Mutagenesis alters sperm swimming speed in Astyanax cavefish

Ethics statements

The fish were cared for according to protocol #05-1235 approved by the New York University Animal Welfare Committee (UAWC). All methods were performed in accordance with current guidelines and regulations. The study is reported in accordance with ARRIVE guidelines. The authors have no competing interests.


Statistical analyzes were performed using Statistica (Tibco). We predicted that ENU treatment would increase the variability of VCL and therefore we calculated the one-tailed probabilities. The probabilities calculated to test the flagellar length hypotheses were two-sided. Statistica’s outlier filtering feature was used to identify outliers in the raw data, which were then excluded from further analysis.


We used commercial sources Astyanax mexicanus cave fish for these experiments as they are inbred, having been held in captivity for many generations since their original collection in 193635. They were about 1.5 years old during the mutagenesis and 2.5 years old when the experiments began.


Mature males were bathed in a system water solution containing 2.5 mM N-ethyl-N-nitrosourea (ENU), according to published protocols.9. Thirty-eight men were initially treated for one hour. There was 16% mortality during treatment. After five weeks of recovery time, the remaining 32 males were treated a second time for 25 min. Twenty-nine of the 32 survived (9% mortality). The fish were then set aside for a full year before the current experiments began. In Daniothe mutagenic effect of ENU treatment is roughly proportional to the product of exposure concentrations and time9. Thus, we quantified the exposure of mutagenized males to ENU in the present experiment at approximately 3.5 mM hours.

The effectiveness of ENU treatment in this experiment can be estimated from published data on other freshwater fish. Sequencing data on three species of Lake Malawi cichlids allowed estimates of spontaneous mutation rates of 3.5 × 10E-9 per base per generation36which we took as a reasonable estimate for the unknown rates in Danio and Astyanax. In Danioa study of ENU-induced mutations (15 mM hour exposure) in four pigmentation genes gave an estimate of a mutation rate per locus of 1.1 × 10E-39. The added coding sequence of these genes (that’s to say, the mutational target size) was estimated at 9189 bases. The exact figure is unknown because a gene used in9“brass”, has not yet been identified in the Danio genome. Its target size was taken as the average of the other three markers (SLC24A5, Kitaand Oca2). Adjusted for difference in ENU exposure and for an estimated proportion of synonymous mutations of 0.3537.38, the ENU-induced mutation rate in our experiments is estimated to be 12.5 times the spontaneous rate. This figure is underestimated because all non-synonymous mutations in9 should have observable phenotypes. Although approximate, the estimate is instructive and indicates that ENU treatment in this study led to a substantial increase in the number of new mutations.

Preparatory treatment and semen collection

Prior to testing, mutagenized and non-mutagenized males were housed in two 57 L aquariums adjacent to each other on an aquarium stand. Both holding tanks had bubblers operated from the same air pump and were of a flow-through design and fed from the same water source. Illuminance and temperature (21 ± 1°C) were equal in both reservoirs. For testing, a small number of individuals (3-6) were transferred out of the holding tanks and into two 21 L treatment tanks adjacent to each other on a nearby rack. They were initially at the same temperature as the retention tanks. Water flow was shut off to the process tanks and 25 watt heaters were turned on in each. Over several hours, the reservoirs warmed to 26 ± 1°C. Final temperatures in the two tanks were close, but not identical, reflecting small differences in heater settings. To minimize the systematic effects of temperature differences, the heaters were switched between treatment tanks at each experimental cycle.

The males were kept at a high temperature overnight to stimulate the production of new sperm for harvest39. For semen collection, males were anesthetized using 0.025% tricaine-methanesulfonate (MS-222, Western Chemical) for approximately 60 s. They were then patted dry, placed supine on a damp nylon wool pad, the gonopore areas were washed with Hanks’ Buffered Saline (HBSS) to suppress sperm activation.40. Subjects were gently squeezed to expel semen, samples of which were taken using fine-tipped transfer pipettes. Samples were placed on ice and used within one hour of collection, either for behavioral study and CASA analyzes or for measurement of sperm flagella length.

CASA analysis

For CASA experiments, we visualized sperm using an Olympus IMT-2 microscope, recording swimming behavior with negative phase-contrast illumination at 100 frames per second. We used a monochrome camera with a resolution of 1280 × 860 pixels (Chameleon3-U3-13Y3M-CS) and FlyCap2 Viewer version, ( For Experiments 1 and 2, we activated sperm by mixing 1 μL of sperm suspension with 7 μL of system water. After activation, 1 μL of well-mixed activated sperm was pipetted into a single well of a 12-well multi-test slide (MP Biomedicals, Irvine, CA, USA) and covered with a coverslip.11.12. Each sample was tested twice. The median numbers of motile sperm tracked in the two experiments were 113 and 163. The objective used was 20X and the planar camera eyepiece was an NFK 3.3.

Based on the results of experiments 1 and 2, we diluted the sperm suspensions for experiment 3 by adding small volumes of HBSS before recording. This reduced lane overlap and collisions. To maintain a sufficient sample size, we increased the number of replicates for each sample. The median number of samples processed per male in Experiment 3 was nine and the median number of motile sperm per recording was 50. The objective used was 10X. Changing the power from 20 to 10X had two helpful effects. First, the field had four times the area in Experiment 3 compared to Experiments 1 and 2. Second, the 10X had increased depth of field and more clearly imaged objects. The result of all the changes was that we tracked more sperm per male in Experiment 3 than in Experiments 1 and 2, but had significantly less crowding, by a factor of eight or more.

For the three experiments, we analyzed the 100 images from 20 s after activation. CASA identifies individual sperm moving from frame to frame and reports multiple swimming phenotypes11.12. Three phenotypes are measures of speed, all of which are highly correlated with each other. Four other phenotypes are derived primarily from the original speed measurements. We focused on the simplest measurement, the curvilinear velocity (VCL). VCL is the average speed during the tracked time interval, for example, the time divided by the total distance traveled.

Sperm Flagellar Measurements

To measure individual sperm tail lengths, immobilized sperm samples were diluted with appropriate volumes of HBSS to obtain uncrowded images. Then, 1 μL of diluted sperm was pipetted into a single well of a 12-well multi-test slide (MP Biomedicals, Irvine, CA, USA) and covered with a coverslip. A Zeiss microscope (Axio, 40X objective) was used to visualize sperm structure, including the sperm head and flagella. Images were processed by FIJI software (NIH) and 50 or more sperm were randomly selected from each male for their flagellum length measurements. Sperm cells were photographed using a 1.3 MP monochrome camera (CMLN-13S2M-CS, 1296 × 964 res, FLIR Inc.) and FlyCap2 software (FLIR Inc.).

Comments are closed.