Generation of mouse models for type 1 diabetes by selective depletion of pancreatic beta cells using toxin receptor-mediated cell knockout.

By using the toxin receptor-mediated cell knockout (TRECK) method, we have generated two transgenic (Tg) murine lines that model type 1 (insulin-dependent) diabetes. The first strain, C.B-17/Icr-Prkdc(scid)/Prkdc(scid)-INS-TRECK-Tg, carries the diphtheria toxin receptor (hDTR) driven by the human insulin gene promoter, while the other strain, C57BL/6-ins2(BAC)-TRECK-Tg, expresses hDTR cDNA under the control of the mouse insulin II gene promoter. With regard to the C.B-17/Icr-Prkdc(scid)/Prkdc(scid)-INS-TRECK-Tg strain, only one of three Tg strains exhibited proper expression of hDTR in pancreatic ß cells. By contrast, hDTR was expressed in the pancreatic ß cells of all four of the generated C57BL/6-ins2(BAC)-TRECK-Tg strains. Hyperglycemia, severe ablation of pancreatic ß cells and depletion of serum insulin were observed within 3days after the administration of diphtheria toxin (DT) in these Tg mice. Subcutaneous injection of a suitable dosage of insulin was sufficient for recovery from hyperglycemia in all of the examined strains. Using the C.B-17/Icr-Prkdc(scid)/Prkdc(scid)-INS-TRECK-Tg model, we tried to perform regenerative therapeutic approaches: allogeneic transplantation of pancreatic islet cells from C57BL/6 and xenogeneic transplantation of CD34(+) human umbilical cord blood cells. Both approaches successfully rescued C.B-17/Icr-Prkdc(scid)/Prkdc(scid)-INS-TRECK-Tg mice from hyperglycemia caused by DT administration. The high specificity with which DT causes depletion in pancreatic ß cells of these Tg mice is highly useful for diabetogenic research.

Selective depletion of mouse kidney proximal straight tubule cells causes acute kidney injury.

The proximal straight tubule (S3 segment) of the kidney is highly susceptible to ischemia and toxic insults but has a remarkable capacity to repair its structure and function. In response to such injuries, complex processes take place to regenerate the epithelial cells of the S3 segment; however, the precise molecular mechanisms of this regeneration are still being investigated. By applying the "toxin receptor mediated cell knockout" method under the control of the S3 segment-specific promoter/enhancer, Gsl5, which drives core 2 ß-1,6-N-acetylglucosaminyltransferase gene expression, we established a transgenic mouse line expressing the human diphtheria toxin (DT) receptor only in the S3 segment. The administration of DT to these transgenic mice caused the selective ablation of S3 segment cells in a dose-dependent manner, and transgenic mice exhibited polyuria containing serum albumin and subsequently developed oliguria. An increase in the concentration of blood urea nitrogen was also observed, and the peak BUN levels occurred 3-7 days after DT administration. Histological analysis revealed that the most severe injury occurred in the S3 segments of the proximal tubule, in which tubular cells were exfoliated into the tubular lumen. In addition, aquaporin 7, which is localized exclusively to the S3 segment, was diminished. These results indicate that this transgenic mouse can suffer acute kidney injury (AKI) caused by S3 segment-specific damage after DT administration. This transgenic line offers an excellent model to uncover the mechanisms of AKI and its rapid recovery.

The cis-regulatory element Gsl5 is indispensable for proximal straight tubule cell-specific transcription of core 2 beta-1,6-N-acetylglucosaminyltransferase in the mouse kidney.

