Drosophila melanogaster model of RVCL-S demonstrates age dependent disease progression

Reviewer #1 (Public review):

Summary:

In this manuscript, the authors describe the generation of a Drosophila model of RVCL-S by disrupting the fly TREX1 ortholog cg3165 and by expressing human TREX1 transgenes (WT and the RVCL-S-associated V235Gfs variant). They evaluate organismal phenotypes using OCT-based cardiac imaging, climbing assays, and lifespan analysis. The authors show that loss of cg3165 compromises heart performance and locomotion, and that expression of human TREX1 partially rescues these phenotypes. They further report modest differences between WT and mutant hTREX1 under overexpression conditions. The study aims to establish Drosophila as an in vivo model for RVCL-S biology and future therapeutic testing.

Strengths:

(1) The manuscript addresses an understudied monogenic vascular disease where animal models are scarce.

(2) The use of OCT imaging to quantify fly cardiac performance is technically strong and may be useful for broader applications.

(3) The authors generated both cg3165 null mutants and humanized transgenes at a defined genomic landing site.

(4) The study provided initial in vivo evidence that human TREX1 truncation variants can induce functional impairments in flies.

Weaknesses:

(1) Limited mechanistic insight.

RVCL-S pathogenesis is strongly linked to mislocalization of truncated TREX1, DNA damage accumulation, and endothelial/podocyte cellular senescence. The current manuscript does not examine any cellular, molecular, or mechanistic readouts - e.g. DNA damage markers, TREX1 subcellular localization in fly tissues, oxidative stress, apoptosis, or senescence-related pathways. As a result, the model remains largely phenotypic and descriptive.

To strengthen the impact, the authors should provide at least one mechanistic assay demonstrating that the humanized TREX1 variants induce expected molecular consequences in vivo.

(2) The distinction between WT and RVCL-S TREX1 variants is modest.

In the cg3165 rescue experiments, the authors do not observe differences between hTREX1 and the V235Gfs variant (e.g., Figure 3A-B). Phenotypic differences only emerge under ubiquitous overexpression, raising two issues:

(i) It is unclear whether these differences reflect disease-relevant biology or artifacts of strong Act5C-driven expression.

(ii) The authors conclude that the model captures RVCL-S pathogenicity, yet the data do not robustly separate WT from mutant TREX1 under physiological expression levels.

The authors should clarify these limitations and consider additional data or explanations to support the claim that the model distinguishes WT vs RVCL-S variants.

(3) Heart phenotypes are presented as vascular defects without sufficient justification.

RVCL-S is a small-vessel vasculopathy, but the Drosophila heart is a contractile tube without an endothelial lining. The authors refer to "vascular integrity restoration," but the Drosophila heart lacks vasculature.

The manuscript would benefit from careful wording and from a discussion of how the fly heart phenotypes relate to RVCL-S microvascular pathology.

(4) General absence of tissue-level or cellular imaging.

No images of fly hearts, brains, eyes, or other tissues are shown. TREX1 nuclear mislocalization is a hallmark of RVCL-S, yet no localization studies are included in this manuscript.

Adding one or two imaging experiments demonstrating TREX1 localization or tissue pathology would greatly enhance confidence in the model.

Reviewer #2 (Public review):

Summary:

The authors used the Drosophila heart tube to model Retinal vasculopathy with the goal of building a model that could be used to identify druggable targets and for testing chemical compounds that might target the disease. They generated flies expressing human TREX1 as well as a line expressing the V235G mutation that causes a C-terminal truncation that has been linked to the disease. In humans, this mutation is dominant. Heart tube function was monitored using OCM; the most robust change upon overexpression of wild-type or mutant TREX1was heart tube restriction, and this effect was similar for both forms of TREX1. Lifespan and climbing assays did show differential effects between wt and mutant forms when they were strongly and ubiquitously expressed by an actin-Gal4 driver. Unfortunately, these types of assays are less useful as drug screening tools. Their conclusion that the primary effect of TREX is on neuronal function is inferential and not directly supported by the data.

Strengths:

The authors do not show that CG3165 is normally expressed in the heart. Further fly heart tube function was similarly restricted in response to expression of either wild-type or mutant TREX1. The fact that expression of any form of human TREX1 had deleterious effects on heart function suggests that TREX1 serves different roles in flies compared to humans. Thus, in the case of this gene, it may not be a useful model to use to identify targets or use it as a drug screening tool.

