CemOrange2 fusions facilitate multifluorophore subcellular imaging in C. elegans

Due to its ease of genetic manipulation and transparency, Caenorhabditis elegans (C. elegans) has become a preferred model system to study gene function by microscopy. The use of Aequorea victoria green fluorescent protein (GFP) fused to proteins or targeting sequences of interest, further expanded upon the utility of C. elegans by labeling subcellular structures, which enables following their disposition during development or in the presence of genetic mutations. Fluorescent proteins with excitation and emission spectra different from that of GFP accelerated the use of multifluorophore imaging in real time. We have expanded the repertoire of fluorescent proteins for use in C. elegans by developing a codon-optimized version of Orange2 (CemOrange2). Proteins or targeting motifs fused to CemOrange2 were distinguishable from the more common fluorophores used in the nematode; such as GFP, YFP, and mKate2. We generated a panel of CemOrange2 fusion constructs, and confirmed they were targeted to their correct subcellular addresses by colocalization with independent markers. To demonstrate the potential usefulness of this new panel of fluorescent protein markers, we showed that CemOrange2 fusion proteins could be used to: 1) monitor biological pathways, 2) multiplex with other fluorescent proteins to determine colocalization and 3) gain phenotypic knowledge of a human ABCA3 orthologue, ABT-4, trafficking variant in the C. elegans model organism.


Transgene integration
Ten transgenic adult animals were placed on 2 separate, 100 mm NGM source plates and allowed to lay eggs for 3 days. One-hundred L4s from the source plates were transferred to one, 100 mm NGM plate. Animals were exposed to 35 Gy radiation using an X-ray irradiator. Animals were allowed to recover for 2 hrs at room temperature or 16˚C overnight. Three irradiated L4 animals were placed on 30 separate, 100 mm NGM plates and grown to starvation (~7-10 days). Animals were collected in 1 ml of PBS and 10 μl was placed on 30 separate, 100 mm NGM plates. These plates were grown to starvation (~5-7 days), washed with 1 ml of PBS and 10 μl of the wash was placed onto 30 separate, 60 mm NGM plates seeded with OP50 and allowed to grow for 2 days. Six transgenic animals from each plate were transferred to 6 separate, 35 mm NGM plates and grown for 3 days. Animals that had 100% transmission of the transgenic marker were considered integrated and stocked.

E. coli OP50 preparation
A single colony of E. coli (OP50) was placed in 5 ml LB broth and incubated at 37˚C with shaking (250 rpm) overnight. This culture was then added to 500 ml of LB broth and incubated at 37˚C with shaking until reaching an OD 600 = 0.75-1.0. The bacteria were collected by centrifugation and washed once with 1:10 original volume of PBS and resuspended in 1:25 original volume in PBS. An equal volume of 50% glycerol was added for long-term storage at -80˚C. After thawing, the bacteria were concentrated by centrifugation, washed and re-suspended in 1:25 original volume in PBS.

Autophagic flux assay
Twenty late L4 animals were placed into 100 μl PBS supplemented with OP50 and the following conditions: no treatment (diluent control), 25 μM fluphenazine (Millipore-Sigma, St. Louis, MO), or 10 mM 3-methyladenine (3-MA; Millipore-Sigma). The animals were incubated for 16 hrs at RT and recovered on an NGM plate for 30 min. Animals were transferred to a 35 mm coverglass bottom petri dish (MatTek, Ashland, MA) for confocal imaging.

