C. elegans were maintained on agar plates with a normal growth medium (NGM) at 20°C with the OP50 strain of E. coli used as food. Strains used in this study include N2 (RRID: WB-STRAIN: WBStrain00000001), OK411 (RRID: WB-STRAIN: WBStrain00031419), and UA57 (RRID: WB-STRAIN: WBStrain00035183).

Synchronization of C. elegans populations was performed through the following protocol: Gravid adult C. elegans were collected from agar plates using M9 buffer (22 mM KH₂PO₄, 42 mM Na₂HPO₄, 85 mM NaCl, and 1 mM MgSO₄ prepared in sterile water) and transferred in 1.5-mL Eppendorf tubes. The tubes were centrifuged at 11,000 rpm for 1 min, the supernatant was removed, and 1 mL of the bleach solution (250 mM KOH and 0.5% hypochlorite prepared in sterile water) was added to each tube. The tubes containing bleach and C. elegans were rested on their side with periodic vortexing until most of the bodies disintegrated. The tubes were centrifuged at 11,000 rpm for 1 min, the bleach supernatant was discarded, and the eggs were washed two times to remove any residual bleach solution by resuspending in 1 mL of M9 and centrifuging at 11,000 prm for 1 min. Following the last wash, the supernatant was removed, and the eggs were resuspended in 100 µL of M9 before transferring to 50-mL conical tubes in a final volume of 2 mL M9 solution. Conical tubes were kept on a shaking incubator at 120 rpm at 20°C overnight to let the eggs hatch to the larval 1 stage (L1). L1 C. elegans were placed on 6-cm agar growth plates (normal growth medium with live E. coli OP50) and placed upside down in a 20°C incubator. For L4 analysis, C. elegans samples were collected 48 h after L1 plating. Day 1 adult C. elegans were analyzed 72 h after L1 plating. Day 2 adult C. elegans were analyzed 96 h after L1 plating. Day 3 adult C. elegans were analyzed 120 h after L1 plating.

For the confirmation of the genotype of the MBIA strain, polymerase chain reaction (PCR) was performed by collecting ∼5 gravid adult C. elegans for analysis using a Sigma Extract-N-AMP™ Tissue PCR Kit (XNAT2) and Apex Taq Red 2× Master Mix. Briefly, C. elegans samples were placed in a lysis buffer and incubated in a thermocycler at 55°C for 10 min and 95°C for 3 min. Following incubation, a neutralization solution was added to each tube to halt the reaction. A primer set was used to confirm the gene deletion of cat-1 with a 701-base pair product in wild-type and a 272-base pair product in cat-1-null C. elegans. The forward-3 primer used was CCG​CGC​AAA​TGA​ATG​ACC​TA. The reverse-3 primer used was GAA​CCG​GAA​GGT​TCA​AGC​AT. The C. elegans lysate was mixed with Apex Taq Red 2× Master Mix, primers, and water and underwent thermocycling at 94°C for 2 min, 94°C (30 s 35×), 55.6°C (30 s 35×), 72°C (1 min 35×), 72°C (5 min), and 4°C (hold). The PCR product was run with a loading buffer on pre-cast 2% gels (Thermo Fisher) using the E-gel Power Snap Electrophoresis System (Thermo Fisher).

C. elegans were imaged on microscope slides with agar pads (3% pure agarose in M9). The agar mixture was brought to a boil before allowing to cool to 58°C. The cooled agar was pipetted onto a glass microscope slide, and a second slide was placed on top to flatten the agar into an even surface. After the agar cooled completely and solidified, the top microscope slide was removed, and 15–30 C. elegans were placed on the agar pad in 5 mM levamisole before cover-slipping. The slides were imaged on an EVOS FL or APEX100 microscope using brightfield or GFP at 4×, 10×, and 20× magnification.

