Autophagy supports generation of cells with high CD44 expression via modulation of oxidative stress and Parkin-mediated mitochondrial clearance.

High CD44 expression is associated with enhanced malignant potential in esophageal squamous cell carcinoma (ESCC), among the deadliest of all human carcinomas. Although alterations in autophagy and CD44 expression are associated with poor patient outcomes in various cancer types, the relationship between autophagy and cells with high CD44 expression remains incompletely understood. In transformed oesophageal keratinocytes, CD44Low-CD24High (CD44L) cells give rise to CD44High-CD24-/Low (CD44H) cells via epithelial-mesenchymal transition (EMT) in response to transforming growth factor (TGF)-ß. We couple patient samples and xenotransplantation studies with this tractable in vitro system of CD44L to CD44H cell conversion to investigate the functional role of autophagy in generation of cells with high CD44 expression. We report that high expression of the autophagy marker cleaved LC3 expression correlates with poor clinical outcome in ESCC. In ESCC xenograft tumours, pharmacological autophagy inhibition with chloroquine derivatives depletes cells with high CD44 expression while promoting oxidative stress. Autophagic flux impairment during EMT-mediated CD44L to CD44H cell conversion in vitro induces mitochondrial dysfunction, oxidative stress and cell death. During CD44H cell generation, transformed keratinocytes display evidence of mitophagy, including mitochondrial fragmentation, decreased mitochondrial content and mitochondrial translocation of Parkin, essential in mitophagy. RNA interference-mediated Parkin depletion attenuates CD44H cell generation. These data suggest that autophagy facilitates EMT-mediated CD44H generation via modulation of redox homeostasis and Parkin-dependent mitochondrial clearance. This is the first report to implicate mitophagy in regulation of tumour cells with high CD44 expression, representing a potential novel therapeutic avenue in cancers where EMT and CD44H cells have been implicated, including ESCC.

ZEB1 and TCF4 reciprocally modulate their transcriptional activities to regulate Wnt target gene expression.

The canonical Wnt pathway (TCF4/ß-catenin) has important roles during normal differentiation and in disease. Some Wnt functions depend on signaling gradients requiring the pathway to be tightly regulated. A key Wnt target is the transcription factor ZEB1 whose expression by cancer cells promotes tumor invasiveness by repressing the expression of epithelial specification markers and activating mesenchymal genes, including a number of Wnt targets such as LAMC2 and uPA. The ability of ZEB1 to activate/repress its target genes depends on its recruitment of corepressors (CtBP, BRG1) or coactivators (p300) although conditions under which ZEB1 binds these cofactors are not elucidated. Here, we show that TCF4 and ZEB1 reciprocally modulate each other's transcriptional activity: ZEB1 enhances TCF4/ß-catenin-mediated transcription and, in turn, Wnt signaling switches ZEB1 from a repressor into an activator. In colorectal cancer (CRC) cells with active Wnt signaling, ZEB1 enhances transcriptional activation of LAMC2 and uPA by TCF4/ß-catenin. However, in CRC cells with inactive Wnt, ZEB1 represses both genes. Reciprocal modulation of ZEB1 and TCF4 activities involves their binding to DNA and mutual interaction. Wnt signaling turns ZEB1 into an activator by replacing binding of CtBP/BRG1 in favor of p300. Using a mouse model of Wnt-induced intestinal tumorigenesis, we found that downregulation of ZEB1 reduces the expression of LAMC2 in vivo. These results identify a mechanism through which Wnt and ZEB1 transcriptional activities are modulated, offering new approaches in cancer therapy.

Parotid secretory protein binds phosphatidylinositol (3,4) bisphosphate.

