Functional analysis of Rfx6 and mutant variants associated with neonatal diabetes.
Mutations in rfx6 were recently associated with Mitchell-Riley syndrome, which involves neonatal diabetes, and other digestive system defects. To better define the function of Rfx6 in early endoderm development we cloned the Xenopus homologue. Expression of rfx6 begins early, showing broad expression throughout the anterior endoderm; at later stages rfx6 expression becomes restricted to the endocrine cells of the gut and pancreas. Morpholino knockdown of rfx6 caused a loss of pancreas marker expression, as well as other abnormalities. Co-injection of exogenous wild-type rfx6 rescued the morpholino phenotype in Xenopus tadpoles, whereas attempts to rescue the loss-of-function phenotype using mutant rfx6 based on Mitchell-Riley patients were unsuccessful. To better define the pleiotropic effects, we performed microarray analyses of gene expression in knockdown foregut tissue. In addition to pancreatic defects, the microarray analyses revealed downregulation of lung, stomach and heart markers and an upregulation of kidney markers. We verified these results using RT-PCR and in situ hybridization. Based on the different rfx6 expression patterns and our functional analyses, we propose that rfx6 has both early and late functions. In early development Rfx6 plays a broad role, being essential for development of most anterior endodermal organs. At later stages however, Rfx6 function is restricted to endocrine cells.
PubMed ID: 21215266
PMC ID: PMC3042741
Article link: Dev Biol.
Grant support: R01 DK077197-03 NIDDK NIH HHS , DK077197 NIDDK NIH HHS , R01 DK077197-05 NIDDK NIH HHS , R01 DK077197 NIDDK NIH HHS
Genes referenced: agr2 foxa2 gcg hhex ins kcnj1 neurod1 nkx2-5 onecut1 pdx1 ptf1a rfx6 sfrp5 sftpc sox17a sst unnamed
Morpholinos referenced: rfx6 MO1 rfx6 MO2
Article Images: [+] show captions
|Fig. 1. Rfx6 expression in the endoderm and isolated gut tissue. (A) Sagittal bisection of an NF15 embryo showing an anteriorly localized stripe of rfx6 expression. (B) Sagittal bisection of an NF20 embryo showing rfx6 expression. (C) Transverse section through rfx6 expressing region of NF24 embryo showing broad expression in anterior endoderm. (D) NF28 embryo showing rfx6 expression restricted to a dorsal (arrow) and ventral (arrowhead) patch in the anterior foregut. (E) NF40 liver and pancreas showing rfx6 expression in the dorsal pancreas. (F) NF42 liver and pancreas showing rfx6 expression spreading to the ventral pancreas. (G) NF46 liver and pancreas showing rfx6 evenly expressed throughout the pancreas. (H) NF42 whole gut showing rfx6 expression in the dorsal pancreas, stomach and intestine. (A, B, D) Anterior is left. (A–D) Dorsal is up. (H) li, liver; p, pancreas; st, stomach. (I) RT-PCR of rfx6 in whole tadpoles at various stages throughout development.|
|Fig. 2. Rfx6 LOF leads to decreased expression of foregut and pancreas markers early in pancreas development. (A) Western blot with anti-flag antibody of in vitro TnT assay of translation. This demonstrates that MO1 prevents translation of rfx6 mRNA, while the mis-match morpholino (MM) does not. Rfx6 mRNA, flag-tagged rfx6 mRNA plasmid; MM, mis-match morpholino. (B) Diagramatic representation of RT-PCR test of MO2 efficiency. (C) RT-PCR performed on wild-type (WT) and MO2-injected tadpoles. WT mRNA is processed leaving the reverse primer unable to bind resulting in no amplification in the Rfx6 panel. MO2 blocks mRNA processing at the intron 2 donor site (5' end of the intron) as demonstrated by the presence of PCR product in MO2 lane. Ef1α is a housekeeping gene used as a loading control. (D, E) Hnf6 expression is reduced in the morphant foregut endoderm at NF25 (24/28 reduced). (F, G) FoxA2 expression is reduced in the morphant foregut endoderm at NF25 (17/28). (H, I) Hnf6 expression is reduced in the morphant foregut at NF35 (10/13 reduced). (J, K) Pdx1 expression is reduced in the morphant (17/19 reduced). (L, M) Ptf1a expression is reduced in both the dorsal (arrow) and ventral (arrowhead) pancreatic buds of the morphant (n = 18). (N, O) Early insulin expression in the dorsal pancreatic bud (arrow) is lost in morphant tissue (56/62 reduced). Control tadpoles were injected with 25 ng mis-match morpholino.|
|Fig. 3. Rfx6 is required for pancreas differentiation. Whole mount in situ hybridizations of NF42 whole guts from morphant (MO1) and mis-match (control) tadpoles. (A–D) Expression of pancreatic transcription factors pdx1 (n = 18) and ptf1a (n = 18) is almost completely abolished in morphant tissue. (E, F) Expression of acinar marker elastase is reduced (n = 16). (G, H) Schematic illustrating organs in the whole gut. Pink, liver; red, gall bladder; blue, pancreas; yellow, stomach, duodenum and intestine. (I, J) β cell marker insulin is reduced (72/75 reduced). (K–N) Glucagon and somatostatin, both markers of endocrine cells in the pancreas and stomach are reduced (glucagon n = 18, somatostatin 15/16 reduced). (O, P) NeuroD, developmental endocrine cell marker, is reduced (n = 17).|
