amino acid sequence (without signal peptide) was 93.2% identical to mp24γ3 and 69.1% identical to mp24γ4. Thus, the cDNA clones isolated with the mp24γ4 probe encode a Xenopus ortholog of p24γ3 ( Fig. 1). The NIL cDNA library was re-screened and 175 hybridization-positive plaques were observed when the filters were first hybridized with a 3′-untranslated probe of Xp24γ3 under stringent hybridization conditions (6×105 plaques screened). A subsequent low-stringency hybridization with the coding sequence of Xp24γ3 revealed no differences in the hybridization pattern, suggesting that only Xp24γ3 and not Xp24γ4 is expressed in Xenopus melanotrope cells. Besides the NIL cDNA library we also screened the whole-embryo cDNA library with a 306-bp mp24γ3 probe under conditions of low-stringency. This led to the identification of 66 positive plaques (4×105 plaques screened), which also remained positive after a more stringent wash (0.1×SSC/0.1% SDS; 50°C). Together with the fact that from 20 of these clones a specific Xp24γ3 fragment was amplified in a PCR reaction, the hybridization-positive clones most likely contain an Xp24γ3 cDNA insert. From these experiments we conclude that a p24γ4 ortholog is not expressed in the Xenopus intermediate pituitary and embryos.
The p24δ subfamily
In Xenopus, the p24δ subfamily has already been characterized in our laboratory, leading to the identification of two p24δ proteins, Xp24δ1 and -δ2, expressed in the melanotrope cells ( Kuiper et al., 2000). Up to then, only one representative of the p24δ subfamily had been described in vertebrates, with mouse p24δ being more related to Xp24δ1 than to Xp24δ2 (amino acid sequence identities of 82.2% and 70%, respectively).
Analysis of p24 sequences
A multiple amino acid sequence alignment of all known Xenopus p24 proteins revealed a low degree of overall amino acid sequence identity between the members of the different subfamilies ( Fig. 1A). However, the topology common to all p24 proteins is also preserved in the Xenopus proteins (an N-terminal signal sequence, a large lumenal domain followed by a transmembrane stretch, and a cytoplasmically exposed C-terminal region of 10-16 residues), as are conserved amino acid residues such as the two cysteine residues that form a disulfide bridge in the N-terminal region, a glutamine residue within the transmembrane domain and a phenylalanine residue that constitutes part of a COPII-binding motif ( Dominguez et al., 1998). Furthermore, heptad repeats of aliphatic amino acids are found in the membrane proximal parts of the lumenal domains of Xp24α2, -α3, -β1, -γ1, -δ1 and -δ2 (but not in Xp24γ2 and -γ3) that have a medium to high propensity to form coiled-coil structures (>0.4 by coils algorithm, version 2.2) ( Lupas, 1996). Coiled-coil interactions between members of different p24 subfamilies are involved in the formation of hetero-oligomeric complexes ( Ciufo and Boyd, 2000). The alternatively spliced form of Xp24γ1 has in this particular region an insertion of ten amino acids (following residue +148), which reduces its propensity to form coiled-coil structures. The cytoplasmic tail sequences are highly conserved between the two Xenopus p24α subfamily members, as they are in the p24δ subfamily and the p24γ1/γ2 subgroup. A classical ER retrieval/COPI-binding motif in the C-terminal region (K(X)KXX) is present only in the Xenopus p24α proteins, whereas members of the other subfamilies show variations of this motif. Binding studies with the cytoplasmic domains of human p24 proteins have revealed efficient COPI binding for hp24α2 and also for hp24δ1 (despite an imperfect COPI-binding motif), while all human p24 proteins analyzed (α2, β1, γ3,γ 4, δ1) have been found to interact with COPII ( Dominguez et al., 1998). Since the critical amino acid residues in the cytoplasmic tails are conserved between the human and Xenopus p24 proteins, similar binding properties are to be expected for the Xenopus p24 proteins.
During evolution, the genome of Xenopus laevis underwent a genome duplication event, causing this species to be tetraploid ( Graf and Kobel, 1991). As a consequence, two highly conserved genes (paralogs) are usually found. This was also the case for every novel p24 protein isolated in this study. The nucleotide sequence identities over the entire ORFs were found to be in the order of 94 to 95.5%, and the nucleotide substitutions were either neutral or led to conservative amino acid substitutions in the deduced primary sequences of the respective p24 proteins. For clarity, the primary sequence of only one paralog is shown in Fig. 1.