Molecular interactions that direct trafficking of secreted proteins are not well-described in salivary glands. Here, we report that the soluble cargo protein Parotid Secretory Protein (PSP) is bound to the membranes of secretory granules isolated from rat parotids. This is apparently due to specific interaction with phosphatidylinositol phosphates (PtdInsP). PSP binds PtdIns(3,4)P(2), 10-fold greater than PtdIns(3,5)P(2) or PtdIns(4)P, and does not bind PtdIns(3)P or PtdIns(5)P. Human PSP synthesized in vitro also binds PtdIns(3,4)P(2). Bacterially expressed rat PSP binds PtdIns(3,4)P(2) with a K(d) of 2.4 x 10(-11) M. Other major secretory proteins (amylase, proline-rich protein) are not bound to isolated granule membranes and do not bind phosphatidylinositol phosphates. Immunofluorescence shows PtdIns(3,4)P(2) at the secretory granules, and fluorescent PtdIns(3,4)P(2) can flip from the outer leaflet to the inner leaflet of the membrane. Binding of PSP to PtdInsPs may contribute to sorting during the formation of the secretory granules, or sorting by retention during maturation of the granules.

Parotid secretory granules: crossroads of secretory pathways and protein storage.

Saliva plays an important role in digestion, host defense, and lubrication. The parotid gland contributes a variety of secretory proteins-including amylase, proline-rich proteins, and parotid secretory protein (PSP)-to these functions. The regulated secretion of salivary proteins ensures the availability of the correct mix of salivary proteins when needed. In addition, the major salivary glands are targets for gene therapy protocols aimed at targeting therapeutic proteins either to the oral cavity or to circulation. To be successful, such protocols must be based on a solid understanding of protein trafficking in salivary gland cells. In this paper, model systems available to study the secretion of salivary proteins are reviewed. Parotid secretory proteins are stored in large dense-core secretory granules that undergo stimulated secretion in response to extracellular stimulation. Secretory proteins that are not stored in large secretory granules are secreted by either the minor regulated secretory pathway, constitutive secretory pathways (apical or basolateral), or the constitutive-like secretory pathway. It is proposed that the maturing secretory granules act as a distribution center for secretory proteins in salivary acinar cells. Protein distribution or sorting is thought to involve their selective retention during secretory granule maturation. Unlike regulated secretory proteins in other cell types, salivary proteins do not exhibit calcium-induced aggregation. Instead, sulfated proteoglycans play a role in the storage of secretory proteins in parotid acinar cells. This work suggests that unique sorting and retention mechanisms are responsible for the distribution of secretory proteins to different secretory pathways from the maturing secretory granules in parotid acinar cells.

Developmental and functional evidence of a role for Zfhep in neural cell development.

The rat Zfhep gene encodes a member of the Zfh family of transcription factors having a homeodomain-like sequence and multiple zinc fingers. We examined expression of Zfhep in the rat forebrain during embryonic and postnatal development. Zfhep mRNA was strongly expressed in the progenitor cells of the ventricular zone around the lateral ventricles on E14 and E16, but showed little expression in cells that had migrated to form the developing cortex. Dual labeling with PCNA demonstrated expression of Zfhep mRNA in proliferating cells. Expression of Zfhep in the ventricular zone decreases during late development as the population of progenitor cells decreases. This pattern is distinctly different from other members of the Zfh family. We also examined the expression of Zfhep protein during retinoic acid-induced neurogenesis of P19 embryonal carcinoma cells. Zfhep is highly expressed in P19 neuroblasts, and expression decreases by the time of morphological neurogenesis. Hence, both P19 cells and embryonic brain demonstrate a loss of Zfhep expression during the transition from proliferating precursor to differentiated neural cells. We investigated a possible link between Zfhep and proliferation by treating human glial cell lines with Zfhep antisense phosphorothioate oligodeoxynucleotides. Two Zfhep antisense oligonucleotides repressed proliferation of either U-138 or U-343 glioblastoma cells more than control oligonucleotides. Based on the expression patterns of Zfhep in vivo and in the P19 cell model of neurogenesis, we suggest that Zfhep may play a role in proliferation or differentiation of neural cells.

T3-activation of the rat growth hormone gene is inhibited by a zinc finger/homeodomain protein.

