Skip to main page content
U.S. flag

An official website of the United States government

Dot gov

The .gov means it’s official.
Federal government websites often end in .gov or .mil. Before sharing sensitive information, make sure you’re on a federal government site.

Https

The site is secure.
The https:// ensures that you are connecting to the official website and that any information you provide is encrypted and transmitted securely.

Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation
. 2002 Sep 17;99(19):12236-41.
doi: 10.1073/pnas.192255099. Epub 2002 Sep 9.

Experimental control of pancreatic development and maintenance

Affiliations

Experimental control of pancreatic development and maintenance

Andrew M Holland et al. Proc Natl Acad Sci U S A. .

Abstract

To investigate the role of the HOX-like homeoprotein PDX1 in the formation and maintenance of the pancreas, we have genetically engineered mice so that the only source of PDX1 is a transgene that can be controlled by the application of tetracycline or its analogue doxycycline. In these mice the coding region for the tetracycline-regulated transactivator (tTA(off)) has replaced the coding region of the endogenous Pdx1 gene to ensure correct temporal and spatial expression of the regulatable transactivator. In the absence of doxycycline, tTA(off) activates the transcription of a bicistronic transgene encoding PDX1 and an enhanced green fluorescent protein reporter, which acts as a visual marker of transgene expression in living cells. Expression of the transgene-encoded PDX1 rescues the Pdx1-null phenotype; the pancreata of these mice develop and function normally. The rescue is conditional; doxycycline-mediated repression of the transgenic Pdx1 throughout gestation recapitulates the Pdx1 null phenotype. Moreover, application of doxycycline at mid-pancreogenesis blocks further development. Adult animals of the rescue genotype that were treated with doxycycline for 3 weeks shut off Pdx1 expression, decreased insulin production, and lost the ability to maintain glucose homeostasis. These results demonstrate the feasibility of controlling the formation of an organ during embryogenesis in utero and the maintenance of the mature organ through the experimental manipulation of a key developmental regulator.