For behavioral analysis, 4–6 C. elegans of a specific genotype and developmental stage were placed in 60 µL of the M9 solution added to microscope slides with hydrophobic circles. A stereomicroscope with a camera was used to analyze the behavior of C. elegans. The microscope was focused such that a well-defined image could be captured, and videos of C. elegans swimming in the M9 solution were collected for 30,000 ms at 18 fps and saved as JPEG images. For each strain, 4–6 replicates were performed. Behavioral analysis was performed using ImageJ. To perform curling analysis, the image sequence was played through, and the number of times an individual C. elegans was in a curled position was counted. We defined curling as the position in which the tail of a C. elegans was directly crossing its body (Sohrabi et al., 2021). To perform thrashing analysis, we ensured that animation options were set to 18 fps so that the image sequence played a 30-s video. A stopwatch starting at 0 s was prepared prior to starting the video. The percentage of time for which a C. elegans was thrashing over a 30-s time period was measured. We defined thrashing as the motion of a C. elegans aggressively moving side to side (Sohrabi et al., 2021).

Individual C. elegans were placed in one well of a clear U-shaped 96-well plate filled with 100 µL of a 20% bleach solution (in M9) per well. The plate was observed in brightfield at 4× and 10× using an EVOS FL microscope. Once the C. elegans body dissolved and eggs were released, the number of eggs was counted and recorded as a measure of eggs in utero.

C. elegans at the L4 and day-1 adulthood developmental stages were analyzed using the COPAS FP-250 large particle flow cytometer (Union Biometrica, Massachusetts). C. elegans were collected from agar growth plates using 5 mM levamisole to anesthetize the animals. The COPAS cytometer recorded the extinction and time of flight of each object, which was exported in .csv files and analyzed using RStudio. Raw data files from each strain were stored in separate variables and then combined into one .csv file per experiment containing data for all strains using the “rbind” function. A few methods of data cleaning were performed to exclude objects that were not C. elegans (e.g., contaminants, pieces of agar, and dust particles). First, the data were analyzed for the number of objects in a sample to determine if excessive contamination was likely due to excessive objects detected and linear data distribution not representative of typical data spread. Any contaminated datasets were removed from analysis. Of the 48 samples compiled for the COPAS analysis, three contaminated samples were removed. Next, the objects were filtered based on extinction measurements to remove objects with extinction values smaller than 250 and greater than 4,000. As reported by other groups, the analysis of particles outside of this range by microscopy revealed objects with extinction values < 250 to be C. elegans body fragments, unhatched embryos, pieces of agar, and other contaminants such as dust (Smith et al., 2009). Lastly, any remaining contaminants were removed by determining the average and standard deviation of the compiled strain data. Any values outside the range of mean ±2*StDev were removed. This outlier exclusion resulted in the removal of 4.16% or 423 data points out of 10,158 L4 stage worms and 6.42% or 559 data points out of 8,708 day-1 adulthood stage worms. The EXT and TOF values were imported into GraphPad Prism 10 for statistical analysis, and log-transformations were performed for graphical depiction.

All data were analyzed using the GraphPad Prism 10 statistical software package (San Diego, California). Analysis of more than two groups with discrete variables was done by one-way ANOVA on a mean rank (Kruskal–Wallis [KW]) test with Dunn’s multiple comparisons post hoc test. Analysis of more than two groups with continuous variables was done for for normality and lognormality; however, all datasets failed normality tests and, thus, underwent one-way ANOVA on the mean rank (KW) test with Dunn’s multiple comparisons post hoc test. Analysis of the neuron area underwent a log-transformation; however, it failed a test for normality and, thus, underwent a Mann–Whitney (MW) U test. Data are represented by a truncated violin plot extending to minimum and maximum values, with dashed lines demarking the median and dotted lines demarking quartiles. Groups within each dataset with distinct letter assignments show statistically significant differences between the groups. Groups within each dataset sharing letter assignments show no statistically significant differences between the groups.