Intracellular organelle labeling
LysoTracker Deep Red (LTDR; Life Technologies Corp, Carlsbad, CA) were diluted to 1 μM in PBS. About 20-30 adult stage animals were placed in the solution and incubated at room temperature for 30 min. To chase excess dye prior to imaging, animals were placed on a fresh NGM plate seeded with OP50 for 15 min.
MitoTracker Deep Red (Life Technologies Corp) was resuspended in DMSO to a stock concentration of 1.25 mM. Prior to use, the stock solution was diluted to 500 nM in PBS. About 20-30 adult stage animals were placed in 10 μl staining solution and incubated at room temperature for 30 min. To chase excess dye prior to imaging, animals were placed on a fresh NGM plate seeded with OP50 for 15 min.
Bovine serum albumin (BSA)-Alexa Fluor conjugate (Life Technologies Corp) was resuspended in PBS to a concentration of 5 mg/ml. About 20-30 adult stage animals were placed in 10 μl BSA and 5 μl OP50 and incubated for 16hrs. To chase excess dye prior to imaging, animals were placed on a fresh NGM plate seeded with OP50 for 60 min.
BODIPY FL C 5 -Ceramide (C5 FL) complexed to BSA (Life Technologies Corp) was resuspended in water to a stock concentration of 500 μM. Prior to use, the stock solution was diluted to 50 μM with PBS supplemented with OP50. About 20-30 adult stage animals were placed in 10 μl staining solution and incubated at room temperature for 1hr. To chase excess dye prior to imaging, animals were placed on a fresh NGM plate seeded with OP50 for 30 min.
CellMask Deep Red (Life Technologies Corp) plasma membrane stain was diluted 1:100 in PBS. About 20-30 adult stage animals were placed in 10 μl staining solution and incubated at room temperature for 1 hr. To chase excess dye prior to imaging, animals were placed on a fresh NGM plate seeded with OP50 for 30-60 min.

Microscopic imaging and analysis
To prepare animals for imaging, 6-10 μl of 100 mM NaN 3 (Millipore-Sigma) in PBS was placed on the middle of a 35 mm coverglass bottom dish (MatTek). Approximately, 10-15 adult stage animals were transferred to the NaN 3 solution and covered with a 12 mm circular coverslip and then a 25 mm square coverslip.
Confocal images were taken with a Lecia SP8X tandem scanning confocal microscope with a white light laser using either a 40x 1.3 NA or 63x 1.4 NA oil PlanApo objective over �20 zplanes and a pinhole size of 1.00 (Leica Microsystems, Buffalo Grove, IL). Images were displayed as single XY planes, except nuclear markers which were displayed as maximum intensity projections. Images were rendered and analyzed using LASX (Leica Microsystems) and Volocity (v6.3; Quorum Technologies, CAN) software.
Spectral data were obtained using a supercontinuum white light laser and prism-based spectral detector. Excitation spectra were determined by measuring the intensity of the sample while exciting every 3-5 nm. Sample excitation was started~100 nm lower of reported excitation maximum (Ex max ), up to a detection range (~40-50 nm higher than the emission maxima (Em max )). The Em spectra were determined by exciting the sample 50-100 nm lower than the Ex max and measuring in a range of 20 nm in 5 nm steps.
Colocalization and quantification of data were obtained using the Volocity cellular imaging and analysis software (v6.3; Quorum Technologies). Either Pearson correlation or Manders correlation coefficient (MCC) was used to assess the positional relationship between two objects [23]. Colocalization and quantification of vesicular or punctate structures were obtained using the Volocity Colocalization or Volocity Quantification modules, respectively, with thresholding of fluorescence intensity and size to remove background noise. A minimum of 5 individual animals was used in the analysis.

Statistical analyses
Two-tailed (heteroscedastic) t-tests were performed on quantified data using Microsoft Excel. Experiments were repeated at least two times to ensure reproducibility.

