Electron microscopy reveals cross-kingdom affinities in eukaryotes and roots the Metazoa and Metaphyta (and Fungi)
In fact, the data for a major shake-up were just becoming available (Taylor, 1976, 1978). Beginning in the late 1950s, Irene Manton pioneered the application of the transmission electron microscope (TEM) to study the ultrastructure of photosynthetic protists, and she was soon joined by Dorothy Pitelka (heterotrophic protists) and others more concerned with the techniques themselves. A whole new, rich dataset became available for comparisons across all the eukaryotic group boundaries and, thanks to the observation of slices through many individuals simultaneously, it was possible to get a strong sense of the degree of variability to be expected in each. It also established the homology of many structures disguised under different names. For example, cilia and flagella were found to be identical in basic structure, with an extraordinary conservatism of the ‘9+2’ microtubular arrangement of basal bodies, and centrioles were clearly homologous with them. Bacterial flagella were shown to be not only much smaller but fundamentally different in structure, composition and function. The eukaryotic structures have historical precedence for the name: Flagellaten was a major protozoan cell type ever since Bütschli (1880–1889), the Flagellata being one of four major Superclasses of the Protozoa in Grassé (1952), subdivided into phytoflagellates and zooflagellates. ‘Flagellates' and ‘flagella’ have prevailed, despite attempts to replace the name of the eukaryotic structures with ‘undulipodia’ (all papers by Margulis since the 1970s) or the use of cilia for all of them (Hülsmann, 1992; all papers by Cavalier-Smith over the same period). If ‘flagellum’ is ambiguous, it is no more so than ‘cell’, and the context should leave no doubt as to the structure referred to.
The basic structures of nuclei, mitosis, mitochondria and chloroplasts were soon shown to be essentially similar in all eukaryotes, both protists and multicellular organisms, but differed in details, some of which were group specific, e.g. internal or external spindles (the former apparently being earlier in origin). Golgi bodies and dictyosomes were shown to be fundamentally the same in animal and plant cells, respectively. At the same time, a whole new fine-structural dataset became available for comparative purposes that covered the whole, or most, of the range of eukaryotes.
The outcome of the application of these data, reviewed by Taylor (1994, 1999) and Patterson (1999, 2000) among others, was no less than a revolution in our view of the relationships of the ‘lower eukaryotes’. Many long-standing categories, such as Algae, Fungi, Protozoa, amoebae, heliozoans, Phytoflagellates, Zooflagellates, etc., were shown to be polyphyletic. As a result of the protistological approach, ignoring the old boundaries, new probable relationships were recognized, such as euglenoids with trypanosomes, based initially on their paddle-shaped cristae (Taylor, 1976), but soon reinforced by several other features (Kivic & Walne, 1984). Ciliates were linked with dinoflagellates on the basis of their cortical structural similarities (Taylor, 1976; suggested also by Corliss, 1975). Oomycetes resembled xanthophytes and other protists with compound flagellar hairs (Taylor, 1978). Corliss (1986) dated the ‘re-emergence of the field [protistology] as a respectable interdisciplinary area’ from the mid-1970s, the same period in which the International Society for Evolutionary Protistology (ISEP) was founded (see below).
As lineages became clearer, it was also realized that ultrastructural characters might help to distinguish between competing hypotheses for the protistan roots of plants and animals. It had long been evident that the chlorophytes were the most likely candidates for the origin of plants, given the near-identity of biochemical characters (pigments, storage). Copeland even placed them within the same Kingdom Plantae. Ultrastructure cemented this, showing that that the chlorophyll a+b-containing plastids had only two membranes around them and that starch was located within the plastids, like plants. It even offered clues as to where within the chlorophytes (including prasinomonads) the divergence may have occurred, scaly flagella occurring in the sperm of mosses and ferns. Ultrastructure of choanoflagellates was consistent with the long-held view of them as animal ancestors, based on the resemblance to choanocytes of sponges, but a ciliate ancestry had been proposed by Hadzi and Hanson. Ciliates are strongly tubulocristate (Taylor, 1978). The flat mitochondrial cristae and flagellar roots of sponge choanocytes were consistent with the choanoflagellate origin hypothesis dating back to Saville Kent a century before (see Corliss, 1989b for a review).
In the mycological world, it was widely accepted that slime moulds (especially the acrasiids) and phycomycetes (flagellated ‘water moulds’) seemed to differ from other ‘fungi’ in a number of ways. Electron microscopy strengthened these differences, zygo-, asco- and basidiomycetes lacking any trace of ‘9+2’ structures, including flagella, and having flat mitochondrial cristae. The phycomycetes had tubular cristae with one notable exception: chytrids differed from the other phycomycetes, not only in having a posterior, smooth flagellum, but also in having flat cristae (Taylor, 1978). Molecular sequence comparisons (see below) eventually showed that chytrids were the nearest flagellated sister-group of the ‘true fungi’, the other phycomycetes, with tubular cristae, being linked to the stramenopiles/heterokonts. In another example, mitochondrial cristae can be used as a key criterion in the recognition of natural groups of amoebae, separating the heteroloboseans (discoidal cristae) from the gymnamoebae (tubular cristae, sometimes branching/ramicristate) of Rogerson & Patterson (2000). Time has only strengthened the surprising conservatism and consequent usefulness of crista-type within protist groups in indicating major clusters, such as discicristate, tubulocristate and platycristate protists (Gray et al., 1998; Taylor, 1994, 1999; Patterson, 1999). Well-known exceptions include the size reduction of the cristae in low-oxygen conditions, as in some trypanosomes. The position of cryptomonads continues to be troublesome, with some molecular evidence indicating the placement of cryptomonads (flat cristae) with the chromalveolates (tubular) (Fast et al., 2001; Yoon et al., 2002), although the use of plastid characters to place the ‘host’ phylogenetic position can be questioned. The position of Dictyostelium (tubular cristae) closer to animals (flat) than plants (also flat) (Bapteste et al., 2002) contradicts the monophyly of flat cristae. Time and the use of various criteria, including multiple genes and mitochondrial DNA, will tell whether these and others are undoubted exceptions to the rule.