The significant effects on lifespan and climbing that did show differential effects required ubiquitous overexpression using an actin-gal4 driver that does not allow the identification of tissue-specific effects. Thus, their assertion that the results suggested a strong positive correlation between Drosophila neuromotor regulation and transgenic hTREX1 presence and a negative impact from hTREX1 V235G" is not supported by these data. Also worrisome was the inability to identify the mutant TREX1 protein by Western blot despite the enhanced expression levels suggested by qPCR analysis. Mutant TREX1 cannot exert a dominant effect on cell function if it isn't present.

There are also some technical problems. The lifespan assays lack important controls, and the climbing assays do not appear to have been performed correctly. It is unclear what the WT genetic background is in Figure 1-3, so it is unclear if the appropriate controls have been used. Finally, the lack of information on the specific statistical analyses used for each graph makes it difficult to judge the significance of the data. Overall, the current findings establish the Retinal vasculopathy disease model platform, but with only incremental new data and without any mechanistic insights.

Author response:

Public Reviews:

Reviewer #1 (Public review):

Summary:

In this manuscript, the authors describe the generation of a Drosophila model of RVCL-S by disrupting the fly TREX1 ortholog cg3165 and by expressing human TREX1 transgenes (WT and the RVCL-S-associated V235Gfs variant). They evaluate organismal phenotypes using OCT-based cardiac imaging, climbing assays, and lifespan analysis. The authors show that loss of cg3165 compromises heart performance and locomotion, and that expression of human TREX1 partially rescues these phenotypes. They further report modest differences between WT and mutant hTREX1 under overexpression conditions. The study aims to establish Drosophila as an in vivo model for RVCL-S biology and future therapeutic testing.

Strengths:

(1) The manuscript addresses an understudied monogenic vascular disease where animal models are scarce.

(2) The use of OCT imaging to quantify fly cardiac performance is technically strong and may be useful for broader applications.

(3) The authors generated both cg3165 null mutants and humanized transgenes at a defined genomic landing site.

(4) The study provided initial in vivo evidence that human TREX1 truncation variants can induce functional impairments in flies.

Weaknesses:

(1) Limited mechanistic insight.

RVCL-S pathogenesis is strongly linked to mislocalization of truncated TREX1, DNA damage accumulation, and endothelial/podocyte cellular senescence. The current manuscript does not examine any cellular, molecular, or mechanistic readouts - e.g. DNA damage markers, TREX1 subcellular localization in fly tissues, oxidative stress, apoptosis, or senescence-related pathways. As a result, the model remains largely phenotypic and descriptive.

We thank the reviewers for these suggestions. We are planning to perform experiments addressing the RVCL-S linked cellular deviations. We will examine DNA damage markers on cellular level and perform TUNEL tissue staining to visualize apoptosis, etc.

To strengthen the impact, the authors should provide at least one mechanistic assay demonstrating that the humanized TREX1 variants induce expected molecular consequences in vivo.

Yes, we are planning to demonstrate the distinct effects from TREX1 and TREX1 V235G expression on molecular level.

(2) The distinction between WT and RVCL-S TREX1 variants is modest.

In the cg3165 rescue experiments, the authors do not observe differences between hTREX1 and the V235Gfs variant (e.g., Figure 3A-B). Phenotypic differences only emerge under ubiquitous overexpression, raising two issues:

i) It is unclear whether these differences reflect disease-relevant biology or artifacts of strong Act5C-driven expression.

Thanks for pointing out this issue. We will discuss the differences between two expression models in the revised manuscript.

ii) The authors conclude that the model captures RVCL-S pathogenicity, yet the data do not robustly separate WT from mutant TREX1 under physiological expression levels.

We will provide more details related to the RVCL-S disease development and agerelated manifestations.

The authors should clarify these limitations and consider additional data or explanations to support the claim that the model distinguishes WT vs RVCL-S variants.

We will address the reviewer concerns and re-write the related manuscript sections to provide more clarity.

(3) Heart phenotypes are presented as vascular defects without sufficient justification.

RVCL-S is a small-vessel vasculopathy, but the Drosophila heart is a contractile tube without an endothelial lining. The authors refer to "vascular integrity restoration," but the Drosophila heart lacks vasculature.

We will expand the model justification section and will be more careful with our statements to avoid misunderstanding of the experimental conclusions.

The manuscript would benefit from careful wording and from a discussion of how the fly heart phenotypes relate to RVCL-S microvascular pathology.