The novel C. elegans strain “MBIA” was generated by crossing the two previously existing strains UA57 (cat-2 overexpressing [cat-2^OE ]) and OK411 (cat-1 null [cat-1^null ]) (Figure 1A). In addition to overexpressing cat-2, an enzyme involved in dopamine synthesis, UA57 worms possess fluorescent dopaminergic neurons expressing the green fluorescent protein (GFP). Male UA57 worms were generated using a heatshock protocol. Briefly, 10–15 L4 hermaphrodite UA57 worms were placed on an agar plate with OP50 food in a 30°C incubator for 4 h. The plate was transferred to a standard incubation environment of 20°C, and the worms were allowed to reproduce. Male UA57 worms were identified by phenotype (e.g., fin-shaped tail) and placed on an agar plate with OP50 food with hermaphrodite worms to generate a consistent source of male UA57 offspring. Male UA57 worms were placed on an agar plate with hermaphrodite OK411 worms to promote breeding to generate the novel MBIA strain containing both the UA57 and OK411 genotypes. Individual first-generation (F1) offspring from the UA57 x OK411 cross with fluorescent dopaminergic neurons indicating the presence of the UA57 genotype were selected and placed on agar plates with OP50 food. These individual F1 worms were allowed to hermaphroditically reproduce, generating second-generation (F2) offspring. Based on the principles of Mendelian inheritance, the F2 offspring were hypothesized to be present at a 1:2:1 ratio of homozygous for wild-type cat-1 alleles, heterozygous for cat-1 alleles, and homozygous for cat-1 null. Thus, individual F2 offspring with fluorescent dopaminergic neurons indicating the presence of the UA57 genotype were selected and placed on agar plates with OP50 food and allowed to reproduce. Once the individual F2 worm had reproduced, the parent F2 worm was collected for PCR analysis to confirm the genotype of the strain. Fluorescent dopaminergic neurons confirmed the presence of the UA57 phenotype, and through PCR analysis, the cat-1 null genotype was identified owing to a shift in genome size (Figure 1B).

To determine the effect of dopamine system mutations on body size, C. elegans populations were synchronized and allowed to develop to the L4 stage, day-1 adulthood, day-2 adulthood, and day-3 adulthood stages. At each stage, the worms were transferred from the agar culture plate onto microscope slides with agar pads in a solution of levamisole (5 mM) to immobilize the animals for brightfield imaging at 4×. The body length was determined by processing images in ImageJ with Fiji using the segmented line tool. Lines were drawn through the midline of each worm from head to tail, consistently beginning and ending at the same morphological points on each animal and reported in micrometer (Figure 2A). The trend of body lengths for the four strains (Figure 2B) demonstrates body size in the presence of dopamine system mutations over four developmental stages. As the worms continue to develop and age, the overproduction of dopamine begins to rescue the body length deficit partially on day 1 adulthood, with a full rescue in body length to wild-type levels by day 2 adulthood and by day 3 adulthood, nearly equivalent body lengths across all strains. Statistical analysis was performed at each developmental stage to determine whether the differences in body size were statistically significant. N2s showed a statistically significant higher body length over OK411 by 12.01%, MBIA by 10.08%, and UA57 by 9.70% at the L4 stage (Figure 2C) (one-way ANOVA on mean ranks with Dunn’s post hoc test, n = 47–93 worms per genotype at each developmental stage, p < 0.0001); however, no statistically significant differences were detected in length between the OK411, MBIA, and UA57 strains at the L4 stage. N2 nematodes were significantly longer than OK411 by 12.36%, MBIA by 6.70%, and UA57 by 8.00% at the day-1 adulthood stage (Figure 2D) (one-way ANOVA on mean ranks with Dunn’s post hoc test, n = 47–93 worms per genotype at each developmental stage, p < 0.0001), and OK411 nematodes were significantly shorter than both MBIA (p < 0.01) and UA57 (p < 0.01) by 5.87% and 4.73%, respectively, at the day-1 adulthood stage. No statistically significant differences were detected in length between N2, MBIA, and UA57 strains at the day-2 adulthood stage, and OK411 nematodes remained significantly shorter than N2 by 9.02% (Figure 2E) (p < 0.0001), MBIA by 7.44% (p < 0.0005), and UA57 by 7.64% (p < 0.0001). At the day-3 adulthood stage, no statistically significant differences were observed in length among N2, MBIA, and UA57 strains, and OK411 nematodes remained significantly shorter than the MBIA strain by 7.66% (Figure 2F) (p < 0.0005).

As a secondary measure of body size, C. elegans were analyzed using a COPAS FP-250 large-particle flow cytometer for high-throughput automated measurement. A COPAS-based analysis was included to determine whether a high-throughput method used to measure the body length was able to replicate the findings from the microscopy analysis and to determine whether this method was able to detect more subtle differences in body size due to the larger population analyzed. C. elegans were collected at L4 and day-1 adulthood stages for analysis of length. Body size was reported for each worm as time of flight (TOF) indicating length and extinction (EXT) indicating width. Following data processing, detailed in Methods, to remove any erroneous objects (e.g., contaminants), the values were compiled across experimental replicates and log-transformed.

