Modification of gene activity in mouse embryos in utero by a tamoxifen-inducible form of Cre recombinase

The ability to generate specific genetic modifications in mice provides a powerful approach to assess gene function. When genetic modifications have been generated in the germ line, however, the resulting phenotype often only reflects the first time a gene has an influence on - or is necessary for - a particular biological process. Therefore, systems allowing conditional genetic modification have been developed (for a review, see [1]); for example, inducible forms of the Cre recombinase from P1 phage have been generated that can catalyse intramolecular recombination between target recognition sequences (loxP sites) in response to ligand [2] [3] [4] [5]. Here, we assessed whether a tamoxifen-inducible form of Cre recombinase (Cre-ERTM) could be used to modify gene activity in the mouse embryo in utero. Using the enhancer of the Wnt1 gene to restrict the expression of Cre-ERTM to the embryonic neural tube, we found that a single injection of tamoxifen into pregnant mice induced Cre-mediated recombination within the embryonic central nervous system, thereby activating expression of a reporter gene. Induction was ligand dependent, rapid and efficient. The results demonstrate that tamoxifen-inducible recombination can be used to effectively modify gene function in the mouse embryo.

Identification of an evolutionarily conserved 110 base-pair cis-acting regulatory sequence that governs Wnt-1 expression in the murine neural plate

The generation of anterior-posterior polarity in the vertebrate brain requires the establishment of regional domains of gene expression at early somite stages. Wnt-1 encodes a signal that is expressed in the developing midbrain and is essential for midbrain and anterior hindbrain development. Previous work identified a 5.5 kilobase region located downstream of the Wnt-1 coding sequence which is necessary and sufficient for Wnt-1 expression in vivo. Using a transgenic mouse reporter assay, we have now identified a 110 base pair regulatory sequence within the 5.5 kilobase enhancer, which is sufficient for expression of a lacZ reporter in the approximate Wnt-1 pattern at neural plate stages. Multimers of this element driving Wnt-1 expression can partially rescue the midbrain-hindbrain phenotype of Wnt-1(−/−) embryos. The possibility that this region represents an evolutionarily conserved regulatory module is suggested by the identification of a highly homologous region located downstream of the wnt-1 gene in the pufferfish (Fugu rubripes). These sequences are capable of appropriate temporal and spatial activation of a reporter gene in the embryonic mouse midbrain; although, later aspects of the Wnt-1 expression pattern are absent. Genetic evidence has implicated Pax transcription factors in the regulation of Wnt-1. Although Pax-2 binds to the 110 base pair murine regulatory element in vitro, the location of the binding sites could not be precisely established and mutation of two putative low affinity sites did not abolish activation of a Wnt-1 reporter transgene in vivo. Thus, it is unlikely that Pax proteins regulate Wnt-1 by direct interactions with this cis-acting regulatory region. Our analysis of the 110 base pair minimal regulatory element suggests that Wnt-1 regulation is complex, involving different regulatory interactions for activation and the later maintenance of transgene expression in the dorsal midbrain and ventral diencephalon, and at the midbrain-hindbrain junction.

GDNF induces branching and increased cell proliferation in the ureter of the mouse

The secreted signaling molecule GDNF is expressed in the metanephric mesenchyme and has recently been implicated as a factor necessary for development of the metanephric kidney. We have examined the effects of GDNF on mouse kidney explants. We show that GDNF increases cell proliferation in ureter tips. There is an increase in the number of ureter tips and expansion and fusion of adjacent tips and some tips appear to grow toward the source of GDNF. These events are accompanied by transcriptional upregulation of several genes localized to the tips, including its own receptor, c-ret, the transcription factor Sox9, and the signal Wnt-11. These results support a model in which GDNF supplied by the mesenchyme regulates growth and branching in the metanephric kidney through the local regulation of ureter tip-specific factors.