Since the transcription factor Zfhep is expressed in somatotropes and binds the rat growth hormone (rGH) gene T3-response element (TRE), we investigated whether Zfhep regulates the response of this gene to T3. In cotransfection experiments, Zfhep did not regulate the native rGH promoter in the absence of T3. However, Zfhep repressed T3-mediated activation significantly in either GH(3) or JEG-3 cells. Up to 70% repression was mediated through the rGH TRE in a heterologous promoter (thymidine kinase), but was not observed with the idealized DR4 or chicken lysozyme F2 TREs. Zfhep apparently does not repress T3-mediated activation simply by competition for binding to DNA since the C-terminal DNA-binding domain of Zfhep (which is sufficient for DNA-binding) is not sufficient for repression and since cotransfection of excess thyroid hormone receptor (TR) did not prevent repression by Zfhep. These data indicate that the rGH TRE is a composite element that can integrate Zfhep and T3 regulation.

The DXD motif is required for GM2 synthase activity but is not critical for nucleotide binding.

We tested the importance of the aspartate-any residue-aspartate (DXD) motif for the enzymatic activity and nucleotide binding capacity of the Golgi glycosyltransferase GM2 synthase. We prepared point mutations of the motif, which is found in the sequence 352-VLWVDDDFV, and analyzed cells that stably expressed the mutated proteins. Whereas the folding of the mutated proteins was not seriously disrupted as judged by assembly into homodimers, Golgi localization, and secretion of a soluble form of the enzyme, exchange of the highly conserved aspartic acid residues at position 356 or 358 with alanine or asparagine reduced enzyme activity to background levels. In contrast, the D356E and D357N mutations retained weak activity, while the activity of V352A and W354A mutants was 167% and 24% that of wild-type enzyme, respectively. Despite the major effect of the DXD motif on enzymatic activity, nucleotide binding was not altered in the triple mutant D356N/D357N/D358N as revealed by binding to UDP-beads and labeling with the photoaffinity reagent, P(3)-(4-azidoanilido)uridine 5'-triphosphate (AAUTP). In summary, rather than being critical for nucleotide binding, this motif may function during catalysis in GM2 synthase, as has been proposed elsewhere for the SpsA glycosyltransferase based on its crystal structure.

Disulfide bonds of GM2 synthase homodimers. Antiparallel orientation of the catalytic domains.

GM2 synthase is a homodimer in which the subunits are joined by lumenal domain disulfide bond(s). To define the disulfide bond pattern of this enzyme, we analyzed a soluble form by chemical fragmentation, enzymatic digestion, and mass spectrometry and a full-length form by site-directed mutagenesis. All Cys residues of the lumenal domain of GM2 synthase are disulfide bonded with Cys(429) and Cys(476) forming a disulfide-bonded pair while Cys(80) and Cys(82) are disulfide bonded in combination with Cys(412) and Cys(529). Partial reduction to produce monomers converted Cys(80) and Cys(82) to free thiols while the Cys(429) to Cys(476) disulfide remained intact. CNBr cleavage at amino acid 330 produced a monomer-sized band under nonreducing conditions which was converted upon reduction to a 40-kDa fragment and a 24-kDa myc-positive fragment. Double mutation of Cys(80) and Cys(82) to Ser produced monomers but not dimers. In summary these results demonstrate that Cys(429) and Cys(476) form an intrasubunit disulfide while the intersubunit disulfides formed by both Cys(80) and Cys(82) with Cys(412) and Cys(529) are responsible for formation of the homodimer. This disulfide bond arrangement results in an antiparallel orientation of the catalytic domains of the GM2 synthase homodimer.

Evidence supporting a late Golgi location for lactosylceramide to ganglioside GM3 conversion.