CemOrange2 excitation and emission spectra in vivo
Detectable differences in the Ex/Em spectra of individual FPs permit the visualization of multiple fluorophores within the same physical space (Table 1). Three FPs with different spectra, GFP (Ex488 nm/Em505 nm), YFP(Ex515 nm/Em526 nm) and mKate2(Em585 nm/Ex605 nm) [24][25][26], have been utilized in C. elegans within our laboratory [27,28]. To broaden our repertoire for multiplexing, we synthesized minigene gblocks for mOrange2, mNeptune2.5 and mCardinal2 [17,29]. To enhance expression of these FPs in C. elegans, we codon optimized the cDNA sequence and introduced 3 synthetic introns using the C. elegans Codon Adapter software, with a Codon Adaption Index of 1.0 (https://worm.mpi-cbg.de/codons/cgi-bin/ optimize.py) [19,30]. The C. elegans (Ce) optimized CemOrange2, CemNeptune2.5 and Cem-Cardinal2 gblocks were ligated into a backbone expression vector containing the intestinal promoter nhx-2 in pPD49.26 [31,32]. Correct plasmid construction was confirmed by DNA sequencing. These expression plasmids, along with a co-injection marker (P myo-2 GFP), were introduced to the germline of N2 (wild-type) animals via microinjection. Individual transgenic animals were selected using the co-injection marker, discernable via a fluorescent stereoscope and placed onto single plates. Lines that passed the marker onto the second generation were selected for imaging and spectral analysis. The Ex max /Em max spectra in the intestinal cytosol were determined using a confocal laser-scanning microscope equipped with a supercontinuum white light laser and prism-based spectral detector. The Ex max /Em max were: 1) GFP, Ex488 nm/Em505 nm; 2) YFP, Ex515 nm/Em526 nm; 3) CemOrange2, Ex555 nm/Em562 nm; 4) mKate2, Ex585 nm/Em605 nm; 5) CemNeptune2.5, Ex595 nm/Em620 nm; and 6) CemCar-dinal2, Ex595 nm/Em642 nm ( Fig 1A). Spectral characteristics show that the Ex max /Em max spectra of CemOrange2 were between those of GFP, YFP and mKate2. CemNeptune2 and CemCardinal2 had Ex max spectra shifted~10 nm higher than that of mKate2 and Em max spectra shifted~15 nm and~37 nm higher than that of mKate2, respectively ( Fig 1A). Moreover, CemOrange2 was distinct from autofluoresence found in wild-type N2 animals (S2 Fig) These data suggest that CemOrange2 could be utilized for multifluorophore imaging with other commonly used FPs.
Excitation and emission spectra of fluorescent proteins (FPs) expressed within the cytoplasm of C. elegans intestinal cells: GFP (green), YFP (yellow), CemOrange2 (orange), mKate2 (red), CemNeptune2 (dark red), CemCardinal2 (magenta) FPs were expressed from transgenes driven by the intestinal specific promoter, P nhx-2 (A). Dotted and solid lines represent excitation and emission curves, respectively. Spectral scans in both the excitation (470 nm-670 nm) and emission (480 nm-770 nm) wavelengths were obtained using a supercontinuum white light laser and spectral detectors on a confocal microscope. Intensity was normalized to the maximum fluorescence intensity value (I/I max ). The inset table indicates the maximal excitation (Ex max ) and emission wavelengths (Em max ) determined for each fluorophore. Different FPs exhibit a range of characteristics (Table 1), such as brightness, aggregation and cellular toxicity, depending on the cell or organism in which they are expressed [33]. Interestingly, the three related far-red FPs, mKate2, CemNeptune2.5 and CemCardinal2, had lower Em max in vivo ( Fig 1A) than previously reported. Transgenic animals expressing different cytosolic FPs within the intestine were visualized qualitatively using confocal laser-scanning microscopy at the previously determined Ex max /Em max . GFP, YFP, CemOrange2 and mKate2 were bright, diffuse and uniform (Fig 1B, 1D, 1F and 1H). CemNeptune2.5 expression was diffuse but non-uniform within the cytosol (Fig 1J). CemCardinal2 was neither bright nor uniform in its distribution and exhibited punctate structures suggestive of FP aggregation or concentration within a subcellular structure (Fig 1L). Corresponding DIC images confirmed expression was restricted to intestinal cells. (Fig 1C, 1E, 1G, 1I, 1K and 1M). Taken together, these data suggest that CemOrange2 would be the FP of choice for subcellular imaging.

CemOrange2 fused to targeting motifs or proteins localized to subcellular structures and organelles
CemOrange2 exhibited Ex max /Em max spectra that were distinguishable from those of GFP, YFP and mKate2. Also, CemOrange2 expressed in intestinal cells yielded a diffuse cytoplasmic pattern (i.e., no evidence of aggregation). Taken together, these characteristics suggested that this FP would facilitate multiplexing with unknown FP-tagged proteins by targeting known Subcellular imaging in C. elegans subcellular structures. Thus, the P nhx-2 CemOrange2 construct (pJR2945) was used to generate a transgene expressing the FP fused to different targeting motifs or a C. elegans protein known to localize to specific organelles/structures (Table 2). These structures included the nuclear lamina [34]; the rough ER [15]; the Golgi [35]; early and late endosomes [36]; the lysosome [12,14]; autophagosomes [16]; ubiquitinylated cargos [37,38]; lysosomal related organelles (LRO) [13,39]; the apical, lateral and basal plasma membrane [40]; the nucleus [41]; the mitochondrial matrix [42]; and peroxisomes [43]. All genes were amplified by PCR from wild-type genomic DNA using gene-specific primers and cloned in frame with CemOrange2 at either the N-or C-termini (Table 2). Subcellular targeting subsequences were introduced at either the N-or C-termini by site-directed mutagenesis. The identities of all constructs were confirmed by DNA sequencing.
The selective autophagy receptor, SQST-1::CemOrange2, is a diffuse cytosolic protein with few puncta detectable at lower resolution under normal conditions (Fig 2Y and 2Z, white arrowheads). Similarly, the autophagy marker, CemOrange2::LGG-1, also showed mainly a cytosolic expression pattern with a few puncta in well-fed animals (Fig 2AA and 2BB, white arrowheads). The characteristics of these makers change depending on the state of autophagic flux (vide infra).