We thank the reviewer for pointing to this issue. Justifying Drosophila usage for human disease modelling is always challenging. We will re-write the corresponding parts of the manuscript.

(4) General absence of tissue-level or cellular imaging.

No images of fly hearts, brains, eyes, or other tissues are shown. TREX1 nuclear mislocalization is a hallmark of RVCL-S, yet no localization studies are included in this manuscript. Adding one or two imaging experiments demonstrating TREX1 localization or tissue pathology would greatly enhance confidence in the model.

As suggested by the reviewers,we will add tissue imaging experiments to illustrate the pathological effects of RVCL linked TREX1 expression. We are also planning to utilize CRIMIC line CR70804 to visualize fly TREX1 tissue distribution.

Reviewer #2 (Public review):

Summary:

The authors used the Drosophila heart tube to model Retinal vasculopathy with the goal of building a model that could be used to identify druggable targets and for testing chemical compounds that might target the disease. They generated flies expressing human TREX1 as well as a line expressing the V235G mutation that causes a C-terminal truncation that has been linked to the disease. In humans, this mutation is dominant. Heart tube function was monitored using OCM; the most robust change upon overexpression of wild-type or mutant TREX1was heart tube restriction, and this effect was similar for both forms of TREX1.

Our results are consistent with the human disease nature, RVCL-S carriers and non-carriers are both healthy and asymptomatic at young age; however, the accumulation of physiological stress becomes obvious in midlife, leading to premature death in 40s and 50s. We will expand the discussion section focusing on RVCL-S manifestations in aged animals.

Lifespan and climbing assays did show differential effects between wt and mutant forms when they were strongly and ubiquitously expressed by an actin-Gal4 driver. Unfortunately, these types of assays are less useful as drug screening tools. Their conclusion that the primary effect of TREX is on neuronal function is inferential and not directly supported by the data.

We will revise this experiment discussion and plan to include additional experiments to strengthen the conclusions.

The authors do not show that CG3165 is normally expressed in the heart. Further fly heart tube function was similarly restricted in response to expression of either wild-type or mutant TREX1. The fact that expression of any form of human TREX1 had deleterious effects on heart function suggests that TREX1 serves different roles in flies compared to humans. Thus, in the case of this gene, it may not be a useful model to use to identify targets or use it as a drug screening tool.

We will examine the expression of cg3165, human TREX1 transgenes in whole organism to demonstrate tissue expression profiles, as noted above. We will also expand the relevant manuscript sections to address the systemic manifestations of RVCL.

The significant effects on lifespan and climbing that did show differential effects required ubiquitous overexpression using an actin-gal4 driver that does not allow the identification of tissue-specific effects.

We plan to carry out additional experiments to determine cg3165, and human TREX1 tissue expression profile.

Thus, their assertion that the results suggested a strong positive correlation between Drosophila neuromotor regulation and transgenic hTREX1 presence and a negative impact from hTREX1 V235G" is not supported by these data.

Thanks for pointing this out. We will revise our conclusions appropriately after we include the results from additional new experiments.

Also worrisome was the inability to identify the mutant TREX1 protein by Western blot despite the enhanced expression levels suggested by qPCR analysis. Mutant TREX1 cannot exert a dominant effect on cell function if it isn't present.

We will try to resolve this issue by technical means.

There are also some technical problems. The lifespan assays lack important controls, and the climbing assays do not appear to have been performed correctly.

We would disagree with this statement. We will re-write the method description for better clarity.

It is unclear what the WT genetic background is in Figure 1-3, so it is unclear if the appropriate controls have been used. Finally, the lack of information on the specific statistical analyses used for each graph makes it difficult to judge the significance of the data.

We will provide clearer descriptions of our controls and procedures.

Overall, the current findings establish the Retinal vasculopathy disease model platform, but with only incremental new data and without any mechanistic insights.

We will include additional experiments addressing the mechanism (see previous responses above).

Reviewing Editor Comments:

I (Hugo Bellen) also read your paper and noted that you do not document the expression pattern in the nervous system and other tissues, such as the heart. The stock https://flypush.research.bcm.edu/pscreen/crimic/info.php?CRname=CR70804 may help you do this and should allow you to compare the GAL4 induced expression of the stock you created and this stock. If compatible, you should consider reporting expression patterns.

Thank you for the suggestion. We will obtain the line and will use it for expression visualization.