What are the key innovations that led to the evolution of multicellularity, and what were their precursors in the single-celled microbial life that existed before the metazoa? We can hypothesize at least two distinct kinds of features that had to have preceded true multicellularity.
-
The obvious feature is that cells must stick together; specific adhesion molecules must be present that link cells together, that aren't generically sticky and bind the organism to everything. So we need molecules that link cell to cell. Another feature of multicellular animals is that they secrete extracellular matrix, a feltwork of molecules outside the cells to which they can also adhere.
-
A feature that distinguishes true multicellular animals from colonial organisms is division of labor — cells within the organism specialize and follow different functional roles. This requires cell signaling, in which information beyond simple stickiness is communicated to cells, and signal transduction mechanisms which translate the signals into different patterns of gene activity.
These are features that evolved over 600 million years ago, and we need to use a comparative approach to figure out how they arose. One strategy is to pursue breadth, cast the net wide, and examine divergent forms, for instance by
comparing multicellular plants and animals. This approach leads to an understanding of universal properties, of how general programs of multicellular development work. Another is to go deep and examine closer relatives to find the step by step details of our specific lineage, and that's exactly what is being done in a new analysis of the choanoflagellate genome.
So what is a choanoflagellate? They are members of a diverse and common group of single-celled eukaryotes that possess a flagellum for motility and a collar of slender processes called microvilli that it uses to capture bacterial prey. It's a very successful lifestyle that has allowed them to flourish in both marine and freshwater environments.
Here's the connection to multicellular animals, the metazoa: choanoflagellates are markedly similar to the choanocytes of sponges. These are cells lining the interior channels of the sponge, which beat their flagella to propel water through the animal, and use the microvilli to filter out food particles. Right away, we can see an adaptive reason for the evolution of multicellularity: ancestral choanoflagellate-like organisms that teamed up could more efficiently filter water to extract food. The question here is the identity of the specific molecules they used to form early colonies.
King and others have sequence the entire genome of Monosiga brevicollis and compared what they found there to similar genes in other organisms. The genome itself is about 41.6 megabases, and contains approximately 9,200 genes (about half of what is present in humans). These genes were then compared to those of a suite of organisms to sort out what was held in common with the multicellular animals, the metazoa, and what was different from our distant cousins, the fungi and plants.
So what did they find? Choanoflagellates have a surprisingly rich repertoire of cell adhesion molecules, with many members of families of genes that the metazoa also use. They have at least 23 cadherin genes; cadherins are calcium-dependent cell adhesion molecules that are not found in other multicellular organisms like fungi and plants, and are present in animals where they are used for essential processes in development like cell sorting and polarization, and in regulation of the morphogenetic movements of sheets of tissues … and there they are in the choanoflagellates as well. While some species of choanoflagellates will form clusters and at least transient colonies, M. brevicollis is not known to make such associations, so the function of these molecules in these particular organisms is a bit mysterious. They also contain integrin-α genes and genes with immunoglobulin domains — while you may be familiar with immunoglobulins as key proteins of the vertebrate immune system, the immunoglobulin motif is also a more general cell adhesion domain that is also found in many cells of the nervous system.
While these proteins that metazoans use to mediate interactions between cells are exciting to find in a choanoflagellate, and while their presence opens up new questions about their function, there's another class of genes that are even more peculiar to find in a single-celled organism: genes for proteins that bind to the extracellular matrix. These are important in animals like us; we construct layers of extracellular matrix proteins during our development that are contained within our bodies, and cells bind to them and take advantage of them in embryogenesis. What do choanoflagellates use them for? They may be important in substrate attachment, or possibly these organisms secrete a more complex suite of molecules into their environment than is known.
So we see some remarkable homologies between choanoflagellates and metazoans in the genes that mediate cell adhesion and adhesions between cells and a matrix of molecules in the environment — our single-celled ancestors first built up a collection of tools to make them sticky, a cookbook of glue molecules that would later enable more sophisticated patterns of attachment to one another. The table below also shows that the choanoflagellates and their last common ancestor with the metazoa also evolved some common transcription factors, or gene regulators.
