The glucose transporter (GLUT4) enhancer factor is required for normal wing positioning in Drosophila

Abstract

Many of the transcription factors and target genes that pattern the developing adult remain unknown. In the present study, we find that an ortholog of the poorly understood transcription factor, glucose transporter (GLUT4) enhancer factor (Glut4EF, GEF) [also known as the Huntington's disease gene regulatory region-binding protein (HDBP) 1], plays a critical role in specifying normal wing positioning in adult Drosophila. Glut4EF proteins are zinc-finger transcription factors named for their ability to regulate expression of GLUT4 but nothing is known of Glut4EF's in vivo physiological functions. Here, we identify a family of Glut4EF proteins that are well conserved from Drosophila to humans and find that mutations in Drosophila Glut4EF underlie the wing-positioning defects seen in stretch mutants. In addition, our results indicate that previously uncharacterized mutations in Glut4EF are present in at least 11 publicly available fly lines and on the widely used TM3 balancer chromosome. These results indicate that previous observations utilizing these common stocks may be complicated by the presence of Glut4EF mutations. For example, our results indicate that Glut4EF mutations are also present on the same chromosome as two gain-of-function mutations of the homeobox transcription factor Antennapedia (Antp) and underlie defects previously attributed to Antp. In fact, our results support a role for Glut4EF in the modulation of morphogenetic processes mediated by Antp, further highlighting the importance of Glut4EF transcription factors in patterning and morphogenesis.

Figure 1.—

Adult Drosophila stretched out (stretch) mutants exhibit abnormal wing positioning. (A) The wings of a fly heterozygous for a stretch mutation (stretch^NP7418-GAL4/+) are wild type in appearance and are both positioned over the dorsal abdomen. (B) The wings of a fly homozygous for a mutation in stretch (stretch^NP7418-GAL4/stretch^{TM3 balancer}) are stretched out from the body in an abnormal position. (C) In a rescue line (UAS-D-Glut4EFd; stretch^NP7418-GAL4/stretch^{TM3 balancer}), expression of D-Glut4EFd driven by a GAL4 insertion in D-Glut4EF restores normal wing positioning to a stretch mutant. (D) Percentage of wing-positioning defects in adults heterozygous for the stretch mutation, 10477 (P{PZ}stretch¹⁰⁴⁷⁷, l(3)10477¹⁰⁴⁷⁷), and also heterozygous for wild type (+; n = 159), st, 10477 (n = 63), Df(3R)by62 (n = 160), Df(3R)by10 (n = 70), TM3 balancer (n = 148), TM6B balancer (n = 100), In(3LR)sep (n = 37), A1 (n = 34), NP7418 (n = 50), EY03156 (n = 84), EY04651 (n = 59), Df(3R)swp2 (n = 133), In(3R)Antp^R (n = 35), or In(3R)Antp^B (n = 27) (for full nomenclature see below). In the D-Glut4EFd rescue experiment, flies homozygous for a mutation in stretch (stretch^NP7418-GAL4/stretch^{TM3 balancer}) exhibit normal wing positioning when expressing Glut4EFd using the stretch^NP7418-GAL4 driver (UAS-D-Glut4EFd; stretch^NP7418-GAL4/stretch^{TM3 balancer}; n = 61). The asterisk indicates that a statistically significant difference in the number of stretch mutant flies (stretch^NP7418-GAL4/stretch^{TM3 balancer}) exhibiting outstretched wings is observed when Glut4EFd is expressed in these stretch mutants using the stretch^NP7418-GAL4 driver (UAS-D-Glut4EFd; stretch^NP7418-GAL4/stretch^{TM3 balancer}; n = 61) (chi-square test; P < 0.0001; d.f. = 1; see also supplemental Table S1). (E and F) Genetic complementation and molecular analysis was used to identify a number of different stretch alleles. +, viable; −, lethal; wing, wing phenotype was exhibited in all of the resultant flies; the asterisk indicates not completely penetrant (>50%); N/D, not determined. See also supplemental Table S1. Interestingly, in line with previously published chromosomal breakpoints for the In(3R)Antp^B mutant (breakpoints 84B1–2 and 85E), we find that the In(3R)Antp^B allele is viable over Df(3R)by10 (deleted region 85D8–85E13) while it is lethal over Df(3R)by62 (breakpoints 85D10–11 and 85F1–8). These data position the right breakpoint of In(3R)Antp^B to the right of 85E13 (see also Figure 2A). Furthermore, the In(3R)Antp^B allele is viable over Df(3R)swp2 (a deficiency that removes stretch and several adjacent genes to the right of stretch; see also Figure 2A) but gives a wing phenotype in combination with Df(3R)swp2. Therefore, these results and the others from the complementation matrix indicate that the stretch gene is disrupted in the In(3R)Antp^B fly line. These results also suggest that In(3R)Antp^B may disrupt an adjacent gene(s) to the left of stretch since it is lethal when in combination with Df(3R)by62 but viable when in combination with Df(3R)swp2 (see also Figure 2A). The new nomenclature for these stretch mutant lines is as follows: 10477, P{PZ}stretch¹⁰⁴⁷⁷, l(3)10477¹⁰⁴⁷⁷; st, 10477, st, P{PZ}stretch¹⁰⁴⁷⁷. The full new nomenclature for each of these stretch mutant alleles is listed in brackets: 10477 [P{PZ}stretch¹⁰⁴⁷⁷, l(3)10477¹⁰⁴⁷⁷]; st, 10477 [st, P{PZ}stretch¹⁰⁴⁷⁷]; A1 [P{Dpse82}stretch^A1]; NP7418 [P{GawB}stretch^NP7418-GAL4]; EY03156 [P{EPgy2}stretch^EY03156]; EY04651 [P{EPgy2}stretch^EY04651]; TM3 balancer [In(3LR)TM3, kni^ri-1 p^p sep¹ stretch^{TM3 balancer} l(3)89Aa¹ Ubx^bx-34e e¹]; In(3LR)sep [In(3LR)sep, vvl^sep kni^ri-1 p^p sep¹ stretch^sep]; In(3R)Antp^R [In(3R) Antp^R, Antp^R stretch^Antp-R]; and In(3R)Antp^B [In(3R)Antp^B, Antp^B stretch^Antp-B].

