Evidence has been obtained that the microtubules of the insect sensory dendrite may originate from the nine pairs of peripheral fibrils in the ciliary portion of the dendrite and retain direct continuity with them. The peripheral fibrils separate and each of the single fibrils resulting appears to divide one or more times. The number of microtubules within the dendrite distal to the ciliary region often lies close to 18 or to a multiple of 18, such as 36 or 72. If our interpretation is correct, the basal bodies, their rootlets, the ciliary fibrils and microtubules of the insect sensory dendrite form a continuous structure with the probable function of a cytoskeleton.

During the past 5 years there have been many reports on the substructure of flagellar fibrils (Pease, 1963; André & Thiéry, 1963; Grimstone & Klug, 1966; Ringo, 1967), ciliary fibrils (Renaud, Rowe & Gibbons, 1968), cytoplasmic microtubules (Slautterback, 1963; Ledbetter & Porter, 1963, 1964; Gall, 1965; Gonatas & Robbins, 1965) and the fibres of the mitotic spindle (Barnicot, 1966).

The close resemblance in size and appearance of these several elements, all with a structure that is basically tubular, was emphasized first by Ledbetter & Porter in 1963. Since then other workers have collected information that stresses the similarities of these parts. Recently, Behnke & Forer (1967) have produced evidence that microtubular structures are not all alike and may be divided into at least four classes. These they list as: (1) A-fibrils of cilia and flagella; (2) B-fibrils of cilia and flagella; (3) central tubules and accessory tubules of cilia and flagella; and (4) cytoplasmic microtubules and spindle tubules. The behaviour of the tubules at different temperatures, their reaction to treatment with colchicine, to pepsin digestion and to negative staining were the means used to separate the groups. A comprehensive listing and review of earlier literature is included in the paper of Behnke & Forer.

It is apparent, then, that the relationships between cytoplasmic microtubules, the fibrils or tubules of cilia and flagella and those of the mitotic spindle are a subject of continuing interest at this time. The antennal sense organs of insects, with which the present authors have been working for about 10 years, provide an unusual opportunity for investigating certain of these relationships. In 1964 Slifer & Sekhon suggested that the microtubules—or neurofilaments as they were then called—of the outer parts of the insect sensory dendrite may arise by the separation and subsequent division of the nine pairs of peripheral fibrils of the ciliary region of the dendrite. We now believe that we have evidence that will strengthen this theory.

The antennae of adult earwigs, Forficula auricularia Linnaeus, were used in the present study. They are covered with sense organs and a description of them has been published recently (Slifer, 1967 a). The antennae were cut into small pieces in icecold 3% glutaraldehyde solution buffered to pH 7·4 with 0·1M phosphate buffer (Sabatini, Bensch & Barrnett, 1963). After several rinses in phosphate buffer for a period of 2 h, the tissue was post-fixed for 1 h in a 0·1M phosphate-buffered (7·4) solution of osmium tetroxide. Subsequently the tissue was dehydrated rapidly in a series of cold ethanols and embedded in Epon 812 (Luft, 1961). Dr Barbara Stay, then at the University of Pennsylvania, kindly provided facilities for the embedding procedure. Sections were doubly stained with uranyl acetate (Watson, 1958) and lead citrate (Reynolds, 1963) and were examined with an RCA EMU-3E electron microscope.

The chemoreceptors of insects, often present in very large numbers on the antennae, are usually innervated by a group of neurons situated in the epidermis. The group may contain from 2 to over 60 neurons in different species and in different kinds of chemoreceptors. Rarely, such a sense organ has only one neuron. Tactile sense organs and proprioreceptors, on the other hand, usually have a single neuron. The neurons are bipolar and each sends a dendrite towards the hair, peg, plate or other specialized cuticular part of the sense organ. As it leaves the cell body of the neuron each dendrite contains the elements commonly present in cytoplasm, such as mitochondria, endoplasmic reticulum, Golgi complexes, microtubules, etc. A few microns above this the dendrite suddenly narrows and assumes the structure of a cilium (Figs. 2, 3). Beyond this point no formed cytoplasmic structures other than ciliary fibrils and microtubules can be identified. Two basal bodies lie just below the base of the cilium (Figs. 2, 4). Typically, the basal bodies are arranged in tandem although there is some variation in different species. Rootlets with a periodic structure arise from the inner end of the distal basal body and pass around the proximal basal body to end deep within the cell body and not far above the nucleus. Above, the ciliary structure is retained for only a short distance and the dendrite increases in diameter as it approaches the antennal wall. At this point the dendrites of a group usually converge and enter the inner, open end of a cuticular sheath invaginated from some point on the surface of the h’air or peg of the sense organ. Not all insect sense organs have such a sheath. Near the base of the hair or peg of the thin-walled chemoreceptor, which is especially useful in these studies, the dendrites usually leave the sheath through small openings in its wall and pass upward into the peg lumen (Slifer, 1967b). Here each divides several times so that the lumen is filled with dendrite branches. They are bathed in fluid and no other cells, or parts of cells, are present in the lumen. The wall of the peg is perforated by many minute openings, sometimes less than 0·01 μ in diameter, and clusters of pore filaments leave the sides of the dendrites and end in the pores. Here their tips are exposed to the air and to any odorous substance present in it. The dendrites of thick-walled chemoreceptors, in contrast, do not branch but go directly to a single opening at the tip of the peg. Tactile sense organs have but one dendrite. This is enclosed within a cuticular sheath that is fastened to the wall of the hair lumen a short distance above its base. All of the dendrites so far examined, whether in mechanoreceptors or chemoreceptors, have a ciliary region. Above it only microtubules are present in the cytoplasm.