PubMed Disclaimer

Figures

Fig 1.
Fig 1.
Strategy for doxycycline-regulated expression of Pdx1. (A) The tTA knockin construct, the organization of the targeted Pdx1 locus, and the tetO-Pdx1/EGFP transgene. The positions of the 5′ and 3′ external hybridization probes (5′EP and 3′EP) and the internal probe (IP) are shown relative to restriction endonuclease sites and the two Pdx1 exons. The tTA knockin construct contains the HSV thymidine kinase gene (HSV tk), 4.5-kb upstream and 1.3-kb downstream Pdx1 homology regions, the tTA-coding region, and the neomycin-resistance gene (Neo). The lengths of the diagnostic restriction fragments [BglII (Bg), BamHI (Ba), XbaI (X), and NotI (N)] are shown for the normal and the targeted Pdx1 locus for the genomic Southern hybridizations of B. The tTA-responsive Pdx1-EGFP transgene contains seven repeats of the tTA-binding tet operator (tetO), the minimal cytomegalovirus (CMV) promoter with a short region of the cytomegalovirus 5′ untranslated region, a Pdx1 minigene (shortened by deleting an internal region of the sole intron), the internal ribosome entry site (ires) derived from the encephalomyocarditis virus, the coding region of EGFP, and the 3′ untranslated region from the bovine growth hormone gene containing a poly(A) addition signal. The black flags show the positions of the PCR primers used to follow the transmission of the knock-in and transgene loci and to assay their expression by RT-PCR. (B) Homologous recombination of the Pdx1-tTA knockin construct in ES cells. The results of the Southern hybridization are shown for the R1 parent line and the two ES cell clones used to generate mice. The BamHI fragment detected by the 5′EP is 8.7 kb for Pdx1+ and 7.7 kb for Pdx1tTA. The BglII fragment detected by the 3′EP is 3.2 kb for Pdx1+ and 5 kb for Pdx1tTA. The hybridization results from all three probes indicate a single insertion and only at the Pdx1 locus. (C) PCR genotyping of nonchimeric mice for six lines generated from the two Pdx1tTA knockin ES cell clones distinguishes the Pdx1+ and Pdx1tTA alleles. (D) tTA and Pdx1 RNAs were present selectively in the pancreas of adult Pdx1+/tTA mice of line 8-2 (C). Templates from cDNA synthesis reactions with and without reverse transcriptase (RTase) were used to show that amplified products were not caused by genomic DNA contamination of the RNAs. The PCR products for the tTA and Pdx1 RNAs were detected with gene-specific oligonucleotide hybridization probes.
Fig 2.
Fig 2.
Activation of the tetO-Pdx1-EGFP transgene (TgPdx1) by the Pdx1tTA locus. (A) PCR genotyping identified mice with one wild-type and one tTA-disrupted allele of Pdx1 and the presence of the transgene. (B) Pancreatic expression of TgPdx1 is dependent on tTA. Reverse-transcriptase-dependent detection of RNA transcripts for tTA, the EGFP region of the transgene, the Pdx1 region of the transgene, all versions of Pdx1 (endogenous locus and transgene), and cytoplasmic actin by PCR. (C) Detection of the transgenic mRNA for nonpancreatic organs was performed as in B with primers for the EGFP region of the transgenic mRNA. L, liver; K, kidney; S, spleen; SG, salivary gland; P, pancreas.
Fig 3.
Fig 3.
The tTA-activated Pdx1 transgene rescues the apancreatic phenotype of Pdx1-deficient mice. Low- (A, C, and E) and high- (B, D, and F) magnification views of viscera dissected from 2-day postpartum pups (P2) include stomach (st), spleen (sp), a caudal lobe of the liver (cl), duodenum (d), and pancreas (p). Wild type: The normal pancreas is organized as a collection of small lobules (B) spread throughout the duodenal loop with lobes that extend to the spleen and around the pyloric region of the stomach (A). Pancreatic tissue from mice lacking the transgene does not fluoresce (B′). Knockout: Viscera from newborn Pdx1tTA/tTA mice lack pancreatic tissue, but otherwise seem normal (C). A very small epithelial remnant is present at the normal attachment site of the dorsal lobe to the duodenum (arrowhead in D). No EGFP fluorescence is detectable in the remnant (arrowhead in D′). Endogenous fluorescence from the duodenum (D′) is independent of the TgPdx1 transgene, has a different hue than EGFP, and arises from the luminal content of partially digested milk, which varies among neonates. Rescue: The pancreas of the rescue genotype is present in the duodenal loop and has splenic and gastric lobes (arrowheads in E and p in F). The pancreas has EGFP fluorescence (F′).
Fig 4.
Fig 4.
The tetracycline-regulated Pdx1 transgene supports the formation of a normal adult pancreas. (A) Normal gross morphology of the pancreas from a 9-week-old adult mouse with the rescue genotype (Pdx1tTA/tTA;TgPdx1). (Bar = 1 mm.) (B and C) Two magnified views showing widespread EGFP fluorescence directed by the tetO-Pdx1/EGFP transgene. (Bars: 200 μm in B, 50 μm in C.) The arrows in C indicate clusters of fluorescing cells that may be islets of Langerhans. (D and D′) Bright-field (D) and fluorescence (D′) views show EGFP fluorescence in most cells of an islet isolated from a rescued adult pancreas. (Bar = 25 μm.) Islets were isolated as described (39) and cultured for 48 h in high-glucose medium to reduce autofluorescence. (E) Insulin is present in the islet and amylase in acini with normal morphology. The punctate amylase-immunofluorescence detects zymogen granules segregated near the apical lumens of acinar cells. Consecutive sections through an islet from a rescue mouse show the presence of an insulin-staining core and more peripheral glucagon (F) and somatostatin (G) cells. (Bars = 25 μm.)
Fig 5.
Fig 5.
Treatment of pregnant mice with doxycycline blocks pancreogenesis. (A) Viscera from a Pdx1-null (Pdx1tTA/tTA) neonate (st, stomach; sp, spleen; cl, a caudal lobe of the liver; duo, duodenum). (B) Viscera from a neonate of the rescue genotype after doxycycline treatment of the mother from the day of conception. The arrowheads of A and B indicate the pancreatic remnants. (C) Viscera from a Pdx1+/tTA neonate with a normal pancreas. The animals of AC were littermates. (A′ and B′) Enlarged bright-field views of the pancreatic remnants of A and B, respectively. Tracings of the epithelial caeca of the pancreatic remnants in A′ and B′ are shown in the image at the Middle Right. (D and E) An extended pancreatic ductal remnant (arrowhead) of a neonate treated with doxycycline from embryonic day 11.5. Aborted ductal budding at the distal end of the duct structure can be seen in E. (F and G) Histology of the ductal remnant (F) stained with hematoxylin and eosin compared with that of a normal neonatal pancreas (G). Crude ductal structures in the remnant replace the plump eosinophilic acini of the normal pancreas.
Fig 6.
Fig 6.
PDX1 deficiency caused by treating mice of the rescue genotype with doxycycline creates a diabetic phenotype. (A) Doxycycline decreased the level of transgenic mRNA in adult Pdx1tTA/tTA;TgPdx1 mice. Before doxycycline treatment, pancreatic tissue samples were obtained by laparotomy from three adult males. After 7 days of recovery, doxycycline was administered for 6 days. RNA was isolated from the pancreatic tissue obtained before and after doxycycline treatment. The level of transgenic mRNA was measured by RT-PCR for both the Pdx1 and EGFP regions of the bicistronic mRNA. (B) Doxycycline causes the loss of β-cell PDX1 and insulin after 21 days. In the absence of doxycycline treatment, the normal high level of insulin causes bleeding of the immunostain into adjacent acinar tissue. (C) Treatment with doxycycline for 21 days lowers the levels of mRNAs for exocrine enzymes as well as for islet hormones. mRNA levels were quantified as for A. (D) Glucose handling in doxycycline-treated adults is greatly compromised. All values are for adults treated for 14 days with doxycycline, except for the values designated with the open triangles and dashed lines, which are for the rescue animals before doxycycline treatment. The values marked with numbers exceeded the uppermost limit of the glucometer (33.3 mM) for four (1), one (2), and one (3) of the six animals, so that average values could not be calculated.

References

    1. Srivastava D. & Olson, E. N. (2000) Nature (London) 407, 221-226. - PubMed
    1. Scully K. & Rosenfeld, M. G. (2002) Science 295, 2231-2235. - PubMed
    1. Ohlsson H., Karlsson, K. & Edlund, T. (1993) EMBO J. 12, 4251-4259. - PMC - PubMed
    1. Pollard S. L. & Holland, P. W. H. (2000) Curr. Biol. 10, 1059-1062. - PubMed
    1. Wessells N. K. & Cohen, J. H. (1967) Dev. Biol. 15, 237-270. - PubMed

MeSH terms

LinkOut - more resources