Live applications of norbormide-based fluorescent probes in Drosophila melanogaster

In this study we investigated the performance of two norbormide (NRB)-derived fluorescent probes, NRBMC009 (green) and NRBZLW0047 (red), on dissected, living larvae of Drosophila, to verify their potential application in live cell imaging confocal microscopy. To this end, larval tissues were exposed to NRB probes alone or in combination with other commercial dyes or GFP-tagged protein markers. Both probes were rapidly internalized by most tissues (except the central nervous system) allowing each organ in the microscope field to be readily distinguished at low magnification. At the cellular level, the probes showed a very similar distribution (except for fat bodies), defined by loss of signal in the nucleus and plasma membrane, and a preferential localization to endoplasmic reticulum (ER) and mitochondria. They also recognized ER and mitochondrial phenotypes in the skeletal muscles of fruit fly models that had loss of function mutations in the atlastin and mitofusin genes, suggesting NRBMC009 and NRBZLW0047 as potentially useful screening tools for characterizing ER and mitochondria morphological alterations. Feeding of larvae and adult Drosophilae with the NRB-derived dyes led to staining of the gut and its epithelial cells, revealing a potential role in food intake assays. In addition, when flies were exposed to either dye over their entire life cycle no apparent functional or morphological abnormalities were detected. Rapid internalization, a bright signal, a compatibility with other available fluorescent probes and GFP-tagged protein markers, and a lack of toxicity make NRBZLW0047 and, particularly, NRBMC009 highly performing fluorescent probes for live cell microscopy studies and food intake assays in Drosophila.


Introduction
Norbormide [5-(α-hydroxy-α-2-pyridylbenzyl)-7-(α-2-pyridylbenzylidene)-5-norbornene-2,3-dicarboximide] (NRB) is a selective rat toxicant that exhibits little or no non-target effects [1], and was developed and commercialized as an ecologic pesticide in the 1980s. Evidence suggests that the rat-selective action of NRB is mediated by a generalized vasoconstrictor effect a1111111111 a1111111111 a1111111111 a1111111111 a1111111111 that has only been observed in the rat peripheral blood vessels, both in vivo and in vitro. In contrast, NRB displays a vasorelaxant action in arteries from non-rat species, as well as in rat aorta and extravascular smooth muscle, that has been proposed to be the result of a reduction of Ca 2+ entry through L-type Ca 2+ channels [2,3]. The molecular mechanism underlying NRBinduced vasoconstriction is not known, however, it has been proposed that the compound acts on rat vascular myocytes where it activates the PLC-IP3-PKC pathway [2], a signaling cascade stimulated by most receptor-coupled vasoconstrictor agents [4]. In an attempt to identify the cellular targets of NRB, we previously developed fluorescent derivatives of the parent compound by linking it to either nitrobenzoxadiazole (NBD) or boron-dipyrromethene (BODIPY FL) fluorophores, and found that both were able to clearly label intracellular structures such as endoplasmic reticulum (ER), Golgi apparatus, mitochondria, and lipid droplets (LDs) in various cell lines, in the absence of cytotoxic effects. Based on these results, we proposed NRB as a scaffold for the development of new, high performing, non-toxic fluorescent probes for live cell imaging [5,6].
Drosophila melanogaster is an animal model widely used to investigate the biochemical pathways and the cellular/subcellular morphological alterations that characterize human diseases [7][8][9][10][11][12]. Confocal fluorescent microscopy live cell imaging is a particularly informative methodology in this model, especially when fluorescent probes are used in combination with genetic tools (e.g. mutant flies, RNA interference, fluorescently marked proteins, Gal4/UAS activation system) [13,14], allowing the visualization of dynamic biological processes in living systems without the artifacts often generated by sample fixation procedures [15].
In this study, we investigated the performance of two NRB-derived fluorescent probes, the previously developed NRB MC009 (green fluorescence) and the newly developed NRB ZLW0047 (red fluorescence), on dissected, living third instar larvae of Drosophila melanogaster, to assess their potential application in live cell imaging confocal microscopy. In particular, we were able to characterize the distribution of NRB MC009 and NRB ZLW0047 to cellular structures and organelles in tissues of wild-type Drosophila larvae, as well as in tissues of mutant lines exhibiting morphological alterations of endoplasmic reticulum and mitochondria. Finally, we explored if these probes could be useful for studying fruit fly feeding and gut morphology.

