α-actinin accounts for the bioactivity of actin preparations in inducing STAT target genes in Drosophila melanogaster

Essential revisions:

1) It is fortunate that the same authors can identify themselves that this was not actin but α-actinin the elicitor. Their conclusion would be reinforced if another laboratory could reproduce independently their observations. Was the injection experiment carried out independently in the Wood, Dionne or Teixeira lab (order the product and monitor upd3 upon injection)? This should be the case and this should be mentioned in the paper.

As requested, we asked for the experiment to be independently reproduced by the Dionne and Teixeira lab. Both labs carried out experiments as per our protocol but using α-actinin that they purchased from Cytoskeleton (independently from each other and independently from the Reis e Sousa lab). Both the Dionne and Teixeira labs reproduced the observation that intrathoracic injection of α-actinin results in upregulation of STAT-responsive genes such as TotM (assessed by both labs) and TotA (additionally assessed by the Teixeira lab). A sentence was added to the manuscript mentioning these experiments (Results and Discussion, first paragraph). The results of one of two independent experiments per lab are shown below in Author response image 1 and 2 (figure legends provided by the originators).

Author response image 1

Download asset Open asset

Author response image 2

Download asset Open asset

2) Many studies have shown that upd3 is produced by hemocytes not the fat-body (Agaisse, 2003; Chakrabarti, 2016). C564 is a Gal4 driver that target both hemocyte and fat body. I would recommend that the authors test their hypothesis using more specific drivers: hmlGal4 (hemocyte) and Lpp-Gal4 (fat body) to reinforce their conclusion that fat body not hemocytes is sensing actinin.

As suggested by the reviewers, we have repeated knockdown of Nox, Src42A or Shark using additional fat body drivers (Lpp-Gal4 and r4-Gal4), as well as a haemocyte driver (HmlΔ-Gal4). Knockdown with all three fat body drivers (c564-Gal4, Lpp-Gal4 and r4-Gal4) resulted in complete or nearly complete loss of TotM induction upon injection of α-actinin, whereas knockdown using HmlΔ-Gal4 did not have an appreciable effect. These data are now included in Figure 3 of the manuscript. We tried to measure upd3 induction by RT-qPCR, but did not obtain clear cut results – we discussed our inability to convincingly measure upd3 transcripts in our previous study (Srinivasan et al., 2016). However, in our previous study we knocked down upd3 specifically in haemocytes (or depleted haemocytes altogether) and found that this had no impact on induction of STAT-responsive genes (Srinivasan et al., 2016 – Figures 5C, D). Although those experiments utilised purified actin and not α-actinin, we believe that those data, together with the experiments included in the present study, support the notion that haemocytes are dispensable for the response to cytoskeletal protein injection, which involves the fat body.

3) The three reviewers consider that the present paper is more a correction (of a mistake) than an extension. This should be mentioned in the Abstract and text (Introduction, 'corrected' instead of 'extend').

We are uncomfortable with stating that we are “correcting a mistake” as it can be read to mean that either the data in the previous paper were incorrect or that we made an actual mistake (such as misread actin for actinin on a tube label). In actual fact, we are revising the interpretation of those data and refining the conclusions. We therefore have altered the Abstract and Introduction, as requested, to state that we revise our earlier conclusion/revise interpretation of the earlier data.