LED based real-time survival bioassays for nematode research

Nematode bioassays are extensively conducted worldwide, either for screening anthelmintic drugs or for assessing the toxicity of drug candidates. Recently, the US Environmental Protection Agency mandated the use of invertebrate models including nematodes especially Caenorhabditis elegans, for toxicity testing as an alternative to rodent models. The significance of nematode bioassays in the biological sciences is escalating, but no standardized protocol is available to assess nematode mortality in a liquid medium. Manual counting under white light is the only approach currently practiced, which exhibit large variabilities and false positive results. Here, we describe an innovative counting strategy that employs light-emitting diode (LED) technology. We found that the nematodes stopped moving under white light (360–760 nm) when administered with sub-lethal dosage (LC50) of a toxic drug, whereas they responded rapidly to blue (450–490 nm) and ultraviolet (UV) (100–400 nm) LED lights. Furthermore, paralyzed nematodes responded in less than 5 seconds to a LED pulse. The response to the LED stimulus was distinctively noted in C. elegans dauers, which squirmed away from illuminated sites within seconds. LED produced an incoherent beam, and uniformly distributed light across the sampling area. In conclusion, this method is more accurate than the conventional counting techniques, and enables us to differentiate paralyzed and dead nematodes virtually in real-time. Furthermore, this technique would appear to be suitable for incorporating a motion-sensor based automated system.


Results and Discussion
Nematode bioassays are frequently performed in a liquid medium by manually counting the live or dead nematodes using visible light. Here, we described an improved method of nematode counting that employs LEDs. The technique relies on the instant responses of nematodes to the dissipation of heat induced by the LEDs. The microscopic imaging interface used is presented in Fig. 1.
LED bioassays on C. elegans produced consistent results. Initially, synchronized C. elegans juvenile stages were treated with abamectin. Using regular microscopic counting methods with white light, the sub-lethal dose of abamectin was estimated as 5 µg/mL ( Fig. 2A). Few nematodes were inactive, and to appeared dead under white light when treated with abamectin at sub-lethal dosage. Mean survival percentage was 38 ± 9% under normal light. But, when treated nematodes treated with blue LED and UV LED light, mean survival percentages were 80 ± 8% (an error rate of ~40%) and 90 ± 4% (an error rate of ~50%), respectively (Fig. 2B). Nematodes rapidly responded to blue and UV LED lights in 10 and 2 seconds, respectively (Fig. 2C). At a sub-lethal dosage, nematodes behaved differently under white light by remaining immotile for a considerable period, so they were recorded as dead. However, these nematodes responded rapidly to LED stimuli with motion detected within 2 seconds under UV LED exposure. Time-lapse images were acquired to confirm the inactivity of adult nematodes.  against C. elegans juveniles (Graphs are plotted as the means ± SEMs. **P < 0.01 vs. the non-treated controls), (B) estimation of the survival rates of juvenile exposed to different light sources (Graphs are plotted as the means ± SEMs. **P < 0.01 vs. the white light controls, and ***P < 0.001 vs. the white light controls), and (C) a C. elegans juvenile visualized under three different light sources (nematodes responded to exposure in 300 s for white light, 10 s for blue LED, and in 2 s for UV LED light). Scale bar: 10 µm. The adult C. elegans treated with abamectin (5 µg/mL) shown in Fig. 3A, indicates the nematode did not respond to white light (360-750 nm). Subsequently, we assessed and recorded the motions of this inactive C. elegans when exposed to a short ten-second pulse of blue LED light. It is also important to note that nematodes survive harsh treatments, and are rendered inactive which makes them appear dead under the experimental conditions. The dauer shown in Fig. 3B remained motionless after exposure to white light for several minutes. But, when exposed to blue LED light, rapid thrashing movements were observed after 10, 20, and 30 seconds of illumination. The responses of dauers to a green LED (510 nm) and UV LED light were also assessed. The dauers did not respond to green LED, but showed rapid flexing when exposed to UV LED light ( Supplementary Fig. S1). Their response time probably depends on the dosage of the drug administered, although generally inactive nematodes showed movements within 2 seconds after exposure to UV LED light. During time-lapse imaging, the LED light source was automatically turned off every 10 seconds while capturing the image. Nematodes that underwent ecdysis, a slow process of shedding its old cuticle, were also examined under LED lights. We observed that LED lights triggered rapid movements in these nematodes as well ( Supplementary Fig. S2).