Fluorescence spectra
Excitation and emission spectra of NRB MC009 and NRB ZLW0047 were obtained by diluting the stock solution in Milli-Q water to reach the final concentration of 2 μM. To verify that the medium used for Drosophila live imaging (HL3) had no effect on the emission spectra, a comparison was made between probes diluted in water and probes diluted in HL3-no variation in fluorescence was observed.

Larval dissection
Fly larvae having reached the third instar stage were pinned between the posterior spiracles and above the mouth hooks in a Sylgard dissection dish, and cut along the dorsal midline. Hemolymph-like (HL3) saline (70 mM NaCl, 5 mM KCl, 1.5 mM CaCl 2 , 20 mM MgCl 2 , 10 mM NaHCO 3 , 5 mM trehalose, 115 mM sucrose, 5 mM sodium HEPES, pH 7.2, all supplied by Sigma-Aldrich) was added and the lateral flaps were fastened with four needles to stretch the body wall. All larval organs were left on the muscle fillet for whole larval acquisition, whereas for single tissue acquisition the unnecessary organs were removed.

Three-choice preference assay
To test larval food preference, a Petri dish was divided in three quadrants filled with a warm liquefied standard food solution. In two quadrants, the food contained either NRB MC009 or NRB ZLW0047 (both 20 μM), in the third the food was left probe-free. At the beginning of the experiment, a group of 10 larvae was placed in the middle of the assay plate and after 5 min the number of larvae on each quadrant was counted [20]. The experiment was repeated five times.

Food intake test
The food intake test was conducted in third instar larvae starved for 1 hour in 20% sucrose dissolved in PBS. The starved larvae were then divided into three groups, separately fed with brilliant blue R dye 0.08%, NRB MC009 , or NRB ZLW0047 (both at a concentration of 20 μM) dissolved in liquid food (sucrose 20% and 20% dry yeast in PBS) for 30 min [21], frozen at -80C and imaged using a Leica MZ 16 FA microscope. To estimate food intake, the area of dyelabeled gut, visible through the cuticle, was quantified relative to larval total body area. 10 larvae for each experiment were used and the experiment was repeated three times. Image analysis was performed with ImageJ 1.52h software.

Analysis of gut fluorescence in Drosophila chronically fed with NRB MC009 and NRB ZLW0047
Third instar larvae, grown in NRB MC009 -and NRB ZLW0047 -enriched food as described in the chronic toxicity assay, were collected and food debris was removed by washing in PBS for 5 minutes and 70% ethanol for 1 minute and then dissected, being careful to not cut the digestive tract. Images were acquired using a Zeiss Axio Observer Z1 inverted microscope equipped with a Zeiss Plan-Apochromat 5x/0.15 ph1 or 40x/0.95 objectives.

Chronic toxicity assay
The toxicity of NRB-derived probes was investigated by exposing the flies to NRB MC009 or NRB ZLW0047 20 μM over the entire life-cycle (mating, eggs maturation, pupal development, eclosion). Male and female w [1118] flies were placed in a tube with standard food containing vehicle, NRB MC009 , or NRB ZLW0047 20 μM, and grown at 25˚C. After 5-6 days, parent flies were removed and analyzed to verify probe intake. Eggs were left to develop in food containing one of each of the NRB fluorescent probes until the eclosion. Two parameters were considered in the evaluation of toxicity: 1) eclosion rate (percent of emerged flies versus the total number of pupae, including the dead ones); 2) morphological alterations of eclosed adults. Morphological alterations of male and female adult body, eyes, wings, and legs were evaluated and imaged using a Leica MZ 16 FA microscope.

Statistical analysis
Analysis of colocalization was performed using Pearson's correlation coefficient calculated with Coloc2 plugin of Fiji [22]. All values are expressed as means ± standard error of the mean (SEM) of n observations (n�10). The significance of the probes fluorescence intensity versus background subtracted fluorescence intensity was calculated using unpaired t test. The significance of the three-choice assays, food intake tests and eclosion rate was calculated using One-way ANOVA test followed by Tukey's Multiple Comparison Test, using GraphPad Prism 6.01.

