The effects of high temperature on aborted pollen production, pollen morphology and pollen viability.
Male buds of P. canescens developed slowly and anthers became dry and brown after high temperature treatment. Some buds were dead after they were exposed to high temperatures, which precluded pollen collection.
High temperature-induced pollen was collected from all surviving male buds. A high proportion of aborted pollen was observed in addition to some induced 2n pollen (Figure 1a). Spontaneously aborted pollen was rarely observed in the control group (Figure 1b). The percentage of aborted pollen in different treatments is shown in Table 2. The average percentage of high temperature-induced aborted pollen varied from 9.45 to 25.11%. GLM-univariate analysis of the percentages of high temperature-induced aborted pollen showed that dominant meiotic stage (F = 5.75, P<0.001), temperature (F = 25.71, P<0.001), and duration of treatment (F = 17.82, P<0.001) significantly affected the percentages of high temperature-induced aborted pollen. Dominant meiotic stage × temperature (F = 3.96, P= 0.004) and dominant meiotic stage × duration (F = 5.48, P<0.001) interactions also significantly affected the percentage of high temperature-induced aborted pollen. Temperature × duration (F = 2.40, P= 0.128) and dominant meiotic stage × temperature × duration (F = 1.21, P= 0.319) interactions had no significant effect on the percentage of high temperature-induced aborted pollen. LSD multiple comparison tests showed that the differences in induced aborted pollen production were significantly higher at the diplotene stage than at the leptotene, pachytene, diakinesis, and metaphase I stages (α = 0.05). The percentage of high temperature-induced aborted pollen was higher at 41 °C than at 38 °C. The percentage of high temperature-induced aborted pollen was higher for samples that were treated after 6 hr than after 3 hr.
Pollen morphology was studied by scanning electron microscopy. The pollen grains within the control group was uniform, spherical, with few corrugations and granular surface (Figure 2a, d and g). There was no aperture on surface. Few aborted pollen grains (arrows) were induced (Figure 2b and c) by 38 °C high temperature for 3 or 6 hr. The morphology (Figure 2e, f, h and i) and the ectexine deposition (Figure 2d and g) of induced pollen were similar to the control group, suggesting that high temperature had no significant effects on the pollen morphology.
For evaluation of induced pollen viability, Fresh high temperature-induced pollen grains were supplied for germination test on the medium containing 0.7% agar, 50 mg/L calcium chloride, 120 mg/L boric acidm. Some germinated pollen grains were, respectively, observed for the control (Figure 3a), treatment at 38 °C for 3 hr (Figure 3b) and treatment at 38 °C for 6 hr (Figure 3c) groups. Pollen germination rates are showed in Figure 3d. After 6 hr of culture, the average gerimation rate in the control group was 26.95%, which was slightly higher than that of induced pollen after treatments at 38 °C for 3 hr (21.52%) and at 38 °C for 6 hr (20.29%), indicating that the frequencies of aborted pollen in the treatment groups were slightly higher than that in the control group. Among the treatments, the average germination rate of induced pollen via 38 °C treatment for 3 hr slightly higher than that of induced pollen via 38 °C treatment for 6 hr. However, the GLM-Univariate analysis of germination rates revealed that high temperatures had no significant effect on induced pollen germination rates.
Induced meiotic abnormalities by high temperature.
In male flower buds that were exposed to 38 °C for 3 or 6 hr at the diploptene stage, all treated flower branches continued to be hydroponically cultured until pollen was released from the anthers. At the anaphase I, anaphase II, and tetrad stages, five flower buds per treatment were sampled to determine the effect of high temperature on the meiosis of PMCs. In the control group, the PMCs underwent normal meiosis, and a few lagging chromosomes in some PMCs were observed at anaphase I (Figure 4a) or anaphase II (Figure 4b). The number of lagging homologous chromosomes varied from one to eight, and the percentage of PMCs with lagging homologous chromosomes was 22.7 ± 4.2% at anaphase I. The number of lagging sister chromosomes varied from one to ten, and the percentage of PMCs with lagging sister chromosomes was 13.3 ± 5.0% at anaphase II (Table 2).
In some PMCs exposed to 38 °C, meiosis was abnormal, and a large number of lagging chromosomes were observed at anaphase I or II (Figure 4d, f, g and h). After 3 hr of treatment at 38 °C, the number of lagging homologous chromosomes varied from 1 to 13, and the percentage of PMCs with lagging homologous chromosomes was 40.7 ± 3.1% at anaphase I. The number of lagging sister chromosomes varied from one to 20, and the percentage of PMCs with lagging sister chromosomes was 33.3 ± 8.0% at anaphase II (Table 3). The percentage of PMCs with lagging homologous or sister chromosomes was significantly higher after treatment at 38 °C for 3 h compared with the control group. When PMCs were treated at 38 °C for 6 hr, the number of lagging homologous chromosomes varied from one to 28, and the percentage of PMCs with lagging homologous chromosomes was 55.3 ± 4.2% at anaphase I. The number of lagging sister chromosomes varied from one to 23, and the percentage of PMCs with lagging sister chromosomes was 48.7 ± 3.1% at anaphase II (Table 3). The percentage of PMCs with lagging homologous or sister chromosomes was slightly higher after treatment at 38 °C for 6 hr compared with PMCs treated at 38 °C for 3 hr.
