Contribution
The authors had earlier found Hodgkinia splitting into lineages with complementary gene sets (refs 19–22). They had also guessed that this would make transmission to eggs risky. This paper tests the guess. It measures how many symbiont cells reach each egg across cicadas with one, six and “perhaps dozens” of Hodgkinia lineages. It asks whether every egg gets every lineage, and whether the transmission route has changed.
Key insight
A host whose essential symbiont has split into uneven parts faces a sampling problem at each generation. The rarest essential lineage sets the number of cells that must be sent. The answer found is not a new sorting mechanism, but more of the same: the host sends more Hodgkinia cells and not more Sulcia. The authors frame this as the host compensating for changes in its symbiont, and as a cost the symbiont’s changes impose. They add a twist: a larger transmitted population means weaker bottlenecks, which could make further splitting easier. In their words, “the compensatory changes cicadas have evolved in response to increasing Hodgkinia complexity might themselves make the problem of splitting worse”.
Assumptions
- The simulation (Fig. 1). Cells are drawn at random, with replacement, from a mother’s pool whose lineage proportions are taken from genome-assembly coverage. An egg is viable if it gets at least 1 (or 50, or 100) cells of every lineage. For Magicicada tredecim, every finished genomic circle is treated as a separate cell lineage. The authors say this, and the 1-cell rule, make the result an upper bound.
- Cell counts (Fig. 2). Sulcia cells are counted in a box of about 50 × 50 × 10 μm³ and scaled to the volume of the symbiont ball, taken as a sphere or spheroid. Hodgkinia is then estimated from the Hodgkinia:Sulcia ratio on one confocal slice. Cell shape may cause double counting, which the authors expect to affect all species alike. Egg age is uncontrolled; only eggs with an intact symbiont ball were counted.
- Lineage presence (Fig. 3). One marker gene stands for the lineages: rpoB in two Tettigades species, where every cellular lineage carries it; etfD in M. septendecim, where no gene is on every circle and two homologues were tracked. Abundances from amplicons (over 25 PCR cycles) are taken as less accurate than metagenomic ones.
- Egg nests. All eggs in a nest are assumed to come from one female. For Magicicada, nests in a series whose genotypes did not match were dropped.
Key results
- Simulation (Fig. 1).
- With lineage abundances from M. tredecim, the minimum number of cells that guarantees every egg all lineages grows “by more than 2,000-fold” as lineages multiply and grow uneven.
- For T. chilensis (6 lineages, 69-fold abundance range), 1 cell of each lineage needs over 500 cells per egg, 50 cells over 8,000, and 100 cells over 15,000.
- M. tredecim (30 putative lineages, 74-fold range) needs 4 to 5 times as many as T. chilensis.
- Counts in eggs (Fig. 2). Five or six eggs per species.
- Sulcia: 2,572 (M. septendecim) to 5,643 (Diceroprocta semicincta), about 2-fold. Only that pair differs significantly (P = 0.0005).
- Hodgkinia: 4,889 (T. ulnaria, 1 lineage) to 30,154 (M. septendecim), about 6-fold. M. septendecim is higher than every other species (P < 0.001). “The differences in other pairwise comparisons were not significant.” So T. chilensis, with six lineages and about 12,000 cells, is not shown to differ from the single-lineage species.
- Hodgkinia:Sulcia ratios: about 1:1 with one lineage, 2.4:1 with six, 11.2:1 in M. septendecim. Within a species, Hodgkinia exceeds Sulcia significantly in T. chilensis (P = 0.018) and M. septendecim (P < 0.0001).
- Combining counts with the simulation: T. chilensis eggs should get 50 to 100 cells of the rarest lineage, about 80 by coverage; M. septendecim fewer than 50.
- Every egg, every lineage (Fig. 3).
- All Tettigades eggs carried all rpoB genotypes. All M. septendecim eggs carried both universally present etfD genotypes.
- Lineage proportions differ between egg nests (T. chilensis: F₁₀,₆₈ = 33.88; T. limbata: F₅,₃₇ = 30.16; both P < 0.001), and eggs of one nest cluster together. The rare lineage 5 of T. limbata ranged from 0.25% to 2.72%.
