Widely Used Mutants of eiger, Encoding the Drosophila Tumor Necrosis Factor, Carry Additional Mutations in the NimrodC1 Phagocytosis Receptor.

The process of apoptosis in epithelia involves activation of caspases, delamination of cells, and degradation of cellular components. Corpses and cellular debris are then rapidly cleared from the tissue by phagocytic blood cells. In studies of the Drosophila TNF, Eiger (Egr) and cell death in wing imaginal discs, the epithelial primordia of fly wings, we noticed that dying cells appeared to transiently accumulate in egr 3 mutant wing discs, raising the possibility that their phagocytic engulfment by hemocytes was impaired. Further investigation revealed that lymph glands and circulating hemocytes from egr 3 mutant larvae were completely devoid of NimC1 staining, a marker of phagocytic hemocytes. Genome sequencing uncovered mutations in the NimC1 coding region that are predicted to truncate the NimC1 protein before its transmembrane domain, and provide an explanation for the lack of NimC staining. The work that we report here demonstrates the presence of these NimC1 mutations in the widely used egr 3 mutant, its sister allele, egr 1 , and its parental strain, R egg 1GS9830 As the egr 3 and egr 1 alleles have been used in numerous studies of immunity and cell death, it may be advisable to re-evaluate their associated phenotypes.

The Drosophila genome encodes a single TNF homolog, known as Eiger (Eda-like cell death trigger, Egr) (Igaki ; Moreno ; Narasimamurthy et al.). Egr is expressed in many different tissues and plays various roles in cellular processes such as the immune response, energy homeostasis, and JNK-dependent cell death. Since its identification, numerous studies on cell death and immunity have utilized the egr and egr alleles, which were generated by imprecise excisions of the Regg10 P element and resulted in deletions of the first coding exon of the gene (Igaki ). Both egr and egr strains are homozygous viable and considered severe loss-of-function alleles.

Dead cells in Drosophila are commonly removed from tissues by phagocytic engulfment by plasmatocytes, the most abundant of the circulating hemocytes in the larva (Abrams ; Sonnenfeld and Jacobs 1995; Franc ; Shklyar ). Plasmatocytes carry cell surface receptors for the recognition and rapid engulfment of bacteria, dead cells and cellular debris, such as Eater (Kocks ; Chung and Kocks 2011), NimrodC1 (NimC1) (Kurucz et al. 2007; Honti ) and Draper (Manaka ). A frequently used marker for plasmatocytes in Drosophila is positivity for NimC1, a transmembrane protein characterized by the presence of a special type of EGF repeat known as the NIM repeat, located immediately proximal to a conserved CCxGY motif (Somogyi ). The gene is part of a cluster of four NimC (NimC1-4) genes in the midst of several other related Nimrod genes at 34E on chromosome 2. Nimrod proteins contain 2–16 NIM repeats as well as additional conserved residues at their amino termini. The Nimrod proteins, together with Eater and Draper, form a conserved superfamily of 12 proteins in Drosophila, and Nimrod proteins are also encoded in the C. elegans and mammalian genomes (Melcarne ). Loss of any Nimrod protein diminishes the capacity of hemocytes to fight microbes. For example, RNAi against has implicated it in bacterial phagocytosis (Kurucz ), while complete loss of demonstrated cooperativity between NimC1 and Eater in the recognition and phagocytosis of bacteria (Melcarne ). In addition, Egr has been reported to have a role in regulating phagocytosis of certain bacteria (Schneider ). Here we report that egr and egr, two commonly used alleles, and their parental strain Regg1, carry additional mutations in the phagocytosis receptor gene.

Materials and Methods

Drosophila genetics and husbandry

Flies were raised at 25° on standard cornmeal-molasses food supplemented with fresh dry yeast. The following strains were used: egr, egr and egrRegg (Igaki ), egr (gift of H. Kanda), egr and egr (generated in this work), HmlΔ-DsRed (gift of K. Brückner; (Makhijani )), yw hsflp (gift of G. Struhl), act > y+Gal4 and tub > Myc > Gal4 (de la Cova ), OregonR (Bloomington Drosophila Stock Center).