Trafficking of CemOrange2 fusion proteins
The transgenic lines described above demonstrated subcellular distribution profiles consistent with those described in previously published reports. However, we sought an independent means to confirm that CemOrange2 markers trafficked to the correct positions by performing colocalization studies with a different set of previously reported genes, target peptides or fluorescent stains known to localize to specific organelles. Colocalization was assessed using the Manders Correlation Coefficient (MCC) with a value � 0.5 showing no colocalization, 0.5-0.8 demonstrating partial colocalization and � 0.8 corresponding to high colocalization [23].
MitoTracker Deep Red is a fluorescent probe that accumulates within mitochondria. Transgenic animals expressing CemOrange2 with the N-terminal mitochondrial localization tag ( mt CemOrange2, Fig 3P) were stained with MitoTracker Deep Red and examined by confocal microscopy. The mt CemOrange2 protein colocalized with MitoTracker Deep Red (Fig 3R; white arrowheads) with an average MCC = 0.82 (Fig 3BB).
The apical plasma membrane is labeled using CellMask, an amphipathic molecule that exhibits both lipophilic and hydrophilic moieties. AQP-1::CemOrange2 ( Fig 3Y) and CellMask ( Fig 3Z) colocalized (Fig 3AA, magenta) with an average MCC = 0.91 (Fig 3BB). Taken together, these data suggested that CemOrange2 did not affect the trafficking of fusion proteins or peptides described in this report.

The biological application of CemOrange2 fusion proteins
To assess the bioapplicability of the subcellular organelle markers, transgenic strains expressing a CemOrange2 fusion protein were exposed to different stressors. Lysosomal membrane permeabilization (LMP) is the hallmark of necrotic cell death in the C. elegans enterocyte [50,51]. LMP was induced in wild-type animals expressing LMP-1::CemOrange2 by the ROS SKL, a C-terminal peroxisomal targeting peptide (U, V); mt, an N-terminal mitochondrial target peptide (W, X); SQST-1 and LGG-1, autophagosome-related proteins (Y-BB); and AQP-1, an apical and basolateral plasma membrane protein (CC, DD). All transgenes were expressed using the intestinal specific promoter, P nhx-2 . Images were collected over >20 z-planes using a 40x PlanApo oil immersion objective (N.A. 1 Subcellular imaging in C. elegans inducer, tert-Butyl hydroperoxide (t-BOOH) [50] or the lysosomotropic agent Leu-Leu-OMe (LLOMe) [52]. After exposure to 5% t-BOOH (Fig 4B and 4D) or 500 mM LLOMe (Fig 4F and  4H), the LMP-1::CemOrange2 staining lost the punctate staining pattern and became faintly diffuse cytosolic fluorescence (Fig 4B and 4F) as compared to the untreated controls (Fig 4A  and 4E) suggesting a loss of lysosomal integrity. These studies indicated that LMP-1::CemOr-ange2 can be used to monitor lysosomal injury.

