topic::
It is proposed that the lamella edge cells in Amanita are not cheilocystidia and that they are analogous to the stipe context of the Amanitaceae or to the supporting hyphae of the glutinous universal veil on the stipe in Limacella.
posted by:: R. E. Tulloss
annotation:: second draft posting
date of post:: 12 feb 2010
date last edited:: 13 feb 2010
original extended annotation:: This posting was written in response to a question from Mr. Yadwinder Singh (Research Scholar, Department of Botany, Punjabi University, Patiala-147002, Punjab, India): "What is the difference between cheilocystidia and the lamellae edge cells present in the genus Amanita?  Both are sterile cells; both are present at the 'knife edges' of gills."
This topic addresses a question of choice of correct terminology.  The topic has some history.  While specialists in the Amanitaceae have consistently held for some time that cystidia of any sort are absent in that family, a number of individual authors have used the term "cystidium" in reference to cells found on the stipe or on the gill edges.  As a reviewer of a number of papers describing or revising amanitas, I have had numerous occasions to offer reviewer's comments discouraging the inappropriate usage of the term "cystidium" in describing cells that are specialized and unique to Amanita or to the Amanitaceae as a whole (e.g., the longitudinally oriented, inflated cells—acrophysalides—of the stipe context that is a defining element of the family OR the chains of easily dissociated inflated cells on the edges of lamellae in Amanita).  The latter example is the topic discussed here.

What is a cheilocystidium?

This discussion utilizes definitions from Ainsworth & Bisby's Dictionary of the Fungi (8th ed., 1995), the 1971 revised edition of Snell and Dick's A Glossary of Mycology, and the glossary by E. C. Vellinga in vol. 2 of the Flora Agaricina Neerlandica (1990).

From these sources, it is clear that "cheilo-" is a position-defining prefix originally proposed by Buller.  Specifically, a cheilocystidium is a cystidium found on the edge of a lamella. So what is a cystidium?

It is not so much that there is disagreement on this point but that the definitions are generally not complete.  The cited resources agree that a cystidium is

  1. a sterile cell
  2. distinguished from cells adjacent or near by to it by its form (e.g., by one or more of the following: shape, size, content, wall-thickness, chemistry, surface decoration, contents, and/or relative prominence)
  3. found only on a surface of a basidiome
  4. not foreign to that basidiome (i.e., is determined entirely by the genome of the fungus producing the basidiome) [clearly intended, but not often made explicit].

Among the definitions, that of Vellinga in the Flora Agaricina Neerlandica is unique in making explicit that a cystidium is a (singular) terminal cell. Illustrative cystidia are always drawn this way (e.g., see Dict. Fungi 8: 121, fig. 12 OR Glossary Mycol.: pl. XI-XII), but Vellinga puts this critical point into words.

Hence, a cheilocystidium is a (single) sterile, terminal cell determined by the genome of a basidiomycete, found [only] on the edges of lamellae of basidiomes of such a fungus, and differentiated from adjacent or nearby cells on the lamella edge of that fungus.

Purely based on this definition, lamella edge cells in Amanita basidiomes are not cheilocystidia.  The lamella edge cells are formed in chains arising from a cable-like fascicle of undifferentiated filamentous hyphae that runs along the edge of the lamella.  Hence, lamella edge cells are NOT single terminal cells.  The chains of cells are closely packed and make up a uniform continuous band along the lamella edge—the inflated cells of the lamella edge tissue in Amanita are NOT isolated, and differentiated from, other cells that also appear on the lamella edge.  Indeed, as we will see below, the lamella edge cells are part of a layer of cells with a function unique to Amanita among all agarics in that Amanita is the only genus in which the ontogeny of the basidiome is schizohymenial.