Note: sticky proteins and transcription factors are not unique to choanoflagellates and metazoans — bacteria, plants, fungi, etc. all also have them. What this work is showing is that the choanoflagellates and metazoa share an idiosyncratic, special set of sticky molecules and transcription factors.
At the beginning of this article I said that there were two properties essential to multicellularity: adhesion and signaling. Choanoflagellates have the molecular precursors needed for metazoan-style adhesivity, but they lack metazoan-specific signaling pathways. This makes sense. A general property like adhesion may well have utility to a single celled organism, but the specific pathways that would trigger region- and tissue-specific differentiation would be a later innovation.
Even in the case of these unrepresented elements of the metazoan genome, though, we see premonitions in the choanoflagellate. While complete, recognizable homologs of important signaling genes like hedgehog and
Notch are not found, fragments of them are found scattered about. They didn't appear out of nowhere — rather, there was a process of domain shuffling during metazoan evolution that built new signaling molecules by recombining elements present in the ancestral genome. So, while choanoflagellates reveal that the ancestor almost certainly did not have a true Notch gene, we can find 3 genes that contain pieces of Notch: one has the EGF domain, another the NL domain, and another the set of ankyrin repeats … and it's easy to see that the Notch gene was not generated ex nihilo, but was assembled by splicing bits of pieces of extant genes into a novel protein.
Transitional fossils always get all the attention — and you've got to admit, a new collection of old bones is a sexy thing, an arresting attention grabber that has a lot of visual appeal. I think the more powerful modern evidence for evolution, though, are these examples of molecular transitions in which we can reconstruct the details of ancient changes. Not to belittle the fossil evidence, but changes within a single narrow lineage within a single phylum aren't quite as dramatic or impressive as the kind of radical evolutionary event we see here — not just the reshaping of a femur, for instance, but the acquisition of whole new capabilities, the novel potential to build a femur in the first place. This is big stuff, a peek into core innovations that led to insects and jellyfish and grasshoppers and snails and cows, and that are held in common among all of us.
Yet despite the magnitude of the potential evolutionary consequences, we can also see revealed the mechanisms underlying them, and that they are small and simple changes, an expansion of capabilities present in miniscule, single-celled creatures. When evolving such fundamental and revolutionary features as multicellularity is such a patently feasible and explainable event, it does seem absurd that some people can still question relatively minor transformations, such as between varieties of ape.
King N, Westbrook MJ, Young SL, Kuo A, Abedin M, Chapman J, Fairclough S, Hellsten U, Isogai Y, Letunic I, Marr M, Pincus D, Putnam N, Rokas A, Wright KJ, Zuzow R, Dirks W, Good M, Goodstein D, Lemons D, Li W, Lyons JB, Morris A, Nichols S, Richter DJ, Salamov A, Sequencing JG, Bork P, Lim WA, Manning G, Miller WT, McGinnis W, Shapiro H, Tjian R, Grigoriev IV, Rokhsar D. (2008) The genome of the choanoflagellate Monosiga brevicollis and the origin of metazoans. Nature 451(7180):783-788.
"What are the key innovations that led to the evolution of multicellularity, and what were there precursors in the single-celled microbial life that existed before the metazoa?"
Do you mean THEIR precursors?
Fascinating.
As an aside, this flagella talk reminds me that yesterday I noticed Behe's Edge of Evolution is now on the Bargain Book page at Amazon. You can pick one up for like six bucks. Not exactly flying off the shelves, I would assume.
Creationists declare frontloading in 5...4...3...
Fascinating and well-written. When your article gets to the part about "premonitions (of signalling) in the choanoflagellate" (my parenthesis), it felt like the rush to the buried treasure at the end of a good story.