Three-color imaging
To determine the utility of the CemOrange2 FP in multifluorophore imaging, we generated a transgenic line harboring three different nhx-2-driven transgenes expressing proteins fused to three different FPs; mKate2::LGG-1, LMP-1::CemOrange2, and sGFP::ATZ [27]. ATZ is a mutant form of the human serpin, μ1-antitrypsin/SERPINA1. The Z mutation causes ATZ to misfold in the ER and accumulate as polymers and aggregates [55,56]. Using a modified Colocalization of CemOrange2 probes with differentially labeled subcellular markers. Transgenic animals expressing the LMP-1, CUP-5, RAB-5, RAB-7, SKL, mt, TRAM-1, AMAN-2 or AQP-1 markers fused to CemOrange2 were labeled with previously described subcellular organelle markers (colocalization marker) and examined by confocal microscopy. Images were acquired over >20 z-planes using a 40x PlanApo oil immersion objective (N.A. 1.3) at wavelengths Ex555 nm/Em565-590 nm and a single XY-plane of representative images of animals are shown. To determine the subcellular localization, LMP-1::CemOrange2 (A; red) and CemOrange2::CUP-5 (D; red) expressing transgenic C. elegans were incubated with LTDR (B and E, blue) to label acidic compartments such as the lysosome. LMP-1 and CUP-5 colocalized with LTDR staining (C and F; magenta, arrowheads). Note, GLO-1 did not colocalize with LTDR (S1 Fig CemOrange2 spectra in both the excitation and emission spectrum (Ex545 nm/ Em560-590 nm) and sequential scanning, the CemOrange2 FP was distinguished from both GFP and mKate2 (Fig 6A-6H). This combination of probes also confirmed several features related to the degradation of ATZ. Macroautophagy is a major degradation pathway for ER-retained sGFP::ATZ. At higher resolution, sGFP::ATZ was detected in LGG + puncta (i.e., autophagosomes) as evident by the merge of GFP and mKate2 signals (Fig 6G and 6H, yellow arrowheads). The colocalization of autophagosomes with lysosomes is evident in the merge of mKate2 and CemOrange2 signals (Fig 6G and 6H; white arrowheads). The absence of GFP from these latter structures likely reflects the instability of GFP in acidic vesicles.
In another example, a transgenic C. elegans strain expressing both P vha-6 LMP-1::GFP and P nhx-2 GLO-1::CemOrange2 was stained with LTDR and imaged as described above (Fig 6I-6P). Representative images show there is minimal overlap between the fluorophores. As expected, there was extensive overlap between LTDR + vesicles merged with GFP + lysosomes (Fig 6O and 6P; arrowheads). Taken together, these data demonstrated that CemOrange2 was useful for live-animal multifluorophore imaging.

Pathological variant validation by using co-localization markers
Model organisms, such as C. elegans, D. rerio and D. melanogaster, have proven to be essential systems in determining whether human variants of unknown significance (VUS) are pathologic [10,11]. In C. elegans, genotype-pathologic phenotype correlations can be determined by comparing wild-type versus mutant human transgenes or by generating homologous mutations in orthologous genes. One means to assess these phenotypes is by using microscopic analysis to determine whether the variant of interest perturbs cellular dynamics or by altering A live-animal sensor for lysosomal membrane integrity. P nhx-2 lmp-1::CemOrange2 transgenic C. elegans were exposed to either PBS (A, C, E, G) or 5% tert-butyl hydroperoxide (t-BOOH; B and D) in PBS for 30 min or 500 mM LLOMe (F and H) in PBS for 30 min. Animals were visualized by confocal microscopy using a 40x PlanApo oil immersion objective (N.A.1.3) at Ex555 nm/Em565-590 nm over >20 z-planes. Representative confocal images of animals shown as a single XY plane. Transgenic C. elegans exposed to PBS alone (A, C, E, G) have punctate LMP-1::CemOrange2 expression (A and E, red). The same animals exposed to 5% t-BOOH (B) or 500 mM LLoMe (F) lose punctate LMP-1::CemOrange2 labeled structures. DIC images (C, D, G, H) displayed for reference. Scale bars = 25 μm.
https://doi.org/10.1371/journal.pone.0214257.g004 the proteins subcellular distribution [2,3]. The generation of the CemOrange2 organelle marker toolbox described here will be useful in identifying the subcellular phenotypic changes associated with a pathological VUS. As an example, human ABCA3 is a phospholipid transporter required for assembly of pulmonary surfactant in lamellar bodies and lamellar body biogenesis in type 2 pulmonary alveolar epithelial cells [57]. Different types of ABCA3 mutations disrupt surfactant synthesis and cause neonatal respiratory failure or childhood interstitial lung disease (chILD) in older infants and children [57][58][59][60][61]. One mutation, L101P, results in protein misfolding and accumulation within the ER. To determine if this pathologic variant can be detected in C. elegans, a mutated transgene (the conserved leucine is at position 162 in C. elegans) containing the worm orthologue of ABCA3, abt-4, fused to the N-terminus of mKate2 was introduced into wild-type animals (Fig 7A, asterix). Wild-type ABT-4::mKate2 was trafficked to the intestinal cell membrane (Fig 7B, 7D and 7F) however, similar to that LGG-1 CemOrange2 fusion proteins monitor autophagic flux. P nhx-2 sqst-1::CemOrange2 (A, C, E; red) and P nhx-2 CemOrange2::lgg-1 (B, D, F; red) transgenic animals were treated with a diluent control (0.1% DMSO; A and B), 3-methyladenine (10 mM 3-MA, C and D), or 25 μM fluphenazine (E and F) for 16 hrs in liquid culture and analyzed using confocal microscopy over � 20 z-planes Ex555 nm/Em565-590 nm. Representative confocal images of animals shown as a single XY plane. Scale bar = 25 μm. Multiple confocal images of transgenic animals (n � 5 animals) expressing SQST-1::CemOrange2 (G) or CemOrange2::LGG-1 (H) after treatment with the above compounds, were quantified for number of puncta per μm 2 imaged using the Quantification module in the Volocity image analysis software (v6.3). Statistical analysis of the drugtreated animals relative to diluent control was performed using an unpaired, 2-tailed t-test ( � p<0.05).

