Tissue-specific expression in the salivary glands of transgenic mice.

Using a DNA construct, named Lama, derived from the murine parotid secretory protein (PSP) gene, we have obtained salivary gland specific gene expression in transgenic mice. Lama is a PSP minigene and allows analysis of the PSP gene 5' regulatory region by transgenesis. We show here that the regulatory region included in Lama with 4.6 kb of 5' flanking sequence is sufficient to direct expression specifically to the salivary glands. The expression level in the parotid gland is only about one percent of the PSP mRNA level, while that of the sublingual gland is near the PSP mRNA level. This suggests significant differences in the PSP gene regulation in the two glands. In addition, Lama is a secretory expression vector in which cDNAs or genomic fragments can be inserted. We demonstrate that the Lama construct can direct the expression of a heterologous cDNA encoding the C-terminal peptide of human factor VIII to salivary glands and that the corresponding peptide is secreted into saliva.     

Coordination of murine parotid secretory protein and salivary amylase expression.

PSP, parotid secretory protein, and salivary amylase are the major secretory proteins of mouse parotid gland where they appear in a constant ratio. Here we describe the isolation of the PSP gene and show through expression analysis on this and the salivary amylase gene that the two genes are transcribed in a coordinate fashion in adult animals, whereas the activation profiles are different during postnatal development. An explanation is put forward that involves activation of the genes at different stages of the acinar cell differentiation, leading in adults to the maximal and thus proportionate expression.     

Molecular cloning of mouse PSP mRNA.

PSP is the most abundant translation product of mouse parotid glands where its production is co-ordinated with that of salivary amylase. The synthesis of these two proteins apparently is restricted to this tissue. In order to enable us to study common regulatory elements in the genes of the two proteins, double stranded cDNA, synthesized for parotid gland poly (A)+ RNA, was cloned. DNA sequencing of three clones complementary to the most abundant messenger indicated overlap and resulted in a total sequence of 867 nucleotides. Translation of this sequence revealed that at one end the amino acid sequence was the same as the N-terminal sequence of PSP. The sequence contains 60 nucleotides coding for part of or the complete signal peptide, 645 nucleotides coding for the PSP protein, and 162 nucleotides that apparently are not translated. Southern blot analysis suggests a simple structure for the PSP gene in mouse and man.     

Concentration and distribution of insulin and insulin-like protein in the organs of normal and diabetic mice

Presence of insulin or insulin-like protein has been studied in mouse liver, kidneys, lungs, duodenum, jejunum, submandibular and parotid salivary glands, in femoral, diaphragmal and abdominal wall muscles by means of the immunofluorescent method. In order to understand the role of the extrapancreatic insulin for compensation of the insular insufficiency, corresponding organs have been examined in mice with alloxan diabetes. The immunoreactive insulin is proved to be present only in cells of the granular parts of the salivary tubules of the submandibular and striated ducts of the parotid glands. As demonstrates microfluorometry, a relative amount of insulin in the submandibular gland cells is 1.5 times and in the parotid gland cells--2 times as small as in beta-cells of the pancreatic glands. Under alloxan diabetes insulin content in the salivary gland cells decreases by 1.3-1.9 times (in the beta-cells--by 2.7 times). This may designate that the extrapancreatic insulin (or insulin-like protein) participates in compensation of hypoinsulinemia. In mice with alloxan diabetes, immunoreactivity of insulin is also revealed in hepatocytes.     

The BSP30 salivary proteins from cattle,LUNX/PLUNC and von Ebner's minor salivary gland protein are members of the PSP/LBP superfamily of proteins