As a sometime paleo-artist, I agree, fossil evidence is sexy, but in a glimpse-of-a-well-turned-ankle kind of way. This type of modern evidence is more like thigh-high boots. Amazing.
Great stuff.
A cool, if kinda random and chaotic, experiment would be to transform M. brevicollis with modern-day Notch and Delta and see what, if anything, happens.
Amazing stuff!
We are so lucky to be able to recognise such a "transitional form" mechanism still being used in living organisms so long after metazoans used the same trick to form multicellular organisms so long ago.
"-- not just the reshaping of a femur, for instance, but the acquisition of whole new capabiiities, the novel potential to build a femur in the first place."
I look forward to reading up on this paper, in part because I teach a module on these organisms. However, I have a couple of points of contention.
The obvious feature is that cells must stick together; specific adhesion molecules must be present that link cells together, that aren't generically sticky and bind the organism to anything. So we need molecules that link cell to cell. Another feature of multicellular animals is that they secrete extracellular matrix, a feltwork of molecules outside the cells to which they can also adhere.
First, many bacterial cells have specific adhesion molecules to link self-to-self or self-to-us or self-to-other. Also, many if not all the bacteria can form a biofilm of extracellular polymer. So while they are simpler, bacteria seem to be capable of, at least, pseudo-multicellular life based on these criteria.
A feature that distinguishes true multicellular animals from colonial organisms is division of labor -- cells within the organism specialize and follow different functional roles. This requires cell signaling, in which information beyond simple stickiness is communicated to cells, and signal transduction mechanisms which translate the signals into different patterns of gene activity.
Again these paradigms have been defined in the bacteria, one obvious example being the quorum sensing signaling systems. Clearly, the bacteria lack some of the metazoan-specific systems (by definition), but the fundamental machinery necessary for multi-cellularity existed and was used similarly long before the animalia.
It would have been more surprising if these things were lacking from the choanoflagellate genome rather than that they are there. One thing this paper does allows us to do is see how deep into the evolutionary shrub these things are found. None of my issues effects the gist of the paper or PZ's post, but I wanted to point out an underlying-metazoist perspective often found in there types of genome papers and the discussions therein.
The figure showing how to build Notch from a kit of parts is quite simply one of the more arresting I have ever seen. I worked on the regulatory regions of mouse Myf5 where metazoan evolution is written on the genome with discrete regulatory elements coopted to drive gene expression in each new domain. This with coding sequences is exactly the same. Wonderful.
One question/point, I found the lack of any sponge sequences in the trees curious, were any present elsewhere in the paper?
Yeah, notice the paragraph beginning "Note: sticky proteins and transcription factors are not unique to choanoflagellates and metazoans — bacteria, plants, fungi, etc. all also have them". This paper describes not the evolution of adhesion proteins, signaling mechanisms, etc. in general, but the history of that subset of the genes for such proteins that are used in the metazoa,
No, nothing much is said about sponges in the paper. I wondered about that, too -- that would have been an interesting column in that table of genes above.
Awesome post! There's a couple of mutated, non-coding regions though:
If PZ ever stops blogging on peer-reviewed research, I'll quit the internet.
Although, apparently, I should stop trying to write these things at 1am (but that was the only free time I had this week!)
PZ, you mention two capabilities required for multicellular life...but isn't there a third, the ability to transport oxygen and nutrients to cells on the interior? (And CO2 and wastes away?)
I think that sponges overcome this by having thin layers of cells and porous spaces, so that no cell is too far from water...but, that may be just a fuzzy memory.
Outlier said:
This is only an issue for multicellular organisms whose bodily arrangements prevent every cell from having access to the exterior environment for it's own gas exchange/ feeding e.t.c which is not the case if an organism is very small or has a very large surface area.
Ofcourse when you get complicated and internally bulky enough for this not to be feasible the need for oxygen and nutritional transport systems arise but this is not a necessity for a multicellular organism per se.
Wow. Very cool.