Cell death assays

Eggs from appropriate crosses were collected on yeasted grape plates for 2-4 hr and allowed to develop at 25° in a humid chamber for 24 hr. Prior to collecting the eggs, two 30 min pre-collections were carried out to allow females to void any developing eggs. At hatching, larvae were transferred to food vials supplemented with fresh yeast paste at densities of less than 50 larvae/vial to prevent crowding. To generate ‘loser’ clones in a competitive context, a tub > myc, y+ > Gal4 cassette (> represents a FLP-recognition target (FRT) site) was used to excise the >myc, y+ cassette and generate tub > Gal4, UAS-GFP expressing “loser” cells in WT and in egr mutants. FLP recombinase, under heat shock (HS) control, was activated by HS of larvae in a 37° water bath for 10 min, at 48 hr after egg laying (AEL). Post-HS, larvae were allowed to develop at 25° for 24 or 48 hr. To generate Myc-expressing clones, act > y+Gal4 was used to generate act > Gal clones that expressed UAS-GFP and UAS-Myc in WT and in egr mutants. act > Gal4 clones were induced in larvae with a HS in a 37° water bath for 6 min. Clones were allowed to grow in the wing discs for 24 or 48 hr, as described (de la Cova ; Meyer ; Alpar ). The wing discs were then dissected from larvae at the indicated times after clone induction (ACI). A detailed protocol is available upon request.

Larval dissection and imaging

Wing imaginal discs were dissected from third instar larvae as indicated above, and fixed in 4% paraformaldehyde in phosphate-buffered saline (PF-PBS) for 20 min at room temperature, and washed 3-5 times for 20 min each with 0.01% Tween-20 in PBS (PBTw). Larval carcasses were stained with Rabbit anti-Cleaved Caspase-3 (Cas-3) at 1:100 (Cell Signaling Technology, cat. # 9661). Secondary Alexa Fluor 555 Goat anti-Rabbit IgG antibodies (1:600) were purchased from Molecular Probes (cat.# A-21429). Lymph glands were stained with plasmatocyte-specific P1 antibodies (Mouse anti-NimC; 1:100) (Kurucz ) and subsequently, Alexa Fluor 555 (Invitrogen) secondary antibodies. Hoechst 33258 (Sigma) was used to stain DNA. Wing discs and lymph glands were mounted in VectaShield Antifade (Vector Laboratories, Cat# H-1000) on glass slides. Images were acquired with a Zeiss Axiophot microscope with Apotome and processed using ImageJ and Adobe Photoshop.

Hemocyte immunohistochemistry and image processing

Hemocytes were collected from 10-20 experimental larvae, by bleeding from a small tear in the posterior cuticle into a 10-fold volume of PBS. Cells were then transferred to a coverslip and allowed to settle for 30 min at room temperature in a humidity chamber. All subsequent steps were performed directly on the coverslip. Cells were fixed in 4% PFA in PBS for 7 min at room temperature, washed 3 times in PBS, permeablized for 5 min in 0.1% Triton in PBS (PBT), blocked for 5 min in 10% normal goat serum (NGS) in PBT and then incubated with primary antibody in 10% NGS in PBT either overnight at 4° or for 30 min at room temperature. Washes were carried out in PBT and secondary incubation was performed for 30 min at room temperature in 10% NGS in PBT. Cells were then washed 2 times in PBT, followed by 2 washes in PBS. A final 5 min incubation with DAPI in PBS was performed. Coverslips were mounted in Slowfade Gold Antifade (Molecular Probes, cat # S36937). Primary antibodies used were plasmatocyte-specific Mouse anti-P1 (1:100) (Kurucz ). Secondary antibodies were Alexa Fluor 555 (1:500, Invitrogen). Images were collected on a Zeiss Axio Imager M1 and were processed using ImageJ and Adobe Photoshop.

Sequencing of egr mutant strains

Genomic DNA was isolated from homozygous adult female flies. 100 ng of DNA was amplified using the primer sets as described in Honti . The 355bp deletion, 5bp insertion and 6bp microdeletion were found in the fragment amplified by P11189fw (CGCAGGAGCCTACGATAATC) and P11189rev (AAGGAATGTGGACACCATAG). The fragments were cloned into a pCR4-TOPO-TA vector for sequencing using the common sequencing primers M13. Primer sequences are listed in Supplementary Table 3.