Ganglioside GM2 synthase and other enzymes required for complex ganglioside synthesis were localized recently to the trans Golgi network (TGN). However, there are conflicting reports as to the location of GM3 synthase; originally this enzyme was detected in the early Golgi of rat liver but a recent report localized it to the late Golgi. We have used chimeric forms of ganglioside GM2 synthase to determine if the location of lactosylceramide (LacCer) to GM3 conversion in Chinese hamster ovary (CHO) cells was the early or late Golgi. Our approach tested whether GM3 could be utilized as a substrate by GM2 synthase chimeras which were targeted to compartments earlier than the trans Golgi, i.e., GM3 produced in the cis Golgi should be utilized by GM2 synthase located anywhere in the Golgi whereas GM3 produced in the trans Golgi should only be used by GM2 synthase located in the trans Golgi or TGN. Comparison of cell lines stably expressing these chimeras revealed that the in vivo functional activity of GM2 synthase decreased progressively as the enzyme was targeted to earlier compartments; specifically, the percentage of GM3 converted to GM2 was 83-86% for wild type enzyme, 70% for the medial Golgi targeted enzyme, 13% for the ER and cis Golgi targeted enzyme, and only 1.7% for the ER targeted enzyme. Thus, these data are consistent with a late Golgi location for LacCer to GM3 conversion in these cells.

N- and C-terminal domains direct cell type-specific sorting of chromogranin A to secretory granules.

Chromogranins are a family of regulated secretory proteins that are stored in secretory granules in endocrine and neuroendocrine cells and released in response to extracellular stimulation (regulated secretion). A conserved N-terminal disulfide bond is necessary for sorting of chromogranins in neuroendocrine PC12 cells. Surprisingly, this disulfide bond is not necessary for sorting of chromogranins in endocrine GH4C1 cells. To investigate the sorting mechanism in GH4C1 cells, we made several mutant forms removing highly conserved N- and C-terminal regions of bovine chromogranin A. Removing the conserved N-terminal disulfide bond and the conserved C-terminal dimerization and tetramerization domain did not affect the sorting of chromogranin A to the regulated secretory pathway. In contrast, removing the C-terminal 90 amino acids of chromogranin A caused rerouting to the constitutive secretory pathway and impaired aggregation properties as compared with wild-type chromogranin A. Since this mutant was sorted to the regulated secretory pathway in PC12 cells, these results demonstrate that chromogranins contain independent N- and C-terminal sorting domains that function in a cell type-specific manner. Moreover, this is the first evidence that low pH/calcium-induced aggregation is necessary for sorting of a chromogranin to the regulated secretory pathway of endocrine cells.

A zinc finger homeodomain transcription factor binds specific thyroid hormone response elements.

Thyroid hormone receptors can act through response elements (TREs) having a wide variation of sequence. We screened for transcription factors that bind the rat (r) glycoprotein hormone alpha-subunit TRE (alpha-sub), and isolated a cDNA, termed zinc finger homeodomain enhancer-binding protein (Zfhep), which encodes two separate zinc finger domains (ZD1 and ZD2), and a region similar to homeodomains. DNA-binding assays show that ZD1 or ZD2 can bind the alpha-subunit, rat growth hormone, or thyrotropin beta (TSHbeta) gene TREs, but do not bind DR4 or palindromic (pal) TREs. Methylation interference footprinting demonstrates that Zfhep binds the alpha-sub overlapping the TR-binding site. Similarly, the ZD1 protein footprints over TR-binding halfsites of the rat growth hormone (rGH) and TSHbeta TREs. Hence, Zfhep binding is dependent on sequences within and outside the AGGTCA TR-binding halfsite. Interactions of non-receptor transcription factors (such as Zfhep) with certain TREs are important to modify gene-specific regulation by thyroid hormones.

Two soluble glycosyltransferases glycosylate less efficiently in vivo than their membrane bound counterparts.