I've just been reading _Endless Forms Most Beautiful_ (don't give me any spoilers--I'm only half way through) and the question of where all the switch-binding proteins *come* from has been eating me alive with curiosity. Of course, by "where they come from" I also want to know what they did there--they weren't just hanging around waiting for the curtain to go up on multicellularity. So this article is a tantalizing glimpse, but now I want *more*!
Great post, and an awesome example of the power of comparative studies to enlighten.
Quick note, though. Anyone notice that one of the authors of the study is "JGI Sequencing". That's a reference to the DoJ's Joint Genome Institute. Could "JGI Sequencing" have read and approved the paper, or defend it in public? I think not. What makes it an author? Individuals from JGI are already included in the author list.
That is "DoE's Joint Genome Institute", not DoJ's.
Cat Faber:
Sorry to ruin it for you, but it turns out that God is in fact nothing more than a cis regulatory element, and that is why things are so beautiful.
I know, crazy isn't it?
;)
Wow, that's small for a eukaryote. Escherichia coli is 1 x 1 x 3 µm.
Would be stupid of them -- Monosiga, existing today, is not one of our ancestors. But I suppose that's your point... :-)
No, it's a different kind of stupid. See, the common ancestor of Monosiga and humans and all other eukaryotes contained all the information for making all the descendant species (sort of like homunculi in sperm), since information can only be lost, not gained, by natural processes.
caveat: this question is really naive, not snarky nor trying to give fodder to jerks or something.
Did I miss something that rules out the hypothesis that these animals, rather than being a sister clade to sponges, are descended from sponges, such that they could have degraded adhesion and signaling rather than precursor?
Oh. Yeah.
Yes, a few molecular phylogenetic analyses for example.
Oops -- they aren't the sister-group of the sponges, we are. Choanoflagellates and animals are sister-groups or nearly so.
Can someone recommend some good books on the early evolution of cells? I'm curious as to how this stuff works.
Obviously PZ hasn't read (Han and Warda, "Multicellularity, the missing link between body and soul: Proteomic prospective evidence." Proteomics, January 23, 2008)
I learned in high school that we had cells with microvilli lining our intestines to increase surface area for absorbtion. Any connection to the microvilli of choanoflagellates?
In response to why there are no sponge sequences in the figures - its because no assembly of the forthcoming sponge-genome has been made public. To date, only EST sequences are available. It will be very interesting to get a closer look at sponge signaling, adhesion and transcription factor genes, but remember that choanoflagellates are not more closely related to sponges than to any animal. However, my money says that sponges have retained many ancestral genome characteristics, at least relative to some model systems.
In response to #26 (Monado) - nobody knows whether the morphology of intestinal epithelial cells is directly homologous to choanoflagellates and the feeding cells of sponges (choanocytes), but the prospect is certainly exciting. Ultimately, ALL animal cells are derived from, and therefore homologous to, choanoflagellate-like cells. Furthermore, to my mind (and this is my field of study), all animal epithelia are probably homologous with the rudimentary epithelia seen in sponges (which include choanocyte cells), adding another layer of homology between the two.
Mark @ 21:
I'm wondering the same thing. It seems possible that both modern sponges and choanoflagellates descended from a common sponge-like ancestor. Which could mean that Notch genes were essentially scrambled over time in the choanoflagellates as they were no longer needed (of course, I have no idea whether or not sponges express Notch). Plus, I like the idea of selection and/or genetic drift resulting in a return to unicellularity. Being multicellular isn't always better!
Either way, it's always cool to see gene conservation across such disparate species.
As I understand it, multicellularism took a long time to arise. So something was difficult to achieve.
...or "we" would be, if the "sponges" weren't turning out to be so damn paraphyletic. We can be the sister-group to calcareous sponges, how's that?
I think this blogging on peer-reviewed research may be one of the best things y'all at scienceblogs (or whomever came up with it) have done.