To examine the reliability of the LED based nematode bioassays, we tested the effect of abamectin (10 µg/ mL) on B. xylophilus (another parasitic nematode), using white light or LED lights, to determine the differences between worm variances. The count for synchronized J2s under regular white light indicated a mean survival rate of 24 ± 4%, whereas under blue LED light the mean survival rate was found to be 51 ± 2%, which translate to an error rate of about 27% (Fig. 4A). The experimental groups treated with abamectin were exposed to UV LED indicated survival rates of these J2s were 61 ± 2%, that is, an error rate of ~37%. The results indicate that manual counting with white light is inaccurate and leads to false positive recordings 6,8 . We demonstrated the difficulty of distinguishing between paralyzed and dead nematodes with white light (Figs 4C i & ii, and 5B). Hence, the paralyzed nematodes might be counted as dead under the conventional counting process. On the other hand, our LED-based counting strategy was found to reduce the inadequacies of the conventional white light method. Figure 4B shows an adult B. xylophilus responding to LED lights. In Fig. 4C, we show a J2 stage B. xylophilus treated with a sub-lethal dosage of abamectin. Abamectin is known to kill the instar stages of nematodes and paralyze the adults, so J2 is the best stage for analysis. The paralyzed J2 stage was exposed to blue LED for 300 seconds to record any movement. Under white light, J2s (paralyzed and dead) did not move and appeared dead, while under LED illumination, motions were recorded and blurred images were obtained due to nematode movements, while clear images were obtained for dead nematodes (Fig. 4C). The response times of paralyzed nematodes under blue LED and UV LED lights were 10 and 2 seconds, respectively. All other stages of B. xylophilus responded similarly to LED stimuli exposure.
Assays with other nematicides such as sodium azide, ivermectin, and levamisole were conducted on both C. elegans and B. xylophilus larval stages. Sodium azide treatment resulted in instant paralysis of both the nematodes. Sodium azide is an anesthetic agent, and a respiratory inhibitor causing both the nematodes (C. elegans and B. xylophilus) not to respond to LED stimuli. After a 24 h, the nematodes in the sodium azide solution were removed and suspended in sterile distilled water or an M9 buffer solution for 30 minutes, and counted under UV LED stimulus ( Supplementary Fig. S3). The nematodes treated with 50 mM (3250 µg/mL) sodium azide did not recover and were considered dead, which is in line with the results of a previous study 18 . The nematodes that survived (still paralyzed or partially paralyzed) in 10 mM (650 µg/mL) sodium azide treated groups began to respond to LED stimuli, while the dead ones were still unresponsive (Figs 5 and S4).
Ivermectin (0, 10, 25, and 50 µg/mL) treated nematodes revealed similar phenotypes like abamectin treatments. Both the drugs belonged to macrocyclic lactone anthelmintics used for treating parasitic nematodes 19 . When tested at lower concentrations (25 µg/mL), nematodes resembled straight dead conformations ( Supplementary Fig. S5). Further, levamisole treatment paralyzed the nematodes but did not kill them at the tested concentrations. Most of the nematodes were coiled or showed curly phenotypes ( Supplementary Fig. S4). Like sodium azide, levamisole (1-10 mM) is often used to immobilize C. elegans for microscopic imaging 20 . We show that paralyzed nematodes (C. elegans and B. xylophilus) treated with levamisole (1 mM (204 µg/mL)) also recovered well after 24 h and responded to LED stimuli ( Fig. 6A and B). Importantly, a few dead ones were intact and motionless, thus differentiating the paralyzed from the dead (Fig. 6B).
Levamisole and sodium azide are drugs that paralyze the nematodes at a sub-lethal dosage and do not necessarily kill them altogether, suggesting that a few of the paralyzed nematodes can recover, when suspended in M9 buffer. It is established that sodium azide, levamisole, and tricaine are anesthetic drugs that stop motility in the worms, and can be subsequently rescued form the anesthetic drugs by transferring them to a fresh M9 media for one hour or less 21,22 . Our LED based platform can reliably distinguish dead nematodes from paralyzed nematodes, thus minimizing the source of potential errors in counting.
