A total of 2,535 were used in experiments. The following stocks were used elav-Gal4, OK371-Gal4, UAS-eGFP, ubi-Gal80^ts as previously described (27). Interfering RNA (IR) stocks UAS-anxB11^IR HMS01775, was obtained from Bloomington Drosophila Stock Centre (BDSC) and GD36186 was obtained from Vienna Drosophila Resource Center (VDRC)). UAS-TAR DNA-binding protein wild-type (TDP43^WT (human TDP43 Wild-Type) and UAS-SOD1^G85R (Human SOD1 Gly85Arg mutant) were a gift from Dr Jemeen Sreedharan (King's College London). UAS-TBPH was a gift from Dr Frank Hirth (King's College London).

Human ANXA11 transgenes were generated by PCR cloning the human open reading frames from previously described vectors (5) adding a Myc tag at the C terminal. PCR products were purified with QIAquick PCR purification kit (Qiagen GmbH), cut with XbaI and XhoI enzymes and cloned in the corresponding sites in the multi cloning site in pUAST-attB vectors (a gift from Dr Joe Bateman (King's College London)]. Presence of the correct ANXA11 mutation was validated by Sanger sequencing performed by Eurofins Genomics. Insertions were generated by injection in the attP40 Drosophila stocks at the Cambridge Fly Facility (Cambridge, UK) injection service. The following Oligos from Eurofins Genomics were used for PCR and cloning (restriction sites are underlined, Myc tag encoding sequence in lower case, ANXA11 sequences in italics): Forward, 5'-GACTCGAGATGAGCTACCCTGGCTATCC-3' and reverse, 5'-AGTCTAGATTAc agatc ctct tctgagatgagtttttgttcGTCATTGCCACCACAGATCTTCAGC-3'.

All fly stocks were routinely maintained at 18˚C in an incubator with 60% humidity and standard 12/12-h light cycle on a standard fly food mixture of yeast, agar and cornmeal with nipagin and propionic acid.

Flies were monitored over the course of their life cycle to quantify death across genotypes as previously described (27). Briefly, newly eclosed flies were collected daily for 3 days and transferred to a 29˚C incubator in batches of 20 flies/vial (equal mix of males and females). Three times/week the flies were counted and transferred into a new vial with fresh fly food using CO₂ to anaesthetise the flies. The number of flies still alive was recorded each time and flies that escaped or were stuck in the food were censored (attributed a value of 0 on the Day). A dead fly was attributed the value 1 on the day of death.

Flies were kept at 29˚C and age-matched female flies of the genotypes were placed into empty 70 mm tubes (10 flies/tube). When flies are tapped to the bottom of a vial, they immediately climb to the top of the vial due to their innate negative geotaxis abilities (27). To assess negative geotaxis, flies were tapped to the bottom of the vial after acclimatization and distance climbed by the flies in 2-min intervals over five trials was measured as previously described (27). The number of flies climbing to each cm increment was scored and a genotype mean was calculated for each time point across five trials. Flies that jumped or did not perform a vertical climb in one movement burst were excluded from that trial.

Whole mount larval ventral nerve cords were dissected from wandering third instar larva fixed in 4% paraformaldehyde (Sigma) for 45 min on ice, blocked for 1 h at room temperature (RT) in Phosphate Buffer Saline (PBS) complemented with 0.2% Tritox-X100 (Sigma) and 10% Normal Goat Serum (NGS, Gibco) and stained with a rabbit anti-GFP (1:300, cat. no. A11122; Thermo Fisher Scientific, Inc.) and a mouse monoclonal antibody against Myc tag (1:100, cat. no. 9E10; Roche Diagnostics, Ltd.). Additionally, flies were aged for 12 days at 29˚C, sacrificed and whole brains were dissected, fixed and blocked as aforementioned. The mouse primary antibody against TDP43 (used at 1:500) was a gift by Dr Jemeen Sreedharan (King's College London). Brains were imaged on a Nikon A1R inverted confocal microscope and analysis was performed in NIS Elements (Nikon 5.21). Secondary antibodies used were Alexa 555 anti-mouse and Alexa 488 anti-rabbit (Thermo fisher Scientific, Inc.; cat. nos. A21422 and A11008 and) at 1:200. All antibodies were diluted in Blocking solution. Primary antibodies were incubated over night at 4˚C, while secondary antibodies were incubated 1 h at RT. For TDP43 localization, data was normalised for cell area and expressed as a ratio of cytoplasmic TDP43 over nuclear TDP43.