|Fig. 4. Rfx6 mRNA rescues the morpholino knockdown phenotype. (A) Control whole gut stained for elastase expression (n = 9). (B, C) Whole guts from embryos injected with 40 ng MO2, elastase expression is reduced (n = 47, 6 moderate, 41 severe). (B) Moderate MO2 phenotype, (C) Severe MO2 phenotype. (D) Whole gut from embryo injected with 40 ng MO2 and 100 pg rfx6-GR mRNA. No dexamethosone was added. Elastase expression is reduced (n = 32, 8 normal, 15 moderate, 9 severe). (E) Whole gut from embryo injected with 40 ng MO2 and 100 pg rfx6-GR mRNA. 2 μg/mL dexamethosone was added at NF25 to activate Rfx6-GR protein. Elastase expression is restored (n = 36, 20 normal, 15 moderate, 1 severe). (F) Rescue results from all experimental whole guts. Severe phenotype refers to complete absence of elastase expression and gut malformation; moderate phenotype refers to reduced elastase expression and some gut malformation; normal refers to whole guts with normal levels of elastase expression and normal overall gut morphology.|
|Fig. 5. Mutant rfx6 mRNA cannot rescue the morpholino knockdown phenotype. (A) Control whole gut stained for elastase expression (n = 17). (B) Whole gut from embryo injected with 40 ng MO2, elastase expression is reduced (n = 26, 4 normal, 8 moderate, 14 severe). (C) Whole gut from embryo injected with 40 ng MO2 and 100 pg rfx6R181Q-GR mRNA. No dexamethosone was added. Elastase expression is reduced (n = 12, 1 moderate, 11 severe). (D) Whole gut from embryo injected with 40 ng MO2 and 100 pg rfx6R181Q-GR mRNA. 2 μg/mL dexamethosone was added at NF25 to activate Rfx6R181Q-GR protein. Elastase expression remains reduced (n = 16, 3 normal, 4 moderate, 9 severe). (E) Rescue results for three rfx6 mutants: R181Q, S217P, and EX7 (Rfx6 is truncated after exon 7). Severe phenotype refers to complete absence of elastase expression and gut malformation; moderate phenotype refers to reduced elastase expression and some gut malformation; normal refers to whole guts with normal levels of elastase expression and normal overall gut morphology. S217P −dex n = 16, 2 normal, 2 moderate, 14 severe. S217P + dex n = 16, 2 moderate, 14 severe. EX7 −dex n = 13, 1 normal, 1 moderate, 11 severe. EX7 + dex n = 16, 1 normal, 1 moderate, 14 severe.|
|Fig. 6. LOF microarray experiment schematic and RT-PCR verification of selected genes. (A) Schematic diagram of LOF microarray experiment. Eight-cell stage embryos were injected with 25 ng of either Rfx6 start site morpholino (MO) or mis-match control (MM). Samples were collected at NF30, NF40 and NF44, RNA was extracted and hybridized to Affymetrix Xenopus laevis GeneChip 2.0. (B–D) RT-PCR verification of selected differentially expressed genes from the LOF microarray analysis. (B) NF30. (C) NF40. (D) NF44. +, microarray experiment showed upregulation; −, microarray experiment showed downregulation, NA, this probe was not on the microarray but was shown to be downregulated in in situ experiments, NC, the microarray experiment detected no change. Differential expression in RT-PCR findings are consistent with results seen in the microarray experiment. MM, mis-match control; MO, MO1; -RT, control reaction without reverse transcriptase.|
|Fig. 7. In situ hybridization confirms differential expression of organ marker genes for stomach, lung, heart and kidney in LOF microarray. (A, B) Isolated whole gut NF42 showing expression of stomach marker agr2 almost entirely abolished in morphant tissue (n = 23). (C, D) Expression of stomach marker frp5 was reduced (n = 18). (E–J) Whole tadpoles at NF42 with ventral skin removed for in situ hybridization. (E, F) Lung marker surfactant C (SurfC) is decreased in morphant tissue, closed arrows (10/14 reduced). (G, H) Heart marker nkx2.5 is decreased in morphant tissue, open arrow (8/14 reduced). (I, J) Kidney marker kcnJ1 is increased in morphant tissue, arrowhead (8/11 increased). (K,L) Expression of liver marker hex is unaffected by rfx6 knockdown (n = 32). (M, N) Expression of gall bladder and duodenum marker hnf6 is reduced in duodenum but not gall bladder (17/18). (O, P) Gall bladder marker sox17α was not reduced (18/20 showed normal expression).|
|rfx6(regulatory factor X, 6) gene expression in Xenopus laevis embryos, NF stage 15, as assayed by in situ hybridization. Lateral view of bissected embryos: anterior left, dorsal up.|
|rfx6(regulatory factor X, 6) gene expression in Xenopus laevis embryos, NF stage 28, as assayed by in situ hybridization. Lateral view: anterior left, dorsal up.|
|rfx6(regulatory factor X, 6) gene expression in Xenopus laevis embryos, NF stage 42, as assayed by in situ hybridization. Lateral view of dissected foregut: anterior left, dorsal up.|