Systemic overexpression of IFN-gamma and IL-5 exacerbates early phase reaction and conjunctival eosinophilia, respectively, in experimental allergic conjunctivitis.

AIMS/BACKGROUND: To investigate how systemic overexpression of IL-4, IL-5 and IFN-gamma affects the severity of experimental conjunctivitis (EC) in mice. METHODS: The tibialis anterior muscle of naive BALB/c mice was electroporated with IL-4, IL-5, IFN-gamma or a control gene, and then the mice blood and conjunctivas were harvested to measure the eosinophil content in these tissues. To evaluate the effects of cytokine gene electroporation on the early-phase reaction (EPR), cytokine gene-electroporated ragweed (RW)-immunised mice were intravenously injected with Evans Blue (EB) and then challenged with RW in eye-drops. Thirty minutes later, their conjunctivas were harvested to extract EB and evaluate the EPR. Additionally, 24 h after RW challenge, conjunctivas were harvested from cytokine gene-electroporated RW-immunised mice, which had not received intravenous injection of EB, to measure conjunctival eosinophilia. RESULTS: Significantly more eosinophils were detected in the blood and conjunctivas of IL-5-electroporated mice in which EC was not induced. The intensity of the EPR was significantly greater in IFN-gamma-electroporated mice. Significantly greater eosinophil infiltration was seen in the conjunctivas of IL-5-electroporated EC-developing mice. CONCLUSIONS: It appears that systemic IL-5 and IFN-gamma play different roles in the development of EC.

Stimulation of cAMP signalling allows isolation of clonal pancreatic precursor cells from adult mouse pancreas.

AIMS/HYPOTHESIS: Duct cells of the pancreas are thought to include latent progenitors of islet endocrine cells that can be induced to differentiate by appropriate morphogens. Here we developed a method for isolating pancreatic ductal epithelial cells from adult mice that overcomes the shortcomings of previous methods. MATERIALS AND METHODS: Pancreatic ductal cells were grown in serum-free DMEM/F12 medium in the presence of cholera toxin or 8-bromo-cyclic adenosine monophosphate, which is known to be an intracellular cAMP generator. Single cell cloning was performed by limiting dilution in serum-free medium. RESULTS: The isolated clonal cells expressed high levels of cytokeratin and Ipf1 (formerly known as Pdx-1). Adenovirus-mediated expression of ngn3 (also known as Neurog3) and Ptf1a in these cells induced expression of insulin and somatostatin, and of carboxypeptidase A, respectively. Furthermore, albumin production was induced by dexamethasone or by long-term culture in serum-containing medium. CONCLUSIONS/INTERPRETATION: Stimulation of the cAMP-dependent signalling allowed us to isolate clonal pancreatic ductal cells from adult mice. These cells are able to partially differentiate into endocrine cells, exocrine cells and hepatocyte-like cells and are therefore considered to have the characteristics of endodermal progenitor cells.

Development of autoimmune diabetes in glutamic acid decarboxylase 65 (GAD65) knockout NOD mice.

AIMS/HYPOTHESIS: Type 1 diabetes mellitus, a T-cell-mediated autoimmune disease, results from the selective destruction of insulin-producing pancreatic beta cells. Autoantibodies against beta-cell components are used clinically as sensitive markers of this disease; however, their physiological role has not been clear. To investigate the role of glutamic acid decarboxylase 65 (GAD65) in the development of the Type 1 diabetes of non-obese diabetic (NOD) mice, we analysed and characterised NOD mice with targeted disruption of the GAD65 gene. METHODS: GAD65-deficient mice were previously established. After backcrossing the knockout mutation onto the NOD genetic background for up to eight generations, female littermates of the three resulting genotypes were produced by intercrossing: GAD65 +/+ (n=23), GAD65 +/- (n=62), and GAD65 -/- (n=31). RESULTS: The cumulative incidence of autoimmune diabetes showed no significant difference among the three groups in longitudinal studies using the Kaplan-Meier method. Islet morphology showed that the progression of islet infiltration did not differ significantly between the three groups. CONCLUSION/INTERPRETATION: The cumulative incidence of autoimmune diabetes was not influenced by the GAD65 deficiency. These data suggest that GAD65 is not a major regulatory target of beta-cell autoimmunity in NOD mice.