Chemical testing on parasitic nematodes is a leading area of research in the field of parasitology and nematology for developments of eco-friendly nematicides. C. elegans serves as a prominent invertebrate model for toxicity assessments and reduces requirements for rodents. In fact, there is a report indicating that C. elegans shares significant genomic similarities with rodents 23 . Recently, C. elegans has been used as an alternative to rodents as a biomedical tool for disease modeling, drug discovery, and toxicity assessments 11,24 . Recently, a research group developed a straightforward counting protocol to minimize the potential sources that may introduce errors 8 . This group standardized the nematode counting methods in a liquid media by minimizing the potential variability sources, which included the priming of pipette tips, shaking of media, sampling location, and operator-to-operator variances. Here we revealed the influence of different drug dosages on nematodes behavior towards LED lights which is also a potential source of variability faced during nematode counting methods.
The behavioral patterns of nematode stages are difficult to understand. Flat and straight nematodes are considered dead during conventional counting. Nematodes in early stages of molt also remain stiff and appear dead. As proof-of-concept, we show that flat and straight nematodes either cease motion during molting (Supplementary xylophilus larval stages were treated with 10 mM (650 µg/mL) sodium azide for a 24 h period before commencing with the recovery process. The figures for both sets indicate that dead nematodes did not respond to exposure to any of the light stimuli, while paralyzed nematodes in both case showed response to blue LED light, and more rapid response to UV LED light, but did not respond to white light exposure. Scale bar: 50 µm. Fig. S2) or remain inactive when exposed to sub-lethal dosages of toxic chemicals 6 , but exposure to LED light (Fig. 2B) trigger nematodes to revive or at least move. Nematode staining provides a reliable alternative for differentiating live and dead nematodes during counting, but sample staining in liquid culture is laborious, and increases the degree of experimental difficulties and errors 8 . Staining often brings the live or inactive nematodes under stress. On the other hand, LED based counting provides rapid, accurate, and reliable results. Also, we show the effects of few standard nematicides with different mode of actions namely abamectin and ivermectin (glutamate-gated chloride ion channel (GluCL) agonists 25 ), levamisole (nicotinic acetylcholine receptors (nAChRs) agonist 26 ), and sodium azide (respiratory inhibitor 18,22 ) on two different nematode models (C. elegans and B. xylophilus) to validate our findings.
Illumination with blue or UV LEDs induced automatic responses in nematodes. Healthy nematodes moved swiftly when exposed to LED light stimuli to avoid exposure. Previously it was demonstrated that laser ablation effectively kills nematode cells on skin 27 . When a laser microbeam is fired on a specimen, it is thought that the extreme heat and pressure generated denature proteins and DNA 27 . Similarly, prolonged LED exposure has also been shown to cause harmful effects on C. elegans and to cause retinal damage in rats [28][29][30] . It has also been reported that C. elegans possess LITE-1 photoreceptors which are sensitive to laser and UV light stimuli which are responsible for inducing negative photo-tactic responses 31 .
LED light differs from laser light concerning its coherence (Fig. 7). LED light illuminates specimen by distributing light uniformly across the sample area. The power distribution of LEDs cover a large area and affect several local tissues on the nematodes. The power distribution of LED light is advantageous for live/dead nematode counting, as samples containing ~50 nematodes or more per field can be counted accurately in a short timeframe. The power produced by LED light is sufficient to make the inactive nematodes move from the illumination point. Nematodes response to LED and other noxious stimuli is still a grey area and have to be explored further.

Future Perspective and Conclusion
LED based nematode counting can be performed automatically using algorithms and softwares to identify moving nematodes. The work-flow of this type of motion technology presented in Fig. 8 involves the use of image acquisition software equipped with a real-time motion sensor. This system can be also implemented without white light illuminations as well. Microscopic methods have been devised to detect the motion of small molecules 32 , and several automated methods have been devised to detect nematode thrashing 33 . In fact, high-throughput automated devices like 'Wormwatcher' utilize optical microscopic techniques 34,35 . The present study demonstrates the need to merge such techniques with LED illumination. Tested nematicides. The commercial nematicides, abamectin (Purity-98.7%), levamisole (Purity-98%), and ivermectin (Purity-98%) were purchased from Sigma (USA). Sodium azide (Purity-97%) was purchased from Duksan Pure Chemicals Co. Ltd (Republic of Korea). Abamectin and ivermectin were dissolved in dimethyl sulfoxide (DMSO), while levamisole and sodium azide were dissolved in sterile distilled water. 0.1% DMSO or sterile distilled water served as the negative control. For nematode bioassay a sub-lethal dosage (LD 50 ) of abamectin was used (5 µg/mL for C. elegans and 10 µg/mL for Bursaphelenchus xylophilus).
