Introduction

Palombiella stephensoni (Palombi, 1938) is a marine triclad (Phylum Platyhelminthes, Class Turbellaria, Order Tricladida) of the family Bdellouridae. The morphology of the body wall and musculature of this animal as seen by light microscopy of histological preparations has been described (Wineera, 1969). In the present study the methods of investigation employed allowed a much better characterisation of certain of the body wall constituents. The tissues of flatworms are notoriously difficult subjects for study, and the methods of "conventional" histology, such as the examination of stained paraffin sections frequently give poor results. This is particularly true for the parenchyma. In this tissue nuclei can be seen, but cell boundaries are indistinct or invisible. Extraction of fats during tissue processing, and distortion of tissues causes the appearance of irregular vacuoles which may be mistaken for intercellular spaces (Pedersen, 1961a). The walls of such vacuoles may appear as connective tissue fibres. Moreover, as Pedersen further states, none of the classical page 2 staining methods gives meaningful histological pictures, and the need for thin sections precludes the use of celloidin or frozen sections. These factors contributed to the parenchyma being called syncytial.

The methods of tissue processing employed for electron microscopy have recently (Pedersen, 1961a; Richardson, Jarett and Finke, 1960) been found to be excellent for the study of thin (0.25 to 3μ.) sections by light microscopy. One of these methods is employed in this study, and promises to be an important tool in future cytological and histological work.

Material and Methods

The collection of animals and their maintenance in the laboratory has been described (Wineera, 1969). Worms were fixed in buffered 4% formaldehyde (Pease, 1964, p.52) for 18 hr. Some of these worms were then postfixed in 1% osmium tetroxide in distilled water for 4 hrs. at room temperature. Specimens fixed in these two ways were embedded in araldite according to the method of Richardson, Jarett and Finke, (1960). Propylene oxide (two changes of 15 mins. each) was used between dehydration and resin infiltration stages. Polymerisation was carried out at 60°C for 8-12 hrs. Blocks were cut on a LKB ultramicrotome using a glass knife. Section thickness of approximately 0.25μ. and 0.5μ was obtained by advancing the block manually after each cut by means of the "microfeed" control. The sections produced were placed on microslides. Unstained sections were examined by phase contrast microscopy, and sections were subjected to the following staining procedures: (i) Methylene blue-borax/Azure II (Richardson et al, 1960)^*; (ii) The Falg technique (Gurr, 1965).

Remarks: Most conventional staining methods give negative results with tissues fixed by osmium, and tissue constituents which are usually distinguished by a characteristic staining reaction may prove difficult to identify. However, the alkaline methylene blue/Azure II staining method is designed for use with osmium fixed material, with which it gives excellent results, even in extremely thin (0.25μ.) araldite sections. Results with thin formaldehyde fixed material are poor, but by comparing (i) formalin fixed sections stained with methylene blue/Azure II with those stained by more conventional methods such as the Falg technique and (ii) osmium and formaldehyde fixed sections stained with methylene blue/Azure II it is possible to identify in osmium fixed material the "eosinophil" and "basiphil" substances of conventional staining methods.

Some worms fixed in buffered formaldehyde were embedded in 20% gelatin and sectioned longitudinally at 10μ. on a freezing microtome. The sections produced were placed on slides, mounted in glycerogel and were examined by phase contrast microscopy.

For observation of lving cells, a modification of the disassociation procedure of Pedersen (1961a) was used. Worms were placed on cavity slides in a few drops of 0.5% trypsin in sea water. The worms were then cut into small pieces using fine needles and the preparations were covered page 3 by a petri dish and left at room temperature for 3-4 hrs. After this time the small pieces of worm were placed on microslides in a drop of the trypsin/sea water solution and were gently squashed beneath a coverslip. The preparations were ringed with petroleum jelly and were examined by phase contrast and interference contrast microscopy. This technique resulted in fairly good cellular dissociation and the cells remained alive (as judged by contraction of muscle cells and the beating of cilia on epithelial cells and in flame cells) for 4-5 hrs.