Fig. 1.

Diagrams to show various interpretations of the Y-shaped figures present in longitudinal sections of insect sensory dendrites. A, separation of the members of a pair of ciliary fibrils; B, same as A but with paired fibrils lying in one plane and separated fibrils lying in a plane perpendicular to the first; C, one fibril branching to form two; D, two fibrils crossing.

Fig. 1.

Diagrams to show various interpretations of the Y-shaped figures present in longitudinal sections of insect sensory dendrites. A, separation of the members of a pair of ciliary fibrils; B, same as A but with paired fibrils lying in one plane and separated fibrils lying in a plane perpendicular to the first; C, one fibril branching to form two; D, two fibrils crossing.

Fig. 2.

Electron micrograph of longitudinal section through a sensory dendrite from the antenna of F. auricularia at the point where dendrite narrows and assumes a ciliary structure. Distal and proximal basal bodies lie below the cilium and rootlets extend downward from the former. The ciliary region (cr), lying in a vacuole (u) containing fine granular material, widens above as the dendrite enters a dense cuticular sheath (CS). Ciliary fibrils are shown separating (upper arrow, left) and apparently branching (upper arrow, right). Compare the ciliary fibrils with the microtubules in the wider part of the dendrite above with those in wider part below the ciliary region (lower arrow, right) and those in the sheath cell (lower arrow, left). × 33000.

Fig. 2.

Electron micrograph of longitudinal section through a sensory dendrite from the antenna of F. auricularia at the point where dendrite narrows and assumes a ciliary structure. Distal and proximal basal bodies lie below the cilium and rootlets extend downward from the former. The ciliary region (cr), lying in a vacuole (u) containing fine granular material, widens above as the dendrite enters a dense cuticular sheath (CS). Ciliary fibrils are shown separating (upper arrow, left) and apparently branching (upper arrow, right). Compare the ciliary fibrils with the microtubules in the wider part of the dendrite above with those in wider part below the ciliary region (lower arrow, right) and those in the sheath cell (lower arrow, left). × 33000.

Fig. 3.

Electron micrograph of cross-section through the ciliary region of a sensory dendrite from the antenna of F. auricularia just above the distal basal body. Cilium lies in a vacuole that contains a granular material and is enclosed by a sheath cell. Compare cross-sections of ciliary fibrils in dendrite with cross-sections of microtubules (arrow) in cytoplasm of sheath cell. × 6600.

Fig. 3.

Electron micrograph of cross-section through the ciliary region of a sensory dendrite from the antenna of F. auricularia just above the distal basal body. Cilium lies in a vacuole that contains a granular material and is enclosed by a sheath cell. Compare cross-sections of ciliary fibrils in dendrite with cross-sections of microtubules (arrow) in cytoplasm of sheath cell. × 6600.

Fig. 4.

Electron micrograph of cross-section of 5 dendrites from a short, thick-walled chemoreceptor of F. auricularia. (In this unusual sense organ the dendrites expand to form a sphere several microns across before narrowing to become a cilium.) In the dendrite at extreme left the section passes through the distal basal body (arrow) with its 9 triplets. Dendrite at upper right is sectioned through the rootlets that extend below the basal bodies (arrow). Other three dendrites are cut at points between the two designated. ×33000.

Fig. 4.

Electron micrograph of cross-section of 5 dendrites from a short, thick-walled chemoreceptor of F. auricularia. (In this unusual sense organ the dendrites expand to form a sphere several microns across before narrowing to become a cilium.) In the dendrite at extreme left the section passes through the distal basal body (arrow) with its 9 triplets. Dendrite at upper right is sectioned through the rootlets that extend below the basal bodies (arrow). Other three dendrites are cut at points between the two designated. ×33000.

Nine pairs of peripheral fibrils are present in the ciliary region of the dendrite but a central pair is absent (Fig. 3). Occasionally, an aberrant cilium with fewer or more than the usual number of fibrils is seen. Structural differences between the two members of a pair, such as arms on one and none on the other, have not been observed. As the dendrite widens above the ciliary region the members of the pairs of fibrils separate. There is some variation in the point at which this occurs and paired fibrils and single fibrils are often both present in the same cross-section (Fig. 5). Eighteen fibrils would be expected after separation is complete. However, more than 18 are commonly present. Stuart & Satir (1968) who examined the campaniform sense organs of termites noted that the ciliary fibrils are replaced in the outer part of the dendrite by microtubules and that there are more than 18 of the latter. In the cross-section shown in Fig. 7 of their paper at least 36 microtubules can be identified and more are probably present. Guthrie (1966), in an electronmicroscope study of the cephalic airflow receptors of the locust, includes a crosssection (Fig. 3) of two sensory dendrites. One of them contains about 142 microtubules and the other approximately 73. The presence of two dendrites in a mechanoreceptor is unexpected and these sense organs deserve further study.