C. elegans culture conditions. C. elegans wild-type Bristol strain N2 was provided by Professor Eleftherios Mylonakis from Brown University. The strain was maintained on nematode growth medium (NGM) with E. coli OP50 as feed at 25 °C, and synchronization was also performed as previously described 36 . C. elegans in NGM plates were collected and bleached with 2% sodium hypochlorite and 0.5 N sodium hydroxide to obtain eggs, which were allowed to hatch to the L1 stage in M9 buffer and then transferred to fresh NGM plates containing E. coli OP50 lawns. Dauers (morphologically characterized by a slender appearance and pointed tail) were obtained by passaging the L1 stage in a minimal feed medium. Alternatively, C. elegans eggs were collected in M9 buffer, washed, and transferred to 96-well microtitre plate. Once the eggs settled to the bottom of the plates, the wells were washed repeatedly with M9 buffer to remove any other worms and replaced with fresh M9 buffer, and  incubated at 25 °C for 24 h for hatching process. Synchronized C. elegans larval stages were thus obtained, using abamectin (0, 1, 2, 5, and 10 µg/mL), ivermectin (0, 1, 2, 5, 10, and 25 µg/mL), sodium azide (0, 10 mM (650 µg/ mL), and 50 mM (3250 µg/mL)) 18 and levamisole (0, 10, 25, 50, 100, and 200 µg/mL), and incubated at 25 °C for a 24 h period to estimate the LD 50 and LD 90 values.
Bursaphelenchus xylophilus and Botrytis cinerea culturing. Pinewood nematodes (Bursaphelenchus xylophilus (supplied by Professor Hanhong Bae, Yeungnam University)) were sub-cultured by inoculation onto potato dextrose agar (PDA) plates containing fully grown mycelium of the neurotropic fungus Botrytis cinerea, as previously described 13 . The nematodes were allowed to reproduce and proliferate on plates for 7-8 days. The plates were then washed with sterile distilled water and the nematodes were further diluted to obtain approximately 100 nematodes per 100 µL. Initially, toxicity testing was performed using mixed stages of nematodes.
B. xylophilus J2 mortality assay. J2s mortality assays were assessed with slight modification to the method previously described 13 . Briefly, adult nematodes were collected and transferred to microtitre plates, and a day later, nematodes were selectively removed using a pipette and discarded. Eggs, which adhered to plate bases, were resuspended in sterile distilled water and allowed to hatch. The J2s thus obtained were used for mortality assays. Synchronized J2s (~50 nematodes) were treated with abamectin at a sub-lethal concentration (LD 50 ) and incubated at 22 °C for 24 h. Additionally, synchronized the larval stages were treated with ivermectin (0, 10, and 25 µg/ mL), sodium azide (0, 10, and 50 mM) and levamisole (0, 10, 25, 50, 100, and 200 µg/mL) to estimate the LD 50 and LD 90 values. The experiments were conducted three times and images were acquired using an iRiS ™ Digital Cell Imaging System. The survival rate of the J2s was determined by counting the live nematodes, and dead nematodes under white light (350-760 nm), blue LED (450-490 nm), and ultraviolet (UV) LED (100-400 nm) light stimuli. The error percentages were calculated by comparing the nematode survival rates ascertained under white light, blue LED light, and UV LED light, respectively.

Microscope and imaging interfaces.
A simple fluorescent microscope equipped with a camera, a monitor, and integrated image analysis software (The iRiS ™ Digital Cell Imaging System, Korea) was used for real-time monitoring. The microscope offered three fluorescence illumination functions such as blue, green and UV, provided by high power, long-lasting LEDs and hard-coated optical fibers. The LED filter cubes in the microscope provide maximum illuminations with uniform distributions. Nematodes exposed to a nematicides (abamectin, ivermectin, sodium azide, and levamisole) were placed in the microscopic stage and initially illuminated with white light to capture bright-field images. Subsequently, images were acquired using time-lapse images (one image per 10 s) captured over on or 2 min, and then following specimens were exposed to LED light (blue (450-490 nm), green (495-570 nm) and UV (100-400 nm)) to visualize nematode movement. The acquired images were either viewed at 40X (4X objective and 10X eyepiece), 100X (10X objective and 10X eyepiece), and/or 200X (20X objective and 10X eyepiece) magnification. Data availability. All data generated or analyzed during this study are included in this published article (and its Supplementary Information files).