IGF-I gene transfer by electroporation promotes regeneration in a muscle injury model.

The goal of this study was to determine whether insulin-like growth factor-I (IGF-I) gene delivery by electroporation promotes repair after muscle injury. An injury-repair model was created using mice in which a hamstring muscle was cut and sutured. A total of 50 microg of IGF-I DNA or green fluorescent protein (GFP) DNA (both in pCAGGS) was injected into the lesion and introduced into muscle cells by electrostimulation using an electric pulse generator. The number of regenerating muscle fibers in the IGF-I DNA group was significantly more than that in the GFP DNA group at 2 weeks after injection. The diameter of regenerating muscle fibers from the IGF-I DNA group was larger than that of the GFP DNA group at 4 weeks after injection. There was no significant difference in the serum IGF-I concentration between the IGF-I DNA group and the GFP DNA group at 1, 2, and 4 weeks after injection. However, muscle IGF-I concentration in the IGF-I DNA injection group was significantly greater than that in the GFP DNA injection group at 2 weeks after injection. These results demonstrated that the effects of enhanced IGF-I production were local and limited to the injected area. The ratio (injected/uninjected; intact) of the amplitude of compound muscle action potentials (CMAP) in the IGF-I DNA injection group was greater than that in the GFP DNA injection group at 4 weeks after injection and of the control group. In conclusion, IGF-I gene transfer by electroporation proved to be a simple, safe, inexpensive, and effective method to promote the regeneration of injured muscles in our injury model.

Usefulness of double gene construct for rapid identification of transgenic mice exhibiting tissue-specific gene expression.

Identification of transgenics still requires PCR and genomic Southern blot hybridization of genomic DNA isolated from tail pieces. Furthermore, identification of transgene-expressing transgenics (hereafter called "expressor") requires mRNA analyses (RT-PCR and Northern blot hybridization) or protein analysis (Western blotting and immunohistochemical staining using specific antibodies). These approaches are often labor-intensive and time-consuming. We developed a technique that simplifies the process of screening expressor transgenics using enhanced green fluorescent protein (EGFP), a noninvasive reporter recently utilized in a variety of organisms, including mice, as a tag. We constructed a MNCE transgene consisting of two expression units, MBP-NCre (termed "MN") and CAG-EGFP (termed "CE"). MN consists of a myelin basic protein (MBP) promoter and NCre gene (Cre gene carrying a nuclear localization signal (NLS) sequence at its 5' end). CE consists of a promoter element, CAG composed of cytomegalovirus (CMV) enhancer and chicken beta-actin promoter, and EGFP cDNA. Of a total of 72 F0 mice obtained after pronuclear injection of MNCE at 1-cell egg stage, 15 were found to express EGFP when the tail, eye, and inner surface of the ear were inspected for EGFP fluorescence under UV illumination at weaning stage. These fluorescent mice were found to possess MNCE and to express NCre mRNA in a brain-specific manner. Mice exhibiting no fluorescence were transgenic or nontransgenic. Mice carrying MNCE, but exhibiting no fluorescence, never expressed NCre mRNA in any organs tested. These findings indicate that (i) direct inspection of the surface of mice for fluorescence under UV illumination enables identification of expressor transgenics without performances of the molecular biological analyses mentioned above, and (ii) systemic promoters such as CAG do not affect the tissue-specificity of a tissue-specific promoter such as MBP promoter, which is located upstream of CAG by approximately 2 kb.

Cytokine gene therapy for myocarditis by in vivo electroporation.