Many Golgi glycosyltransferases are type II membrane proteins which are cleaved to produce soluble forms that are released from cells. Cho and Cummings recently reported that a soluble form of alpha1, 3-galactosyltransferase was comparable to its membrane bound counterpart in its ability to galactosylate newly synthesized glycoproteins (Cho,S.K. and Cummings,R.D. (1997) J. Biol. Chem., 272, 13622-13628). To test the generality of their findings, we compared the activities of the full length and soluble forms of two such glycosyltransferases, ss1,4 N-Acetylgalactosaminyltransferase (GM2/GD2/ GA2 synthase; GalNAcT) and beta galactoside alpha2,6 sialyltransferase (alpha2,6-ST; ST6Gal I), for production of their glycoconjugate products in vivo . Unlike the full length form of GalNAcT which produced ganglioside GM2 in transfected cells, soluble GalNAcT did not produce detectable GM2 in vivo even though it possessed in vitro GalNAcT activity comparable to that of full length GalNAcT. When compared with cells expressing full length alpha2,6-ST, cells expressing a soluble form of alpha2,6-ST contained 3-fold higher alpha2,6-ST mRNA levels and secreted 7-fold greater alpha2,6-ST activity as measured in vitro , but in striking contrast contained 2- to 4-fold less of the alpha2,6-linked sialic acid moiety in cellular glycoproteins in vivo . In summary these results suggest that unlike alpha1,3-galactosyltransferase the soluble forms of these two glycosyltransferases are less efficient at glycosylation of membrane proteins and lipids in vivo than their membrane bound counterparts.

Beta 1,4 N-acetylgalactosaminyltransferase (GM2/GD2/GA2 synthase) forms homodimers in the endoplasmic reticulum: a strategy to test for dimerization of Golgi membrane proteins.

Many Golgi membrane-bound glycosyltransferases exist as intermolecular disulfide bonded species, some of which have been demonstrated to be homodimers. Evidence for homodimer formation has come primarily from radiation inactivation experiments. We utilized an alternative strategy to test for homodimer formation of the cloned beta 1,4 N-acetylgalactosaminyltransferase (GalNAcT) responsible for synthesis of the glycosphingolipids GM2, GD2, and GA2. We stably transfected CHO cells with myc epitopetagged GalNAcT, which localizes primarily to the Golgi, and a hemagglutinin (HA) epitope-tagged GalNAcT fusion protein in which the cytoplasmic domain of GalNAcT was replaced by an ER retention signal. We then sought evidence for dimer formation between the two forms of GalNAcT. Immunoprecipitation with anti-myc or anti-HA co-immunoprecipitated the HA-tagged form or the myc-tagged form, respectively, providing evidence for the physical association of the two forms of GalNAcT. As a result of this association, GalNAcT/myc increased in the ER as demonstrated by Western blots and immunofluorescence. The rapid formation of dimers provided further evidence for dimer formation occurring in the ER. In summary, these results demonstrate that GalNAcT forms homodimers as a result of intermolecular disulfide bond formation in the ER. Furthermore, this ER motif strategy is potentially useful for demonstrating homodimer formation of other Golgi enzymes.

Expression of multiple thyroid hormone receptor mRNAs in human oocytes, cumulus cells, and granulosa cells.

Thyroid hormones have diverse effects on ovarian function. We examined the expression of thyroid hormone receptor (TR) mRNAs (including TRalpha-1, TRbeta-1, TRbeta-2, and c-erbAalpha-2 isoforms) in three types of cells from human follicles, and determined the concentration of free tri-iodothyronine (T3) present in human follicular fluid. Human failed-fertilized oocytes, granulosa (GC) and cumulus (CC) cells from patients of the in-vitro fertilization (IVF) programme at Alliant Hospital Fertility Center were used to detect TR mRNA expression using reverse transcription-polymerase chain reaction (RT-PCR) followed by Southern blot analysis. Human spermatozoa were also analysed to determine whether results obtained with CC would be affected by the presence of spermatozoa. beta-Actin mRNA was amplified in each cell type as a positive control for the RT-PCR. Our results show that human oocytes express TRalpha-1, TRbeta-1, TRbeta-2, and c-erbAalpha-2 mRNAs and that these same isoforms are expressed in both human granulosa cells and cumulus cells. No differences were detected in the apparent amounts of RT-PCR products when comparing GC with CC, suggesting a similar pattern of expression of these RNAs. beta-actin mRNA was detected in spermatozoa, but TRalpha-1 expression was not detectable. The concentrations of free T3 measured in follicular fluid were similar to, or slightly below, those in serum of euthyroid patients. These data demonstrated that several isoforms of TR mRNA are expressed in the human oocyte, and hence thyroid hormone may have direct affects on the oocyte, as well as on GC and CC. In addition thyroid hormone may have indirect effects on the oocytes via the CC.