One of the things that came out of a conference I was a part of a couple years ago--with regard to social movement scholarship and social movement activity--was the lack of connection between practitioners and researchers. People just often don't know what those of us engaged in research enterprises do. I often view my work as a teacher as providing a road map to the broader field of investigation (I use the metaphors of either tour guide or translator to describe what I do), but I love the ways this new communications technology is being adopted to fill what is often a split between the research world and communication of technical knowledge to broader publics.
Plus, as someone in one of those broader publics, whose speciality lies elsewhere, this shit is just fun to learn.
Scott thanks for the clarity on the sponge genome issue, it nicely explains the obvious absence of those data, they simply weren't available.
As for epithelial homology, my brain vaguely remembers reading something about how old the epithelial junctions like tight and gap are and it was pretty old. The choanoflagellates have obviously got cell polarity sorted, so no need to evolve that post epithelialisation.
Wisaakah #28, if choanoflagellates had notch then lost it then there would likely be a relic pseudo gene in the genome which the screen for Notch sequences would have found before the elements in other genes. We see pseudo genes in all sorts of organisms and they are a problem because it can be tough to prove that they are still pseudo and not active. They are particularly common after genome duplications, ask the zebrafish people ;-)
Peter Ashby, I do not agree that screening for Notch would likely pick up relic pseudo genes. For that to happen, the gene would have to have been "lost" relatively recently. Also, as you said, such genes are common after genome or gene duplication events where some redundancy has been introduced. This may be a case where absence of evidence IS most likely evidence of absence.
"That is "DoE's Joint Genome Institute", not DoJ's.
Oh good! I was getting worried for a moment, given current politics.
"As a sometime paleo-artist, I agree, fossil evidence is sexy, but in a glimpse-of-a-well-turned-ankle kind of way. This type of modern evidence is more like thigh-high boots."
So, Tiktaalik's kinda like Jane Austen, while the choanoflagellate genome is kinda like the Pussycat Girls? Huh.
That's awesome, though - fossil evidence is occasionally a glimpse of a well-turned ankle. Ha!
Rightsaid, that depends on the searching doesn't it? Think about it, to search for Notch in that situation you do not search the naked nucleic acid sequence. At that level what is conserved is the amino acid sequence and that is degenerate at the dna level. This change inevitably means you will pick up any dead pseudo genes, and remember pseudo genes change only randomly.
What a great post. But I must agree there could be more attention paid to discussion of the "choanoflaggelates are unicellular descendants of sponges" issue. It is an obvious concern.
Also -
It strikes me that this may be wrong. I've never heard of a nucleated eukaryotic cell that was much less than 10 microns in approximate diameter. A human RBC without a nucleus is crudely about 7 microns. This should be checked.
No, the scale bars are correct. The cells are very small. Choanocyte cells of sponges are equally small (maybe even slightly smaller).
As for choanoflagellates being derived from sponges - all of the current evidence suggests that they are not. Phylogenetic analyses strongly support them as the sister group to metazoa, so they are not derived within any modern group of sponge. As to whether they could have still had a sponge-like ancestor - sure, but it is not the most parsimonious explanation. Furthermore, there are several other unicellular lineages clustered around choanoflagellates on the tree of life, suggesting that the ancestor of choanoflagellates and animals really had a free-living unicellular life history stage, even if it was capable of forming colonies like some modern choanoflagellates.
Note that sponges develop much like other animals and have a very animal like gene compliment. The real story from the choanoflagellate genome is that they aren't very animal-like, but they provide insight into the origin of a few important animal genes.
Yes, or even to a part of the calcareous sponges, the Homo...basalomorpha they are called, I think. We're living in interesting times!
There are plenty. The record is a spherical marine green alga 0.8 µm in diameter -- contains not only a nucleus (with a single pore), but also a chloroplast and a (single) mitochondrion. I'll post a link later today. Or google "picoplankton". Brown tides consist of tiny eukaryotes, for example...
