Ploidy dynamics in aphid host cells harboring bacterial symbionts

General observation and methods for ploidy analysis on aphid bacteriome cells

Consistent with previous observations^9,21,22,40, the bacteriome of viviparous aphids consisted of two types of cells: bacteriocytes and sheath cells (Fig. 2). Bacteriocytes contained Buchnera cells and were much larger than sheath cells. Sheath cells exhibited a flattened morphology and surrounded the bacteriocytes. Both cell types possessed a single nucleus. Bacteriocytes had a single prominent nucleolus, which was not stained using DAPI, but using “Nucleolus Bright Red” staining (Fig. 2). Most sheath cells also had a single nucleolus, yet a small number had two. “Nucleolus Bright Red” also stained the peripheral region of Buchnera, probably because of the richness of RNA around Buchnera cells.

To determine the most suitable methods for ploidy analysis of aphid bacteriocytes, three types of methods, flow cytometry, Feulgen densitometry, and fluorometry were compared. First, flow cytometry successfully detected the nuclei of bacteriome cells and heads, and distinct peaks were present (Fig. S3). There were several peaks, which can be categorized as ploidy classes based on head peaks, assuming that the smallest peaks correspond to a diploid population. We recognized peaks up to 256C (256-ploidy) cells but could not distinguish cell types (i.e., bacteriocytes or sheath cells) in this method due to a lack of cytological information. Note that “C” means haploid genome size, for example, 2C = diploid and 8C = octoploid. Second, Feulgen densitometry also showed several ploidy levels of up to 128C (Fig. S4) in bacteriocytes. Sheath cells mainly consisted of 16-32C cells. However, we found that many cells were lost during the experimental procedures, probably due to the repeated washing processes and the long incubation time.

We found the third method, image-based fluorometry for isolated nuclei, the best for quantitative ploidy analysis of aphid bacteriocytes (Fig. 3). Fluorometry showed distinct peaks of integrated fluorescence intensity, and they could be categorized as each ploidy class based on the intensity of the smallest peak in head cells (diploid population). The results were consistent with other methods; ploidy levels were 32C-256C in bacteriocytes and 16C-32C in sheath cells. In this analysis, the nucleolus size was used to discriminate between cell types. During cytological observation, we obtained the size distribution of the nucleolus, and it was revealed that the nucleolus of bacteriocytes was always larger than that of sheath cells (Fig. S5). Based on the results, we determined the threshold of the size of the nucleolus. More specifically, in viviparous females, nuclei that have nucleoli larger than 20 μm² were categorized into bacteriocytes. Note that the peaks of sheath cells were not distinct or reliable for categorizing their ploidy class; therefore, we showed results focusing on bacteriocytes in the following sections.

Cellular features of bacteriome cells in viviparous and oviparous females, and males

The cellular features were generally consistent among young adults (within 5 days of adult eclosion) of three morphs, viviparous and oviparous females, and males (Fig. 2). Nevertheless, Buchnera-absence zones in the cytoplasm of bacteriocytes, which are considered to be degeneration of Buchnera⁴⁵, and bacteriocytes degeneration⁴⁶ were both observed more frequently in male bacteriocytes than in females (Fig. 2). The cell size of bacteriocytes was significantly different among morphs (LM with type II test, F = 286.15, df = 2, p < 0.001, Fig. S6). Viviparous females had significantly larger bacteriocytes (Tukey’s test, p < 0.05, Fig. S6). The size of nucleoli was significantly different between bacteriocytes and sheath cells, regardless of aphid morphs (LM with type II test; viviparous females, χ² = 618.4, df = 1, p < 0.001, oviparous females, χ² = 1,430.4, df = 1, p < 0.001, males, χ² = 261.37, df = 1, p < 0.001, Fig. S5). There was no overlap in the nucleolus size between cell types (Fig. S5). Based on these data, we determined the threshold of the size of the nucleolus to discriminate between bacteriocytes and sheath cells. Specifically, in viviparous and oviparous females, and males, nuclei that have nucleoli larger than 20 μm²_, 20 μm²_, 8 μm², were categorized into bacteriocytes, respectively.

Ploidy analysis on the bacteriocyte of viviparous and oviparous females, and males

Ploidy analysis of the adult bacteriocytes revealed that the cells were highly polyploid (from 32 to 256C) in all phenotypes (Fig. 4). We found variation in the level of ploidy; bacteriocytes of viviparous females, oviparous females, and males mainly consisted of 64-128C (45% for each), 64C (70%), and 32-64C (30% and 47%), respectively. There were significant differences in the degree of polyploidy (median ploidy level) in bacteriocytes among the three aphid phenotypes (Brunner–Munzel test with Bonferroni adjustment; viviparous females vs. oviparous females, p < 0.05; viviparous females vs. males, p < 0.05; oviparous females vs. males, p < 0.05; Fig. 4).

Fecundity and longevity of viviparous and oviparous females

Viviparous females in this strain laid 95.25 ± 24.75 (mean ± SD, n = 12) nymphs during their lifetime, while oviparous females oviposited 28.83 ± 6.52 (n = 12) eggs (Fig. S7. The total number of nymphs laid by viviparous females was significantly higher than that of eggs oviposited by oviparous females (GLMM with type II test, χ² = 449.74, df = 1, p < 0.001). Viviparous females lived longer [44.33 ± 4.38 (mean ± SEM) days] than oviparous ones (26.08 ± 2.31 days), yet there was no interaction between female types and their lifetime (GLMM, with type II test; female type, χ² = 118.13, df = 1, p < 0.001, lifetime, χ² = 69.32, df = 1, p < 0.001, and the interaction χ² = 0.74, df = 1, p = 0.39). Viviparous females started reproducing from days 2–3 of adulthood and the rate of larviposition peaked during days 3–20 but slowed down during days 21–28. They lived at most 50–55 days, although most of them stopped the larviposition after day 30. In oviparous females, first oviposition and mating with males were observed on days 3–4. They actively laid eggs until day 14, but their death was observed almost simultaneously (Fig. S7).