The body size of C. elegans acquired by the COPAS can be visualized by the scatter plot with an individual worm’s TOF and EXT values, which demonstrates clustering of worms from each genotype (Figure 3A). At the L4 stage, the N2 strain had the largest TOF values compared to OK411 (p < 0.0001), MBIA (p < 0.0001), and UA57 (p < 0.0001) strains (Figure 3B) (one-way ANOVA on mean ranks with Dunn’s post hoc test, n = 5,396 total worms compiled across 5–6 experimental replicates, p < 0.0001). The OK411 strain had an average TOF value of 68.08% of N2 values, MBIA had an average TOF value of 67.45% of N2 values, and UA57 had an average TOF value of 87.90% of N2 values. While the UA57 strain had statistically significant larger TOF values than OK411 (p < 0.0001) and MBIA (p < 0.0001), there was no statistically significant difference in TOF values detected between the OK411 and MBIA strains. At the day-1 adulthood stage, the N2 strain again had the largest TOF values compared to OK411 (p < 0.0001), MBIA (p < 0.0001), and UA57 (p < 0.0001) strains (Figure 3C) (one-way ANOVA on mean ranks with Dunn’s post hoc test, n = 4,373 total worms compiled across 4–6 experimental replicates, p < 0.0001). The OK411 strain had an average TOF value of 76.38% of N2 values, the MBIA strain had an average TOF value of 86.90% of N2 values, and the UA57 strain had an average TOF value of 74.64% of N2 values. However, at the day-1 adulthood stage, the MBIA strain had significantly larger TOF values than OK411 (p < 0.0001) and UA57 (p < 0.0001), and there were no statistically significant differences in the TOF values detected between OK411 and UA57.

The EXT values at the L4 developmental stage mirrored the TOF values, with the N2 strain having the largest EXT values compared to OK411 (p < 0.0001), MBIA (p < 0.0001), and UA57 (p < 0.001) (Figure 3D) (one-way ANOVA on mean ranks with Dunn’s post hoc test, n = 5,390 total worms compiled across 5–6 experimental replicates, p < 0.0001). The OK411 strain had an average EXT value of 69.50% of N2 values, the MBIA strain had an average EXT value of 66.99% of N2 values, and the UA57 strain had an average EXT value of 95.28% of N2 values. While the UA57 strain had statistically significantly larger EXT values than OK411 (p < 0.0001) and MBIA (p < 0.0001), there was no statistically significant difference in EXT values detected between the OK411 and MBIA strains. At the day-1 adulthood stage, the EXT values of each strain were statistically significant from those of all other strains (Figure 3E) (one-way ANOVA on mean ranks with Dunn’s post hoc test, n = 4,456 total worms compiled across 4–6 experimental replicates, p < 0.0001). The OK411 strain had an average EXT value of 66.07% of N2 values, the MBIA strain had an average EXT value of 83.18% of N2 values, and the UA57 strain had an average EXT value of 77.98% of N2 values. The MBIA strain had statistically significantly larger EXT values than OK411 (p < 0.0001) and UA57 (p < 0.0001), and the UA57 strain had statistically larger EXT values than OK411 (p < 0.0001).

To distinguish whether the observed differences in body size had any effects on development or reproduction, synchronized populations of C. elegans were evaluated for vulval development, reproduction timeline, and egg-laying deficits. At the time of analysis (48 h post-L1 plating), the majority of N2 C. elegans had fully developed vulvas corresponding to stages 8 and 9; however, the OK411, MBIA, and UA57 strains all showed significantly delayed vulval development at this time point of analysis, with the majority of OK411 at stages 5 and 6, the majority of MBIA at stages 5 and 6, and the majority of UA57 at stages 4 and 5 (Figure 4A) (one-way ANOVA on mean ranks with Dunn’s post hoc test, n = 24–32 worms per strain, p < 0.0001). To determine whether this delay in vulval development impacted the amount of time post-L1 plating before eggs were laid, a reproduction assay was developed to monitor individual worms for 24 h beginning at L4 to calculate the amount of time post-L1 plating until eggs were laid (Figure 4B). Statistical analysis revealed a significant p-value of p < 0.05 by one-way ANOVA on mean ranks; however, Dunn’s post hoc test did not identify a significant difference in the number of hours until egg laying between any of the strains (one-way ANOVA on mean ranks with Dunn’s post hoc test, n = 20–29 worms per strain, p < 0.05).