Synthesis and excitation/emission spectra of fluorescent probes NRB MC009 and NRB ZLW0047
The structures and synthetic routes to generate NRB MC009 and NRB ZLW0047 are summarized in Fig 1A and 1B, respectively. The excitation and emission spectra of the dyes in water, and in the physiological buffer (HL3) solution, which was used to perform live imaging experiments in Drosophila tissues, are presented in Fig 1C and 1D, respectively, and summarized in Fig 1E. The excitation spectra analysis showed that the probes can be excited using the common laser lines 561 nm (NRB ZLW0047 ) and 488 nm (NRB MC009 ), respectively.

Tissue distribution of NRB MC009 and NRB ZLW0047 in Drosophila larvae
To characterize NRB MC009 and NRB ZLW0047 fluorescence distribution in Drosophila tissues, we dissected a wild type larva and exposed the whole body to the fluorescent probes. Figs 2A and 3A are confocal images of dissected third instar larvae in which all tissues were left intact and labeled with 500 nM NRB MC009 or 1 μM NRB ZLW0047 , respectively. As shown in the pictures, both dyes localized to most of the tissues, with an apparent brighter signal detected in imaginal discs, tracheal system, salivary glands, and fat body (Fig 2B, 2D, 2H and 2J, respectively, and Fig 3B, 3H and 3J, respectively), and a less intense signal being detected in muscular tissues, oenocytes, the entire digestive tract, the ring gland, and epidermal cells (Fig 2C, 2E, 2F, 2G, 2I and 2K, respectively, and Fig 3C, 3E, 3F, 3G, 3I and 3K, respectively). NRB MC009 and NRB ZLW0047 were unable to label the central nervous system i.e. ganglion, brain lobes, and nerves (dotted boxes on Figs 2A and 3A). In order to exclude possible autofluorescence disturbance, all tissues were captured under the same laser conditions without probe labeling; as shown in S2A-S2J Fig, tissues autofluorescence is almost null and does not interfere with the emission intensity of NRB MC009 and NRB ZLW0047 (S2K and S2L Fig).
The labeling properties of the two probes were compared by co-loading larval tissues with both dyes, and merging the corresponding images for colocalization analysis. The results, reported in Fig 4A-4D, indicate that both dyes were able to penetrate the cells of the tissues, and effectively label the intracellular structures. Cell internalization of the dyes was very rapid, allowing a clear visualization of the intracellular structures in less than 2 minutes (S1 and S2 Videos and S3 Fig). Pearson's coefficient results (Fig 4E and 4F) confirmed that NRB MC009 and NRB ZLW0047 recognized the same intracellular structures in most of the tissues investigated. However, a clear difference in subcellular expression between NRB MC009 and NRB ZLW0047 was observed in the fat bodies, in which lipid droplets (LDs) were selectively stained by NRB MC009 (Fig 4B). This behavior may reflect a different binding capacity of the two probes to the constituents of LDs, i.e. neutral lipids, mainly triacylglycerols and sterol esters, and phospholipids [23,24]. The intracellular distribution of NRB MC009 and NRB ZLW0047 was analyzed in more detail in Drosophila larval musculature in which we investigated the binding of these NRB fluorescent derivatives to ER, mitochondria, lysosomes and LDs. The larval body wall muscles provide a relatively simple system to study development of muscles, cytoskeleton dynamics, intracellular trafficking and neuromuscular junction dysfunction. In fact, besides the well-known actin and myosin filaments and their associated proteins, muscles also contain a cytoskeleton, intracellular organelles of the endo-lysosomal pathway, and well-defined endoplasmic reticulum and mitochondrial networks [25]. We focused on these particular structures because we recently showed that they contained a significant density of binding sites for NRB MC009 [6]. To study the intracellular localization of NRB MC009 we co-loaded it into larval muscle with organellespecific red fluorescent dyes. To confirm NRB ZLW0047 localization we profiled it in larval muscles labelled with organelle-selective green fluorescent proteins.
These data show that both compounds, but particularly NRB MC009 , can be used to visualize and distinguish most organs/tissues of the dissected living larvae, and allow good definition of their intracellular structures. In addition, the co-localization studies revealed that both NRB MC009 and NRB ZLW0047 labelled subcellular organelles, that they preferentially targeted the ER and mitochondria, and that they were totally absent from the nuclei, plasma membranes and lysosomes, which is in agreement with data reported in mammalian cell studies [6]. Moreover, the efficiency of the probes was tested with two different approaches: 1) in combination with commercially available dyes for live imaging, and 2) together with green fluorescent proteins that label specific cellular structures. In both the experimental backgrounds the NRB based probes allowed the identification of endoplasmic reticulum and mitochondria structures, making both probes useful new markers in Drosophila studies. Based on the capability of these dyes to recognize the same intracellular structures in both mammalian and fruit fly cells, their potential use in more complex animal models is anticipated. Subsequently, our ongoing work is focused on the development of these probes as tools to allow live imaging studies to be conducted in mouse and rat tissues.