Results

The Epidermis and "Basement Membrane": The epidermal "channels" (Wineera, 1969) are well shown by phase contrast microscopy of frozen sections (Pl. 1 Fig. 1.EC). They are seen to extend, in many cases, to the bases of epidermal cells, and often appear continuous with small "cavities" situated between the bases of epidermal cells (Pl. 1, Fig. 1, arrow). These cavities are more clearly seen in thin (0.5μ) osmium fixed sections stained with MBAII (Pl. 1 Fig. 2, C). The cells through which subepidermal eosinophil glands open to the surface are devoid of cilia (Pl. 1, Fig. 2, A), but this is not the case in the cells through which subepidermal basiphil glands open.

IN MBAII stained sections, the bases of epidermal cells can be traced by a thin dense line (Pl. 1, Fig. 2, Arrow). Below this line is material of the "basement membrane". Small vacuoles, or spaces are common within this latter structure, as are sections of the ducts of subepidermal glands.

The Muscles: The longitudinal muscle layers differ in size at the dorsal and ventral surfaces. At the dorsal surface the muscle fibres constitute a layer from 5-7μ thick. At the ventral surface this layer is thicker often approaching 25μ. Both in section (Pl. 2, Fig. 1, C) and as seen by dissociation preparations (Pl. 2, Figs. 2 and 3, C) the cell body lies up to away from the contracticle part of the muscle cell, and is joined to it by a thin cytoplasmic connection. In osmium fixed MBAII stained sections the muscle cell body colours a purplish blue while the contractile part colours a light blue grey. The nucleus is often irregularly shaped and contains a characteristic pattern of densely staining chromatin. This pattern is made up of relatively large chromatin clumps situated at regular intervals around the periphery of the nucleus, together with chromatin clumps of similar size in the interior of the nucleus. From circular muscle fibres seen in transverse section, and from dorso-ventral muscle fibres which approach the "basement membrane", fine processes (Pl. 1, Fig. 3, arrow; Pl. 1, Fig. 4, arrow), appear to connect with the "basement membrane" and with other muscle fibres. Muscle cells up to 200μ in length have been seen. Fine processes can sometimes be observed along the length of the muscle cells (Pl. 2, Fig. 2, arrows). The Parenchyma: In the parenchyma the cells fit very closely together. In 0.25μ thick sections many vacuoles of varying size are present. None of these vacuoles can definitely be identified as being extracellular. Also, connective tissue fibres cannot be seen between cells in these thin sections. As well as containing neoblast cells (Pedersen, 1959, 1961a; Wineera, 1969) the parenchyma contains largish cells 10-12μ in diameter. In osmium fixed sections stained with MBAII these cells have cytoplasm page 4 which is partly or wholly composed of moderately dense cytoplasmic strands. The strands are parallel to each other, and are arranged in curvilinear patterns (Pl. 3, Fig. 1, arrows). Other cells of similar size and staining reaction contain numerous vacuoles or pale granules (Pl. 3, Fig. 1, G) and yet others are seen in which both vacuoles and stranded cytoplasm are present. Pale staining large nuclei are present in the parenchyma but it is not possible to identify the limits of the cells to which they belong. Neoblast cells are visible in dissociation preparations, but fixed parenchyma cells (Pedersen, 1961a) have not been positively identified.

The Subepidermal Glands: Basiphil subepidermal glands are easily seen in osmium fixed MBAII stained sections. The cytoplasm is dark blue to purplish blue in colour. These glands open to the surface through the epithelial cells. In the cephalic and caudal regions of the animal the basiphil secretion granules often occur as aggregates in sac like structures within the epidermis.

Eosinophil subepidermal glands are either not stained by the above technique, or are stained a very pale greyish colour. Sometimes they colour with a faint yellowish tinge. In 0.5μ sections their secretion is seen to be made of discrete closely packed granules approximately 1.0μ in diameter. In the epidermis the granules colour a little more deeply, but do not approach the staining intensity of basiphil glands.

Pigment: In all dissociation preparations pigment granules are located in large rounded cells (Pl. 2, Figs. 4A & B, P). They may be few or many in number and may be scattered irregularly throughout the cytoplasm of these cells but often occur in groups several of which may be present in any one pigment cell. In incompletely dissociated cell masses from the dorsal region of the animal, pigment granules occur densely between muscle cells (Pl. 2, Fig 5). In these conditions pigment cell boundaries can not be discerned, but groups of pigment granules may be seen (Pl. 2, Fig 5, arrow).