Figure 2.—

Stretch mutations disrupt a large novel Drosophila gene that codes for at least four overlapping transcripts. (A) Genetic organization of the D-Glut4EF locus. The large D-Glut4EF locus covers ∼115 kb of DNA and is adjacent to the MICAL locus that covers ∼40 kb of DNA. The positions of several P elements and deficiencies (solid lines) are indicated on the basis of our complementation analysis, mapping, and molecular data (see Figure 1F and supplemental Figure S1): Df(3R)GB104 (deleted region 85D12–85E10), Df(3R)by10 (deleted region 85D8–85E13), Df(3R)by62 (breakpoints 85D10–11 and 85F1–8), and Df(3R)swp2 (Terman et al. 2002). Dashed lines indicate that the end points of deficiencies are not molecularly defined. Sizes are in kilobases (kb); noncontinuous sequence is indicated by //. (B) The D-Glut4EF genomic locus (black). cDNAs encoding four splice variants were identified that contained overlapping exons and were named D-Glut4EFa, D-Glut4EFb, D-Glut4EFc, and D-Glut4EFd. The positions of several P elements are indicated. Flies containing the 5′-most P elements (EP3681 and BG02024) do not show a wing-positioning defect (supplemental Table S1, supplemental Figure S1; data not shown). Flies containing any of the other P elements (10477, EY03156, NP7418^GAL4, or EY04651) exhibit a wing-positioning defect. This single gene had been previously annotated as CG12418, CG12802, CG11676, and CG32469. CG11676 and CG32469 have now been combined and annotated into CG33975 (which we find to be D-Glut4EFd). (C) D-Glut4EF protein splice variants. cDNAs containing four different splice variants of D-Glut4EF were isolated. Each one contained overlapping exons coding for a region of the protein (black region). Three of the splice variants contained the gray region. Each splice variant had a unique C-terminal tail. The nuclear localization signal (NLS) was mostly intact in isoforms b, c, and d. Only the d isoform contains the zinc-finger C2H2 region and the CRARF/CR domain. Interestingly, a stretch of amino acids in the N terminus of isoforms a and b (brackets) showed similarity to human Glut4EF orthologs [PBF (HDBP2)], suggesting that similar isoforms may also exist for mammalian Glut4EF family members.

Figure 3.—

Sequence of the D-Glut4EFd isoform. The DNA sequence of the D-Glut4EFd isoform with the amino acid sequence is shown. The nuclear localization signal (NLS) is underlined, the zinc-finger C2H2 domain is indicated with a dashed line, and the CRARF/CR domain is indicated with a double underline. The insertion position of the EY04651 P element [the stretch allele P{EPgy2}stretch^EY04651] in relation to the amino acid sequence is indicated (black triangle).