Gsl5 regulates the expression of a glycolipid and glycoproteins that contain the Le(X) epitope in the mouse kidney through tissue-specific transcriptional regulation of the core 2 beta-1,6-N-acetylglucosaminyltransferase (core 2 GnT) gene. The core 2 GnT gene has six exons and produces three alternatively spliced transcripts. Gsl5 regulates only the expression of the kidney-type mRNA, which is transcribed from the most 5'-upstream exon. By introducing a 159-kb bacterial artificial chromosome (BAC) clone that carries the mouse core 2 GnT gene and its 5'-upstream region into DBA/2 mice that carry a defective Gsl5 allele, we were able to rescue the deficient phenotype. The BAC clone was subsequently engineered to replace the core 2 GnT gene with the sequence of enhanced green fluorescent protein (EGFP) as a reporter by an inducible homologous recombination system in Escherichia coli. The transgenic mice derived from the modified BAC clone expressed EGFP in the kidney, which suggests that the candidate Gsl5 is in the 5'-upstream region of the core 2 GnT gene. Sequence analysis of the 5'-upstream regions of the BAC clone and DBA/2 genomic DNA revealed a candidate sequence for Gsl5 at about 5.5 kb upstream of exon 1. This sequence consisted of eight repeats of two GT-rich units in the wild-type mice, whereas it consisted of only one pair of GT-rich units with a minor modification in the DBA/2 mice. Transgenic mice produced with the EGFP reporter gene construct that included this candidate sequence expressed EGFP exclusively in the proximal straight tubular cells of the kidney. These results indicated that this unique repeat is indeed the Gsl5, and it is a cis-regulatory element responsible for proximal straight tubule cell-specific transcriptional regulation.

Core 2 GlcNAc transferase and kidney tubular cell-specific expression.

The expression of glycan chains is precisely regulated in a time- and space-dependent manner. We summarize here our recent work on the kidney tubular cell-specific regulation of core 2 beta-1,6-GlcNAc transferase. Gsl5 gene was first identified by genetic analysis on the basis of polymorphic expression of kidney glycolipids among inbred strains of mice and turned out to be a regulatory gene controlling the level of mRNA of kidney-specific core 2 beta-1,6-GlcNAc transferase. This kidney-specific core 2 GlcNAc transferase takes glycolipids having Gal beta 1-3GalNAc at their termini, Gal beta 1-3GalNAc alpha 1- and beta 1-oligosaccharide derivatives, and glycoproteins having core 1 structure, as substrates. Immunohistochemistry with anti-core 2-Le( x ) monoclonal antibody demonstrated that vesicles located just below the microvillous membrane of proximal tubule cells were clearly stained in a Gsl5 -wild type mouse. Western blotting with the monoclonal antibody detected a major glycoprotein with a molecular mass of 500 kDa in the microsomal fraction of the wild type mouse kidney. In situ hybridization with anti-sense cDNA of kidney-specific core 2 GlcNAc transferase confirmed that Gsl5 gene controls the expression of the core 2 beta-1,6-GlcNAc transferase mRNA in a proximal tubular cell-specific manner. The 5' upstream sequences of the kidney-specific core 2 GlcNAc transferase gene in inbred and wild-derived strains of mice were analyzed, and the phylogenetic analysis of these sequences suggests that functional Gsl5 gene might be produced by the time of subspeciation of M. musculus, about one million years ago.

Phylogenetic development of a regulatory gene for the core 2 GlcNAc transferase in Mus musculus.

In our previous study, we identified a mouse gene, Gsl5, that controls the expression of a glycolipid, GL-Y [Galbeta1-4(Fucalpha1-3)GlcNAcbeta1-6(Galbeta1-3)GalNAcbeta1-3Galalpha1-4Galbeta1-4Glcbeta1-Cer], and the core 2 structure of O-linked glycans of glycoproteins, GlcNAcbeta1-6(Galbeta1-3)GalNAcalpha-Ser/Thr, in a kidney tubular cell-specific manner through the regulation of UDP-GlcNAc beta-1,6-GlcNAc transferase (GNT). Regulation by the Gsl5 gene occurs at the level of GNT mRNA and the recessive allele of Gsl5 is rare and carried by DBA/2 and its related strains. Here, we report a sequence comparison of the 5' flanking region of the GNT gene among 5 laboratory strains and 10 wild-derived strains, demonstrating that the DBA/2 allele sequence is similar to the sequence carried by Asian Mus m. musculus and differs substantially from the East European M. m. musculus. These results suggest that the DBA/2 allele of Gsl5 was introduced into laboratory mouse strains by Asian wild-derived mice. Phylogenetic comparison of the 5' flanking region sequences between the recessive and dominant Gsl5 alleles indicates that mutations to create a functional Gsl5 gene occurred approximately one million years ago during the subspeciation of M. musculus, and provides a case for studies on the creation of functional genes involved in tissue-specific transcriptional regulation.