Discussion

The Epidermis: In a previous paper (Wineera, 1969) it was stated that unicellular basiphil glands occur scattered throughout the epidermis. From the results of the present study it appears, however, that the basiphil secretions which were described in the epidermis are in reality the thin gland cell "necks" described by other workers (Hyman, 1959; Pedersen, 1963 as belonging to the subepidermal basiphil glands.

The epidermal "channels", and the cavities, or spaces between the bases of epidermal cells are interesting features. However, the relationship between these structures cannot be decided at present. Török & Röhlich (1959) state that planarian epidermal cells are attached to each other only superficially, and are surrounded by intercellular spaces of varying size, and Skaer (1961) and Pedersen (1961a) have described fluid filled cavities between the "basement membrane" and the bases of epidermal cells in fresh water triclads. Skaer states that the cavities appear to form a continuous system ramifying above the basement membrane, and suggests that their function may be hydrostatic or associated with the extrusion of granular secretion products through the epidermal cells.

The present study supports the view of Wineera (1969) that the "basement membrane" as described in triclads is not a true basement membrane. That is, it is not a basal lamina (Fawcett, 1966). Plate 1, Figures 2-4 show two components in the "basement membrane": One is the thin dense line immediately beneath the epidermis, and the other is the thicker, less dense layer beneath the thinner one. It seems probable that the former represents the true basement membrane (basal lamina) of the epidermis or at least approaches it more closely than any other structures visible with the light microscope, while the thicker layer is the "connective tissue layer" or "connective layer" as previously described (Wineera, 1969).

The Muscles and Parenchyma: The muscle fibres of P. stephensoni appear to resemble closely those described by Skaer (1961) for a fresh water triclad. In both instances the fibres give off fine processes along their length, and the nuclei may be situated some distance from the contractile part of the cell. According to Skaer the fine processes consist of aggregates of mitochondria bounded by the plasma membrane of the cells. It is probable that these processes account for some of the fibres described as connective tissue fibres (Wineera, 1969).

Three recent studies on turbellarians give slightly different pictures regarding the parenchyma and the occurrence within this tissue of connective tissue fibres. Two of these studies were carried out on triclads: In one (Skaer, 1961) no mention is made of the occurrence of fibres in the parenchyma; in the other (Pedersen, 1961a) it is stated that the cells of the parenchyma are closely packed together leaving a gap only a few hundred angstroms wide between two opposing cell membranes. Occasionally this gap is widened, especially around muscle cells, and filamentous material is often found in this narrow intercellular space. The third paper is a study of a marine polyclad turbellarian (MacRae, 1965), in which is shown the occurrence between muscle fibres and around muscle fibre bundles of extracellular fibrillae, presumed by MacRae to be an early form of collagen. Measurements of the regions of these fibrillae from MacRae's electron micrographs show them to be from 800-2,800 A in diameter. These larger regions possibly would be visible with the light microscope as thin fibres. It is possible that the connective tissue fibres associated with muscles as described by Wineera (1969) may be regions of extracellular fibrillae similar to those described by MacRae. Pedersen (1961a, p.597) suggests that the filaments (extracellular fibrillae of MacRae) occurring between muscle cells might be identical to those occurring in the basement membrane (Pedersen's "basement membrane" is equivalent to the basement membrane plus the connective layer of the present study) and says that "... there often seems to be a continuity between the subepidermal basement membrane and the intercellular space around the outer circular muscle cells." This same feature is recorded by Wineera (1969), and is seen in the present study. However, the present study has shown that the terminology used by Pedersen to describe this feature should be modified: It is not the basement membrane but the connective layer which projects between the muscles (Pl. 1, Fig. 4, arrow).

None of these papers mentions the origin or insertion of these muscle fibres or fibre bundles. It is axiomatic that muscle fibres or fibre bundles have to be attached in at least two places some distance apart in order to effect movement of body parts. In higher animals it is connective tissue which provides anchorage. MacRae (1965) suggests that a parallel exists between the filamentous material surrounding muscle fibres in the planarians and the endomysium round the muscle fibres of higher animals; and the material round planarian muscle fibre bundles and perimysium in higher animals. In the present study fine processes arising from muscle fibres are seen to join the connective layer (i.e. the basement membrane of other workers) (Pl. 1, Figs. 3, 4, arrows). It is suggested that these processes are analagous to tendons and similar structures of higher animals serving as attachments to the "connective tissue skeleton" supplied by the thick connective layer ("basement membrane").