Expression of the homeobox-containing genes EN1 and EN2 in human fetal midgestational medulla and cerebellum

Homeobox-containing genes En-1 and En-2 have been implicated in the control of pattern formation during development of the central nervous system in experimental animals. In order to determine whether the expression of homologous human EN genes can be used as a developmental genetic marker of the arcuate nucleus of the medulla (a putative precerebellar nucleus that shows developmental deficiency in a subset of sudden infant death syndrome [SIDS]), we performed in situ hybridization with human EN1 and EN2 RNA probes in human fetal midgestational medulla and cerebellum (18-21 weeks gestational age, n=4). Expression of EN genes was demonstrated in all neuronal groups of the medulla and throughout the cerebellum. The RNA signal for both EN1 and EN2 was strongest in the cerebellar granule cell layers, white matter of the vermis and flocculus, inferior olive, arcuate nucleus, caudal raphe nuclei, corpus pontobulbare and nucleus ambiguus. Most of the structures that showed the strongest EN signal originate in the rhombic lip. Some of these structures are functionally interconnected, and show pathologic changes in the syndrome of infantile olivopontocerebellar hypoplasia/atrophy. Strong expression of EN signal in the arcuate nucleus could be used as a genetic marker of this nucleus in further developmental studies of the arcuate nucleus in SIDS. Although EN expression is not specific to the arcuate nucleus or to the rhombic lip derivatives, our results suggest that rhombic lip derivatives have the highest levels of EN RNA message among the medullary structures at midgestation.

Proteoglycans are required for maintenance of Wnt-11 expression in the ureter tips

Development of the metanephric kidney requires the concerted interaction of two tissues, the epithelium of the ureteric duct and the metanephric mesenchyme. Signals from the ureter induce the metanephric mesenchyme to condense and proliferate around the ureter tip, reciprocal signals from the mesenchyme induce the ureter tip to grow and to branch. Wnt genes encode secreted glycoproteins, which are candidate mediators of these signaling events. We have identified three Wnt genes with specific, non-overlapping expression patterns in the metanephric kidney, Wnt-4, Wnt-7b and Wnt-11. Wnt-4 is expressed in the condensing mesenchyme and the comma- and S-shaped bodies. Wnt-7b is expressed in the collecting duct epithelium from 13.5 days post coitum onward. Wnt-1l is first expressed in the nephric duct adjacent to the metanephric blastema prior to the outgrowth of the ureteric bud. Wnt-l1 expression in Danforth's short-tail mice suggests that signaling from the mesenchyme may regulate Wnt-ll activation. During metanephric development, Wnt-11 expression is confined to the tips of the branching ureter. Maintenance of this expression is independent of Wnt-4 signaling and mature mesenchymal elements in the kidney. Moreover, Wnt-ll expression is maintained in recombinants between ureter and lung mesenchyme suggesting that branching morphogenesis and maintenance of Wnt-ll expression are independent of metanephric mesenchyme-specific factors. Interference with proteoglycan synthesis leads to loss of Wnt-ll expression in the ureter tip. We suggest that Wnt-11 acts as an autocrine factor within the ureter epithelium and that its expression is regulated at least in part by proteoglycans.

Topical iodine and neonatal hypothyroidism

OBJECTIVES

To determine whether skin care practices with iodine-containing disinfectants are putting patients in the neonatal intensive care unit at risk for primary hypothyroidism. Cutaneous exposure to povidone-iodine antiseptic solutions may be a cause of primary hypothyroidism in neonates.

DESIGN

Prospective pilot study.

SETTING

Level III neonatal intensive care unit of a university-affiliated hospital.

PARTICIPANTS

Sequential sample of 47 medial and surgical patients admitted to the neonatal intensive care unit who received cutaneous povidone-iodine applications in preparation for invasive or surgical procedures.

METHODS

Seven to 10 days after iodine exposure, capillary blood samples were obtained on filter paper blots for thyroid function testing and urine samples were collected to determine quantitative iodine concentrations. A plasma creatinine level was determined for each subject.

