Terminal Developments in Ediacaran Embryology

See allHide authors and affiliations

Science  23 Dec 2011:
Vol. 334, Issue 6063, pp. 1655-1656
DOI: 10.1126/science.1216125

Ever since Darwin there has been a disturbing void, both paleontological and psychological, at the base of the Phanerozoic eon. If his theory of gradualistic evolution be true, then surely the pre-Phanerozoic oceans must have swarmed with living animals—despite their conspicuous absence from the early fossil record. Thus, the 1998 report of fossilized animal embryos in the early Ediacaran Doushantuo Formation of South China (1) was met with almost palpable relief. It was indeed the fossil record that had let us down, not the textbooks, and certainly not the exciting new insights from molecular clocks. All was not as it seemed, however, and new data from Huldtgren et al. on page 1696 of this issue (2) look set to revoke the status of these most celebrated Ediacaran fossils.

Originally described as colonial green algae (3), the Doushantuo fossils were reinterpreted as animals based largely on the recognition of a developmental pattern involving serial cell division without accompanying size increase, a process known as palintomy (see the first figure). The comparison with early animal development was compelling enough to induce a flurry of follow-up studies, which yielded sightings of later-stage embryos and even adult metazoans (4). Such claims proved controversial, however, with particular concerns raised over the interpretation and biological origin of key diagnostic features (5).

In 2006 Hagadorn et al. (6) presented synchrotron-based tomographic evidence for nuclei within individual cells, although the authors also worried about the absence of a differentiated outer layer of cells (epithelium) on later-stage embryos, one of the hallmarks of extant animals. Perhaps these fossils were mere “stem-group” metazoans, yet to acquire the full suite of characters expected in the last common ancestor of living forms. The phylogenetic nadir came in 2007 with a report documenting the same size and arrangement of cells in the modern sulfur-oxidizing (and phosphate-concentrating) bacterium Thiomarginata (7). Not animals, not embryos, not even eukaryotes, it seemed.

Palintomic cell division.

“Embryo” fossils from the Doushantuo Formation of South China have been interpreted as the oldest known animal remains, but the key diagnostic feature—palintomic cell division—occurs widely within eukaryotes, including various “nonmetazoan holozoans” (2) and the volvocacean green algae. (A to E) “Embryo fossils” exhibiting serial palintomy (division without accompanying growth), beginning with a single large cell [the “embryo fossil” in (E) shows marked differences in cell sizes]. These specimens are all roughly 400 to 700 µm in diameter. (F to J) Palintomic cell division in asexual Volvox carteri f. nagariensis; note the appearance of reproductive gonidia beginning at the 64-cell stage (I). These specimens are all roughly 50 to 60 µm in diameter. The comparison between the fossils and modern Volvox is far from exact, but nonetheless reflects a similar grade of organization.

Morphological analogs.

There are close morphological similarities between Doushantuo acritarchs such as Meghystrichosphaeridium (A) and the encysted zygotes of various Volvox species (B), although the fossils are typically larger than the modern zygocysts. Intriguingly, the zygotes of V. aureus undergo palintomic cell division within the zygocyst (18), a pattern directly compatible to “embryo fossils” preserved within various Doushantuo acritarchs (8). The fossil in (A) was first described in (17); the images in (B) are reproduced from (14). Scale bar represents 75 µm for (A) and 30 µm for (B).


The story wasn't over, however, and reports of embryo-like fossils within the spine-bearing cyst-like walls of unambiguously eukaryotic microfossils (8) put animal interpretations firmly back on the rails; indeed, the record of Ediacaran metazoans could now be extended to include forms represented solely by these (much more common) cyst-like microfossils (9). It's at this stage that Huldtgren et al.'s study comes into the picture, offering a rather more sober interpretation for the Doushantuo “embryos.”

Hultdgren et al. once again document convincing eukaryotic nuclei within individual cells, but their key contribution lies in tracking the fate of these cells into later developmental stages. What they find is an ontogenetic trajectory entirely foreign to the Metazoa. Instead of developing a differentiated epithelium, the constituent cells simply continue to divide palintomically, giving rise to thousands of tiny cells—interpreted as reproductive propagules—while the overall structure develops local constrictions and outgrowths, possibly related to propagule release. Although unquestionably eukaryotic, the fossils are not metazoan, or even properly multicellular by all appearances.

So what exactly are they? Huldtgren et al. wisely refrain from making absolute claims, but are more than intrigued by a similar style of palintomic division and overall deformation exhibited by certain kinds of “nonmetazoan holozoans,” that mixed bag of mostly unicellular eukaryotes that evolved after the last common ancestor of animals and fungi, but before the last common ancestor of living (that is, crown-group) animals (see the first figure).

In terms of progressivist storytelling, this all seems a little too good to be true, particularly as the best morphological comparison appears to be with the obligate fish parasite Ichthyophonus (although this lacks anything comparable to the ornamented cyst-like walls of the fossils). It is conceivable that the palintomic cell division of metazoan embryos was inherited from ancestral holozoans, but—as Huldtgren et al. acknowledge—the much broader distribution of this habit undermines its utility as a phylogenetic marker: A substantial range of ciliates, dinoflagellates, and volvocacean green algae have independently evolved this same type of cell division (2) (see the first figure).

The absence of a close modern analog for the Doushantuo fossils calls for a more flexible consideration of phylogenetic hypotheses. In the case of Volvox carteri f. nagariensis, for example, a large asexual reproductive cell divides palintomically up to 12 times within a tough glycoproteinaceous vesicle (10), yielding a range of forms strikingly similar to those seen in the Doushantuo populations (see the first figure). Unlike the fossils, however, the embryonic stages of V. carteri retain multiple cytoplasmic bridges between sister cells (2, 10) and form hollow spheroidal colonies (11). Are these distinctions sufficient to rule out any affiliation between the fossils and green algae?

Given the almost trivial genomic distinction between multicellular V. carteri and its unicellular sister group Chlamydomonas (12), it is not difficult to imagine multiple evolutionary forays into this grade of organization. Indeed, Volvox is now known to have evolved from multiple unicellular ancestors (13), with a major subsection characterized by the absence of cytoplasmic bridges (14). The Volvox life cycle also involves various types of cellular differentiation, as well as the production of large, often ornamented zygotes with a distinctive three-layered wall (14). Morphological analogs for such stages are well documented in the Doushantuo fossil assemblage (see the second figure). Interpretation at this level is inevitably impressionistic, but to my eye there is still a case for identifying the Doushantuo fossils as embryos, albeit algal rather than animal.

Wherever the Doushantuo fossils eventually end up, it will clearly not be within “crown-group” Metazoa. Does this then mean there were no early Ediacaran animals? Not at all. No fossil assemblage, however well preserved, provides a full account of past diversity, particularly when the local conditions are so extraordinary as to fossilize nuclei and other intracellular constituents. The “exceptional” fossil record is, by any measure, woefully unrepresentative and incomplete (15). When it comes to assessing the first appearance of early, difficult-to-preserve animals, the most reliable signal will be found in the conspicuous coevolutionary features that they have induced in other organisms with higher preservation potential (16). Early Ediacaran microfossils—with or without included “embryos”—may no longer include animals, but their dramatic radiation clearly marks the arrival of this revolutionary new clade.


View Abstract

Navigate This Article