Functional Divergence of the Tribolium castaneum engrailed and invected Paralogs

Abstract

Engrailed (en) and invected (inv) encode paralogous transcription factors found as a closely linked tandem duplication within holometabolous insects. Drosophila en mutants segment normally, then fail to maintain their segments. Loss of Drosophila inv is viable, while loss of both genes results in asegmental larvae. Surprisingly, the knockdown of Oncopeltus inv can result in the loss or fusion of the entire abdomen and en knockdowns in Tribolium show variable degrees of segmental loss. The consequence of losing or knocking down both paralogs on embryogenesis has not been studied beyond Drosophila. To further investigate the relative functions of each paralog and the mechanism behind the segmental loss, Tribolium double and single knockdowns of en and inv were analyzed. The most common cuticular phenotype of the double knockdowns was small, limbless, and open dorsally, with all but a single, segmentally iterated row of bristles. Less severe knockdowns had fused segments and reduced appendages. The Tribolium paralogs appear to act synergistically: the knockdown of either Tribolium gene alone was typically less severe, with all limbs present, whereas the most extreme single knockdowns mimic the most severe double knockdown phenotype. Morphological abnormalities unique to either single gene knockdown were not found. inv expression was not affected in the Tribolium en knockdowns, but hh expression was unexpectedly increased midway through development. Thus, while the segmental expression of en/inv is broadly conserved within insects, the functions of en and inv are evolving independently in different lineages.

Keywords: Tribolium; engrailed; gene paralogs; invected; segment-polarity; sequential segmentation.

Figure 1

Description of clones used in this study. Location of dsRNA fragments within the en and inv genes. The Tribolium en and inv gene are present in a closely linked tandem duplication shown at the top (as previously described in [24]). Exons (E1,2,3) are colored yellow, the homeobox encoding region is colored red, location of the regions cloned and used for dsRNA experiments is shown in blue.

Figure 2

Mild phenotypes resulting from double dsRNA knockdown of en and inv. (A) Dorsal side of buffer injected control; (B) ventral side of buffer injected control; (C) Category B larva, primarily normal, with incomplete dorsal closure on first and second thoracic segments (outlined in red); (D) Category B larva, primarily normal with missing (arrowhead) or abnormal appendages (highlighted in purple). All scale bars = 50 µm.

Figure 3

Variety of severe phenotypes resulting from knockdown of both en and inv. (A) Category C larva with fusing and minimized gnathic and thoracic segments (arrowheads); (B) Category C larva with a reduced number of segments and incomplete dorsal closure (red circled area) but complete urogomphi (circled in red). (C) Category D asegmental larva with fused spiracles (blue). Moreover, note the absence of long bristles, mouthparts, and legs; antennae (circled in yellow) and labrum (circled in purple) were present. (D) Category D asegmental larva with similar cuticular features to C but more complete dorsal closure; (E) Category E larvae with secreted cuticle and minimal cuticular features. All larvae are oriented anterior to the left. Larvae had been injected with 0.5 µg/µL of both Tc-en and Tc-inv, except for C, which was injected with (1 µg/uL each of Tc-en and Tc-inv). Scale bar = 50 µm.

Figure 4

Phenotypic categories resulting from increasing concentration of Tc-en/inv dsRNA injected. Category A: phenotypically normal; Category B: fully segmented, primarily normal with incomplete dorsal closure on the anterior thorax, reduced limbs, and minor disruptions of bristle pattern; Category C: larvae with fused and/or missing segments, a greater degree of incomplete dorsal closure, and missing appendages; Category D: asegmental larvae with fused tracheal pits, reduced bristle pattern, and failed dorsal closure; Category E: asegmental, unhatched larvae with minimal or no cuticular features. Sample sizes are indicated at the top of each bar; concentrations in µg/µL of dsRNA injected of either en/inv, en, or inv dsRNA are indicated below the bars.

Figure 5

En/Inv and Ubx-adbA protein expression in wild type and en/inv RNAi knockdowns. (A) En/Inv expression in wild-type and (B) en/inv RNAi knockdowns. (C,C’) Ubx-abdA expression in wild-type embryos. (D,D’) Ubx-abdA expression in en/inv RNAi knockdowns. Ubx-adbA protein expression extends from the posterior of the third thoracic segment through 8 abdominal segments in both wild-type and knockdown embryos. (C’,D’) The expression is shown in a single focal plane. The arrowhead in (D) points to fused spiracles. Both WT and knockdown embryos were reared to 48–50 h AEL. Scale bars = 50µm.

Figure 6

Reduction in length as a function of dsRNA injected. All concentrations of dsRNA used in single or double knockdowns led to a significant reduction in length (p < 0.05) relative to controls. Within each group, only en 0.5 µg/µL relative to either en 1.0 µg/µL or en 1.5 µg/µL led to significant differences in length, indicated with an asterisk. No inv or en/inv dsRNA concentrations led to larvae significantly different in length from one another. dsRNA concentrations are shown in µg/µL.

Figure 7

Model for loss of repeated segmental elements and segment fusions. (A) Dorsal view, buffer-injected control. Spiracles are circled in green; (B) en/inv dsRNA injected (1 µg/µL) category D larvae with repeating rows of short bristles missing intervening regions of the naked cuticle. Fused spiracles are circled in green. Ant, antennae, Lb, labrum. (C) Diagram of loss of cuticular phenotypes in the severely affected double knockdowns.

Figure 8

Representative images of embryos that failed to secrete cuticle. (A–F) Embryos stained with DAPI. Embryo (F) is shown within the vitelline membrane, the remaining panels are cropped to isolate the embryo. Dotted lines in (D) indicate unusual orthogonal axes, also shown visibly in (F). Scale bars = 50 µm.

Figure 9

No qualitative differences between Tc-en and Tc-inv single knockdowns. (A) 2.0 µg/µL Tc-en dsRNA injected larva with abdominal bristle fusions (yellow* and small arrowheads) and minor failure to complete dorsal closure (large arrowhead); (B) 1.5 µg/µL Tc-inv dsRNA injected larva with fusing segments A4-A5 and A5-A6 (yellow*); (C) 2.0 ug/uL Tc-en dsRNA injected larva with misshapen limbs (T2 legs colored); (D) 0.5 µg/µL inv dsRNA injected larva with misshapen limbs (T1 legs colored); (E) 2.0 µg/µL Tc-en larva with misshapen limbs, failed dorsal closure and missing abdominal segment; (F) 1.5 µg/µL Tc-inv dsRNA injected larva with segmental fusion, missing abdominal segment; (G) 2.0 µg/µL Tc-en dsRNA injected amorphous cuticle highlighted in green; (H) 1.5 µg/µL Tc-inv dsRNA injected larva very reduced in size, with the vestigial abdomen.

Figure 10

Effects on expression of inv and hh in en knockdowns. en knockdown is initiated at 4 h AEL and examined at progressively later stages during which the phenotype is more pronounced. At each time point, en expression is reduced (black columns) and inv (red) is not affected, but hh (blue) levels are statistically increased in the 31 h sample (error bars are SD and significance (*) is p < 0.05).