It has been previously reported that cat-1-null C. elegans have egg-laying deficits, resulting in an increase in eggs in utero (Duerr et al., 1999; Bradner et al., 2021). To determine whether overexpression of cat-2 could rescue egg-laying deficits observed in cat-1-null animals, synchronized populations underwent a fecundity assay to quantify the number of eggs in utero at day-1 adulthood, day-2 adulthood, and day-3 adulthood stages (Figure 4C). Overall trends show the fewest number of eggs in utero in the UA57 strain over multiple days of adulthood and the highest amount of eggs in utero in the MBIA strain at day-1 and day-2 adulthood stages. At the day-1 adulthood stage, UA57 had significantly fewer eggs in utero than N2 by an average of 4.52 eggs (p < 0.0001), OK411 by an average of 2.97 eggs (p < 0.01), and MBIA by an average of 6.19 eggs (p < 0.0001), and OK411 had significantly fewer eggs in utero than MBIA by an average of 3.22 eggs (p < 0.01) (Figure 4D) (one-way ANOVA on mean ranks with Dunn’s post hoc test, n = 45–61 worms per genotype at each developmental stage). At the day-2 adulthood stage, UA57 had fewer eggs in utero than MBIA by an average of 5.41 eggs (p < 0.0005), with no other significant differences between strains (Figure 4E) (one-way ANOVA with Dunn’s post hoc test, n = 45–61 worms per genotype at each developmental stage). By day-3 adulthood, no statistically significant differences were detected in the number of eggs in utero between any strains (Figure 4F) (one-way ANOVA on mean ranks with Dunn’s post hoc test, n = 45–61 worms per genotype at each developmental stage, p = n.s.).

As many behaviors in C. elegans have been previously demonstrated as mediated by dopamine, the effect of dopamine-related genetic disruptions on C. elegans behavior was measured by quantifying curling and thrashing in 30-s videos of swimming behavior performed at four developmental stages. Curling was defined as the complete overlap of a worm’s tail to a part of its body and quantified as the number of curls in 30-s intervals (Figure 5A) (Sohrabi et al., 2021). Overall trends showed the highest number of curls at L4 in a 30-s interval in the MBIA strain, followed by OK411 and UA57, and the lowest number of curls in N2 (Figure 5A). With development into adulthood, all strains showed fewer curls in a 30-s interval, reaching near equivalent levels in adulthood (Figure 5A). Overall trends showed more time spent thrashing in a 30-s interval in OK411 than in all other strains at L4, with UA57 spending less time thrashing compared to all other strains (Figure 5B). This trend persists generally over time as the worms develop through adulthood (Figure 5B).

Looking at individual differences in each behavior at each developmental stage, at L4, MBIA had, on average, 1.42 more curls than N2 (p < 0.01) and 1.08 more curls than UA57 (p < 0.05) (Figure 5C) (one-way ANOVA on mean ranks with Dunn’s post hoc test, n = 187 total worms compiled across three experimental replicates, p < 0.001). At the day-1 adulthood stage, no statistically significant differences were detected between any of the strains in the number of curls (Figure 5D) (one-way ANOVA on mean ranks with Dunn’s post hoc test, n = 222 total worms compiled across three experimental replicates, p = n.s.). At the day-2 adulthood stage, OK411 had significantly more curling than N2 by an average of 0.91 curls (p < 0.05), and UA57 had significantly more curling than N2 by an average of 0.67 curls (p < 0.05) (Figure 5E) (one-way ANOVA on mean ranks with Dunn’s post hoc test, n = 278 total worms compiled across three experimental replicates, p < 0.05). At the day-3 adulthood stage, UA57 had significantly more curling than MBIA by an average of 0.80 curls (p < 0.05) despite the one-way ANOVA on mean ranks having a non-significant p-value of > 0.05 (Figure 5F) (one-way ANOVA on mean ranks with Dunn’s post hoc test, n = 268 total worms compiled across three experimental replicates, p = n.s.).