Cytokines are important pathophysiologic and pathogenic factors in cardiovascular disorders, including viral myocarditis. We attempted to treat viral myocarditis with cytokine gene therapy by transferring an inhibitory cytokine, IL-1 receptor antagonist (IL-1ra) or viral IL-10 (vIL-10), by in vivo electroporation, a new method for gene transfer into muscle. Four-week-old male DBA/2 mice were inoculated intraperitoneally with 10 PFU of encephalomyocarditis virus. Immediately after virus inoculation, an expression plasmid carrying IL-1ra or vIL-10 was injected into tibialis anterior muscles followed by electroporation. Serum levels of IL1ra and vIL-10 reached 10.5 and 2.3 ng/ml, respectively, on day 5, when gene expression reached its peak. Histopathological examination showed that myocardial cellular infiltration was improved in mice treated with IL-1ra or vIL-10 compared with the control group. On day 14 after the onset of myocarditis, transfer of IL1ra or vIL-10 expression plasmid had significantly improved the survival rates of the animals. The expression of TNF-alpha was decreased to 0.60-fold (p < 0.005) and inducible nitric oxide synthase (iNOS) 0.43-fold (p < 0.005) by IL-1ra treatment, and the expression of IFN-gamma in the heart was decreased to 0.35-fold (p < 0.05), and iNOS 0.21-fold (p < 0.005), by vIL-10 relative to the controls. These results show that gene therapy with IL-1ra or vIL-10 expression plasmid was effective in the treatment of viral myocarditis, and in vivo electroporation may be a useful method by which to deliver cytokine therapy in cardiovascular diseases.

Liver regeneration on chicken chorioallantoic membrane.

To explore how the liver regenerates, liver pieces from 15-day-old chicken embryos were grafted onto the chorioallantoic membrane (CAM) of 9-day-old chicken embryos and cultured for 11 days at the longest. The cultured liver pieces were examined histologically. The liver implants were gradually engulfed into the CAM and underwent necrosis of hepatocytes, except in their peripheral areas, during the first 1-4 days after grafting. Surviving cells in the peripheral areas began to proliferate 4 days after grafting. Thereafter, the cells were assembled into normal liver tissues and represented almost all the areas of the implants 9 days after grafting. Only after penetration of blood vessels from CAM did the liver implants enter a phase of rapid regeneration to form well-organized liver tissues. At the early stage of regeneration, the cells at the peripheral areas did not produce albumin, but reproduced it in the regenerated liver tissues, implying that hepatocytes restored their functions that were temporarily lost in the process of regeneration. Thus, we concluded that the liver pieces from 15-day-old chicken embryos had the ability to form normal liver tissues on CAM and that the blood supply played an important role in liver regeneration.

Loss of hippocampal CA3 pyramidal neurons in mice lacking STAM1.

STAM1, a member of the STAM (signal transducing adapter molecule) family, has a unique structure containing a Src homology 3 domain and ITAM (immunoreceptor tyrosine-based activation motif). STAM1 was previously shown to be associated with the Jak2 and Jak3 tyrosine kinases and to be involved in the regulation of intracellular signal transduction mediated by interleukin-2 (IL-2) and granulocyte-macrophage colony-stimulating factor (GM-CSF) in vitro. Here we generated mice lacking STAM1 by using homologous recombination with embryonic stem cells. STAM1(-/-) mice were morphologically indistinguishable from their littermates at birth. However, growth retardation in the third week after birth was observed for the STAM1(-/-) mice. Unexpectedly, despite the absence of STAM1, hematopoietic cells, including T- and B-lymphocyte and other hematopoietic cell populations, developed normally and responded well to several cytokines, including IL-2 and GM-CSF. However, histological analyses revealed the disappearance of hippocampal CA3 pyramidal neurons in STAM1(-/-) mice. Furthermore, we observed that primary hippocampal neurons derived from STAM1(-/-) mice are vulnerable to cell death induced by excitotoxic amino acids or an NO donor. These data suggest that STAM1 is dispensable for cytokine-mediated signaling in lymphocytes but may be involved in the survival of hippocampal CA3 pyramidal neurons.

Electroporation-mediated interleukin-12 gene therapy for hepatocellular carcinoma in the mice model.