Outcrossing and genotyping of egr alleles

egr and egr alleles (hereafter ) were treated identically. mutant virgin females were crossed to OregonR (OreR) males and the resulting F1 heterozygous /+ virgin females were backcrossed to OreR males. Ten F2 males (either /+ or +/+) were singly crossed to virgin OreR females, killed and used for single-fly PCR (Gloor ) to identify males carrying the alleles, but no longer carrying the 355bp deletion in . A single -positive, -negative line for each allele was carried forward by crossing F3 virgin females back to OreR males. This process was repeated four times, after which alleles were re-isolated by crossing to +; Sco/CyO actin-GFP, a chromosome II marker/balancer strain similarly crossed into an OreR background. These alleles were named egr and egr. Genotyping of egr and egr used primers flanking the reported deletion in each strain: Egr_F1 (CCAGAGCCCACTGTATCACC) and Egr_R3 (TCACCTCCTTTTGGAACTCG) amplify a ∼1500bp and ∼2000bp fragment in egr and egr, respectively. Genotyping of used primers flanking the 355bp deletion and 5bp insertion between nucleotides 1582 to 1937 (Honti genome annotation). Nimrod_del_F1 (CCGGGCTACGTAATGAGAAA) and Nimrod_del_R1 (CAATTTGAGTGCGGAACCTC) amplify a 656bp fragment in WT animals and a ∼300bp fragment in animals bearing the deletion. Primer sequences are listed in Supplementary Table 3.

Statistical analysis

T-tests were carried out using two tails of unequal variance.

Data availability

Strains and plasmids are available upon request. Supplementary Table 1 lists genotypes and sources of all Drosophila strains and sources and identifiers of other reagents that were used. Supplementary Table 2 lists the genotypes used in Figures 1, 2 and 3. Supplementary Table 3 contains sequences for the primers used in the study. All of this information is also included in the Reagent Table.

Figure 1

Dying cells transiently accumulate in egr mutants. A. Clones in A and B were examined 24 hr after clone induction (ACI). Clones of loser cells in a wildtype (WT) wing disc, expressing GFP (cyan) and cleaved caspase 3 (Cas-3); Cas-3 channel is shown in A’. Larval genotype: yw hsflp/w;; tub > myc, y+Gal4, UAS-GFP/+. Genotype of the WT loser clones: yw hsflp/w;; tub > Gal4, UAS-GFP/+. B. Loser clones, expressing GFP and Cas-3, in a egr mutant wing disc; Cas-3 channel is shown in B’. Larval genotype: yw hsflp/w; egr/egrtub > myc, y+Gal4, UAS-GFP/+. Genotype of these loser clones: yw hsflp/w; egr/egrtub > Gal4, UAS-GFP/+. C. Quantification Cas-3-positive, GFP-positive (loser) cells per disc at 24hr ACI from A-B, and from the same genotypes (WT or egr3) at 48 hr ACI. Numbers of wing discs examined for each genotype is shown in the bars. P values were determined using 2-tailed ttests with unequal variance. D. Image of a Myc-expressing clone, marked by expression of GFP (cyan), in a wildtype wing disc. Cas-3 positive cells are shown in red. D’ shows Cas-3 as a single channel. Larval genotype: yw hsflp/w; act > CD2Gal4, UAS-GFP/UAS-Myc. Clone genotype: yw hsflp/w;; act > Gal4, UAS-GFP/UAS-Myc. E. Image of a Myc-expressing clone, marked by expression of GFP (cyan), in an egr mutant wing disc. Cas-3 positive cells are shown in red. E’ shows Cas-3 as a single channel. Larval genotype: yw hsflp/w; egr/ egrGal4, UAS-GFP/UAS-Myc. Clone genotype: yw hsflp/w; egr/ egrGal4, UAS-GFP/UAS-Myc. Clones in D and E were examined 48 hr ACI. Scale bars represent relative size.