Cellular features of bacteriome cells in each stage of post-embryonic development

At 16 °C, viviparous (and apterous) aphids reached adult stage approximately 14 days after birth [13.73 ± 0.32 (mean ± SEM), n = 16, Fig. S8a]. In particular, N1, N2, N3, and N4 periods lasted for 3.2 ± 0.11, 3.0 ± 0.09, 3.3 ± 0.12, and 4.2 ± 0.14 days (mean ± SEM, n = 16, Fig. S8a], respectively. Adult aphids started reproducing 2 or 3 days after eclosion (molt for an adult) and continued larviposition for approximately 4 weeks (Fig. S7). Based on these data, A7 aphids (7 days after eclosion) could be categorized as actively reproducing individuals. A21 aphids (21 days after eclosion) were categorized as senescent individuals, although they continuously produced offspring. During the nymphal stages of viviparous aphids, the morphology of bacteriome cells was generally consistent; all bacteriocytes and most sheath cells were uninuclear (Fig. S8b), but very few of the latter cells had several small nuclei. Notably, there were drastic morphological changes in adult stages; bacteriocyte and sheath cell nuclei of A21 individuals were irregularly shaped in comparison with those of young (A0) and reproducing (A7) individuals. Furthermore, in A21, we frequently observed bacteriocytes in which the signals of DAPI and Nucleolus Bright Red signals on Buchnera were weak (Fig. S8b). These changes were consistent with symptoms of Buchnera degeneration and cell senescence, which have been previously reported ^45,46. Developmental stages had a significant effect on bacteriocyte size (LMM with type II test, χ² = 338.73, df = 6, p < 0.001). During post-embryonic development, the size of bacteriocytes consistently increased (Tukey’s test: N1 = N2 = N3 ≤ N4 < A0 = A7 = A21, p < 0.05, Fig. S9). The volume of bacteriocytes was positively correlated with those of their nuclei (Simple correlation analysis with LM; p < 0.001, R² = 0.83, Fig. S10). The size of nucleoli was significantly different between bacteriocytes and sheath cells, regardless of the post-embryonic developmental stages (LMM with type II test; N1, χ² = 891.82, df = 1, p < 0.001, N2, χ² = 294.04, df = 1, p < 0.001, N3, χ² = 842.31, df = 1, p < 0.001, N4, χ² = 817.18, df = 1, p < 0.001, old adults, χ² = 1,405.6, df = 1, p < 0.001, Fig. S11). There was no overlap in the nucleolus size between cell types (Fig. S11). Based on these data and the data from young adults, we determined the threshold of the size of the nucleolus for ploidy analysis (in N1 and N2, 10 μm², and later stages, 25 μm²).

Ploidy dynamics of aphid bacteriocytes along with post-embryonic development

During post-embryonic development of viviparous females, the ploidy level of bacteriocytes gradually increased; bacteriocytes were 16-32C at the time of birth (N1) and reached the highest ploidy level in actively reproducing adults (A7, 128–256C) (Fig. 5). All stages of viviparous females except senescence stage A21 showed significant differences in ploidy levels (Brunner–Munzel test with Bonferroni adjustment, N1 < N2 < N3 < N4 < A0 < A21 < A7, p < 0.05, Fig. 5). The highest dominant ploidy class was observed in A7 aphids (256C, 43%) (Fig. 5). A similar pattern was observed in oviparous aphids (Fig. S12), and the highest dominant ploidy was also observed in A7 individuals (but 128C, 47%).

The size of the nuclei and nucleolus, and ploidy levels in aphid bacteriocytes

The size of bacteriocyte nuclei was positively correlated with the ploidy class in all aphid categories [adult viviparous females, adult oviparous females, adult males, and all stages of viviparous/oviparous females (N1-N4 and A0, A7, A21 were pooled)] (Simple correlation analysis with LM; p < 0.001) (details in Fig. S13). There were significant effects of ploidy class on the size of the nucleolus in adult bacteriocytes of each morph (LM with type II test; viviparous females, F = 62.94, df = 2, p < 0.001, oviparous females, F = 23.97, df = 2, p < 0.001; males, F = 6.44, df = 3, p < 0.001, Fig. 6a). Note that 16C and 256C viviparous bacteriocytes were excluded from the analysis due to their small number. Similarly, 16C and 8C cells of females and males, respectively, were excluded from the analysis. In viviparous females, the size of the nucleolus consistently increased from 32 to 128C (Tukey’s test, p < 0.001). In oviparous females, 128C cells had larger nucleoli than 32C and 64C cells (p < 0.001 each), yet the difference between 32 and 64C cells was marginally non-significant (p = 0.06). In males, the size of the nucleolus of 128C cells was significantly larger than that of 16C and 32C cells (p < 0.001 each), but we did not find any significant difference among other comparisons (16C vs. 32C, p = 0.71, 16C vs. 64C, p = 0.10, 32C vs 64C, p = 0.08, 64C vs 128C, p = 0.18) (Fig. 6a). A significant effect of ploidy class on the nucleolus size was also detected in the data from each developmental stage of viviparous aphids (LMM with type II test; χ² = 788.83, df = 5, p < 0.001, Fig. 6b). Note that 4C and 512C bacteriocytes were excluded from the analysis because of the small number of observations. The size of the nucleolus consistently increased from 8 to 256C (Tukey’s test, p < 0.05, Fig. 6b).

Source: Ecology - nature.com