Applications of nonviral vectors for gene transfer into tumors in vivo have been limited by the relatively low expression levels of the transferred gene. The aim of this study is to evaluate the efficacy of electroporation-mediated interleukin-12 (IL-12) gene therapy for hepatocellular carcinoma (HCC). First, we investigated the optimal conditions of electric pulses (voltage, pulsing duration, numbers of shocks) of in vivo electroporation for gene transfer into HCC established by s.c. implantation of MH134 cells to C3H mice. This process made use of plasmid DNA that express the luciferase gene. We concluded that the optimal conditions for the electric pulses are as follows: voltage at 150 V; pulsing duration at 50 ms; nonpulsing duration at 950 ms; and the number of shocks at 10. Second, we tried to treat s.c. HCC by electroporation using plasmid DNA that expresses the murine interleukin-12 (mlL-12) gene. Intratumoral administration of the mIL-12 vector elevated serum IL-12 and IFN-gamma and significantly inhibited the growth not only of HCC into which the mIL-12 vector had been directly transferred, but also of the distant HCC. In addition, intratumoral administration of the mIL-12 vector inhibited spontaneous lung metastasis and delayed establishment of HCC injected 3 days after mIL-12 gene therapy. The IL-12 gene therapy induced more lymphocyte infiltration by NK cells, CD3+ cells, and Mac-1 positive cells into the tumor and reduced the number of microvessels. Therefore, more terminal deoxynucleotidyl transferase-mediated dUTP nick end labeling-positive tumor cells were found. These results demonstrate that gene therapy for HCC by electroporation in vivo using IL-12 is very efficient and is thus promising for further clinical trial.

The gradual expression of troponin T isoforms in chicken wing muscles.

Troponin T (TnT) is one of the muscle regulatory proteins and is thought to be related to unique contractile properties in diverse muscles and also to myogenesis. The expression of TnT isoforms in the chicken wing muscles was investigated by two-dimensional gel electrophoresis and immunoblotting with an antiserum against fast skeletal muscle TnT. The upper arm muscles, M. biceps brachii and M. triceps brachii, showed striking differential expression of high molecular weight B type and low molecular weight L type TnT isoforms in the proximal, middle and distal portions of each muscle. The ratio of the total quantity of B type TnT isoforms to that of L type TnT isoforms decreased gradually along the proximo-distal axis of the wing. The upper arm muscles, M. deltoideus and M. tensor patagii longus, and most of the lower arm muscles examined in this study did not show such differential expression. The lower arm muscles, M. flexor carpi ulnaris and M. extensor carpi radialis longus, showed some gradual expression of TnT isoforms, but the B/L ratio in the former slightly increased along the proximo-distal axis. The gradual expression in M. biceps brachii was not found in the 1-day-old chick, but was established by 30 days post-hatching. The biological significance of differential expression of TnT isoforms in different muscles and even in single muscles was speculated upon with respect to muscle contractile regulation and myogenesis.

Insulin secretion and differential gene expression in glucose-responsive and -unresponsive MIN6 sublines.

We have established two sublines derived from the insulin-secreting mouse pancreatic beta-cell line MIN6, designated m9 and m14. m9 Cells exhibit glucose-induced insulin secretion in a concentration-dependent manner, whereas m14 cells respond poorly to glucose. In m14 cells, glucose consumption and lactate production are enhanced, and ATP production is largely through nonoxidative pathways. Moreover, lactate dehydrogenase activity is increased, and hexokinase replaces glucokinase as a glucose-phosphorylating enzyme. The ATP-sensitive K(+) channel activity and voltage-dependent calcium channel activity in m14 cells are reduced, and the resting membrane potential is significantly higher than in m9 cells. Thus, in contrast to m9, a model for beta-cells with normal insulin response, m14 is a model for beta-cells with impaired glucose-induced insulin secretion. By mRNA differential display of these sublines, we found 10 genes to be expressed at markedly different levels. These newly established MIN6 cell sublines should be useful tools in the analysis of the genetic and molecular basis of impaired glucose-induced insulin secretion.

Cre-loxP-mediated bax gene activation reduces growth rate and increases sensitivity to chemotherapeutic agents in human gastric cancer cells.