Thrashing was defined as increased spasm-like behavior based on side-to-side movement during swimming behavior and was recorded as the number of seconds spent thrashing during a 30-s interval (Figure 5C) (Sohrabi et al., 2021). At L4, OK411 spent 94.89% of the 30-s interval thrashing, which was significantly higher than that spent by N2 (86.38% of the 30-s interval) (p < 0.0001), MBIA (87.05% of the 30-s interval) (p < 0.0001), and UA57 (67.16% of the 30-s interval) (p < 0.0001) (Figure 5G) (one-way ANOVA on mean ranks with Dunn’s post hoc test, n = 220 total worms compiled across three experimental replicates, p < 0.0001). No statistically significant differences were detected in the amount of time spent thrashing between N2 and MBIA; however, both N2 and MBIA spent more time thrashing than UA57 (p < 0.0001 and p < 0.0001, respectively) (Figure 5G) (one-way ANOVA on mean ranks with Dunn’s post hoc test, n = 220 total worms compiled across three experimental replicates, p < 0.0001). At the day-1 adulthood stage, OK411 continued to display significantly more time spent thrashing, with 96.04% of the 30-s interval spent thrashing compared to N2 (82.07% of the 30-s interval) (p < 0.0001), MBIA (84.51% of the 30-s interval) (p < 0.0001), and UA57 (70.34% of the 30-s interval) (p < 0.0001), and UA57 spent significantly less time thrashing than N2 (p < 0.05) and MBIA (p < 0.005) (Figure 5H) (one-way ANOVA on mean ranks with Dunn’s post hoc test, n = 218 total worms compiled across three experimental replicates, p < 0.0001). This trend continued at the day-2 adulthood stage, where OK411 displayed significantly more time thrashing, with 96.39% of the 30-s interval spent thrashing compared to N2 (91.60% of the 30-s interval) (p < 0.0001), MBIA (91.69% of the 30-s interval) (p < 0.0001), and UA57 (85.85% of the 30-s interval) (p < 0.0001), and UA57 spent significantly less time thrashing than N2 (p < 0.001) and MBIA (p < 0.0001) (Figure 5I) (one-way ANOVA on mean ranks with Dunn’s post hoc test, n = 200 total worms compiled across three experimental replicates, p < 0.0001). At the day-3 adulthood stage, the only significant differences between the genotypes were significantly more time spent thrashing in OK411, with 96.81% of the 30-s interval spent thrashing, compared to the UA57 (76.44% of the 30-s interval) (p < 0.0001), and less time spent thrashing in the UA57 compared to MBIA (94.17% of the 30-s interval) (p < 0.01) (Figure 5J) (one-way ANOVA on mean ranks with Dunn’s post hoc test, n = 246 total worms compiled across three experimental replicates, p < 0.0001).

Behavioral deficits resulting from dopamine dysfunction may be due to deficient dopaminergic neurotransmission or the degeneration of dopaminergic neurons. To determine whether the observed behavioral deficits resulted from differences in neuronal health, the size of the dopaminergic neurons was quantified using GFP images of UA57 and MBIA C. elegans from the L4 to day-3 adulthood stages. Neuron size was determined by processing GFP images in ImageJ with Fiji using the thresholding and ROI area tools. Thresholding was performed to create a mask over dopaminergic neurons, and the area of the ROI was quantified (Figure 6A). A log-transformation was performed on the raw neuron area values to improve data visualization. No statistically significant differences were observed in neuronal size observed at L4 (MW test, p = n.s., U = 8,853) (Figure 6B) and day-1 adulthood stages (MW test, p = n.s., U = 5,331) (Figure 6C); however, at the day-2 adulthood stage, the MBIA neurons were significantly smaller than UA57 by 3.56% (Figure 6D) (MW test, p < 0.05, U = 19,714), and at the day-3 adulthood, the MBIA neurons were significantly larger than UA57 neurons by 5.62% (MW test, p < 0.005, U = 7,729) (Figure 6E).