- Mothers “seem to avoid substantial genotype abundance shifts” when transmitting, “at least in the single generation we measured”.
- Amplicons found variation the metagenomes had missed: one-base variants in T. chilensis, and in M. septendecim many minor etfD genotypes, present in different combinations in each adult and nest.
- The route (Fig. 4). By light microscopy, transmission in T. lacertosa (3 lineages) and M. septendecim looks the same, and matches Buchner’s early description. Hodgkinia leaves the syncytium in “large, nucleated subcellular compartments”; Sulcia leaves from peripheral bacteriocytes. Both cross the follicular cells and gather in an invagination of the oocyte, the “symbiont ball”. In Magicicada, more cells migrate.
- Comparison with other insects (Table 1). Section-based estimates for symbiont balls of many hemipterans. Buchnera in eggs of three aphids, including Acyrthosiphon pisum at 1,872 ± 524, from earlier work. T. lacertosa 11,895–18,314 and M. septendecim 55,071–84,787 (both symbionts together), from this study. The M. septendecim figure is 1.7 to 2.6 times the FISH-based total of Fig. 2 (about 32,700), so the two methods disagree in absolute terms. That comparison is mine; the paper does not make it.
Claims
| id | claim | strength | support |
|---|---|---|---|
| C1 | As Hodgkinia lineages multiply and grow uneven, random transmission needs many more cells per egg to keep every lineage | moderate | simulation under stated assumptions; the 1-cell rule and one-circle-one-cell make the size an upper bound |
| C2 | Cicadas with more complex Hodgkinia transmit more Hodgkinia cells per egg, up to about 6-fold, but not more Sulcia | moderate for M. septendecim, weak as a trend | four species, five or six eggs each; only M. septendecim differs significantly; one species per complexity level, with no phylogenetic control |
| C3 | All eggs receive all Hodgkinia lineages | moderate | marker genotypes in all sampled eggs of three species; the authors say occasional loss of rare lineages could escape detection |
| C4 | The transmission mechanism is unchanged at the resolution of light microscopy | moderate | sections of two species, compared with Buchner’s description |
| C5 | The raised titre is a host adaptation driven by Hodgkinia’s splitting, and a cost of it | weak | inferred from C2 and Sulcia’s stability; the reverse order (high titre first, enabling splitting) is argued unlikely, not excluded |
| C6 | Larger transmitted populations could weaken bottlenecks, reduce host control and feed further splitting | speculative | the authors’ own word: “While speculative” |
Concepts
- Lineage splitting or fragmentation: one ancestral Hodgkinia lineage becoming several, in separate cells, each with “a subset of the original gene set”, all needed for the ancestral set of nutrients.
- Complexity: here, the number of Hodgkinia lineages and the unevenness of their abundances.
- Symbiont ball: the packet of Sulcia and Hodgkinia cells in an invagination at the base of the oocyte.
- Titre: the number of symbiont cells transmitted to each egg.
- Three solutions the authors name: the host sorts lineages actively (1); the host sends more cells (2); the host accepts some inviable eggs (3). They find evidence for 2, none for 3, and cannot test 1.
- Surplus: single-lineage cicadas already send “substantially more Hodgkinia cells than strictly necessary”, which the authors think buffered the first splits.
Connections
The splitting itself comes from the authors’ earlier papers: Tettigades (refs 19, 22: LIT-1563, LIT-1586) and Magicicada (refs 20, 21: LIT-1612, LIT-1602), all filed and all Unfetched. Background stated here from them: Tettigades has one to six lineages; the seven Magicicada species have “possibly dozens”; splitting began within the last about 4 million years in Tettigades and over 5 to 20 million in Magicicada. The authors place their result among other cases of host compensation: nuclear genes tracking fast mitochondrial evolution in plants and primates, horizontal transfer of bacterial genes to the host nucleus, and trafficking systems between host and symbiont. They end on symbiont replacement as another escape from a degrading symbiont. The claim that small symbiont populations, kept small by bottlenecks, weaken selection among symbionts is cited to refs 48–51 and is not tested here.
