Habitat-Specific Clock Variation and Its Consequence on Reproductive Fitness

Abstract

The circadian clock controls daily activities at the cellular and organismic level, allowing an organism to anticipate incoming stresses and to use resources accordingly. The circadian clock has therefore been considered a fitness trait in multiple organisms. However, the mechanism of how circadian clock variation influences organismal reproductive fitness is still not well understood. Here we describe habitat-specific clock variation (HSCV) of asexual reproduction in Neurospora discreta, a species that is adapted to 2 different habitats, under or above tree bark. African (AF) N. discreta strains, whose habitat is above the tree bark in light-dark (LD) conditions, display a higher rhythmicity index compared with North American (NA) strains, whose habitat is under the tree bark in constant dark (DD). Although AF-type strains demonstrated an overall fitness advantage under LD and DD conditions, NA-type strains exhibit a habitat-specific fitness advantage in DD over the LD condition. In addition, we show that allelic variation of the clock-controlled gene, Ubiquinol cytochrome c oxidoreductase (NEUDI_158280), plays a role in HSCV by modulating cellular reactive oxygen species levels. Our results demonstrate a mechanism by which local adaptation involving circadian clock regulation influences reproductive fitness.

Keywords: adaptation; circadian clock; fitness; habitat; natural variation.

Figure 1.

Habitat-specific clock variation in N. discreta PS4b. (A) Inverted race tube analysis of N. discreta ecotypes under 12-h light/dark (LD) cycles at 25 °C. The black bars indicate dark conditions, while the white bars indicate light conditions. (B) Scattered plot showing the distribution of rhythmicity index scores across samples from different continents: other continents (circle) and North America (square). Error bars represent SEM, Welch’s t test, p < 0.0001.

Figure 2.

FRQ rhythms are uncoupled to circadian regulation of asexual development in North American (NA) strains. (A) Levels of FRQ in 3 representative NA strains (FGSC8579, New Mexico United States; FGSC8568, Idaho, United States; FGSC9981, Alaska, United States) and in 3 other continent strains (FGSC8814, Thailand; FGSC9963, Ivory Coast; FGSC9989, Portugal) in constant darkness (DD) and light induction (LI) conditions. Neurospora crassa FGSC 2225 and frq knockout (frq^ko; FGSC11554) strains were used as positive and negative controls, respectively. the graph below the Western blot analysis is the quantification of FRQ. There was no significant difference in light induction of FRQ between high and low rhythmicity index (RI) ecotypes (t test, p = 0.17). (B) Western blot analysis of FRQ in constant darkness shows the typical progressive phosphorylation and circadian oscillation in NA strains; N. discreta FGSC8579 and N. discreta FGSC9981 are representative NA strains. frq^ko is used as a negative control. The graph below the Western blot analysis is the quantification of FRQ. Open symbols represent NA strains. N. crassa ras-1^bd (circle), N. crassa FGSC2489 (square), N. discreta sensu stricto (triangle), N. discreta FGSC8579 (open inverted triangle), N. discreta FGSC9981 (open diamond). (C-D) The cellular FRQ levels do not segregate with RI in N309. (C) Six representative race tube images of N309 strains with high RI (309–10, −80, and −89) and low RI (309–30, −38, and −50) in LD conditions. (D) FRQ levels in representative N309 strains with high and low RI and parent strains 8831 and 8578. There is no correlation between levels of FRQ and RI score (Pearson correlation; R² = 0.3411, p = 0.1285). (E) Decoupling of the circadian regulation of asexual development. Asexual development of FGSC9980 (NA strain) in race tubes under LD at 25 °C. Error bar represents SEM, n = 6. Upper panel, A significant chemiluminescent molecular rhythm of FRQ:LUC translational fusion protein of the FGSC9980 strain bearing a FRQ:LUC fusion (analysis of variance [ANOVA], p < 0.0001). Measurements of arrhythmic conidiation (ANOVA, p = 0.99) in the transgenic FGSC9980 strain bearing a FRQ:LUC fusion construct from the same chamber where the race tube experiment (lower panel) was performed. Error bars represent SEM, n = 6. Data were collected every 10 min (lower panel) and 30 min (upper panel), respectively, for 3 days.

Figure 3.

North America strains exhibit a fitness advantage in their habitat of constant dark. (A) The number of conidia is counted in African-type (AF) and North American–type (NA) strains in constant dark (DD) and light-dark (LD) cycling conditions, n = 16. AF-type strains demonstrate a significant fitness advantage over NA-type strains under both conditions (Wilcoxon rank-sum test: DD, W = 7, p = 3.3e-05; LD, W = 6, p = 2.2e-05). There was also a significant fitness advantage of NA-type strains in DD over LD conditions (Wilcoxon rank-sum test: W = 26, p = 0.007). (B) Plot of fitness (total number of conidia) as a function of the oscillators’ coupling and sensitivity. For fixed K (coupling strength, SI Appendix, Suppl. Fig. S5 and Methods section), conidiation decreases as the sensitivity increases. Weakly coupled oscillators with low sensitivity (K = 1) are outperforming highly coupled oscillators (K > 5). For large K’s, conidiation is changing mildly as a function of sensitivity.

Figure 4.

Habitat-specific clock variation of reproductive fitness is shaped by the variation in gene expression and cellular reactive oxygen species (ROS) due to allelic variation of NEUDI_158280. (A) Twenty-four-hour time-course mRNA analysis of NEUDI_158280 in high rhythmicity index (RI; N309–89; square) and low RI (N309–50; circle) phenotypes. mRNA expression is significantly rhythmic in high RI phenotypes (Circwave, p < 0.001). (B) Segregation of the levels of expression of NEUDI_158280 and RI phenotypes. (C) Response of NEUDI_158280 gene expression upon H₂O₂ treatment with N309–89 (circle) and 50 (square). N309–89 mRNA expression levels significantly increase with 10 mM H₂O₂ treatment (t test, p < 0.05), which is not observed with N309–50 (t test, p = 0.76). (D) Variation of ROS levels with N309–89 (high RI) compared with N309–50 (low RI). ROS levels are significantly higher with high RI compared with low I strains (t test, p < 0.01). (E) Inhibition of ROS with N-acetyl-L-cysteine reduces RI scores in high RI strain N309–89 but not low RI strain N309–50.