Figure 2

Lymph glands from egr mutant larvae and circulating hemocytes from egr and egr mutant larvae are negative for NimC1 staining. A. HmlΔ-DsRed (red) and NimC1 (green) are expressed in many hemocytes in lymph glands from larvae of the genotype ywhsflp122; HmlΔ-DsRed The ywhsflp122 (abv. yw122) strain served as a wildtype control. Scale bars in A-C and G represent 50 um. B. NimC1 (red) staining in the primary lobes of lymph glands from control larvae. C. Lymph glands from egr mutant larvae have no NimC1-positive hemocytes. D. Circulating hemocytes from the hemolymph of OreR control larvae stain positively for NimC1 (red). Scale bar represents 10um. E-F. Circulating hemocytes from egr transheterozygous larvae (E) and from egr mutant larvae (F) lack positivity for NimC1. Scale bars represent 10um. G. NimC1 staining (red) in the egr mutant lymph glands, in which the locus was restored to WT.

Figure 3

Summary scheme of mutations in the locus in various mutants. A. Schematic representation of Chromosome 2, where the locus is located at on the left arm (2L) and the gene on the right arm (2R). B. Representation of the locus from yw122 flies, used as a WT strain. This sequence, and that from the outcrossed egr strain, is identical to the reference genome (D. melanogaster version r5.23). PCR genotyping suggests that the OreR WT strain is also wild type at the locus. Numbering as in https://flybase.org/decoratedfasta/FBgn0259896. C. Representation of the locus in the egr precise excision allele, the parental allele egr, and the egr and egr daughter strains. Two deletions, of 355 bp and 6bp, and an insertion of 5 bp, were found in egr and egr. PCR genotyping in egr and the parental strain, egr indicates that they also carry the 355 bp deletion (E); although not notated here, it is likely that they also carry the 5 bp insertion and 6 bp deletion. D. Representation of the locus from Honti et al., 2013. Note that the locus numbering is slightly different than in B-D, presumably due to an earlier genome annotation. The mutations are identical to those found in egr and egr and similar to egr and the parental line, egr (C). E. Gel electrophoresis of the results of PCR genotyping of in the mutants indicated at top, using primers flanking the 355 bp deletion and 5 bp insertion between nucleotides 1582 to 1937 (Honti et al. genome annotation). Primer sequences are listed in Supplementary Table 3 and included in the Reagent Table.

Results

In the course of studying the role of Egr in cell competition, where apoptosis is induced in so-called “loser” cells, we found that dying cells appeared to accumulate in wing discs from egr mutant larvae (Figure 1A-E). That cell death was still induced in egr mutant loser cells suggested that Egr is not required for the cells to die under the two conditions we examined. However, because dead cells are typically cleared within 2-4 hr from wild-type wing imaginal disc epithelia (Milan ), the accumulation of Cas-3 positive loser cells that we observed in egr mutants suggested that loss of might impair corpse clearance. This prompted us to examine plasmatocytes in the lymph glands, the major larval hematopoietic organ, from WT and egr mutant larvae. We immunostained the lymph glands from both genotypes with anti-NimC1 antibodies, a mixture of P1a and P1b antibodies that specifically recognizes the phagocytic plasmatocytes of the larva (Kurucz et al. 2007). As a control, we also examined larvae that carried HmlΔ-DsRed, consisting of a hemocyte-specific enhancer/promoter from the gene fused to red fluorescent protein (DsRed) that identifies larval hemocytes (Makhijani et al. 2011) (Figure 2A). NimC1 is expressed at high levels on the plasma membrane of numerous cells in the primary lymph gland lobes from WT controls, and anti-NimC1 staining overlapped with many HmlΔ-DsRed positive cells (Figure 2A, B). Strikingly, however, no NimC1 positive cells were evident in lymph glands from egr larvae (Figure 2C). As the egr and egr alleles were derived from the same parental strain (Igaki ), we also tested lymph glands from egr larvae, and again found no detectable NimC1 expression (data not shown). To examine circulating plasmatocytes, we isolated hemocytes from larval hemolymph. Although NimC1 was readily observed in circulating hemocytes from OregonR (OreR) controls (Figure 2D), we detected no NimC1-positive hemocytes in the hemolymph from egr/egr transheterozygous larvae, or from egr larvae (Figure 2E-F). egr is a precise excision of the Regg10 P-element present in the parental strain. Thus all of the alleles derived from the Regg1 strain lacked circulating and lymph gland resident plasmatocytes that expressed NimC1.