In summary, the melanotrope cells in the intermediate pituitary of Xenopus express one member of the p24α (α3), one of the p24β (β1), two of the p24γ (γ2, γ3) and two of the p24δ (δ1, δ2) subfamily. Two members, Xp24α2 and -γ1, show a tissue-specific distribution in that they were expressed in Xenopus embryos but not in the intermediate pituitary.
Expression of the various Xenopus p24 proteins in the pituitary gland
To study the expression of the various Xenopus p24 proteins, polyclonal antisera directed against sequences in the lumenal domains of these p24 proteins were generated. Western blot analysis revealed that the various antisera recognize specifically the corresponding Xp24 proteins of the melanotrope cells ( Kuiper et al., 2000) (data not shown), except for the anti-δC antiserum, which recognizes both Xp24δ proteins (δ1 andδ 2). To study the expression of the p24 proteins in the pituitary, lobes from black- and white-adapted Xenopus were manually dissected into the NIL and the anterior lobe (AL). Together with tissue extracts from brain and hypothalamus, NIL and AL lysates were separated with SDS-PAGE and analyzed by immunoblotting (three times more protein was loaded for brain and hypothalamus). In NIL lysates of black-adapted animals, the antibodies directed against epitopes in the lumenal domains of Xp24α3, -β1, -γ2, -γ3, -δ1, and -δ2 recognized proteins with a molecular mass of ∼25, 21, 24, 25.5, 19 and 21 kDa, respectively ( Fig. 2, data not shown for anti-δ1 and anti-δ2). With an antibody against Xp24γ1 (residues 18-194), a 22 kDa protein was identified in tissue lysates from brain and hypothalamus, while no protein was detected in the NIL ( Fig. 2), in line with the finding that a p24γ1 cDNA could be isolated only from a Xenopus embryo but not from a NIL cDNA library. At present it is not clear if the 23 kDa product in the anterior lobe of the pituitary detected with the anti-γ1L antiserum corresponds to the alternatively spliced form of Xp24γ1, or is due to nonspecific binding to the similarly sized and highly expressed anterior pituitary hormones prolactin/growth hormone (judged by staining of the western blot with Ponceau S). Interestingly, immunoblot analysis of cellular extracts showed that the protein levels of Xp24α3, -β1, -γ3, and -δ2 were highly upregulated in NILs of black-adapted Xenopus when compared with that in white-adapted animals (∼20-30-fold). By contrast, the level of Xp24δ1 expression was only slightly induced (∼threefold) in NIL lysates of black-adapted animals, and no change was observed for Xp24γ2 ( Fig. 2). As previously demonstrated for POMC ( Holthuis et al., 1995a), the physiologically induced changes in the expression levels of Xp24α3, -β1, -γ3 and -δ2 were strictly confined to the melanotrope cells of the NIL and did not occur in cells from the AL of the pituitary ( Fig. 2). Semiquantitative reverse transcription (RT)-PCR revealed a three- to fivefold increase in the levels of Xp24α3, -β1, -γ3 and -δ2 mRNA in the melanotropes of black-adapted animals, whereas the levels of Xp24δ1 and -γ2 mRNA were similar (data not shown).
p24 protein expression in Xenopus pituitary. The neurointermediate lobe (NIL) was manually dissected from the anterior lobe (AL) of the pituitary of black- or white-adapted Xenopus. 20 μg of NIL and AL extracts were resolved together with tissue extracts from brain and hypothalamus (∼60 μg each) by SDS-PAGE. For immunoblotting, antisera directed against sequences in the lumenal domains of the respective p24 proteins or the anti-δC antiserum recognizing Xp24δ1 and -δ2 were used. The asterisk indicates an AL protein band presumably resulting from crossreactivity of the anti-Xp24γ1 antiserum with an abundant 23 kDa protein (likely growth hormone/prolactin).