NRB MC009 and NRB ZLW0047 cellular distribution in pathologic mutationrelated phenotypes
In consideration of the preferential distribution of NRB-derived fluorescent probes to ER and mitochondria we next verified if they could be developed into tools to highlight phenotypic modifications of the ER and mitochondrial networks in Drosophila muscles. A large number of human disease genes are conserved in Drosophila and its genome can be easily manipulated to recreate and study human pathologic phenotypes [26]; subsequently, Drosophila is widely used as a model to study muscle growth, degeneration and correlated diseases [27][28][29][30].
CMTd Drosophila phenotype was obtained by inducing a downregulation of Marf, the fruit fly orthologue of the human gene Mitofusin2, which encodes for a GTPase that, together with Opa1, fuses mitochondria; mutations of this gene are implicated in CMT disease [33]. The depletion of this protein in Drosophila is known to cause fragmented and clustered mitochondria in neuronal cell bodies and to disorganize the typical sarcomeric location of mitochondria in the larval muscles, clumping them mainly around the nuclei [34]. Fig 6B shows the fluorescent distribution of NRB MC009 and NRB ZLW0047 in muscles of larvae in which Marf had been ubiquitously downregulated (UAS-Marf RNAi /Tubulin-Gal4). Since these fluorescent images are comparable with those previously reported with the mitochondrial marker UAS-Mito-GFP [34] in the same model, when considering the mitochondrial labeling properties of NRB MC009 and NRB ZLW0047 (this study), it can be argued that fluorescent derivatives of NRB are able to also stain altered mitochondria, and be able to highlight pathologic mitochondrial phenotypes.
HSP Drosophila phenotype was obtained by inducing a downregulation of atlastin. Atlastins are membrane-bound dynamin-like GTPases implicated in ER network morphogenesis, and mutations in atlastin1 gene are involved in the onset of a common form of HSP (SPG3A). Drosophila holds a unique highly conserved atlastin orthologue, and its downregulation elicits a fragmented ER in neurons and an enrichment of ER punctae localized in the proximity of nuclei, and visualized using UAS-KDEL-GFP [18]. Labeling of larva fillets with NRB MC009 and NRB ZLW0047 revealed a different pattern between muscles of wild type (Fig 6A) and atlastindownregulated (UAS-Atl RNAi /Tubulin-Gal4) larvae (Fig 6C), in which a brighter perinuclear signal, compatible with the previously described HSP phenotype, is observed.
Taken together, these results indicate that NRB MC009 and NRB ZLW0047 could be useful tools for Drosophila live imaging to highlight phenotypes attributable to mutations in, and/or downregulation of genes implicated in mitochondria and/or endoplasmic reticulum network modifications. In addition, the short time it takes for these probes to permeate and label tissue and their general ease of use, means that both could be used as tools in compound screening studies to identify candidates that would help alleviate any network malfunction due to genetic modification.