Dysregulation of apoptosis may be closely related to the development of cancer and its chemoresistance. Overexpression of Bax, an inducer of apoptosis, has led to increased cell death in a variety of cancer cell lines. In this study, we investigated the effect of Bax overexpression in two gastric cancer cell lines, MKN-28 and MKN-45, using a Cre-loxP-mediated inducible expression system. After induction of bax, both cell lines showed decreased proliferation, partially due to increased cell death. Furthermore, Bax-expressing MKN-28 cells were more sensitive to cisplatin. These results indicate that up-regulation of the bax gene may provide a novel strategy for the treatment of gastric cancer.

Novel tissue culture method: skeletal muscle implantation under gizzard serous membrane of a chick.

A novel method for a long-term culture of skeletal muscle is described. Skeletal muscle pieces from young chicks were implanted under the gizzard serous membrane of the same chicks. Following muscle degeneration, new well-grouped muscle fibers were formed by the fusion of myocytes that differentiated from surviving satellite cells, and the regenerated muscle tissues were maintained in position for longer than 60 days. The implants were in the vital circulatory system, receiving trophic and oxygen supplies, and are completely free from motor nerve innervation and cell contamination with exogenous muscle cells, not as in intra-muscular implantation. Therefore, this tissue culture method should be useful for studying skeletal muscle regeneration and maturation over a long period. Furthermore, osteogenesis and feather development were also found in the implants of embryonic limbs by using the same method. These observations showed that not only skeletal muscle tissues but also other tissues could be cultured under the gizzard serous membrane.

Immunohistochemical studies on regulation of alternative splicing of fast skeletal muscle troponin T: non-uniform distribution of the exon x3 epitope in a single muscle fiber.

Troponin T (TnT) isoforms of chicken fast skeletal muscle are classified into two types, breast-muscle-type (B-type) and leg-muscle-type (L-type) isoforms. These isoforms are produced from a single gene by differential alternative splicing of pre-mRNA. We investigated immunohistochemically the distribution of B-type TnT isoforms in chicken leg muscle (musculus biceps femoris), using anti-exon x3 that was raised against a synthetic peptide corresponding to exon x3 and recognized B-type, but not the L-type, TnT isoforms. Mosaic patterns of immunostaining showing locally different expression of B-type TnT isoforms in a single fiber were observed among fibers, and the non-uniform distribution of the isoforms was also detected in sectioned fibers and myofibrils from the muscle. The results indicated that regulation of pre-mRNA splicing of fast skeletal muscle TnT was different not only among the muscle fibers but also within a single fiber, suggesting that heterogeneous myonuclei in regulation of alternative splicings occur in a single muscle fiber.

New approach to cell lineage analysis in mammals using the Cre-loxP system.

The Cre-loxP site-specific recombination system was used for cell lineage analysis in mammals. We constructed an expression plasmid, pCETZ-17, which consists of cytomegalovirus enhancer/chicken beta-actin promoter (CAG), a portion of the rabbit beta-globin gene, loxP-flanked DNA sequence (containing enhanced green fluorescent protein (EGFP) cDNA), and lacZ gene encoding E. coli beta-galactosidase (beta-gal). When circular pCETZ-17 plasmid DNA was microinjected into the pronuclei of fertilized eggs and these eggs were allowed to develop to two-cell stage, 62.8% (59/94) of the two-cell embryos exhibited distinct fluorescence in one or both blastomeres, but never expressed lacZ protein, as evaluated by histochemical staining with X-Gal, a substrate for beta-gal. When both circular plasmids, pCETZ-17 and pCAG/NCre (containing CAG and DNA sequences encoding nuclear location signal and Cre), were co-injected into fertilized eggs, almost all (87.0%, 47/54) embryos exhibited low or no fluorescence, but 51.9% (27/52) exhibited positive staining for beta-gal activity. This indicates that transient expression of the Cre recombinase gene removed the loxP-flanked DNA sequence in pCETZ-17 and then caused expression of the downstream lacZ sequence. We next microinjected pCETZ-17 into the pronuclei of fertilized eggs, cultured these injected eggs for 1 day, and collected only two-cell embryos expressing EGFP in both blastomeres. One blastomere of the EGFP-expressing two-cell embryos was microinjected with pCAG/NCre, and these treated embryos were cultured for 1 day up to four-cell stage. When the developing four-cell embryos were subjected to staining with X-Gal, cell lineage-related staining pattern for beta-gal activity was observed in most (77.8%, 7/9) embryos. These findings were further confirmed using two-cell embryos derived from a transgenic mouse line carrying CETZ-17 transgene. Thus, our system, which is based on transient expression of the Cre recombinase gene directly introduced into nuclei of embryonic cells by microinjection, is a powerful means for cell lineage analysis in mammals.