Next, we tried to confirm our results using immunocytochemistry. The p24 proteins Xp24α3, -β1, -γ3 and -δ2, shown by western blot analysis to be differentially regulated in the NIL, displayed intense staining in the melanotrope cells of the intermediate pituitary of a black-adapted animal, whereas their expression in a white-adapted animal was low ( Fig. 3). Conversely, for Xp24γ2 and -δ1, a low degree of expression was observed in the melanotrope cells, independent of background adaptation. The differential regulation of the various p24 proteins was again confined to the melanotrope cells, because no differences in the expression patterns were observed in the anterior lobes of black- and white-adapted animals. Furthermore, the homogenous distribution of the p24 proteins throughout the intermediate lobe suggests that they are expressed in all melanotrope cells. Antisera directed against Xp24γ1 showed no immunoreactive staining in the melanotrope cells, whereas in the anterior pituitary some immunoreactive material was detected, again likely due to nonspecific cross-reactivity as was observed during Western blot analysis. In conclusion, we find that during background adaptation only a subset of the p24 proteins (Xp24α3, -β1, -γ3, and -δ2) expressed in the melanotrope cells of the intermediate pituitary is coordinately expressed with the prohormone POMC.
Immunocytochemical analysis visualizing the expression of p24 proteins in the pituitary gland of Xenopus laevis adapted to a black or white background. Shown are sagital sections stained with the peroxidase-anti-peroxidase method for Xp24α3 (affinity-purified anti-α3L, 1:300 dilution), Xp24β1 (anti-β1L, 1:200 dilution), Xp24γ1 (anti-γ1L, 1:600 dilution), Xp24γ2 (affinity-purified anti-γ2L, 1:40 dilution), Xp24γ3L (anti-γ3L, 1:200 dilution), Xp24δ1 (affinity-purified anti-δ1, 1:50 dilution), and Xp24δ2 (anti-δ2, 1:1500 dilution). NL, neural lobe; IL, intermediate lobe; AL, anterior lobe. Bar, 200 μm.
p24 expression in Xenopus tissues
The distributions of the various Xp24 proteins in tissues other than the pituitary was studied at the level of RNA (Northern blot analysis) and at the protein level (Western blotting). For Northern blot analysis, total RNA was isolated from a number of Xenopus tissues and hybridized with [32P]-labelled probes covering the entire ORF of Xp24α3, -γ1, -γ2, -γ3 or with a 364-bp fragment of Xp24β1 (nucleotides +77 to +364). The applied method was not sensitive enough for the detection of Xp24γ1 and -γ2 transcripts, whereas more than one transcript was found for Xp24α3, -β1 and -γ3 ( Fig. 4A). Xp24α3, -β1 and -γ3 mRNAs were found to be expressed in brain, liver, kidney, spleen, heart and lung, as was previously found for Xp24δ1 and -δ2 mRNAs ( Holthuis et al., 1995b) (data not shown). Owing to the tetraploid nature of the Xenopus genome, a pair of closely related genes (paralogs) is expressed that often gives rise to transcripts with different sizes. Xp24α3 is represented by two transcripts of about 1.2 and 2.5 kb, ubiquitously expressed in all tissues examined. The size of the 1.2 kb transcript corresponds to the insert sizes of two full-length cDNA clones encoding paralog A of Xp24α3. The transcript length expected for paralog B of Xp24α3 is not known and may be represented by the 2.5 kb transcript. With an Xp24β1 probe, a predominant mRNA transcript of 2.5 kb and a weak one of 1.2 kb were detected. The Xp24β1-encoding cDNAs for the two paralog genes were derived from a 2.5-kb transcript and thus the 1.2-kb transcript may arise from alternative splicing of nuclear RNA or from the use of an alternative polyadenylation signal. Three transcripts (2.3, 2.5 and >4 kb) with similar intensities were observed for Xp24γ3 mRNAs. Since the sizes of two transcripts (2.3 and 2.5 kb) correspond to full-length Xp24γ3 cDNA clones, the presence of the >4 kb transcript indicates that Xp24γ3 gene expression may also include alternative splicing and/or alternative usage of polyadenylation signals.
p24 expression in Xenopus tissues. (A) Nothern blot analysis of Xenopus p24α3, -β1, and -γ3 mRNAs. Equal amounts of RNA were loaded and Xenopus GAPDH was used as a control for RNA loading and integrity. The positions of the 18S and 28S Xenopus ribosomal RNAs are indicated. (B) Western blot analysis of the Xenopus p24 proteins using antibodies against lumenal epitopes of the respective p24 proteins.