Beta1,4-N-acetylgalactosaminyltransferase (GM2 synthase) is released from Golgi membranes as a neuraminidase-sensitive, disulfide-bonded dimer by a cathepsin D-like protease.

Many Golgi membrane-bound glycosyltransferases are released from cells in a soluble form. To characterize this release process, we stably transfected Chinese hamster ovary cells with three myc epitope-tagged forms of cloned beta1, 4-N-acetylgalactosaminyltransferase (GalNAcT); two of these forms resided in the Golgi, while the third was retained in the ER. GalNAcT was released into the culture medium from cells transfected with the Golgi forms but not with the ER form of the enzyme. The medium from cells transfected with the Golgi forms contained disulfide-bonded dimers of GalNAcT, which carried neuraminidase sensitive, complex N-linked carbohydrate chains. This soluble species represented the major degradation product of cellular GalNAcT, which turned over with a half-time of about 1.7 h. The soluble species consisted of a mixture of truncated GalNAcT molecules, the major form of which was produced by cleavage near the boundary between the transmembrane and lumenal domains between Leu-23 and Tyr-24. This cleavage site fits the sequence pattern for sites cleaved by cathepsin D (van Noort, J.M., and van der Drift, A. C.M. (1989) J. Biol. Chem. 264, 14159-14164). These findings suggest that GalNAcT is converted from a membrane-bound to a soluble form as a result of cleavage by a cathepsin D-like protease in a compartment late in the Golgi secretory pathway.

Alternative splicing gives rise to two isoforms of Zfhep, a zinc finger/homeodomain protein that binds T3-response elements.

We have previously isolated a cDNA for a transcription factor referred to as Zfhep (zinc finger homeodomain enhancer-binding protein) containing two separate zinc finger domains, ZD1 and ZD2, each of which binds DNA, and a homeodomain. The rat Zfhep cDNA lacks a 5'-methionine codon, present in some homologs from other species. Hence, the aim of this work was to isolate the 5'-end of the rat Zfhep cDNA. Zfhep-2 cDNA was isolated, having a total length of 2.5 kbp, including more than 1.1 kbp of novel sequence followed by 1.4 kbp identical to the Zfhep-1 clone. The 1.1 kbp of novel sequence contains multiple stop codons in all reading frames, suggesting that it represents the 5'-untranslated (5'-UT) region of the rat Zfhep-2 mRNA. However, the Zfhep-2 clone does not contain the extreme 5'-exon(s) of the Zfhep-1 coding sequence, possibly due to alternative splicing of Zfhep RNA. To distinguish between a splice junction versus an intron-exon junction, the polymerase chain reaction (PCR) with rat genomic DNA and junction-flanking primers from the Zfhep-2 sequence was conducted. No bands were amplified from the genomic DNA by two different pairs of primers, indicating that the Zfhep-2-specific sequence is not intronic. Ribonuclease protection assays were performed to investigate the expression of multiple Zfhep mRNAs. Two protected bands were detected, and both were identified in total RNA or mRNA of rat ovary, hindbrain, forebrain, heart, kidney, small intestine, and GH4C1 cells. Zfhep-2 represents about 20% of the Zfhep RNA in each tissue. Hence, two mRNAs are expressed in these tissues, confirming the alternative splicing. To confirm independently the presence of both Zfhep-2 and Zfhep-1 mRNAs, reverse transcriptase (RT)-PCR was done using primers that span the Zfhep splice site. Specific bands representing both RNAs were obtained. The Zfhep-2 PCR product was subcloned and DNA sequence analysis confirmed the absence of ATG codons near the 5'-end of the open reading frame. The theoretical translation of the Zfhep-2 clone predicts a smaller protein than Zfhep-1. In vitro translation in reticulocyte lysates showed that Zfhep-2 is about 40 kD smaller than Zfhep-1. Hence, Zfhep-2 apparently lacks most of the first zinc finger domain (ZD1) of Zfhep-1. Because the two zinc finger domains bind different DNA sequences, Zfhep-2 is predicted to bind to only a subset of genes recognized by Zfhep-1.