Continuous erythropoietin delivery by muscle-targeted gene transfer using in vivo electroporation.

It has been demonstrated that gene transfer by in vivo electroporation of mouse muscle increases the level of gene expression by more than 100-fold over simple plasmid DNA injection. We tested continuous rat erythropoietin (Epo) delivery by this method in normal rats, using plasmid DNA expressing rat Epo (pCAGGS-Epo) as the vector. A pair of electrodes was inserted into the thigh muscles of rat hind limbs and 100 microg of pCAGGS-Epo was injected between the electrodes. Eight 100-V, 50-msec electric pulses were delivered through the electrodes. Each rat was injected with a total of 400 microg of pCAGGS-Epo, which was delivered to the medial and lateral sides of each thigh. The presence of vector-derived Epo mRNA at the DNA injection site was confirmed by RT-PCR. The serum Epo levels peaked at 122.2 +/- 33.0 mU/ml on day 7 and gradually decreased to 35.9 +/- 18.2 mU/ml on day 32. The hematocrit levels increased continuously, from the preinjection level of 49.5 +/- 1.1 to 67.8 +/- 2.2% on day 32 (p < 0.001). In pCAGGS-Epo treated rats, endogenous Epo secretion was downregulated on day 32. In a control experiment, intramuscular injection of pCAGGS-Epo without subsequent electroporation did not significantly enhance the serum Epo levels. These results demonstrate that muscle-targeted pCAGGS-Epo transfer by in vivo electroporation is a useful procedure for the continuous delivery of Epo.

Stable overexpression of the glucose-6-phosphatase catalytic subunit attenuates glucose sensitivity of insulin secretion from a mouse pancreatic beta-cell line.

Glucose-6-phosphatase (G-6-Pase) hydrolyzes glucose-6-phosphate to glucose, reciprocal with the so-called glucose sensor, glucokinase, in pancreatic beta cells. To study the role of G-6-Pase in glucose-stimulated insulin secretion from beta cells, we have introduced rat G-6-Pase catalytic subunit cDNA and have established permanent clones with 3-, 7- and 24-fold G-6-Pase activity of the mouse beta-cell line, MIN6. In these clones, glucose usage and ATP production in the presence of 5.5 or 25 mM glucose were reduced, and glucose-stimulated insulin secretion was decreased in proportion to the increased G-6-Pase activity. In addition, insulin secretory capacity in response to d-fructose and pyruvate was unchanged; however, 25 mM glucose-stimulated insulin secretion and intracellular calcium response were completely inhibited. In the clone with 24-fold G-6-Pase activity, changes in intracellular NAD(P)H autofluorescence in response to 25 mM glucose were reduced, but the changes with 20 mM fructose and 20 mM pyruvate were not altered. Stable overexpression of G-6-Pase in beta cells resulted in attenuation of the overall glucose-stimulated metabolic responses corresponding to the degree of overexpression. This particular experimental manipulation shows that the possibility exists of modulating glucose-stimulated insulin release by thoroughly altering glucose cycling at the glucokinase/G-6-Pase step.

Decreased food intake and body weight in pancreatic polypeptide-overexpressing mice.

