Expression of Fu during mouse embryonic development. (A to C, G to I, and M to O) Whole-mount in situ hybridization using digoxigenin-labeled riboprobes on wild-type embryos at 8.5 (A and G), 9.5 (B and H), and 10.5 (C, I, and M) dpc and Fu^ΔE1-5 embryos at 10.5 dpc (N and O). All views are lateral, with the exception of M and N, which are dorsal views of the forelimb. Shh is expressed in several signaling centers, while Fu is broadly expressed in wild-type embryos, including in domains of Hh-responsive cells. Only background Fu signal could be detected in Fu^ΔE1-5 embryos. Embryos in panels A and G; B and H; C, I, and O; and M and N were photographed at the same magnification, respectively. (D to F, J to L, and P to R) Isotopic in situ hybridization using [³³P]UTP-labeled riboprobes (pink) on paraffin sections of 10.5-dpc wild-type (D and J) and Fu^ΔE1-5 (P) embryos at the forelimb-heart level, 13.5-dpc wild-type (E and K) and Fu^ΔE1-5 (Q) embryos, and 18.5-dpc wild-type (F and L) and Fu^ΔE1-5 (R) lungs. Panels D, J, and P; E, K, and Q; and F and R were photographed at the same magnification, respectively. Panel L was photographed at a higher magnification than panels F and R. Fu signal is absent in Fu^ΔE1-5 embryos. The Fu probe used for in situ hybridization is derived from Fu cDNA sequences that correspond approximately to exons 2 to 9. Probes derived from several other regions of Fu cDNA, including exons 2 to 5 and exons 27 and 28, gave identical expression patterns (data not shown). The Fu sense probes did not yield signals above the background (data not shown). nt, neural tube; fp, floor plate; nc, notochord; lg, lung; lv, liver.

Alternative splicing of Fu transcripts during mouse embryogenesis. (A) Northern blot analysis of poly(A)⁺ RNA isolated from 7-, 11-, 15-, and 17-dpc wild-type mouse embryos (Clontech). Two major species of Fu transcripts, approximately 4.7 and 4.1 kb, respectively (arrows), were detected. In adult mice, Fu is mainly expressed in testis (data not shown). The Fu probe used for hybridization is derived from Fu cDNA sequences that correspond approximately to exons 2 to 9. The same membrane was rehybridized with a GAPDH probe, which serves as the loading control. (B) RT-PCR using RNA derived from wild-type 10.5-dpc mouse embryos in combination with primers spanning the indicated exons. Alternative splicing of the region containing exons 11 to 16 was detected by PCR. This was subsequently verified to be alternative splicing of exons 13 to 15 by sequencing. The white arrow points to the PCR product without exons 13 to 15, which is much less abundant than the upper band containing exons 13 to 15. No alternative splicing of exon 8 or exon 23, as reported for human Fu transcripts, was detected in this assay. A PCR product was amplified with primers derived from exons 1 and 5, respectively. Sequence analysis of the PCR product revealed that it contains exons 1 and 3 to 5, and whether it represents the dominant form during mouse embryogenesis remains to be further investigated. RT-PCR using primers derived from exons 2 and 5 failed to produce a product (data not shown). (C) Schematic diagram of the Fu genomic locus and possible alternative splicing. The Fu genomic locus consists of 28 exons (E1 to E28). The ATG translation start codon is predicted to reside in the third exon (E3), and the TGA stop codon is in the 28th exon (E28). Alternative splicing of exons 13 to 15 is labeled in red and blue, respectively, and potential alternative splicing of exons 1 and 2 is labeled with lime and olive, respectively. Splicing of common exons is not labeled. The longer Fu transcript, containing exons 13 to 15, appears to constitute the major species during mouse embryonic development based on the relative abundance of the transcripts.

Targeted disruption of the mouse Fu gene. The schematic diagram shows the Fu genomic locus, the targeting vector, and the mutant allele. The top line shows a partial restriction map of the mouse Fu genomic locus on chromosome 1. The Fu genomic locus consists of 28 exons (E1 to E28), and only the first 7 exons are shown for simplicity. The translation start ATG is predicted to reside in the third exon (E3). The regions between the dotted lines represent the 5′ and 3′ regions of homology used in gene targeting, respectively, and the symbol × indicates events of homologous recombination. Germ line-transmitting chimeric males carrying the targeted Fu locus were mated with β-actin::Cre mice to remove sequences between the two loxP sites (including the PGK-neo-pA selection cassette) and generate the Fu^ΔE1-5 allele in which the first five exons of Fu are removed. As a result, β-galactosidase was brought under the control of putative upstream Fu regulatory elements. The locations of the fragments used as the 5′ or 3′ external probes in Southern blotting are shown, as well as the sizes of the restriction fragments detected for wild-type and targeted Fu^ΔE1-5 alleles. The Rnf25 (ring finger protein 25) gene is divergently transcribed immediately upstream of Fu, suggesting that targeted disruption of Fu could potentially remove regulatory elements that control Rnf25 expression.

