The Drosophila TIPE family member Sigmar interacts with the Ste20-like kinase Misshapen and modulates JNK signaling, cytoskeletal remodeling and autophagy

Abstract

TNFAIP8 and other mammalian TIPE family proteins have attracted increased interest due to their associations with disease-related processes including oncogenic transformation, metastasis, and inflammation. The molecular and cellular functions of TIPE family proteins are still not well understood. Here we report the molecular and genetic characterization of the Drosophila TNFAIP8 homolog, CG4091/sigmar. Previous gene expression studies revealed dynamic expression of sigmar in larval salivary glands prior to histolysis. Here we demonstrate that in sigmar loss-of-function mutants, the salivary glands are morphologically abnormal with defects in the tubulin network and decreased autophagic flux. Sigmar localizes subcellularly to microtubule-containing projections in Drosophila S2 cells, and co-immunoprecipitates with the Ste20-like kinase Misshapen, a regulator of the JNK pathway. Further, the Drosophila TNF ligand Eiger can induce sigmar expression, and sigmar loss-of-function leads to altered localization of pDJNK in salivary glands. Together, these findings link Sigmar to the JNK pathway, cytoskeletal remodeling and autophagy activity during salivary gland development, and provide new insights into TIPE family member function.

Keywords: Autophagy; Cytoskeleton; Eiger; Msn; TIPE; TNFAIP8.

Fig. 1.. Creation and analysis of a genetic model of the Drosophila TIPE family member sigmar.

(A) The sigmar gene has three transcripts (RA, RB, RC) that encode two different proteins. Sigmar-CDS-RA and Sigmar-CDS-RB encode the same protein (188 aa) and Sigmar-CDS-RC encodes a protein that is slightly larger (210 aa). The sigmar and l(2)dtl genes are separated by a 78 bp region. The strain EY06821 contained a P element inserted in the 3′UTR region of sigmar (arrowhead). Broken lines indicate the genomic region deleted in the sigmar^Df-C23 strain generated by the imprecise excision of EY06821. Primers l(2)dtl 2723R and Sigmar 1136R were used to detect the deletion PCR product of sigmar^Df-C23. The sigmar^Df-C23 strain has a deletion in both l(2)dtl and sigmar gene coding regions. (B) QRT-PCR analysis indicates that the sigmar^S strain does not express sigmar but does express l(2)dtl transcripts similar to the control OreR and w¹¹¹⁸ strain. Fold expression shown (mean ± s.d.) is relative to expression levels in OreR females using rp49 as the reference RNA. (C) QRT-PCR results show that sigmar expression increased dramatically (approximately 185 fold) prior to the death stage (23 hours APF at 18°C) in OreR larval salivary glands. Fold change in expression (mean±s.d.) is relative to the 16 hours APF time point using rp49 as the reference RNA. (D) In situ hybridization of larval salivary glands confirms that sigmar expression is not detectable at 16 hours APF at 18°C but is abundant at the death initiating stage, 23 hours APF. Scale bar equals 100 microns. (E) sigmar expression is increased in eiger-expresssing salivary glands. Salivary glands were dissected at 16 hours APF, 20 hours APF and 23 hours APF (at 18°C) from wild-type (OreR) animals and from a strain ectopically expressing eiger in salivary glands (D59-Gal4>eiger). The mRNA expression levels of eiger and sigmar were determined using QRT-PCR and are shown as fold changes in expression (Log scale, mean±s.d.) relative to OreR at 16 hours APF. Results show that sigmar expression was increased 7 fold and 3 fold in the eiger-expressing salivary glands relative to wild type at 20 hours APF and 23 hours APF, respectively.

Fig. 2.. Sigmar localizes to cellular projections in S2 cells growing on CC2 coated slides.

A–H show untreated S2 cells, while I–P show S2 cells treated with latrunculin B (actin-disrupting) for 15 minutes and Q–X show cells treated with vinblastine (microtubule-disrupting) for 35 minutes. A cell stained for actin with rhodamine-phalloidin (A, red) and Sigmar (B, green) shows some co-localization of actin and Sigmar (C,D; yellow-orange) in the extended processes (D, arrow); ‘V’ = vacuolar structure. A cell stained for tubulin, a component of microtubules (MTs) (E, red), and Sigmar (F, green) also shows co-localization, especially in the long extended protrusions (G) and around the vacuoles (panel G, labelled ‘V’ and with an arrowhead). Sigmar appears as patches along the protrusions and at the tips (H, arrow). Latrunculin B treatment disrupts actin (I, red), however, Sigmar (J, green) still localizes in the lamellar areas (‘L’) and along the protrusions (panels K and L, arrows). (M–P) A latrunculin B treated cell stained for tubulin (M, red) and Sigmar (N, green) depicts Sigmar protein co-localizing to intact tubulin particularly surrounding vacuoles (O, arrowhead) and in cellular extensions (P, arrows) similar to untreated cells. Disruption of tubulin with vinblastine does not affect the actin filaments (Q, red) and Sigmar (R, green) is still present in the central area of the cells but little Sigmar protein was observed in the extensions (R, arrow). The overlay (S) and magnified view of actin extensions (T) are also shown. Vinblastine treatment disrupted the majority of tubulin filaments (U, red) but some tubulin filament remnants remained (U). Sigmar (V, green) co-localized with these remnants (panel W–X; arrows) but appeared reduced. Scale bars in A–C,E–G,I–K,M–O,Q–S,U–W equal 5 microns. Scale bars in D,H,L,P,T,X equal 25 microns. Boxed areas in C,K,O,S,W show the regions magnified in subsequent panels.