BACKGROUND & AIMS: Pancreatic polypeptide (PP) is a 36-amino acid hormone produced by F cells within the pancreatic islets and the exocrine pancreas. The definitive function of PP in mammalian physiology remains to be determined. This study examined the effects of chronic overexpression of PP through the development of PP transgenic mice. METHODS: PP transgenic mice were created by using mouse PP complementary DNA under the control of the cytomegalovirus immediate early enhancer-chicken beta-actin hybrid promoter (pCAGGS expression vector). RESULTS: A unique line of transgenic mice was created that overexpresses PP in the pancreatic islets with low levels of expression in other tissues including the brain. Plasma PP concentrations were more than 20 times higher than those of control littermates. However, PP overproduction led to postnatal lethality in half of the pups because of markedly decreased milk intake. The remaining PP transgenic mice gained less weight with specifically reduced food intake and fat mass compared with controls, a result that was more evident in male than in female mice. The transgenic mice exhibited a reduced rate of gastric emptying of a solid meal but had normal oxygen consumption and fasting leptin levels. Immunoneutralization with anti-PP antiserum reversed the phenotypic changes of transgenic animals. CONCLUSIONS: PP could be involved in feeding and body weight regulation partly through regulation of gastric emptying.

Identification of a portable repression domain and an E1A-responsive activation domain in Pax4: a possible role of Pax4 as a transcriptional repressor in the pancreas.

Pax4 is a paired-domain (PD)-containing transcription factor which plays a crucial role in pancreatic beta/delta-cell development. In this study, we characterized the DNA-binding and transactivation properties of mouse Pax4. Repetitive rounds of PCR-based selection led to identification of the optimal DNA-binding sequences for the PD of Pax4. In agreement with the conservation of the optimal binding sequences among the Pax family transcription factors, Pax4 could bind to the potential binding sites for Pax6, another member of the Pax family also involved in endocrine pancreas development. The overexpression of Pax4 in HIT-T15 cells dose dependently inhibited the basal transcriptional activity as well as Pax6-induced activity. Detailed domain mapping analyses using GAL4-Pax4 chimeras revealed that the C-terminal region of Pax4 contains both activation and repression domains. The activation domain was active in the embryonic kidney-derived 293/293T cells and embryonal carcinoma-derived F9 cells, containing adenoviral E1A protein or E1A-like activity, respectively but was inactive or very weakly active in other cells including those of pancreatic beta- and alpha-cell origin. Indeed, the exogenous overexpression of type 13S E1A in heterologous cell types could convert the activation domain to an active one. On the other hand, the repression domain was active regardless of the cell type. When the repression domain was linked to the transactivation domain of a heterologous transcription factor, PDX-1, it could completely abolish the transactivation potential of PDX-1. These observations suggest a primary role of Pax4 as a transcriptional repressor whose function may involve the competitive inhibition of Pax6 function. The identification of the E1A-responsive transactivation domain, however, indicates that the function of Pax4 is subject to posttranslational regulation, providing further support for the complexity of mechanisms that regulate pancreas development.

MesP1 is expressed in the heart precursor cells and required for the formation of a single heart tube.

The Mesp1 gene encodes the basic HLH protein MesP1 which is expressed in the mesodermal cell lineage during early gastrulation. Disruption of the Mesp1 gene leads to aberrant heart morphogenesis, resulting in cardia bifida. In order to study the defects in Mesp1-expressing cells during gastrulation and in the specification of mesodermal cell lineages, we introduced a (beta)-galactosidase gene (lacZ) under the control of the Mesp1 promoter by homologous recombination. The early expression pattern revealed by (beta)-gal staining in heterozygous embryos was almost identical to that observed by whole mount in situ hybridization. However, the (beta)-gal activity was retained longer than the mRNA signal, which enabled us to follow cell migration during gastrulation. In heterozygous embryos, the Mesp1-expressing cells migrated out from the primitive streak and were incorporated into the head mesenchyme and heart field. In contrast, Mesp1-expressing cells in the homozygous deficient embryos stayed in the primitive streak for a longer period of time before departure. The expression of FLK-1, an early marker of endothelial cell precursors including heart precursors, also accumulated abnormally in the posterior region in Mesp1-deficient embryos. In addition, using the Cre-loxP site-specific recombination system, we could determine the lineage of the Mesp1-expressing cells. The first mesodermal cells that ingressed through the primitive streak were incorporated as the mesodermal component of the amnion, and the next mesodermal population mainly contributed to the myocardium of the heart tube but not to the endocardium. These results strongly suggest that MesP1 is expressed in the heart tube precursor cells and is required for mesodermal cells to depart from the primitive streak and to generate a single heart tube.