The cells in the parenchyma which show curvilinear patterns in their cytoplasm are particularly interesting. At first sight these cells bring to mind the picture of a cell with a highly developed granular endoplasmic reticulum in which the cisternae and their attached ribosomes are arranged in parallel arrays. This specific arrangement of granular reticulum is characteristic of cells elaborating protein rich secretions for export from the cell (Mercer, 1961; Fawcett, 1966). Garnier (1897, 1899; quoted in Haguenau, 1958) noted the occurrence of basiphil "filaments" or "rods" in the basal cytoplasm of sumaxillary salivary gland cells. For these structures he coined the term "ergastoplasm" (Greek "ergazomai"—to elaborate and transform) for he believed in their importance in the secretory activity of the cells. The electron microscope reveals that "ergastoplasm" consists of the particular arrangement of cisternae and ribsomes described above. Bloom and Fawcett (1966, p.10) show an electron micrograph of pancreas at a magnification of 2,700 x. The appearance in this micrograph of the granular reticulum is very similar to that of the strands of basiphil cytoplasm of these parenchyma cells of P. stephensoni. Moreover, measurements of the cisternae plus attached ribsomes from Bloom & Fawcett's micrograph show them to be in excess of 0.2μ wide. This would make them visible (theoretically) in the light microscope, providing the sections were thin enough and the ribsomes were present in sufficient quantity to confer the needed basiphilia on the cytoplasm around them. It is seen from the micrograph of Bloom & Fawcett that at this magnification the lumen of each cistern is obliterated, so that each cistern plus its attached ribsomes appears as a single dense line. It is possible, then, that the "strands" seen in these cells in P. stephensoni may be dense lines representing cisternae plus ribsomes of a very well developed and regularly arranged granular endoplasmic reticulum. However, for the present the term "ergastoplasm" must be used to describe these structures since this term refers to cytoplasmic features visible with the light microscope. Pedersen (1963) shows that a highly developed granular reticulum occurs in subepidermal eosinophil gland cells in fresh water triclads in the early stages of the secretion cycle. Measurements from an electron micrograph published by Pedersen of such a cell show that the cisternae plus ribsomes of the reticulum range from 0.1 to 0.35μ wide. It is, then, possible that in the present study the cells in which the ergastoplasm occurs are eosinophil gland cells at an early page 7 stage of secretion. The cells of similar size and staining reaction which contain pale granules would then possibly be eosinophil gland cells at a later stage of the secretion cycle containing partly formed secretion granules. It is noticeable that, as has already been mentioned, the eosinophil secretion granules stain very faintly or not at all with MBAII.

Two other possible explanations of these cells must be noted. The first explanation is that the strands may represent an extremely well developed Golgi apparatus. However, the extent of the strands in relation to the size of the cells tends to negate this suggestion; also it is not understood what would bind the stain if these strands were part of a Golgi apparatus. A second "explanation" is that the "strands" are manifestations of some other cellular constituent, such as microtubules, or are arrangements of particular substances. This is a question which probably could be solved with the electron microscope.

Pigment: In a previous study (Wineera, 1969) it was concluded that the pigment granules of P. stephensoni were distributed at the sites of connective tissue fibres particularly where these fibres course between the muscle fibres of the dorsal body wall. This view can no longer be held.

It is now clear from this present study that pigment is present in distinct pigment cells which are located in the parenchyma. Pigment cells "in vitro" are rounded in shape (Pl. 2, Fig. 4A, 4B). However, the shape of the cells "in vivo" probably is more stellate, and they probably ramify between muscle cells in order to appear as they do in sections. Support for this view is given by Needham (1965, p.210) who states that pigment in Polycelis nigra (a fresh water triclad) occurs in "fairly orthodox asteroid cells". However, Skaer (1961) states that Polycelis pigment occurs within muscle cells and nerve cells.