Fig. 3.. Salivary glands of sigmar^S mutants are morphologically abnormal.

(A) Salivary glands from control (w¹¹¹⁸) and (B) sigmar^S null mutant pupa (2 hours APF). Compared to wild-type (w¹¹¹⁸) salivary glands in A, sigmar^S salivary glands in B are comprised of abnormally enlarged cells that appeared to contain large empty vacuoles (black arrows) that were clearly evident after 1–2 hours APF at 18°C. Note also the abnormal arrangement of nuclei (B, DAPI) in the mutant gland at this stage. Images were captured with a Zeiss fluorescence microscope (×10 objective). Differential interference contrast (DIC) images are shown with the corresponding DAPI stained salivary glands below. (C) Salivary gland histolysis is delayed in sigmar^S mutants compared to w¹¹¹⁸ controls. White pre-pupae were incubated at 18°C for indicated time period. Salivary glands were harvested and scored for the extent of histolysis (n>10). (D) H&E staining of salivary glands from whole fixed pupae at 24 hours APF in w¹¹¹⁸ controls is shown (n = 6). (E) H&E staining of salivary gland thin section from whole fixed pupae 24 hours APF in sigmar^s mutants. Whole salivary glands appear enlarged with vacuole-like structures. H&E staining positive regions are apparent (n = 6).

Fig. 4.. Salivary glands from sigmar^S null mutants show defects in the tubulin network.

(A) Salivary glands from w¹¹¹⁸ control and (B) sigmar^res strain control (rescue strain for sigmar^S) appear normal. (C) Salivary glands from the sigmar^S null-mutant stained for tubulin using an anti-beta tubulin antibody, appeared abnormal with some grossly enlarged cells (C, arrows). (D) Magnified view of a cell in control animals shows a normal tubulin network. (E) Magnified view of one of the grossly enlarged salivary gland cells from the sigmar^S mutant shows a fragmented tubulin network (E, arrowhead). (F,G) Some cells in the sigmar^S mutant salivary glands displayed a dense tubulin network pattern with long microtubules that appeared as fibers. Arrows in magnified panel G show these long ‘fiber-like’ microtubules. Salivary glands cells from w¹¹¹⁸ and sigmar^S were categorized as having a normal (H), fragmented (I), sparse (J) or dense (K) tubulin network and the frequency of each cell phenotype from w¹¹¹⁸ (n = 10 glands) and sigmar^S (n = 10 glands) determined. A total of 497 cells were quantified and results presented as a bar graph (L). Scale bars in A–C equal 100 microns, in D–G equal 20 microns. Images A–G were taken with a Nikon confocal microscope and shown is a single Z slice. Images H–K were taken with a Zeiss fluorescence microscope using an ApoTome (Zeiss).

Fig. 5.. Salivary glands from sigmar^S null mutants show defects in autophagy.

(A) In control OreR salivary glands at 24 hours APF (at 18°C), monodansylcadaverine (MDC)-positive autolysosomes appear as punctate structures throughout the cells (arrowheads). (B) In the sigmar^S mutant, some salivary gland cells show MDC-positive autolysosomes (left half of B, arrowheads). However, some enlarged cells show dramatically reduced MDC positive puncta (right half of B) and large vacuole-like structures (arrow). (C) In control w¹¹¹⁸ salivary glands, Ref(2)P protein aggregates were not detected. (D) In the Atg7 mutant salivary glands, Ref(2)P protein aggregates were observed in all cells, indicative of a defect in autophagic flux. (E) In the sigmar^S mutant salivary glands, some of the abnormal cells, typically adjacent to grossly enlarged/vacuolized cells (*), contained abundant Ref(2)P protein aggregates. At 24 hours APF, salivary glands (n = 16) from w¹¹¹⁸ (F) and sigmar^S (G) pupae were dissected and the frequency of MDC+ and MDC- cells (corresponding to either normal sized or grossly enlarged cellular phenotypes; X axis) were quantified from a total of 159 cells and 155 cells per genotype, respectively, and the data is presented as a bar graph. Scale bars in A,B equal 250 microns, and in C,D,E equal 100 microns. Nuclei are marked with ‘N’ in A and B. Images were taken with a Zeiss Axioplan2 microscope.

Fig. 6.. Sigmar is associated with the DJNK signaling pathway.

(A) Sigmar interacts with Msn. S2 cells were co-transfected with C- or N-terminally FLAG-tagged Msn and Myc-Sigmar. Control cells were transfected with Myc-Sigmar only. Immunoprecipitations (IP) 72 hours after transfections with anti-FLAG antibody were performed and Sigmar was detected using anti-Myc antibody. The figure shown is representative of three independent IP replicates. (B) pDJNK localization is altered in sigmar^S mutant salivary glands. OreR salivary gland at 24 hours APF shows pDJNK immunostaining in all nuclei (pDJNK = green, DAPI = red). The sigmar^S salivary gland at 24 hours APF shows diffuse cytoplasmic staining of pDJNK in most cells although some regions (left) show nuclear pDJNK localization. Whole salivary gland images (top panel) were captured using a 10× objective and the zoomed images were taken with a 20× objective (Zeiss Axioplan 2). (C) Quantitation of nuclear-localized pDJNK in OreR and sigmar^S salivary glands at two time points. At both 20 hours APF and 24 hours APF, sigmar^S salivary glands show reduced nuclear localization of pDJNK. At least 10 salivary glands were examined from each genotype at each time point (P<0.003).