Figure 4.—

The Glut4EF (GEF) family of transcription factors. (A) The GLUT4 enhancer factor (Glut4EF) family of proteins. Amino acid identities are indicated among human members and Drosophila D-Glut4EFd (percentages within domains). Our analysis of the database revealed a third human family member (EST ZNF704). (B) All Glut4EF proteins contain a highly conserved C-terminal CRARF/CR domain that is known to be important for transcriptional activation. An alignment of the CRARF/CR domain's presence within Glut4EF family members and TCF/LEF HMG box transcription factors is shown. Consensus is indicated.

Figure 5.—

Mutations in D-Glut4EF are also present in some mutant stocks of the homeobox transcription factor Antennapedia (Antp). (A) Adult flies heterozygous for the Antennapedia mutation, In(3R)Antp^B, have normal wing positioning. (B) Adult flies heterozygous for In(3R)Antp^B and D-Glut4EF (In(3R)Antp^B/stretch^NP7418-GAL4) exhibit a 100% defect in wing positioning. (C) The wing-positioning defects observed in adult flies heterozygous for both In(3R)Antp^B and D-Glut4EF (In(3R)Antp^B/stretch^NP7418-GAL4) are rescued by expressing UAS-D-Glut4EFd using the D-Glut4EF-GAL4 driver (UAS-D-Glut4EF; In(3R)Antp^B/stretch^NP7418-GAL4). (D) Percentage of wing-positioning defects and transformed antennae in adults heterozygous for In(3R)Antp^B (In(3R)Antp^B/+; n = 80), adults heterozygous for both In(3R)Antp^B and D-Glut4EF (In(3R)Antp^B/stretch^NP7418-GAL4; n = 57), or adults heterozygous for both In(3R)Antp^B and D-Glut4EF and also expressing D-Glut4EFd using the D-Glut4EF-GAL4 driver (UAS-D-Glut4EFd; In(3R)Antp^B/stretch^NP7418-GAL4; n = 73). Note that expression of D-Glut4EFd rescues the wing-positioning defects but not the antenna transformation defects in In(3R)Antp^B mutants. The asterisk indicates that a statistically significant difference in the number of flies heterozygous for both In(3R)Antp^B and D-Glut4EF (In(3R)Antp^B/stretch^NP7418-GAL4), which exhibit an outstretched wing phenotype, is observed when Glut4EFd is expressed in these In(3R)Antp^B/stretch^NP7418-GAL4 mutants using the stretch^NP7418-GAL4 driver (UAS-D-Glut4EF; In(3R)Antp^B/stretch^NP7418-GAL4; n = 61) (chi-square test; P < 0.0001; d.f. = 1). A new, more complete nomenclature should now be adopted as written in brackets for both the In(3R)Antp^R and the In(3R)Antp^B mutants to indicate the presence of mutations in both Antp and D-Glut4EF: In(3R)Antp^R [In(3R)Antp^R, Antp^R stretch^Antp-R]; and In(3R)Antp^B [In(3R)Antp^B, Antp^B stretch^Antp-B].

Figure 6.—

D-Glut4EF modifies the morphogenetic function of Antp. (A) Drawing of the Drosophila eye and antenna disc. (B and C) Drosophila eye and antenna disc following in situ hybridization with antisense (B) and sense (C) probes to D-Glut4EFd. Note that high levels of expression of D-Glut4EFd are seen in the eye and the antenna disc. (D) The eyes and antennae (arrows) in a wild-type (+/+) adult fly. (E) The Antp^Ns allele is a dominant allele in which the antennae are transformed to legs and the head is disrupted. Abnormal eyes and antennae are present in an Antp^Ns heterozygote adult. Note the leg-like protrusions emanating from the antennae (arrows) and smaller eyes in these flies. (F) Removing one copy of the D-Glut4EF gene dramatically suppresses this severe defect and returns the fly to a more wild-type appearance. Normal-appearing eyes and antennae (arrows) are present when Antp^Ns heterozygotes are also heterozygous for D-Glut4EF (Antp^Ns/stretch^Df(3R)swp2). (G and H) The percentage of adult flies either heterozygous for Antp^Ns (Antp^Ns/+) or heterozygous for both Antp^Ns and stretch (Antp^Ns/stretch^Df(3R)swp2) in which either two (G) or one (H) antenna were abnormal in appearance. The asterisk in G and H indicates statistically significant differences between the two groups (chi-square test; P < 0.0001; d.f. = 1). Bar in B, 25 μm (also applies to C).