After two rounds of cell division, normal cytokinesis occurred in the control group, resulting in tetrad formation (Figure 4c). However, few micronuclei were observed in some of the treated PMCs (Figure 4f and i), and some polyads formed at the tetrad stage, suggesting that chromosome segregation errors occurred and led to the production of aneuploid gametes.
Delayed tapetum development is not responsible for induced aborted pollen production.
Defective tapetum development is often associated with the disrupted development of meiocytes and/or pollen and reduced/impaired fertility20. We examined the effect of high temperature on tapetum development via sectioning and toluidine blue staining during microspore maturation. At the tetrad stage in the control group, the tapetum was observed in the anther, which surrounded the developing PMCs (Figure 5a). Two days after the tetrad stage, the tapetum degenerated normally by programmed cell death (PCD) (Figure 5b). Four days after the tetrad stage, thinner tapetal layer cells were observed in the anther (Figure 5c). Six days after the tetrad stage, no tapetal layer cells in the anther were observed (Figure 5d). Eight days after the tetrad stage, the anther dehisced, and a large number of pollen grains with good fertility were released (Figure 5e).
At the tetrad stage in the group exposed to 38 °C for 3 hr, the tapetum was observed in the anther, which surrounded the developing PMCs (Figure 5f). Two days after the tetrad stage, tapetum degenerated gradually by PCD (Figure 5g). Four days after the tetrad stage, thinner tapetal layer cells were observed in the anther (Figure 5h). Six days after the tetrad stage, few tapetal layer cells in the anther were observed (Figure 5i), suggesting that the tapetum of the treated anthers degenerated more slowly than the tapetum in the control group. Eight days after the tetrad stage, the anther matured, and more abortive pollen grains were released (Figure 5j). A similar pattern of tapetum degeneration was observed for PMCs exposed to 38°C for 6 hr (Figure 5k-n), except that many more abortive pollen grains were released after anthers matured (Figure 5o).
Induced spindle destabilization results in aborted pollen production.
Because the meiotic microtubular cytoskeleton plays an important role in the segregation of homologous and sister chromosomes during meiosis, we examined the integrity and localization of microtubule structures of meiocytes exposed to 38 °C for 3 or 6 hr using tubulin-α immunocytology. Under control conditions, the microtubular cytoskeleton was distributed regularly in microsporocytes (Figure 6a-h), which allowed for the proper segregation of chromosomes in daughter cells. After PMCs were treated at 38 °C for 3 hr or 6 hr at the diplotene, all meiotic stages exhibited the same microtubular distribution pattern (Figure 6i-x), similar to unstressed meiocytes (Figure 6a-h). However, the high temperature treatment resulted in some defects in the meiotic microtubular cytoskeleton. For example, the microtubules depolymerized and were not present in the cytoplasm. The extent of the treatment affected the degree of microtubule destabilization, and the plasma membrane even ruptured in some cells.
In the metaphase of the first meiotic division, a single bipolar spindle microtubule depolymerized in cells treated at 38 °C for 3 hr (Figure 6j) and was damaged in cells treated at 38 °C for 6 hr (Figure 6r). The damaged spindle microtubule resulted in the slow movement of chromosomes within the cytoplasm. This is consistent with the large number of lagging homologous chromosomes observed at anaphase Ⅰ (Figure 4d and g). Similar to metaphase Ⅰ, a double bipolar spindle microtubule at metaphase Ⅱ depolymerized in cells treated at 38 °C for 3 hr (Figure 6n) and was damaged in cells treated at 38 °C for 6 hr (Figure 6v). Subsequently, several lagging sister chromosomes at anaphase Ⅱ were observed (Figure 4e and h).
Under control conditions, telophase II male meiocytes typically generate a microtubule network that consists of six tetrahedrally arranged phragmoplast-like structures, which are localized between the four haploid nuclei (Figure 6h). However, telophase II male meiocytes treated at 38 °C for 3 hr generated a new complex microtubule network that surrounded more than four nuclei instead of the six tetrahedrally arranged phragmoplast-like structure. Microtubule bundles between two nuclei were much thinner in cells treated at 38 °C for 6 hr than cells treated at 38 °C for 3 hr and control cells.
The expression of PtActin is down-regulated by high temperature.
The F-actin cytoskeleton mediates a variety of essential biological functions in eukaryotic cells, as its dynamic properties drive chromosomal movement during cell division. We hypothesized that the expression of PtActin is affected under heat stress conditions. To test this hypothesis, we monitored the expression of PtActin in male flower buds derived from the treated groups and control group using qRT-PCR. We detected a significant (p < 0.001) decrease in the expression of PtActin when the PMCs of P. canescens at the diplotene were exposed to 38 °C. The relative expression of PtActin was only 9.2 ± 0.9% in PMCs after treatment at 38 °C for 3 hr, which was slightly lower compared with PMCs treated at 38 °C for 6 hr (13.7 ± 1.2%) (Figure 7), indicating that high temperature significantly affects the transcription of PtActin.