Yup, plenty of protists that are barely bigger then the bacteria they eat (in terms of some bacteriovores). Many people in my lab work on organelle evolution in wonky protists and all of the typical ways of trying to isolate mitochondria for instance just don't work.
For anyone interested in this sort of thing (Not what I myself work on) you can check out:
A phylogenomic investigation into the origin of Metazoa.
Ruiz-Trillo I, Roger AJ, Burger G, Gray MW, Lang BF.
Mol Biol Evol. 2008 Jan 9 [Epub ahead of print]
Monosiga ovata was one of the protists used in the EST based analysis dealing with some of metazoans close protistan relatives.
NCBI Link
Peter: I was thinking of the separated domains as a type of pseudogene, though you're right, that doesn't really make sense. I still think it's possible that an obsolete Notch gene would not remain as a pseudogene, depending on how long ago they split off. But in light of the presence of several other related unicellular lineages (re Scott @ 38), I think I'll have to give up my rather uninformed hypothesis!
Are there any known examples of a unicellular descendent from a multicellular ancestor (other than critters with a unicellular stage in their lifecycle)?
Yeast: secondarily unicellular ascomycetes. Some yeasts, like Candida, retain the linear (hyphal) organization, though.
-----------------
Contrary to my assertion, Ostreococcus tauri has several mitochondria. The website I was thinking of, which has nice pictures, isn't among the first 50 Google results anymore.
Brown tides are composed of pelagophytes that are apparently up to 3 µm in diameter.
BTW, I have made the rather stunning experience that there are molecular biologists who sometimes talk of "higher eukaryotes" and mean "everything except yeast"/"plants and animals". They don't know what they are talking about. The term is positively misleading.
Thanks David - I should have known that about yeast. I even get to grade an exam question on yeast genetics (the joys of being a TA)!
Thanks for answers to my naive question, although I'm still a bit confused... maybe I should grab the original paper instead of continuing to question, but...
Isn't it sort of weird that the open dot an the cladogram has only multicellular descendants and siblings except for Mbre? In fact, all the things I'm familiar with in the cladogram sound multicellular, except for slime molds (which are "sorta" multicellular) although I didn't look up the fungi.
#22, it sounds like you know this stuff in detail, what do the "molecular phylogenetic analyses" show? That "open dot" was something more like a slime mold than a primitive metazoan like a sponge, and that plants and fungi developed multicellular assembly in some sort-of-convergent way (albeit using some of the same stickies)? Although I don't doubt that there was a unicellular ancestor, I'm not really clear on why the choanoflagellates are believed to be representative of that ancestor.
Of course, it's known that there were colonial bacterial mats that presumably involved adhesion (I have no idea about signaling) quite a bit earlier (before eukaryotes? I thought so, but I could easily be wrong.)
Are there whole hosts of eukaryotic unicellular things that have been shown by other molecular means to be closely related to choanoflagellates that just aren't shown on the chart, but are siblings around the open-dot area (and, for that matter, other unicellular critters interleaved with all the other clades there)?
New, interesting, and relevant:
Evolution & Development
Volume 10 Issue 2 Page 241-257, March/April 2008
Six major steps in animal evolution: are we derived sponge larvae?