Bearing on the record
- CASE-177. The case’s last bullet and summary come from this
paper’s abstract. Corrections:
- “Eggs still receive the full set, because hosts with more complex Hodgkinia transmit up to six times as many cells.” The full set was tracked by one or two marker genes per species. The “because” is the authors’ inference. The six-fold rise is one species, M. septendecim; the six-lineage T. chilensis did not differ significantly from single-lineage species.
- “Lineages differing in abundance by over a hundredfold” is from the authors’ earlier metagenomes (114-fold in rpoB), not measured here.
- In Magicicada, genomic circles are not shown to be cell lineages. The simulation assumes they are, and the authors suspect that some circles share cells.
- “The host’s response raises reproductive cohesion (R)” overstates it. The host keeps co-transmission complete as the parts multiply. It maintains R; it does not raise it. And the authors suggest that the same response may lower host control (C) and raise selection among symbionts (S).
- QUESTION-277. Host compensation, the route the exchange named
without a source, is here a change in one number, the titre, with the
mechanism unchanged. It moves:
- R: kept, not increased, by the host;
- D: increased, since there are more essential parts;
- G: distributed further among symbiont lineages (from earlier work);
- S and C: possibly worsened, on the authors’ speculation that larger populations mean weaker bottlenecks and “decrease the level of control that the host can exert”. If so, compensation that saves R erodes S, and the dimensions trade off. The inference from the authors’ speculation to the dimensions is mine.
- THEORY-1049. The authors write that “even the largest Hodgkinia genome no longer encodes the machinery to make its own membranes, the host must define Hodgkinia’s envelope”, and draw from it that the host could, in principle, sort lineages. That is the tipping-point argument turned toward control of transmission. Nothing here tests it.
- THEORY-1051. The model’s one-symbiont collectives cannot hold a split symbiont. But this paper reports a related quantity: mothers pass their own lineage proportions to their eggs, so the composition of the symbiont community is inherited at the host level. The comparison is mine.
- CASE-173. Table 1 gives Buchnera counts in aphid eggs (886 to 8,223 across three species, from earlier work), far fewer than in cicadas.
- Filed as THEORY-1106. No instruction for machine-learning practice.
Limitations
- Four species for the comparison, one per complexity level, five or six eggs each, with no phylogenetic correction. Two of the four, the single-lineage species, are tens of millions of years apart. The trend across complexity is one significant species.
- Direction of causation is inferred. The authors consider that a high titre came first and enabled splitting, and argue against it only because some cicadas keep an unsplit Hodgkinia.
- Absolute counts are rough. The authors say so (cell shape, egg age). Table 1’s section-based estimate for M. septendecim is about twice the FISH estimate.
- Lineage presence is shown by markers, and the authors grant that rare lineages “are occasionally lost in certain eggs and that we lack the sensitivity to detect it”.
- Active sorting (solution 1) is untested. The data cannot exclude it. Lineage-proportion variation tolerated across a population shows only that any sorting “is not highly accurate”.
- No data on adult symbiont numbers. The authors’ view that complex cicadas hold larger adult populations is “anecdotal”.
- Slips. One sentence reports the Hodgkinia count “higher in M. tredecim than in any other species”, where the species counted is M. septendecim. The methods print the same sequence for both etfD primers. The reported 11.2:1 ratio differs from the ratio of the reported means (11.7:1), perhaps a mean of per-egg ratios. “∼80” cells of the rarest T. chilensis lineage does not match 0.8% of about 12,000 (96).
Open questions
- Do hosts sort Hodgkinia lineages at all? Single-cell tracking of lineages through the ovariole would show it.
- Does the titre rise smoothly with the number of lineages, or only in Magicicada? More species, with phylogenetic control, would show it.
- Did a high titre precede splitting? Titres in cicadas just before and just after a split, in a clade with several independent splits such as Tettigades, would separate the two orders.
- Do complex cicadas hold more Hodgkinia as adults, and at what cost to Sulcia and to egg number?
- Is the positive feedback real: does a larger transmitted population raise Hodgkinia diversity within hosts, and the rate of further splitting?