Saliva influences rumen function in cattle, yet the biochemical role for most of the bovine salivary proteins (BSPs) has yet to be established. Two cDNAs (BSP30a and BSP30b) from bovine parotid salivary gland were cloned and sequenced, each coding for alternate forms of a prominent protein in bovine saliva. The BSP30 cDNAs share 96% sequence identity with each other at the DNA level and 83% at the amino acid level, and appear to arise from separate genes. The predicted BSP30a and BSP30b proteins share 26-36% amino acid identity with parotid secretory protein (PSP) from mouse, rat and human. BSP30 and PSP are in turn more distantly related to a wider group of proteins that includes lung-specific X protein, also known as palate, lung, and nasal epithelium clone (LUNX/PLUNC), von Ebner's minor salivary gland protein (VEMSGP), bactericidal permeability increasing protein (BPI), lipopolysaccharide binding protein (LBP), cholesteryl ester transfer protein (CETP), and the putative olfactory ligand-binding proteins RYA3 and RY2G5. Bovine cDNAs encoding homologs of LUNX/PLUNC and VEMSGP were isolated and sequenced. Northern blot analysis showed that LUNX/PLUNC, BSP30 and VEMSGP are expressed in bovine salivary tissue and airways, and that they have non-identical patterns of expression in these tissues. The expression of both BSP30a and BSP30b is restricted to salivary tissue, but within this tissue they have distinct patterns of expression. The proximity of the human genes coding for the PSP/LBP superfamily on HSA20q11.2, their similar amino acid sequence, and common exon segmentation strongly suggest that these genes evolved from a common ancestral gene. Furthermore, they imply that the BSP30a and BSP30b proteins may have a function in common with other members of this gene family.     

Proline-rich proteins and glycoproteins: expression of salivary gland multigene families

Our recent research interests have focused on a group of unusual proteins and glycoproteins high in proline content, or the so-called proline-rich proteins (PRPs). The PRPs are tissue-specific expressions of salivary gland multigene families. Normally PRPs are not detected or are present in very low amounts in rat, mouse and hamster salivary glands, but these unusual proteins are dramatically induced by treatment with the catecholamine isoproterenol. The structures and organizations of several PRP mRNAs and PRP genes have been determined. The amino acid sequences of all PRPs show 4 distinct regions, namely, a signal peptide, a transition region, a repeat region and a carboxyl-terminal region. Glycoproteins induced by isoproterenol treatment may be N-glycosylated or O-glycosylated. The N-glycosylated glycoprotein GP-158 from rat submandibular glands has a 12 amino acid glycopeptide which repeats possibly 49 times. Proline-rich proteins of the parotid glands of rats and mice are also greatly induced by dietary tannins. The apparent unique occurrence of PRPs in saliva suggests that one biological role is to neutralize the detrimental effects of dietary tannins and other polyphenols. The upstream regions of the mouse and hamster PRP genes contain cyclic AMP-regulated sequences as demonstrated by deletions and transient transfections. The PRP multigene family members of mouse are all located on chromosome 8.     

Phospholipid transfer protein gene knock-out mice have low high density lipoprotein levels, due to hypercatabolism, and accumulate apoA-IV-rich lamellar lipoproteins

Phospholipid transfer protein gene knock-out (Pltp KO) mice have defective transfer of very low density lipoprotein (VLDL) phospholipids into high density lipoprotein (HDL) and markedly decreased HDL levels (Jiang et al. 1999. J. Clin. Invest. 103: 907-914). These animals also accumulated VLDL- and LDL-sized lipoproteins on a high saturated fat diet. The goals of this study were to further characterize the abnormal lipoproteins of Pltp KO mice and to determine the mechanisms responsible for low HDL levels. A lipoprotein fraction enriched in lamellar structures was isolated from the low density lipoprotein (LDL) region and was shown to be phospholipid- and free cholesterol-rich and to have apoA-IV (55%) and apoE (25%) as major apolipoproteins. The lamellar lipoproteins accumulating in these mice probably represent surface material derived from triglyceride-rich lipoproteins (TRL). The HDL was found to be protein-rich (primarily apoA-I) and specifically depleted in phosphatidylcholine (PC) (28% in wild-type mice (WT) vs. 15% in Pltp KO mice, P < 0.001). Unexpectedly, turnover studies using autologous HDL revealed a profound 4-fold increase in the catabolism of HDL protein and cholesteryl ester in Pltp KO mice compared to wild-type, with minor differences in synthesis rates. In contrast, injection of WT mouse HDL into Pltp KO mice showed only a 2-fold increase in fractional catabolism. Reminiscent of the defect in Tangier disease, the failure of transfer of PC from TRL into the HDL fraction results in dramatic hypercatabolism of HDL. These results suggest that defective phospholipid transfer from TRL into HDL, arising from decreased lipolysis or decreased PLTP activity, could lead to hypoalphalipoproteinemia characterized by hypercatabolism of HDL protein. lipoprotein levels, due to hypercatabolism, and accumulate apoA-IV-rich lamellar lipoproteins.     