•Claus Nielsen
SUMMARY A review of the old and new literature on animal morphology/embryology and molecular studies has led me to the following scenario for the early evolution of the metazoans. The metazoan ancestor, "choanoblastaea," was a pelagic sphere consisting of choanocytes. The evolution of multicellularity enabled division of labor between cells, and an "advanced choanoblastaea" consisted of choanocytes and nonfeeding cells. Polarity became established, and an adult, sessile stage developed. Choanocytes of the upper side became arranged in a groove with the cilia pumping water along the groove. Cells overarched the groove so that a choanocyte chamber was formed, establishing the body plan of an adult sponge; the pelagic larval stage was retained but became lecithotrophic. The sponges radiated into monophyletic Silicea, Calcarea, and Homoscleromorpha. Homoscleromorph larvae show cell layers resembling true, sealed epithelia. A homoscleromorph-like larva developed an archenteron, and the sealed epithelium made extracellular digestion possible in this isolated space. This larva became sexually mature, and the adult sponge-stage was abandoned in an extreme progenesis. This eumetazoan ancestor, "gastraea," corresponds to Haeckel's gastraea. Trichoplax represents this stage, but with the blastopore spread out so that the endoderm has become the underside of the creeping animal. Another lineage developed a nervous system; this "neurogastraea" is the ancestor of the Neuralia. Cnidarians have retained this organization, whereas the Triploblastica (Ctenophora+Bilateria), have developed the mesoderm. The bilaterians developed bilaterality in a primitive form in the Acoelomorpha and in an advanced form with tubular gut and long Hox cluster in the Eubilateria (Protostomia+Deuterostomia).
It is indicated that the major evolutionary steps are the result of suites of existing genes becoming co-opted into new networks that specify new structures.
The evolution of the eumetazoan ancestor from a progenetic homoscleromorph larva implies that we, as well as all the other eumetazoans, are derived sponge larvae.
Wouldn't the next obvious step be to look at the signalling proteins in Metazoa that are only loosely multicellular (i.e. colonial animals)?
I mean, specifically, look at the genomes of coral, the Man'O'War, and other colonies.
Many thanks for this and the other comments on eukaryotic cell size.
That is absolutely fascinating considering that a large virus like the smallpox virus can about 300 nm (0.3 micometers) (I'm sure someone will mention a larger one), let alone bacteria. There is almost a zone of overlap around 500nm to 1 micron where one can perhaps see large viral particles and tiny eukaryotic cells, as well as, of course, many typical prokaryotic cells.
I was trained to think of prokaryotic and eukaryotic cells as more or less dichotomous in size. This information is interesting. It makes me wonder how large the earliest nucleated cells had to be. Of course, that's a whole different topic.
Of course, the vast majority of types of eukaryotic cells are much, much larger than 1 micron. Whether the majority of eukaryotic cells are larger than that depends on how many of those algal cells there are out there :-).
In fact this information was retrievable with google, although I somehow missed it the first time.
In fact, Outlier (#13), Ediacaran fauna seem to have developed a flattish, "quilted" body plan that avoided the need for internal circulation (as far as we can tell). IOW, they're pretty flat. They are distingished by to my layman's eye mainly by their axes of growth.
Wow, that's small for a eukaryote. Escherichia coli is 1 x 1 x 3 µm.
Would be stupid of them -- Monosiga, existing today, is not one of our ancestors. But I suppose that's your point... :-)
Oh. Yeah.
Yes, a few molecular phylogenetic analyses for example.
Oops -- they aren't the sister-group of the sponges, we are. Choanoflagellates and animals are sister-groups or nearly so.
Yes, or even to a part of the calcareous sponges, the Homo...basalomorpha they are called, I think. We're living in interesting times!
There are plenty. The record is a spherical marine green alga 0.8 µm in diameter -- contains not only a nucleus (with a single pore), but also a chloroplast and a (single) mitochondrion. I'll post a link later today. Or google "picoplankton". Brown tides consist of tiny eukaryotes, for example...
Yeast: secondarily unicellular ascomycetes. Some yeasts, like Candida, retain the linear (hyphal) organization, though.
-----------------
Contrary to my assertion, Ostreococcus tauri has several mitochondria. The website I was thinking of, which has nice pictures, isn't among the first 50 Google results anymore.
Brown tides are composed of pelagophytes that are apparently up to 3 µm in diameter.
BTW, I have made the rather stunning experience that there are molecular biologists who sometimes talk of "higher eukaryotes" and mean "everything except yeast"/"plants and animals". They don't know what they are talking about. The term is positively misleading.
As to size variations in eukaryotes: 20 malarial merozoites will fit into one red blood cell, which is a relatively small cell, itself.