Cloned beta 1,4N-acetylgalactosaminyltransferase: subcellular localization and formation of disulfide bonded species.

Cloned human beta 1,4N-acetylgalactosaminyltransferase (GalNAcT) catalyzes the synthesis of the glycosphingolipids GM2, GD2, and gangliotriosylceramide. To determine the subcellular location of this enzyme and whether it exists in intermolecular disulfide bonded species, we stably transfected Chinese hamster ovary (CHO) cells with three myc epitope-tagged forms of the GalNAcT gene: the native enzyme; the lumenal domain of GalNAcT fused to the cytoplasmic and transmembrane domains of N-acetylglucosaminyltransferase I (GNT); and the transmembrane and lumenal domains of GalNAcT fused to the cytoplasmic domain of the Iip33 form of human invariant chain in order to retain the enzyme in the endoplasmic reticulum (ER). Immunoelectron microscopic analysis with anti-myc revealed that GalNAcT/myc was present throughout the Golgi stack, the GNT/GalNAcT/myc form was restricted primarily to the medial Golgi cisternae, and the Iip33/GalNAcT/myc form was restricted to the ER. Cells transfected with each of the three constructs contained high levels of GM2 synthase activity in vitro, but only the GalNAcT/myc form and the GNT/GalNAcT/myc forms were able to synthesize the GM2 product in vivo. The enzyme produced by all three constructs was present in the transfected cells in a disulfide bonded form having a molecular size consistent with that of a homodimer or higher aggregate.

An N-terminal hydrophobic peak is the sorting signal of regulated secretory proteins.

Endocrine and exocrine cells each contain a regulated and constitutive secretory pathway. The presence of two distinct secretory pathways in the same cell type requires a sorting step to direct secretory proteins to the correct pathway. It is thought that regulated secretory proteins contain a specific sorting signal. However, this signal has not been identified. Amino acid sequence comparisons have not revealed any significant similarity between different regulated secretory proteins, suggesting that the sorting signal does not consist of a conserved primary sequence. In the present report, we have analyzed the predicted secondary structures of regulated secretory proteins and identified an N-terminal hydrophobic peak (NHP) which is located approximately from amino acids 9-26, overlaps with a predicted alpha-helix and contains charged amino acid residues. This signal is present in regulated secretory proteins that exhibit an N-terminal sorting sequence, but it is absent from constitutively secreted proteins and proteins where the sorting sequence is not located near the N-terminus. It appears that the NHP is both necessary and sufficient for sorting of many secretory proteins to the regulated secretory pathway.

Cloned beta 1,4 N-acetylgalactosaminyltransferase synthesizes GA2 as well as gangliosides GM2 and GD2. GM3 synthesis has priority over GA2 synthesis for utilization of lactosylceramide substrate in vivo.

Earlier studies reached conflicting conclusions as to the ability of the beta 1,4 N-acetylgalactosaminyltransferase (GalNAc-T) that synthesizes gangliosides GM2 and GD2 to also produce gangliotriosylceramide (GA2). We constructed an experimental system in which to address this question. Wild type Chinese hamster ovary (CHO) cells contain ganglioside GM3 as the most complex glycosphingolipid (GSL), whereas the CHO glycosylation mutant Lec2, which is deficient in sialylation, accumulates lactosylceramide with little GM3 being produced. We transfected both cell types with a plasmid containing a cloned GalNAc-T. Whereas transfected CHO cells produced GM2 as the major complex GSL, the major product in transfected Lec2 cells was GA2. Both types of transfected cells but not the untransfected cells expressed the transfected gene and contained high levels of enzyme activity for synthesizing both GM2 and GA2 in vitro. In summary, these results indicate that this enzyme can in fact synthesize GA2 as well as GM2 and GD2. In addition, these findings suggest that in CHO cells the synthesis of GM3 in vivo has priority over GA2 synthesis for utilization of the substrate lactosylceramide, resulting in little GA2 being produced even though GalNAc-T is present and active. Thus, competition for substrate between glycosylation pathways may have profound effects on the GSL pattern of cells.