The ubiquitous tissue expression of Xp24 proteins observed at the mRNA level was also seen at the protein level ( Fig. 4B). With the anti-α3L antibody, expression of Xp24α3 was found to be high in ovarium and liver, intermediate in hypothalamus, brain, kidney and spleen, and low in heart and lung. This antibody recognized a second protein band in liver and lung that was slightly larger in size and represents the only protein detected in heart tissue. The two protein bands (∼25 and ∼26 kDa) may reflect different glycosylation states of Xp24α3, which is N-linked glycosylated at a single site (data not shown). Tissue-dependent variations in the glycosylation state have also been described for hp24α2 ( Füllekrug et al., 1999). Alternatively, the anti-α3L antibody may recognize with low affinity the highly related Xp24α2 protein, which at least in liver is known to be expressed. The overall tissue distributions of Xp24β1, -γ3, and -δ2 were similar to that obtained for Xp24α3 and the four proteins are predominantly expressed in liver and ovarium. Except for the intermediate pituitary, the Xp24γ1 protein was expressed in all tissues examined ( Fig. 4). When compared to the levels obtained for Xp24α3, -β1, -γ3 and -δ2, the relatively low levels of Xp24γ1 expression in ovarium and liver are remarkable. The tissue distribution of Xp24γ1 is similar to those of Xp24γ2 and -δ1 (the two proteins that are not differentially regulated in the intermediate pituitary), although both Xp24γ2 and -δ1 show higher levels of protein expression in ovarium. Taken together, these experiments indicate that the members of the p24 family are widely expressed but they may display a cell-type specific expression, as was found for Xp24α2 and -γ1 in the melanotrope cells of the intermediate pituitary.
Discussion
We set out to examine the expression profiles of members of the p24 family in a single cell type, namely in the intermediate pituitary melanotrope cell of X. laevis. The primary function of this neuroendocrine transducer cell is the production of POMC and the release of POMC-derived melanophore-stimulating peptides during adaptation of the animal to a black background ( Holthuis et al., 1995a). The melanotrope cells are highly active in black-adapted Xenopus and produce ∼30-fold higher POMC mRNA levels than the biosynthetically virtually inactive melanotrope cells of white-adapted animals. A set of genes has been isolated whose transcripts are coordinately expressed with POMC (differential induction is >10-fold) ( Holthuis et al., 1995a), including POMC processing enzymes and members of the so-called granin family (reviewed by Kuiper and Martens, 2000). Other gene products identified reflect the dynamic changes in the secretory apparatus observed between active and inactive melanotrope cells. Included in this category are the molecular chaperone BiP and TRAPδ (translocon-associated protein subunit δ) as components of the ER translocation machinery, and the cysteine protease ER60 and the chaperone calreticulin as part of the quality control system in the ER. Another differentially expressed gene product was Xp24δ2 (X1262), a member of the p24 family of putative cargo receptors ( Holthuis et al., 1995a) (data not shown). Multiple members of the p24 family have been found in eukaryotes (eight in yeast, five in C. elegans, seven in Drosophila melanogaster, and nine in human and mouse; Fig. 1B), which prompted us to search for Xenopus p24 proteins other than Xp24δ2 and to study their expression in the melanotrope cells. Our extensive Xenopus intermediate pituitary cDNA library screening led to the identification of six p24 proteins, namely one member of the p24α (α3), one of the p24β (β1), two of the p24γ (γ2,γ3) and two of the p24δ (δ1, δ2) subfamily. Two other identified Xenopus p24 proteins, Xp24α2 and -γ1, were expressed in embryos, but not in the melanotrope cells. A phylogenetic tree constructed from the identified Xenopus p24 proteins and from the p24 proteins of mouse, D. melanogaster and C. elegans revealed that each of these species has at least one representative in each subfamily ( Fig. 1B). Members of the p24α subfamily have evolved into two separate branches (α1 and α2/3). The only representative of the α1-branch is gp25L (dog p24α1), but recent database searches indicate that orthologs of dog p24α1 may also exist in mouse and human. Invertebrates seem to have only one representative of the p24α subfamily that belongs to the α2/3-branch. Two closely related but distinct p24 proteins of the α2/3-branch were identified in human and mouse, and two representatives were also found in Xenopus (Xp24α2 and -α3). Vertebrate proteins of theα 2/3-branch may have a cell-type specific expression pattern, since in Xenopus only Xp24α3 but not -α2 