Discussion
The purpose of this study was to generate a panel of FP markers for subcellular structures that could be used for multiplex imaging in C. elegans, one of the premier model organisms for studying cell biological and developmental processes in real-time [2,3,11]. The challenge was to select a FP that did not aggregate spontaneously in vivo, and displayed an Ex/Em spectra that was distinct from the popular FPs already adapted for use in C. elegans (e.g., GFP, YFP, CFP and mCherry) [1,5,[7][8][9]. The ideal fluorescent protein should also have a fast maturation time, high brightness and high photostability. Of the fluorescent proteins that we chose to study, mCardinal has the highest photostability and a fast maturation time (Table 1). However, it possesses approximately half the brightness of EGFP (Table 1). While the maturation time of mNeptune2.5 is similar to EGFP, it has approximately 70% the brightness and the photostability is unknown. Additionally, the reported Ex max /Em max of mNeptune2.5 is closest to both mCherry and mKate2, which would make it more difficult to spectrally separate from these 2 commonly used fluorophores. While mOrange2 has a reported slow maturation time of 270 minutes, it has a brightness similar to EGFP and has a high photostablity of 228 seconds (Table 1). After codon adaptation, we found that CemOrange2, had an Ex max /Em max spectra in vivo that placed it between those of YFP and mKate2, and that the signal from this FP was easily distinguished from C. elegans autofluorescence (S2 Fig). While CemOrange2 could be spectrally distinguished from the far red mKate2 in vivo, this FP spectra would overlap with other commonly used red FPs, such as DsRed (Ex max /Em max 558/553) and dTomato (Ex max /Em max 554/581). In contrast, both the in vivo CemNeptune2.5 and CemCardinal emission and excitation spectra overlapped significantly with mKate2, which would confound colocalization studies. Moreover, the CemOrange2 FP demonstrated a diffuse and homogenous cytoplasmic distribution when expressed in C. elegans intestinal cells. In contrast, CemCardinal2, a monomeric FP, showed both a diffuse and granular distribution pattern. We did not determine whether CemCardinal2 formed aggregates in vivo, or associated with a vesicular structure due to a unique aspect of its structure. Regardless, we were reluctant to utilize CemCardinal2 as this characteristic could confound its use in colocalization studies. Another advantage of CemOrange2 is that it also avoids the phototoxicity associated with the distinct Ex/Em spectra of BFPs [62]. While the confocal setup described here uses a supercontinuum white light laser source, more traditional confocal and widefield systems can be used to to visualize CemOr-ange2. Indeed, CemOrange2 was detectable using a 561nm wavelength excitation laser (S3D Fig). Additionally, there was no observable cross talk with spectral channels visualizing the GFP or mKate2 fluorophore excitation and emission settings (Ex488 nm/Em500-540 nm and Ex594 nm/Em605-640 nm, respectively; S3A-S3F Fig). Moreover, tuning the white light laser (B, D, F, H). Scale bars = 5 μm. mKate2:: LGG-1 and LMP-1::CemOrange2 had multiple colocalization events (G and H; magenta merge, white arrowheads). Alpha-1 antitrypsin Z mutation (ATZ), a polymerizing protein which accumulates in the ER and is degraded, in part, by autophagy, show colocalization with mKate2::LGG-1 (G and H; orange merge, yellow arrowheads). P nhx-2 glo-1::CemOrange2;P vha-6 lmp-1::GFP transgenic C. elegans expressing LMP-1::GFP (I and J, green; Ex488 nm/Em500-540 nm) and GLO-1::CemOrange2 (M and N, red; Ex555 nm/Em560-590 nm) were stained with the acidic organelle fluorescent dye, LTDR (K and L, blue; Ex647 nm/Em660-700 nm). Animals were imaged by confocal microscopy over >20 z-planes and maximum intensity projections are shown (I, K, M, O). Scale bars = 25 μm. Magnified single XY regions (dashed box) are included to highlight colocalization events (J, L, N, P). Scale bars = 5 μm. Colocalization events are seen between LMP-1::GFP and LTDR (O and P, cyan merge, arrowheads) but not with GLO-1::CemOrange2.