NRB MC009 and NRB ZLW0047 in food intake tests
Next we explored the possibility of using NRB MC009 and NRB ZLW0047 as tools to evaluate food intake and to investigate potential gut morphological modifications in Drosophila in vivo. By adopting a three-choice test (behavioral choice test) we verified that NRB MC009 and NRB ZLW0047 , when added to the food, were accepted by the flies. As summarized in Fig 7A, there were no substantial differences in food preference between the standard diet and NRB MC009 -and NRB ZLW0047 -supplemented diets, indicating that the presence of the dyes did not influence the larval food choice. In addition, fluorescence imaging of larvae fed for 30 min with probe-enriched liquid food indicated that both NRB MC009 and NRB ZLW0047 could be clearly detected in the gut ( Fig 7B); a more in depth analysis revealed that gut fluorescence was regulated by the probes contained in the food, since no signal was observed in the gut wall, leading to the conclusion that the strength of the fluorescent signal could be taken as an index of the quantity of ingested food. Fig 7C reports the results of the food intake assay, expressed as a percentage of the gut stained area relative to the total body area-no significant difference was observed between larvae fed with NRB MC009 , NRB ZLW0047 , or brilliant blue dye, a commonly used dye for the evaluation of food intake in Drosophila [21].
The lack of gut labeling by NRB MC009 and NRB ZLW0047 , although a useful outcome for the food intake test, was somewhat unexpected, particularly considering the results obtained in the dissected larvae (see Figs 2 and 3) where the dyes were clearly localized to the intestinal tract. To explain this inconsistency, we hypothesized that the time of exposure (30 min) of the larvae to the probe-supplemented food in the food intake assay was potentially too short to allow an internalization of the dyes to the gut epithelial cells. Therefore, we analyzed the gut wall of larvae grown for 5-7 days in food enriched with NRB MC009 or NRB ZLW0047 . Fig 7D-7F show the clear difference between vehicle-fed ( Fig 7D) and NRB MC009 -and NRB ZLW0047 -fed larvae (Fig 7E and 7F, respectively). The bright fluorescent signal in the digestive tract (mainly midgut and hindgut) indicates that Drosophila larvae readily eat the probe-containing food.
The digestive tract of NRB MC009 -and NRB ZLW0047 -fed larvae were clearly labeled by the dyes, a result that was accentuated by the absence of fluorescence in the rest of the body. The Drosophila intestinal tract is formed by a monolayer of epithelial cells, intestinal stem cells and enteroendocrine cells, surrounded by visceral muscles, nerves and tracheae. Ingested food from the proventriculus is pushed into the midgut, the main region of digestion and absorption, and then to the hindgut where the final absorption process takes place [13,14]. A deeper investigation on dissected Drosophila gut revealed that NRB MC009 and NRB ZLW0047 did not only label the food that was present and visible in the intestinal tract (Fig 7I and 7L), but they also bound to the gut external muscular cells (Fig 7G and 7J) and enterocytes (Fig 7H and 7K).
The bright fluorescence of NRB MC009 and NRB ZLW0047 make these probes eminently suitable for use in food intake tests and chronic feeding assays; as monitoring tools for abnormal gut morphology; and identifying defects in gut functionality during development or screening tests.

NRB MC009 and NRB ZLW0047 toxicity
In an attempt to validate the use of the NRB-derived fluorescent probes for use in chronic assays, we next verified their lack of toxicity in Drosophila by exposing the flies to NRB MC009 or NRB ZLW0047 over their entire life-cycle. The results indicate that male and female flies readily ingested NRB MC009 -and NRB ZLW0047 -supplemented food, mated and laid eggs normally, from which embryos hatched and larvae developed, grew, underwent pupation and eclosed in a similar fashion to non-treated flies. In addition, no difference in eclosion rate and lethality of adult flies was observed between probe-exposed and control flies (Fig 8A). Finally, we could not detect any apparent macroscopic morphological alteration in adult flies treated with either NRB MC009 or NRB ZLW0047 (Fig 8B).
The absence of toxicity and the high level of palatability supportsNRB MC009 and NRB ZLW0047 as potential monitoring tools for long term feeding assays, as markers of intestinal epithelia, and their use in studying Drosophila digestive tract functionality e.g. monitoring the effect of compounds or diet on intestinal performance [35][36][37]. These attributes also make these probes potentially useful as mammalian gastrointestinal (GI) tract markers; for example, the GI tract is one of the most studied tissues in many pathological rodent models [38,39] and the availability of easy-to-use, high performance fluorescent probes to detect intracellular structure abnormalities could be of great benefit.
Overall, this study has investigated live cell imaging applications of NRB MC009 and NRB ZLW0047 in Drosophila melanogaster. The analysis of the fluorescent signals of these compounds reveal that both can label subcellular specific organelles (preferentially ER and mitochondria), in both wild type and pathological phenotypes. The absence of toxicity and the minimal effect on palatability also allows them to be used as potential monitoring tools in feeding assays, and as markers for intestinal epithelia that could be useful in Drosophila digestive tract studies. In summary, the characteristically bright signals of NRB MC009 and NRB ZLW0047 , in combination with their capacity to permeate tissues rapidly, makes them eminently suitable for confocal imaging applications. Our future studies will focus on investigating whether these compounds, with their enhanced attributes, may have potential to be used in invertebrate animal models.