Inducible expression of phospholipid transfer protein (PLTP) in transgenic mice: acute effects of PLTP on lipoprotein metabolism

One main determinant in high-density lipoprotein (HDL) metabolism is phospholipid transfer protein (PLTP), a plasma protein that is associated with HDL. In transgenic mice overexpressing human PLTP we found that elevated plasma PLTP levels dose-dependently increased the susceptibility to diet-induced atherosclerosis. This could be mainly due to the fact that most functions of PLTP are potentially atherogenic, such as decreasing plasma HDL levels. To further elucidate the role of PLTP in lipoprotein metabolism and atherosclerosis we generated a novel transgenic mouse model that allows conditional expression of human PLTP. In this mouse model a human PLTP encoding sequence is controlled by a Tet-On system. Upon induction of PLTP expression, our mouse model showed a strongly increased PLTP activity (from 3.0 ± 0.6 to 11.4 ± 2.8 AU, p < 0.001). The increase in PLTP activity resulted in an acute decrease in plasma cholesterol of 33% and a comparable decrease in phospholipids. The decrease in total plasma cholesterol and phospholipids was caused by a 35% decrease in HDL-cholesterol level and a 41% decrease in HDL-phospholipid level. These results demonstrate the feasibility of our mouse model to induce an acute elevation of PLTP activity, which is easily reversible. As a direct consequence of an increase in PLTP activity, HDL-cholesterol and HDL-phospholipid levels strongly decrease. Using this mouse model, it will be possible to study the effects of acute elevation of PLTP activity on lipoprotein metabolism and pre-existing atherosclerosis.     

Expression and distribution of osteopontin and matrix metalloproteinase (MMP)-3 and -7 in mouse salivary glands

We investigated the expression and distribution of osteopontin in mouse salivary glands. Western blot analysis showed intense positive bands at the predicted molecular mass (about 60 kDa) in mouse parotid and sublingual glands. However, a cross-reacted band around 30 kDa was strongly detected in submandibular glands. Indirect immunofluorescent analysis showed that osteopontin was localized at the luminal (apical) membranes of the acinar cells in parotid and sublingual glands. However, it was not detected in acinar cells of submandibular glands. No expression was found in ductal cells of any glands. We also examined the expression of matrix metalloproteinase (MMP)-3 and -7. In parotid gland, MMP-3 was observed at 57 kDa, indicating a latent form, but MMP-7 was not detected. In contrast, MMP-7 definitely was observed at 28 kDa area in submandibular gland, whereas MMP-3 was not detected. These results suggest that osteopontin localizes at luminal sites of acinar cells and may be associated with saliva secretion in mouse salivary gland. It is also suggested that osteopontin may be cleaved by MMP-7 in mouse submandibular gland.     

Inheritance of a Parotid Secretory Protein in Mice and Its Use in Determining Salivary Amylase Quantitative Variants

Among inbred strains of mice, a major protein, PSP, produced and secreted by the parotid glands, shows variation in electrophoretic mobility and in the peptides produced by cyanogen bromide treatment. This variation is inherited as a single Mendelian factor with two alleles showing co-dominant expression. In genetic crosses, it segregates independently from the amylase complex and is closely linked to the agouti locus on chromosome 2. The protein ratios between amylase and PSP in saliva, obtained by scanning of electrophoretic gel separations, were found to reflect genetic differences in salivary amylase production in strains YBR/Cv and C3H/As.     

The role of protein kinase C on amylase secretion from rat parotid gland

The activities of Ca2+.phospholipid-dependent protein kinase (protein kinase C) in rat salivary gland were assayed using synthetic peptide syntide-2(Pro-Leu-Ala-Arg-Thr-Leu-Ser-Val-Ala-Gly-Leu-Pro-Gly-Lys- Lys) as substrate. Levels of the protein kinase C were less than 0.05 units/g in the parotid and submandibular glands. The protein kinase C inhibitor, H-7, inhibited amylase secretion from rat parotid gland stimulated by PMA or the combination of phosphatidylserine and 1,2-diolein. The results supported the hypothesis of the secretory mechanism that protein kinase C mediates amylase secretion in rat parotid glands.     

