A deletion-generator compound element allows deletion saturation analysis for genomewide phenotypic annotation

Abstract

With the available eukaryotic genome sequences, there are predictions of thousands of previously uncharacterized genes without known function or available mutational variant. Thus, there is an urgent need for efficient genetic tools for genomewide phenotypic analysis. Here we describe such a tool: a deletion-generator technology that exploits properties of a double transposable element to produce molecularly defined deletions at high density and with high efficiency. This double element, called P[wHy], is composed of a "deleter" element hobo, bracketed by two genetic markers and inserted into a "carrier" P element. We have used this P[wHy] element in Drosophila melanogaster to generate sets of nested deletions of sufficient coverage to discriminate among every transcription unit within 60 kb of the starting insertion site. Because these two types of mobile elements, carrier and deleter, can be found in other species, our strategy should be applicable to phenotypic analysis in a variety of model organisms.

Figure 1

The hybrid transposon P{wHy} and its use for generating unidirectional deletions. (A) Scheme for a general compound mobile element containing two markers and a deleter inserted into a carrier element. (B) P{wHy} consists of P5′-w[+mC]-hobo-y[+t7.7]-P3′ (hobo, oriented 5′ to 3′, and the flanking white and yellow markers reside within P element ends). (C) General scheme for P{wHy}-generated unidirectional deletions. A single P{wHy} insertion is adjacent to genes transcribed from both strands of DNA (thick line). After local hobo hopping followed by intrachromosomal recombination between the directly oriented hobo elements, one marker (in this case yellow) along with two proximal genes is excised. The deletion is selected on the basis of the expression of the remaining marker. If the orientation of the second hobo were reversed, an inversion would occur (10).

Figure 2

Distinguishing between mobilization events confined to the P{wHy} transgene and deletion events extending into the adjacent genomic region. (A) A genomic event extends from one end of the hobo element, removes one of the marker genes, one P element terminus, and some adjacent genomic material (indicated with crosshatching). (B) An event confined to the P{wHy} transgene: a small deletion extending from one end of the hobo element and removing part of the yellow transgene, thereby inactivating it. The approximate positions of the two PCR primers that are used to screen for a PCR product diagnostic of P{wHy}-confined events are indicated by the two arrowheads.

Figure 3

Diagrammatic representation of deletion complementation results. Molecularly mapped genes within the 350–400-kb regions subject to analysis are shown at the top of A and B. Tester mutations used for lethal complementation analysis are shown in boldface. The maps are not to scale; discontinuities are indicated by diagonal breaks. Deletions were grouped according to complementation behavior into subsets I–IV for each deletion series. Results of the complementation tests are shown at the right as + (viable), +/− (semilethal), or − (lethal). Vertical lines relate the complementation groups to the molecular map. (A) Complementation tests with P{3′wHy} 01D09Y deletions. The two lethal P insertions are Uba1^s3484 and dap⁰⁴⁴⁵⁴. Relative to the insertion site, the second deletion endpoints map as follows: subset I from 455 to 78,060 bp, subset II from 82,399 to 101,979 bp, subset III from 103,627 to 263,393 bp, and subset IV from 329,337 to 353,318. The semilethality phenotype of l(2) dap⁰⁴⁴⁵⁴ null allele is confirmed (15). A total of 114 nonredundant deletions have been assayed, and 113 are classified as validated. One exception, an endpoint that maps with subset III, is viable over dap⁰⁴⁴⁵⁴. (B) Complementation tests with P{5′wHy}01D01W deletions. The two P lethal insertions are dve⁰¹⁷³⁸ and EP(2)2600 (inserted between the predicted genes CG3633 and CG6339). Relative to the insertion site, the second deletion endpoints map as follows: subset I, from 216 to 42,756 bp, subset II from 46,123 to 46,658 bp, subset III from 53,165 to 211,901 bp, and subset IV from 247,344 to 398,322. A total of 105 nonredundant deletions have been assayed, and of these, 100 deletions are classified as validated deletions. Five exceptions that are viable over both dve⁰¹⁷³² and EP (2)2600 have been found. These exceptional cases may represent complex mobilization events.

Figure 4

Distribution of the extents of the validated and nonredundant deletions of P{3′wHy}01D09Y (113 deletions) and P{5′wHy}01D01W (100 deletions). The solid bars represent deletions (length in kb). (See Methods for GenBank accession nos. of the deletion endpoints.)

Figure 5

Deletions within 60 kb of the P{wHy} CG12930^01D09 and P{wHy} 01D01 insertions discriminate among all transcription units. This diagram zooms in on Fig. 4. The distribution of deletion endpoints (solid bars) and mRNAs (arrows) are compared. The thick line denotes the DNA. The vertical divisions are provided to indicate the number of endpoints falling within (and between) mRNAs based on current knowledge of gene boundaries in the region. The horizontal axis is calibrated every 10 kb. The diagram was made by using GenBank annotation and the Vector NTI Suite program. The intron-exon structure is not shown. (A) Diagram of a gene-poor region near P{wHy} CG12930^01D09. The 80-kb map fully spans CG1794. (B) A gene-rich region near P{wHy} 01D01. Endpoints located in the untranslated terminal regions of the ari-2 and CG3380 transcripts are indicated also.