While cystidia are valued as a significant aid to determination in taxa in which such cells occur, they are only occasionally considered to provide a well-understood function that (for example) benefits fungal reproduction. This is not to ignore the fact that some pleurocystidia are hypothesized to have a lamella-spacing function.  In contrast, the lamella edge cells play a very clear role in facilitating the expansion of an Amanita basidiome—indeed, a role that is an integral (functionally necessary) component of schizohymenial development of an agaricoid basidiome.

I. First, consider the evolutionary history of Amanita.  Morphological and molecular phylogenetic data agree that Amanita and Limacella had a common ancestor.  Of the two sister taxa, Limacella is (by all evidence) the older.  Morphologically the hyphae that support the gluten of the universal veil in Limacella are likely to be analogous to the the chains of inflated cells of the universal veil of the basal taxa in Amanita (members of Amanita subsection Vittadiniae).  The presence of clamps in all known species of Limacella is an additional shared character between that genus and the most basal of the basal Amanita species.  There are many other interesting analogs and shared characters between species of Limacella and the basal taxa of Amanita.  But let us focus on the edges of the lamellae.

The edges of the gills of Limacella are entirely fertile.  There are no known taxa of Limacella that bear any cystidia.  All recent authors (since H.V. Smith in the 1940's) have excluded suspect taxa from Limacella when cystidia were present in those taxa.  In my recent study of the genus Limacella, I could only find two (recently described) taxa that were purported to belong in Limacella and had cystidia reported; in both cases (and I am relying solely on the protologs for the moment), these taxa also have other characters suggesting that they don't belong in Limacella (for example, large spores).  In addition, the descriptions are clearly hurried and produced by authors who seem not to be aware of important papers in the limited literature concerning Limacella.  At any rate, current evidence strongly suggests that the ancestry of the earliest Amanitae included only taxa lacking cystidia.  Therefore, if cystidia were to be found in an Amanita, it unlikely that the production of such cells would have been inherited from an ancestor in common with Limacella.

A major stage in the development of the genus Amanita produced a significant transformation that segregates the genus (by its ontogeny) from all other agarics.  Probably in response to a period of extended drying of [some part of] the earth, the ancestry of Amanita evolved to delay expansion and sporulation to the extent that development of the basidiome can start when water is available and, at least to some degree, stop if the water disappears, and then restart development/expansion when water returns.  Among other things, bearing the basidiome's water on the exterior in a glutinous universal veil was (apparently) replaced by inflating the cells of the volva and containing water therein which served as a protection against dessication.  One of the adaptations of Amanita was its mode of development in which all the usual "agaric parts" develop simultaneously within an enclosed space and complete tissue segregation is put off until near the maturity of the basidiome—the schizohymenial ontogeny of today's Amanitae.

During the hypothetical period of transformation of proto-amanitas to the current form of the genus, new mechanisms developed that permitted such things as the segregation of hymenial surfaces and the segregation of gills from a partial veil or from the stipe surface.  The gill edges ceased to be fertile (the state in Limacella) and the entities we call lamella edge cells developed making possible undistruptive separation of the gill edges from originally adjacent tissue(s).  Once the tissues separate and/or while they are separating, the lamella edge cells collapse, dessicate or gelatinize, and often become deciduous. 

Are there analogs of these edge cells in Limacella? What could they be?

II. Recall that Limacella is considered to completely lack cystidia, and, particularly, cheildocystidia.  Therefore, the analog of the Amanita lamella edge cell is very unlikely to be derived from a cheilocystidium in a common ancestor of Amanita and Limacella. Cheilocystidia arise from tissue lying within the subhymenium or the lamella trama in taxa where cheliocystidia are present.  However, the lamella edge cells arise from an interwoven, cable-like structure that is originally arranged PARALLEL to the similar structures of a partial veil and/or a stipe surface.