was found to be expressed in the melanotrope cells. To date, only one representative of the p24β subfamily has been described, but database searches revealed that two p24β subfamily members exist in D. melanogaster. A single member was found to be expressed in Xenopus melanotrope cells. The primary sequences of human and mouse p24β1 (excluding the signal peptide) are identical, and mouse p24β1 and Xp24β1 share 99.4% identity. Such an exceptionally high sequence conservation was not found between the vertebrate orthologs of the other p24 subfamilies and thus may point to an evolutionarily conserved function for the p24β protein. The phylogenetic tree shows that members of the p24γ subfamily of C. elegans, D. melanogaster, mouse and human can be assigned to two distinct subgroups: theγ 1/γ2-subgroup and theγ 3/γ4-subgroup ( Fig. 1B). Since in vertebrates the two subgroups are only distantly related (amino acid sequence similarity<50% and identity <35%), perhaps a reclassification of these subgroups into separate subfamilies should be considered. The resulting classification into five subfamilies would imply that C. elegans has a single representative in each subfamily, whereas higher-developed organisms have mostly more than one. In mouse and human, the p24γ subfamily has four members, two of the p24γ1/γ2-subgroup (γ1 and γ2) and two of the p24γ3/γ4-subgroup (γ3 andγ 4). We found that in the Xenopus melanotrope cells one member of the p24γ1/γ2-subgroup (Xp24γ2) and one member of the p24γ3/γ4-subgroup (Xp24γ3) is expressed. A second member of the Xenopus p24γ1/γ2-subgroup (Xp24γ1) was found in an embryo cDNA library, while screening for a p24γ4 member of the p24γ3/γ4-subgroup remained negative, despite the fact that the probe used (mouse p24γ4) cross-hybridized even with the Xp24γ3 sequence. Therefore, a Xenopus p24γ4 ortholog appears not to be present in the melanotrope cells and in embryos. The restricted expression of Xp24γ1, namely in embryos but not in the melanotrope cells, was confirmed by immunoblotting. A cell-type specific expression pattern has also been described for the p24γ1 orthologs in human and mouse (T1/ST2 receptor) ( Gayle et al., 1996). Thus far, with the exception of Xenopus, only one representative of the p24δ subfamily has been reported in species of the animal kingdom. The two Xenopus p24δ subfamily members, Xp24δ1 and Xp24δ2, are both expressed in the melanotrope cells, although only Xp24δ2 but not Xp24δ1 is coexpressed with POMC in these cells ( Kuiper et al., 2000). Taken together, these findings show that the complexity of the p24 family is species dependent, with certain organisms having multiple members in distinct subfamilies and others having only one representative.
In the physiologically manipulated Xenopus melanotrope cells, we demonstrated that during black background adaptation the protein levels of Xp24α3, -β1, -γ3 and -δ2 were increased 20 to 30 times, whereas the expression of the Xp24γ2 and -δ1 proteins remained unchanged or increased by only three times, respectively. In yeast and mammals, p24 proteins form tetrameric complexes with a defined complex composition, in which one member from each subfamily is present ( Füllekrug et al., 1999; Marzioch et al., 1999). Furthermore, the steady-state protein level and the intracellular localization of a p24 protein is dependent on the presence or absence of other p24 members that participate in the oligomeric complex ( Denzel et al., 2000; Emery et al., 2000; Füllekrug et al., 1999; Marzioch et al., 1999). Since Xp24α3, -β1, -γ3 and -δ2 show similar dynamics in protein expression in the melanotrope cells and also in other tissues, they may well form a tetrameric p24 complex, whereas Xp24γ2 and -δ1, which are not differentially regulated in the melanotrope cells, could be constituents of other p24 complexes. Interestingly, the apparent composition of the main tetrameric melanotrope p24 complex (Xp24α3, -β1, -γ3, and -δ2) is different from the one previously identified in HeLa cells, where a p24α2/β1/γ3/δ1 p24 complex exists (GMP25/p24/gp27/p23) ( Füllekrug et al., 1999). Variations in p24 complex formation are likely to occur, for example, because of the observed cell-type-specific expression of the various Xenopus p24 proteins. Furthermore, in yeast an Erp1p(yp24α)/Erp2p(yp24γ)/Emp24p(yp24β)/Erv25p(yp24δ) complex is present, in which Erp1p can be substituted by another p24α subfamily member (Erp5p and/or Erp6p) if Erp1p is not expressed ( Marzioch et al., 1999). Also, hp24γ4 has been found to be excluded from the HeLa cellα 2/β1/γ3/δ1 p24 complex ( Füllekrug et al., 1999), while mp24γ4 may well participate in a p24 complex ( Denzel et al., 2000). Therefore, the composition of a tetrameric p24 complex appears to be cell-type specific.