RESULTS

A total of 47 patients were enrolled. The gestational ages of subjects ranged from 26 to 41 weeks (mean, 33.6 weeks); the male-to-female ratio was 28:19; and the birth weights ranged from 0.7 to 5.1 kg(mean, 2.42 kg). The thyroxine level ranged from 20 to 187 nmol/L (1.6 to 14.6 micrograms/dL) (mean, 102 nmol/L [7.9 micrograms/dL];reference, > or = 90 nmol/L [> or = 7 micrograms/dL]; and the thyrotropin level ranged from 0.1 to 16.5 mU/L (mean, 6.4 mU/L; reference, < 20 mU/L). The mean uridine iodine concentration was 2798.0 micrograms/dL (reference, < 40 micrograms/dL), and the mean plasma creatinine level was 60 mumol/L (0.69 mg/dL) (reference, < or = 50 mumol/L [ < or = 0.6 mg/dL] for males and < or = 40 mumol/L [ < or = 0.5 mg/dL] for females).

CONCLUSIONS

There was no documentation of primary hypothyroidism in our subjects despite elevated urine iodine levels. While it is still possible that patients who receive long-term iodine exposure in other settings (eg, cardiac catheterization) are at risk for primary hypothyroidism, our study suggests that the amount of iodine absorbed through routine neonatal intensive care unit procedures does not substantially alter thyroid function during the first 10 days of life. An important confounding variable is that seven patients were receiving dopamine hydrochloride infusions and four were receiving dexamethasone phosphate at the time of sample collection. We therefore cannot rule out the possibility that these medications masked a thyrotropin level elevation that would have occurred in a primary hypothyroid state. We discuss implications for the interpretation of the results of neonatal thyroid function tests.

A single homeodomain binding site restricts spatial expression of Wnt-1 in the developing brain

In this study we investigate the molecular mechanisms that are responsible for the restricted expression of Wnt-1 during embryogenesis. We report that a single homeodomain binding site, HBS1, within the Wnt-1 enhancer contributes to appropriate spatial expression of Wnt-1 in the developing nervous system. This HBS1 site may be required for repressing Wnt-1 expression in the developing forebrain since specific mutations of this site result in an extension of the rostral boundary of Wnt-1/lacZ staining in transgenic embryos. We further demonstrate that a subset of homeodomain proteins expressed in the forebrain (i.e., Dix2, Emx2) interact specifically with HBS1. These findings suggest that these (or related) homeodomain proteins may regulate expression of Wnt-1 during normal brain development by interacting with the HBS1 site in the Wnt-1 enhancer.

Pax-2 expression in the murine neural plate precedes and encompasses the expression domains of Wnt-1 and En-1

In the Drosophila embryo, activation of wingless and engrailed in the parasegment requires paired, a member of the Pax family of transcription factors. We have explored the possible conservation of this regulatory hierarchy in the developing mouse brain. We find that Pax-2 is expressed prior to somite formation in the presumptive mid/hindbrain region. Shortly thereafter, Wnt-1 (the wingless orthologue) and Engrailed-1 are expressed in overlapping regions within the Pax-2 domain. Pax-5 expression commences later, at the 3-somite stage. Thus, the spatial and temporal expression of Pax-2 is consistent with a possible regulatory role in the activation of Wnt-1 and En-1.

Variable electrostatic interaction between DNA and coat protein in filamentous bacteriophage assembly

A restriction fragment carrying the major coat protein gene (gene VIII) was excised from the DNA of the class I filamentous bacteriophage fd, which infects Escherichia coli. This fragment was cloned into the expression plasmid pKK223-3, where it came under the control of the tac promoter, generating plasmid pKf8P. Bacteriophage fd gene VIII was similarly cloned into the plasmid pEMBL9+, enabling it to be subjected to site-directed mutagenesis. By this means the positively charged lysine residue at position 48, one of four positively charged residues near the C terminus of the protein, was turned into a negatively charged glutamic acid residue. The mutated fd gene VIII was cloned back from the pEMBL plasmid into the expression plasmid pKK223-3, creating plasmid pKE48. In the presence of the inducer isopropyl-beta-D-thiogalactoside, the wild-type and mutated coat protein genes were strongly expressed in E. coli TG1 cells transformed with plasmids pKf8P and pKE48, respectively, and the product procoat proteins underwent processing and insertion into the E. coli cell inner membrane. A net positive charge of only 2 on the side-chains in the C-terminal region is evidently sufficient for this initial stage of the virus assembly process. However, the mutated coat protein could not encapsidate the DNA of bacteriophage R252, an fd bacteriophage carrying an amber mutation in its own gene VIII, when tested on non-suppressor strains of E. coli. On the other hand, elongated hybrid bacteriophage particles could be generated whose capsids contained mixtures of wild-type (K48) and mutant (E48) subunits. This suggests that the defect in assembly may occur at the initiation rather than the elongation step(s) in virus assembly. Other mutations of lysine-48 that removed or reversed the positive charge at this position in the C-terminal region of the coat protein were also found to lead to the production of commensurately longer bacteriophage particles. Taken together, these results indicate direct electrostatic interaction between the DNA and the coat protein in the capsid and support a model of non-specific binding between DNA and coat protein subunits with a stoicheiometry that can be varied during assembly.