Different localizations of 21 and 27 kDa gap-junction proteins in rat salivary glands

Antibodies against 21 and 27 kDa gap-junction proteins from rat liver were used to examine the identification and localization of gap-junction proteins in rat salivary glands. Acinar cells of the submandibular glands and parotid glands stained well for the 27 kDa gap junction protein and less intensely for the 21 kDa protein. Acinar cells of the sublingual glands were stained heavily for the 27 kDa gap junction protein and stained well for 21 kDa gap junction protein. No 27 kDa protein was observed in the ducts of the salivary glands. The 21 kDa gap-junction protein was distributed in some of the intercalated ducts in the parotid and submandibular glands. Immunoblotting of an extract of parotid glands with antibodies against 21 and 27 kDa gap-junction proteins revealed the presence of 21 and 27 kDa proteins in the parotid glands. It is concluded that the 27 kDa gap-junction protein in tistributed as a major component of the gap junctions in the acinar cells of all the salivary glands; the 21 kDa protein is localized as a minor component in the acinar cells and some portions of the intercalated ducts in the salivary glands. It is possible that these gap-junction proteins might contribute to the regulation of function of the salivary glands.     

Transgenic mice expressing bacterial phytase as a model for phosphorus pollution control

We have developed transgenic mouse models to determine whether endogenous expression of phytase transgenes in the digestive tract of monogastric animals can increase the bioavailability of dietary phytate, a major but indigestible form of dietary phosphorus. We constructed phytase transgenes composed of the appA phytase gene from Escherichia coli regulated for expression in salivary glands by the rat R15 proline-rich protein promoter or by the mouse parotid secretory protein promoter. Transgenic phytase is highly expressed in the parotid salivary glands and secreted in saliva as an enzymatically active 55 kDa glycosylated protein. Expression of salivary phytase reduces fecal phosphorus by 11%. These results suggest that the introduction of salivary phytase transgenes into monogastric farm animals offers a promising biological approach to relieving the requirement for dietary phosphate supplements and to reducing phosphorus pollution from animal agriculture.     

The Main Regulatory Region in the Murine PSP Gene is a Parotid Gland Enhancer

The murine PSP gene is expressed at a high-level in the parotid glands. To extend the knowledge of parotid gland expression and develop tools for expression of heterologous proteins in this tissue, the regulation of the PSP gene was studied using transgenic mice. High-level parotid gland expression of the PSP gene was indicated to depend on a novel regulatory region situated between –8.0 and –6.5 kb. Together with previous results this indicates that the main regulatory elements in the PSP gene are situated between –8.0 to –3.1 kb. This region was shown to activate a heterologous SV40 early promoter in the parotid glands of transgenic mice, suggesting that the PSP gene is controlled by enhancer sequences. A novel Psp derived 9.7 kb parotid gland expression cassette, Lama IV, carrying all known regulatory regions in the PSP gene was expressed at high-levels in the parotid glands and should prove highly useful for expression of heterologous proteins in the saliva of transgenic mice.     

High levels of GM1-ganglioside beta-galactosidase in the salivary glands and GM1-like-ganglioside storage in parotids of deficient mice.

We have previously demonstrated high levels of GM1-ganglioside beta-galactosidase (beta-gal) in the salivary glands of Swiss-Webster mice (Nowroozi et al., J Craniofac Genet Dev Biol 18:51, 1998), and suggested that this activity reflects an important role for the lysosome in catabolism of salivary glycoconjugates. Here, we characterized and compared activities of lysosomal glycosidases among the salivary glands, spleen, and muscle of C57BL/6 mice, beta-gal hexosaminidase, and beta-glucuronidase activities are high in all three glands relative to muscle. Enzyme activities in the sublingual gland were substantially higher than in the submandibular and parotid glands. Spleen displays levels of activity that are comparable or higher (for beta-glucuronidase) than those in the salivary glands, whereas muscle displays substantially lower levels of these lysosomal glycosidases. In order to investigate the role of beta-gal in the salivary glands, we further characterized the salivary phenotype of knock-out mice deficient in this enzyme, mimicking human GM1-gangliosidosis. In contrast with the relative levels of beta-gal specific-activity among the salivary glands, only the parotid developed severe, generalized, degenerative histopathological changes in beta-gal-deficient knock-out mice. GM1-like-ganglioside, typically found at high levels only in the nerve tissue, where its exact function is still not clear, was demonstrated in storage vacuoles of the parotid glands of the deficient mice by binding of cholera toxin subunit B. Thus, beta-gal activity observed in the parotid gland most likely reflects its role in GM1-ganglioside catabolism, and this ganglioside, never previously reported in the salivary glands, may have a role in parotid exocrine secretory functions. beta-gal may also serve in secretory glycoprotein catabolism in other salivary glands, but this function may be non-essential for these glands.