All data are presented as mean ± SEM and were analysed using Microsoft Excel (office 365, www.microsopht.com) and GraphPad Prism 9 (www.graphpad.com). Lifespan was analysed with the Kaplan-Meyer log-rank (Mantel-Cox) test. One-way ANOVA with Dunnett's multiple comparisons post hoc test was used for comparison of ≥3 groups of normally distributed data. Kruskal-Wallis non-parametric analysis with Dunn's multiple comparisons test was used for comparison of ≥3 groups of non-normally distributed data. Two-way ANOVA with Dunnett's multiple comparisons post hoc test was used for negative geotaxis assays where ≥2 groups of normally distributed data were confounded by a third parameter (timepoints). P<0.05 was considered to indicate a statistically significant difference. All experiments have been performed at least three times.

Sequence comparison revealed that Drosophila ANXB11 was closely related to human ANXA11 than other Drosophila genes were, with two splicing isoforms expressing an ANX protein with a long N-terminal tail (Fig. 1A). While the N-terminus containing the human mutation hotspot (22) around aa36-40 is not conserved, other key residues mutated in ALS, such as G175 and R235, are also conserved in Drosophila (Fig. 1A).

The present study constructed overexpression transgenes for human ANXA11 with G175R and R235Q mutations, and a corresponding wild-type (WT) transgene. All transgenes were knocked into a well-characterised neutral genomic locus to guarantee similar expression levels and neutrality for insertional mutagenesis.

The present study demonstrated the expression of ANXA11 transgenes and notable enrichment of G175R and R235Q mutants around and within the cell nucleus in larval neurons (Fig. 1B and C). In addition, the R235Q mutant protein displayed notable aggregation (Fig. 1B), as reported in human cells (5).

Our previous study showed that knockdown of the fly ANXB11 gene decreases in lifespan and negative geotaxis (28). The introduction of human ANXA11 transgenes successfully rescued the short lifespan observed following pan-neuronal knock-down of the endogenous ANXB11 (Fig. 1D). When analysing the negative geotaxis due to motor neuron knockdown, however, the rescue was more limited, and only the WT A11 construct constantly improved negative geotaxis after 21 days (Fig. 1E).

Despite successful expression, altered subcellular localization and functional alterations of the ANXA11 transgenes, there were no obvious signs of organism-level toxicity when mutant ANXA11 transgenes were expressed in Drosophila neurons, both in terms of lifespan and negative geotaxis. Genetic interactions with other known ALS genes were assessed using transgenic flies expressing the human genes that display toxicity in Drosophila, such as TDP43(24) and SOD1(25).

ANXA11 transgenes markedly worsened lifespan and negative geotaxis deficits in Drosophila expressing TDP43, a protein associated with ALS pathology (Fig. 2A and B). ANXA11 transgenes significantly enhanced the negative geotaxis defects cause by TDP43 after 11 days and made lifespan significantly shorter.

Negative geotaxis response and lifespan analysis showed that OK371-Gal4, TDP43 WT flies illustrate decreased negative geotaxis compared with the ANX mutants alone, which did not significantly affect negative geotaxis. When co-expressed with human TDP43, all ANXA11 transgenes exacerbated the negative geotaxis defects seen with TDP43 alone (Fig. 2B). Annexin A11 transgenes further decreased lifespan compared with TDP43 (Fig. 2B). These effects were not observed in models expressing Drosophila TBPH, an ortholog of TDP43 essential for Drosophila motor neurons (26). TBPH had stronger negative geotaxis defects alone than when co-expressed with human ANXA11 (Fig. 2C) and ANXA11 did not affect TBPH lifespan (Fig. 2D).

To determine if the modulating effect of ANXA11 transgenes occurred in other ALS-associated genes, the association between ANXA11 and SOD1 (a gene implicated in familial forms of ALS) (25) was assessed. There was no discernible genetic interplay between A11 and G85R mutant SOD1 (Fig. S1).

To determine the potential molecular mechanism underlying the interaction with human TDP43 of the ANXA11 transgenes, the present study assessed expression of TDP43 in the adult fly brain. Overexpression of ANXA11 altered the typical distribution of TDP43, increasing its nuclear accumulation (Fig. 3A and B), which may interfere with key nuclear functions.

Furthermore, the extent of TDP43 nuclear accumulation was decreased by G175R and R235Q mutations in ANXA11 compared with WT ANXA11, indicating that ALS-mutations in ANXA11 may cause promote cytoplasmic localisation of TDP43 compared with WT ANXA11 (Fig. 3B). These varying degrees of TDP43 nucleo-cytoplasmic localization suggested that the link between ANXA11 and TDP43 may be sensitive to the structural changes induced by these mutations either by gain or loss of function.