Cloning and expression of the filamentous bacteriophage Pf1 major coat protein gene in Escherichia coli. Membrane protein processing and virus assembly

A restriction fragment carrying the major coat protein gene (gene VIII) was excised from the replicative form (RF) DNA of the class II filamentous bacteriophage Pf1, which infects Pseudomonas aeruginosa. This fragment was cloned into the expression plasmid pKK223-3, where it came under the control of the tac promoter. In transformed Escherichia coli JM101 cells, in the presence of the inducer isopropyl-beta-D-thiogalactoside, the bacteriophage Pf1 gene was strongly expressed. The bacteriophage Pf1 coat protein displays the same pattern of negatively charged N-terminal region, hydrophobic middle region and positively charged C-terminal region as that of its counterpart in the class I bacteriophage fd, which infects E. coli, but otherwise the two proteins have no sequence homology. However, the Pf1 procoat protein was found to undergo processing and insertion into the E. coli cell inner membrane, like its fd counterpart, demonstrating that this part of the assembly process is the same for these different bacteriophages. The complete transcriptional unit, incorporating the tac promoter and rrnB transcription terminators flanking the Pf1 coat protein gene, was excised from the expression plasmid and cloned into the intergenic space of bacteriophage R252, an fd bacteriophage that carries an amber mutation in its own major coat protein gene. The Pf1 coat protein gene was again well expressed in infected E. coli cells but the chimeric bacteriophage had growth properties identical to those of the parent bacteriophage R252 on suppressor and non-suppressor strains of E. coli. The class I bacteriophage Pf1 coat protein evidently cannot be recognized by the class I bacteriophage assembly complex at or in the E. coli cell inner membrane, either at the point of initiation of assembly or during the elongation process.

Interactions between DNA and coat protein in the structure and assembly of filamentous bacteriophage fd

Bacteriophage fd is a class I filamentous virus (others are M13 and f1) that comprises a circular, single-stranded DNA molecule enclosed in a cylindrical protein sheath to form a flexible particle approximately 890 nm long and 7 nm in diameter. The viral DNA contains 6,408 nucleotides incorporating 10 genes, and the protein sheath is composed of about 2,700 major coat protein subunits in a shingled helical array, the symmetry of which is defined by a fivefold rotational axis combined with a twofold screw axis of pitch 3.2 nm. The DNA extends throughout the length of the particle but is not base-paired and has a symmetry different from that of the protein helix. How the DNA is packed remains unclear but the number (2.4) of nucleotides packaged per major coat protein subunit is certainly not integral, in contrast with, say, the packaging of RNA in tobacco mosaic virus. The coat protein subunit is 50 amino-acid residues in length and, in the virus particle, adopts a largely alpha-helical conformation, with the long axis of the helix aligned close to the long axis of the filament. This protein is arranged with its negatively charged N-terminal region on the outside of the filament and its positively charged C-terminal region on the inside abutting the DNA. We report here that positive charge on one of the four lysine side chains in the latter region has a direct effect on DNA packaging, because when this charge is absent, elongated particles are produced with lengths that can be correlated with the residual positive charge in the C-terminal region of the coat protein subunit.