The exact role of the p24 proteins is still elusive, but p24 proteins have been implicated in a number of functions that are all linked to vesicular and protein transport, such as regulation of cargo inclusion in ER vesicles ( Muñiz et al., 2000; Schimmöller et al., 1995), quality control mechanisms in the ER ( Belden and Barlowe, 1996; Wen and Greenwald, 1999), recruitment and regulation of COPI/II vesicle coat assembly ( Bremser et al., 1999; Kaiser, 2000; Kuehn et al., 1998), and generation of vesicular tubular clusters ( Lavoie et al., 1999; Rojo et al., 2000). In yeast, an interaction between the cargo protein Gas1p and p24 proteins has been demonstrated, and loss of function of certain p24 proteins reduces the kinetics of ER-to-Golgi transport of a subset of secretory proteins, whereas resident ER proteins (Kar2p and Pdi1p) are less efficiently retained in the ER ( Elrod-Erickson and Kaiser, 1996; Marzioch et al., 1999; Muñiz et al., 2000; Schimmöller et al., 1995). In C. elegans, reducing the activity of certain p24 proteins restores at least partially the ER transport block of a mutant protein to the plasma membrane ( Wen and Greenwald, 1999). Proper p24 function may thus facilitate the transport of certain cargo molecules and restrict the entry of ER proteins and incorrectly folded proteins into COPII vesicles. Along this line, in the activated Xenopus melanotrope cells a Xp24α3/β1/γ3/δ2 complex could be involved in the inclusion of POMC into transport vesicles, as the four p24 members are coordinately expressed with this prohormone. Unfortunately, extensive crosslinking and co-immunoprecipitation studies using the Xenopus intermediate pituitary cells have not allowed us to establish a direct physical interaction between the Xenopus p24 proteins or between p24 and POMC.
In conclusion, we isolated and characterized the set of p24 proteins expressed in a single cell type (the Xenopus intermediate pituitary melanotrope cell), and revealed that their expression is cell-type specific and can be selectively induced. In the melanotropes, four of the six p24 members are coexpressed and these representatives of the four subfamilies may form a complex that is involved in the efficient ER to Golgi transport of its major cargo protein POMC. Together, our results thus point to an involvement of p24 proteins in the process of selective protein transport within the early secretory pathway.
Acknowledgements
We thank T. Nilsson for antibodies (see Material and Methods), M. King and R. Harland for providing us with Xenopus embryo cDNA libraries, A. J. M. Coenen for technical assistance and R. Engels for animal care. This work was supported by grant ERB-FMRX-CT960023 from the European Union for Training and Mobility of Researchers (EU-TMR).
- Accepted November 28, 2001.
- © The Company of Biologists Limited 2002
References
Article navigation
Related articles
Cited by...
More in this TOC section
Similar articles
Other journals from The Company of Biologists
Welcoming early-career researchers to our Editorial Advisory Board
We are pleased to announce our latest endeavour to support the career development of scientists – the appointment of early-career researchers to our Editorial Advisory Board. Find out more about these seven outstanding scientists and our continuing focus on supporting early-career researchers.
Interview – Xiaolei Gao
In the latest first author interview, Xiaolei discusses her latest paper on the spindle pole body of Aspergillus nidulans, challenges during the project and her motivation to pursue a career in science.
Seven things that shaped preLights in 2019
As 2020 gets well and truly underway, join us as we look back at seven things that shaped preLights in 2019. From launching preLists to getting nominated for an award, it was an exciting year!
Most read of 2019 – Transporting cells over several days without dry-ice
Congratulations to Sally Wheatley and Denys Wheatley - the authors of the most-read paper in JCS of 2019! Read their 'Tools and Resources' article detailing a simple, hazard-free and inexpensive procedure to enable the transport of cultured cells without compromising recovery.