In a development process that seems very likely to be driven by the presence, absence, movement, etc. of hormonal concentration fronts (wet chemistry), it is very easy to understand how interfaces between developing tissues in a solid primoridum could develop a thin layer with a form intermediate to the two tissues such a layer separates.  In the developing stipe context, the tissue is developing longitudinally (vertically) with very strong subparallel alignment of elements.  The predominant elements are filamentous undifferentiated hyphae and inflated cells (in chains in some species, but usually single and terminal) that we call acrophysalides.  Observations of Bas and subsequent authors suggest that solitary terminal acrophysalides are a derived condition and chains of such cells are a "primitive" condition.  While these acrophysalides can be hundreds of micrometers long in the stipe context, they are much smaller at the stipe surface and only a few tens of micrometers long in the partial veil. Hence, the surface acrophysalides and the similar cells in partial veils have lengths very similar to the range of lengths of lamella edge cells.

Considering the orientation of the tissues, we might well suppose that the stipe's acrophysalidic tissue (pre-existing in the common ancestry of both Amanita and Limacella and, indeed, a defining character state of the family Amanitaceae) may be a possible analog for the edge cells of the lamellae. Let us look at the other side of the interface.  In the normal agaric development of a Limacella, the gill edges comprise basidia and basidioles.  They are inflated cells of an appropriate size, shape, and position to be considered as possible analogs for lamella edge cells; HOWEVER, it is much more difficult to explain how basidia and their subhymenium could both be transformed to produce the present day cable-like structure densely set with chains of inflated cells. To shorten this discussion a bit (for the moment only...what is omitted must be made explicit in the future), I think it is much more apparent how the lamella edge cells and their supporting hyphal cable could be an analog of stipe context/surface (there is little difference in form of these tissue-regions) than how subhymenium and basidia could cease to have a reproductive function and become so strongly segregated from the bilateral structure of the lamellae.

III. It might be argued that the hyphal "cable" (from which the lamella edge cells arise) is connected to hyphae that pass into the lamella trama; and, hence, the lamella edge cells could be said to "arise from the lamella trama."  On the contrary, hyphal connections interpenetrating adjacent tissues are to be expected between any pair of adjacent developing tissues in an Amanita primoridum or subsequent "button" because ALL developing tissues of the original primordium are originally intimately connected via hyphae.  Indeed, the word "schizohymenial" comes from the fact that the hyphal connections between what become opposing hymenial surfaces have to be broken for the basidiome to expand into an agaric form.

These considerations give rise to a new question: What is the structure of the interface between the stipe and the layer of tissue we call "lamella edge cells"?  One might hypothesize that it would be a "mirror image" of the arrangement on the lamella edge.  Some observations would be useful.

IV. Experiment.  While Dr. Yang and I have both begun to pay greater attention to the morphology of the cell edge than was done previously, I do not think anyone has examined the morphology of the ridges of cells left on the apex of the stipe and/or on the upper surface of a membranous partial veil in many amanitas after the stipe and lamellae separate. It is commonly observed by collectors of amanitas that these ridges or "lines" on the stipe apex are positioned exactly as though they were matched to the lamellae edge during development of the basidiome. They are (reasonably) assumed to comprise the same tissue as the lamella edge layer by some authors and passed over with few (if any) words by others.  I know of no one who has examined how these "lines" are attached to the stipe surface.

I suggest that careful examination of these lines of tissue might be relevant to the question of the evolutionary origin of the lamella edge cells. I think a thorough investigation of this tissue in a variety of taxa (but especially in basal taxa of Amanita), might produce results bearing strongly on the morphologically-based understanding of how evolution in Amanita has yielded the "success"" of schizohymenial development.

V. Another possibility that cannot yet be posed in detail (because of the very minimal anatomical information available on species of Limacella) is as follows: The analog of the lamella edge cells of Amanita is not the acrophysalides of Amanitaceae stipe context, but (instead) is the analog of the gluten-supporting hyphae of the universal veil on the stipe in Limacella.  Descriptions of these hyphae and (particularly) of their attachment to the stipe is practically nonexistent. Hence, this seems like another good area for anatomical exploration.