Fu^ΔE1-5 animals exhibit growth retardation and postnatal lethality, and Fu transcript levels are barely detectable in Fu^ΔE1-5 embryos. (A) Wild-type and Fu^ΔE1-5 animals photographed at postnatal day 14. The mutant is significantly smaller than its wild-type littermate. Some toes in both animals were clipped for numbering and genotyping. (B) Northern blot analysis of poly(A)⁺ RNA isolated from 10.5-dpc wild-type, Fu^ΔE1-5/+, and Fu^ΔE1-5 embryos. The two Fu probes used for hybridization gave identical results, and they correspond approximately to the last 775 bp of Fu transcripts and the 624-bp genomic sequences (exons 2 to 5) deleted in the gene-targeted Fu^ΔE1-5 allele (data not shown). A ∼4.7-kb upper band with stronger intensity and a ∼4.1-kb lower band were detected in both wild-type and Fu^ΔE1-5/+ embryos but were completely absent in Fu^ΔE1-5 embryos. The expression level of the housekeeping gene GAPDH serves as the loading control with the same membrane reprobed. (C) RT-PCR using RNA derived from 10.5-dpc wild-type (wt) and Fu^ΔE1-5 mouse embryos in combination with primers spanning the exons indicated. While PCR products of the predicted sizes were detected from RNA derived from wild-type embryos, no PCR products were generated using RNA obtained from Fu^ΔE1-5 mutants. The same number of PCR cycles (37 cycles) was employed for all the PCRs shown in this panel. The faint band, corresponding to exons 24 to 28, seen in Fu^ΔE1-5 was not detected when a lower number of PCR cycles (less than 33) was applied (data not shown). PCR products corresponding to other genes such as Shh, Fgf10, and GAPDH (not shown) were detected at similar levels using RNA produced from wild-type or Fu^ΔE1-5 mutants, suggesting that the absence of a Fu signal in Fu^ΔE1-5 was not due to technical difficulties in RNA preparation or RT-PCR.

Development of major tissues and organs does not appear to be affected in Fu^ΔE1-5 mutants. (A to T) Hematoxylin-and-eosin-stained sections of major tissues and organs of wild-type and Fu^ΔE1-5 animals at embryonic dpc 18.5 and postnatal day P8. A to D, coronal sections through the forebrain; E to H, cross sections through the thoracic cavity; I to R, cross sections through the abdominal cavity; S and T, longitudinal sections through the proximal ulna. Although Fu^ΔE1-5 mutants were significantly smaller than their wild-type littermates at day P8, no obvious developmental defects or pathological changes in major tissues and organs were detected in Fu^ΔE1-5 mutants. Multiple wild-type and Fu^ΔE1-5 embryos were examined. In certain regions of the sections, the minor differences in morphology between wild-type and Fu^ΔE1-5 animals were due to planes of sections. sp, spinal cord; lg, lung; h, heart; s, stomach; si, small intestine; li, large intestine; lv, liver; k, kidney; b, bladder. The magnification of each section is indicated.

Expression of Hh targets is not perturbed in Fu^ΔE1-5 mutants. (A to P) Isotopic in situ hybridization using [³³P]UTP-labeled riboprobes (pink) on paraffin sections of wild-type (A, C, E, G, I, K, M, and O) and Fu^ΔE1-5 (B, D, F, H, J, L, N, and P) 10.5-dpc embryos at the forelimb-heart level (A, B, C, D, G, H, K, L, O, and P) or the hind limb level (E, F, I, J, M, and N). The axis for orientation of the specimens is dorsal upward and ventral downward. Shh is expressed in several signaling centers such as the notochord and floor plate, while Hh targets, including Ptch1, Hip1, and Gli1, are expressed in the ventral neural tube and the somite in response to Hh signaling. Shh is also expressed in the epithelium of many developing organs, including the foregut endoderm, while Ptch1, Hip1, and Gli1 are expressed in the surrounding mesenchyme. Multiple sections were examined in multiple rounds of in situ hybridization, and expression of Hh targets does not appear to be affected in Fu^ΔE1-5 mutants. nt, neural tube; fp, floor plate; nc, notochord; s, somite; fg, foregut; h, heart.