Tracheal terminal cells (TTCs) show reduced immune response to natural infection.

Drs-GFP larvae were infected with Erwinia carotovora for 24 h, and GFP fluorescence in the terminal structures of the tracheal system was monitored. Images were taken in the DIC (A–H) and GFP channels (A’-G’, E, H). (A, B) Dorsal TTCs (A) and visceral TTCs (B) show no fluorescence. (C) Dorsal view of the tracheal system, showing the dorsal trunks branching into the dorsal branch (DB) and the dorsal TTCs. (D) Percentage of larvae showing GFP fluorescence in the DB and TTCs. (E–H) TTCs show expression of Drs-GFP. White arrows indicate shortened TTC branches (G). The black arrows marks a melanization site and the arrowhead marks a translucent branch without air filling (H). Dashed lines represent the proximal end of the TTCs. Scale bar, 50 µm.

TTCs do not express the Imd receptor PGRP-LCx.

(A–D) The secreted Imd pathway receptor PGRP-LE is expressed in the main parts of the tracheal system (ppk4>PGRP-LE (GFP-Drs)). The arrows indicate TTCs not expressing GFP. The dashed lines represent the proximal end of the TTC. (A–C) An activated immune response in the larvae is visualized by expression of GFP-tagged Drosomycin (Drs). (D, E) Detailed TTCs were observed in fillet preparations (D) and in the dissected intestine (E) in both the DIC (D, E) and GFP channels (D’, E’). (F, G) Expression of GFP under the control of a PGRP-LCx promoter (PGRP-LCx-Gal4 > UAS-GFP) revealed a lack of promoter activity, and expression of GFP, in TTCs on the cuticle (F’, F = merged image in DIC channel) and intestine (G’, G = merged image in DIC channel), which is in contrast to the rest of the tracheal system. (H, I) The TTCs in the tracheal system are visualized by GFP expression in the cuticle (H) and intestine (I) of wild-type larvae (btl-Gal4; UAS-GFP). Scale bar, 50 µm. dt = dorsal trunk.

Expression of PGRP-LCx by TTCs leads to shrinkage and loss of functionality.

(A, B) Dorsal TTCs in the control (A) and DSRF-driven overexpression of PGRP-LCx in TTCs (B). (C, D) Measurement and quantification of the number (C) and length (D) of branches (n=22–45). Data are presented as the mean ± SD. (E) The hypoxia sensitivity assay was conducted with control and PGRP-LCx-expressing 3rd instar larvae under hypoxic conditions (2–3 % O2, n = 11–14). Data are presented as the means ± SEM. Statistical significance was tested using Mann-Whitney-U test. * p < 0.05, ** p < 0.01, *** p < 0.001, **** p < 0.0001. (F) Culture vials containing control, hid;rpr, and Tak1 larvae (DSRF > hid;rpr/Tak1) at 4 days post-oviposition. (G, H) Transmission light microscopy of dissected intestines from 3rd instar larvae (F, G) or 2nd instar larvae (H) with the connected TTCs of control (G) showing expression of hid;rpr (H) or PGRP-LCx (I). (J, K) Dissected intestines from control larvae (J) or larvae expressing PGRP-LCx in TTCs (K) were stained with an antibody specific for cleaved Drosophila Dcp-1 (purple). (I’, J’) and then counterstained to detect GFP (green). (I’’, J’’) Merged channels. Scale bars, 50 μm.

JNK signaling is associated with impaired TTC branching.

(A) Schematic showing how the PGRP-LC-activated Imd signaling pathway is subdivided into the NF-κB (Relish) and JNK (hep = JNKK, bsk = JNK) pathways. (B–D) Relish (B, Rel68), as well as constitutively active hep (C, hepCA) and bsk (D, bskOE), were expressed in TTCs (DSRF >). (E, F) Measurement and quantification of the number (E) and length (F) of branches in control (w1118) versus PGRP-LCx-expressing TTCs (n=22–45). Data are presented as the mean ± SD. Statistical significance was evaluated using Mann- Whitney-U test, ** p < 0.01, **** p < 0.0001, ns = not significant. (G, H) Dissected intestines from control (DSRF > w1118) and PGRP-LCx-expressing flies (DSRF > PGRP-LCx), in which the TTCs were stained to detect Relish (Rel, G) and pJNK (purple, H). TTCs were counterstained with GFP (green). Arrows mark the TTC nucleus. Merged channels are shown. Scale bars, 50 µm.

The TTC phenotype induced by PGRP-LCx is dependent on the transcription factors kay and foxo.

(A) Schematic illustration of the JNK signaling pathway downstream of Tak1, which includes Ets21C, kay, and Jra (AP1). (B–E) DSRF-driven PGRP-LCxOE in TTCs was combined with the dominant-negative form of Tak1, bsk (B), and kay (C), or with RNAi targeting foxo (D) or Ets21C (E). (F, G) Measurement and quantification of the number (F) and length (G) of branches (n=16–45). (H–K) Measurement and quantification of the number (H) and length (I) of GFP expressing branches in control (w1118) and PGRP-LCx-expressing cells (n=7–45). TTCs overexpressing foxo (J, foxoOE), and kay and Jra (K, kayOE + JraOE). (L, M). TRE-RFP expression in control (L) and PGRP-LCx-expressing TTCs (M). (N) Foxo promoter activity in TTCs (foxo-Gal4 > UAS-GFP). Data are expressed as the mean ± SD. Statistical significance was evaluated using Mann-Whitney-U test, * p < 0.05, *** p < 0.001, **** p < 0.0001, ns = not significant. The color of the asterisk indicates the corresponding comparison. Dashed lines represent the mean control value. Scale bar, 50 µm.

Targeted reduction of foxo expression in TTCs leads to hyperbranching.

(A, B) Representative tracheal branching in control (A) and DSRF-driven foxoRNAi(B) cells under normoxic conditions. (C, D) Representative images showing tracheal branching in control (C) and foxORNAi (D) cells under hypoxic conditions. (E) Quantification of branches in control and DSRF > foxORNAi TTCs under normoxic (21%) and hypoxic (5%) conditions. Scale bar, 50 µm. n=21, Data are presented as the mean ± SD. Statistical significance was evaluated using Mann-Whitney-U t-test, * p < 0.05, **** p < 0.0001, ns = not significant.

Schematic summarizing the JNK-mediated immune/stress response in the trachea and TTCs.

(A) Tracheal infection leads to an immune response involving expression of antimicrobial peptides such as Drosomycin (Drs, green). In most cases, the immune response is restricted to the tracheal trunks and the TTCs are unaffected (bold arrow). In rare cases, TTCs express Drs, resulting in an impaired phenotype (dashed arrow). (B) Imd signaling in the main tracheal trunks is induced by peptidoglycan recognition receptors (PGRP)-LC and -LE. Downstream, the signaling branches into a Relish (Rel) and a JNK signaling pathway. Activation of the pathways mediates airway remodeling (13). However, activation in TTCs is avoided by the absence of PGRP-LC, even though all other JNK signaling pathway components are present. The pathway can be activated by external stressors, resulting in AP-1- mediated cell death. The transcription factor foxo, which is component not only of the JNK signaling pathway but also of the insulin signaling pathway, plays a role in TTC homeostasis and their ability to branch.