We generated a series of plasmids containing a CemOrange2 minigene fused to a targeting sequence or another gene expressing a protein known to target to a unique subcellular organelle or location. Using confocal microscopy, the subcellular distribution patterns of these markers suggested that they were targeted to the correct location. However, FPs can affect the folding and conformation of their fusion partner so we sought an independent means to assure that the CemOrange2 markers were targeting the correct cellular address [63]. In all nine cases tested, the CemOrange2 fusions showed a significant colocalization with a non-overlapping fluorophore targeting the same structure (MCC>0.5). Taken together, these studies showed that CemOrange2 FPs were directed to the correct subcellular addresses and should prove useful for multicolor live-cell imaging. Indeed, this functionality was demonstrated by a transgenic strain expressing human alpha-1 antitrypsin with the Z mutation fused to the Cterminus of GFP. This aggregation-prone protein, is retained in the ER and is partially degraded by macroautophagy [27,28,32,56]. Examination of these animals showed colocalization of ATZ in LGG-1 + (a C. elegans orthologue of LC3) structures (i.e., autophagosomes). The identification in of autophagosomes with a specific cargo is difficult in real-time and underscores the sensitivity of multifluorophore imaging technology.
Knowledge about human variants of unknown significance can be obtained by determining whether the phenotype of the variant differs from the wild-type gene when expressed in model organisms, such a C. elegans, D. rerio, D. melanogaster or S. cerevisiae [2,3,10,11]. Some abnormal phenotypes are straightforward in their presentation at the whole organism level, if they result in, for example, abnormal development of decreased longevity. However, many abnormal phenotypes are subtler in their presentation and are manifest only after the application of a cellular stress or by examining subcellular functions; for example, protein misfolding disorders. Many of these proteostasis disorders only manifest after the aggregation-prone or misfolded proteins accumulate over time. Classical examples are neurodegenerative disorders associated with the aberrant accumulation of Huntingtin, Aß, alpha-synuclein and/or tau [64][65][66][67]. In many cases, alterations in the subcellular distribution of these proteins are detected long before a pathologic phenotype emerges. For this reason, multifluorophore imaging in C. elegans can be a useful adjunct to the analysis of VUS, especially those associated with the subtle effects associated with protein misfolding. We showed that a mutation in the C. elegans ABC transporter, abt-4, resulted in ER retention, which is exactly what occurs in the human orthologous gene, ABCA3, with an identical mutation at a single conserved amino acid [57][58][59][60][61]. Similar results were observed in C. elegans expressing the human pathologic variant Z of alpha-1 antitrypsin [27,28,32]. Taken together, these studies show that the addition of CemOrange2 to the C. elegans FP toolbox expands their ability to assess human VUS behavior by multifluorophore, real-time subcellular imaging.