Honti and colleagues reported that several Drosophila strains that were negative for NimC1 staining carried mutations in the gene, which they postulated were scars of mobile element mobilization (Honti ). Genomic sequencing of these P1-negative strains identified two independent micro-deletions in the gene, including a 6 bp deletion between nucleotides 2264 to 2270 (Honti ). Another deletion of 355 bp was found between nucleotides 1582 to 1937, accompanied by a 5 bp insertion. Together, Honti found that the 355 bp deletion and the 5 bp insertion generated a frameshift mutation in both the RA and RB transcripts, resulting in new sequences and a premature stop codon (Honti ). The alterations were predicted to give rise to a truncated NimC1 protein that lacks the intracellular and transmembrane domains and four extracellular NIM repeats, which would account for its absence on the plasma membrane of hemocytes (Honti ).

To determine whether the egr and egr mutants carried mutations at the locus, we carried out genomic sequencing of a 1254 bp region that encompasses most of the open reading frame (Figure 3). Our data shows that both egr and egr contain identical microdeletions and insertions within the gene, consistent with their common parental origin. Each mutant strain has the same 355 bp deletion, 5 bp micro-insertion, and 6 bp micro-deletion described by Honti at residues 2264-2270 (Figure 3C). In addition, using primers flanking the larger, 355 bp deletion in PCR reactions, we found that both the Regg1 and egr strains carried similar lesions (Figure 3E). Since these mutants were both NimC1 negative (Figure 2E, F), they very likely also carry the premature stop codon generated by the 355 bp deletion and 5 bp insertion. Altogether, these results suggest that these polymorphisms were present in the parental strain (Figure 3C).

To restore the wild-type NimC1 locus to the mutants, we outcrossed both egr and egr to the OreR wild-type strain and isolated recombinants with the WT NimC1 locus and either the egror egr mutation (see Methods). We then sequenced the locus in these outcrossed alleles (hereafter called egr and egr) to verify that the recombination removed the mutant sequences. Both the egr and egr strains lacked the deletions and micro-insertions that characterized the original egr and egr alleles (Figure 3B, E). Consistent with the loss of the deletions, hemocytes from the egr and egr mutants regained NimC1 positivity (Figure 2G and data not shown).

Discussion

Our sequencing data confirm that egr and egr mutants also carry mutations at the locus, similar to those found previously in other Drosophila strains (Honti ). Since the large deletion and micro-insertion in exon 3 of also exist in the original parental line for the egr and egr alleles, egr, and also in egr, a precise excision of the Regg10 P-element, it is highly likely that the mutations in each of these alleles are derived from the parental strain. These mutations are recessive (Honti ), and we speculate that their presence on each mutant chromosome in our experiments might explain the transient accumulation of dying cells; perhaps they also account for the infection susceptibility found previously in egr mutants (Schneider ). Consistent with our sequencing results, the genetic backgrounds of the egr and egr alleles were previously noticed to harbor anomalies that led to -independent susceptibility to infection by Gram-positive bacteria (Narasimamurthy ). Complete deletion of has been reported to prevent phagocytosis of latex beads or yeast zymosan particles by plasmatocytes (Melcarne ), but whether and how phagocytosis of dying cells may be impaired by the mutations we found here remains to be determined. NIM repeats are thought to mediate protein-protein interactions and clustering of receptors is proposed to be key in phagocytic removal of apoptotic cells (Shklyar ). If the truncated mutant NimC1 proteins are aberrantly secreted into the hemolymph, as predicted (Honti ), they could interfere with critical NIM interactions. As the egr and egr alleles have been used in numerous studies of immunity and cell death, it may be worthwhile to re-evaluate some of the phenotypes obtained with these alleles.