Fig. 5.

Electron micrograph of cross-section through 4 dendrites enclosed by cuticular sheath from antenna of F. auricularia. Arrows indicate pairs of ciliary fibrils still in contact. Dendrites contain, respectively, 19, 20, 35 and 35 fibrils or microtubules. × 66000.

Fig. 5.

Electron micrograph of cross-section through 4 dendrites enclosed by cuticular sheath from antenna of F. auricularia. Arrows indicate pairs of ciliary fibrils still in contact. Dendrites contain, respectively, 19, 20, 35 and 35 fibrils or microtubules. × 66000.

Fig. 6.

Electron micrograph of longitudinal section through 5 dendrites of chemoreceptor from antenna of F. auricularia. Two ciliary fibrils separating (lower arrow) in second dendrite from left; fibril apparently branching into two (upper arrow) in middle dendrite, c, cuticle of antennal wall; cs, cuticular sheath. ×33000.

Fig. 6.

Electron micrograph of longitudinal section through 5 dendrites of chemoreceptor from antenna of F. auricularia. Two ciliary fibrils separating (lower arrow) in second dendrite from left; fibril apparently branching into two (upper arrow) in middle dendrite, c, cuticle of antennal wall; cs, cuticular sheath. ×33000.

Fig. 7.

Electron micrograph of cross-section of single dendrite of a tactile hair from antenna of F. auricularia. Cuticular sheath (cs) passes through vacuole (u) containing fine granular material. Sheath cell surrounded by cuticle of antennal wall (c). Note junction of opposite sides of sheath cell (upper right). Compare microtubules of dendrite with those in sheath cell (arrow and elsewhere). × 6600.

Fig. 7.

Electron micrograph of cross-section of single dendrite of a tactile hair from antenna of F. auricularia. Cuticular sheath (cs) passes through vacuole (u) containing fine granular material. Sheath cell surrounded by cuticle of antennal wall (c). Note junction of opposite sides of sheath cell (upper right). Compare microtubules of dendrite with those in sheath cell (arrow and elsewhere). × 6600.

Sections of dendrites sufficiently well fixed and so oriented that every fibril can be counted with certainty are obtained only rarely. Figure 5 shows such a cross-section. Here the four dendrites contain 19, 20, 35 and 35 fibrils or microtubules respectively. All of the tubular structures are very similar in appearance. At least two explanations for the increase in the number of elements are possible: (1) additional fibrils may appear de novo and be in no way associated with the original 18 fibrils; or (2) each fibril may branch to form two and these, in turn, may branch again. Assuming that the second explanation is the correct one and each fibril divided once, 36 fibrils would be expected, and if twice, 72. Actual counts are often close to these numbers. Occasionally, it is difficult to decide whether or not a particular structure is a fibril and this may account for aberrant numbers. Or, some fibrils may fail to divide or, again, since the fibrils appear to branch at different levels the cross-section being examined may lie below the level where branching has occurred in every fibril. Another source of variation might be a larger or smaller number of peripheral fibrils than the usual nine pairs in the ciliary region of the dendrite. The ciliary fibrils are indistinguishable in size and shape from the microtubules in the more distal parts of the dendrite (Figs. 2, 6) and this, in itself, suggests, although it does not prove, that the two are parts of a continuous structure.

The question, then, is: Do the nine pairs of ciliary fibrils, after separating, branch and give rise to the microtubules in the outer portion of the dendrite? Evidence obtained from longitudinal sections adds support to the possibility that they do. Figure 6 shows a longitudinal section through five dendrites enclosed within a cuticular sheath and approaching the base of an olfactory peg. In the second dendrite from the left (lower arrow) two fibrils closely applied to one another are shown separating. This is clearly the separation of a pair of ciliary fibrils as diagrammed in Fig. i A. In the centre dendrite a single fibre (upper arrow) is apparently branching to form two. It will be noted that each of the branches is of the same diameter. It could be argued that such a structure might result from the crossing of two fibrils (Fig. 1 D). This interpretation becomes less plausible when Fig. 5 is re-examined. In it each of the single fibrils or tubules lies well apart from the others and is separated from it by a distance several times its own diameter. It is probable that the section shown in Fig. 6 is too thin to include two fibrils lying one behind the other and separated by such a distance. Another possibility would be that the apparent branches are actually paired fibrils separating but oriented in such a way as to include only one of the two fibrils still in contact. In other words, the paired fibrils would lie in one plane and the two diverging fibrils in a plane perpendicular to the first (Fig. 1 B). Should this be the case, one would expect to see some indication that one fibril lies above the other at the point where they diverge. Immediately to the left is another Y-shaped figure that might also proλ’ide evidence but there are several small granular masses in contact with the right edge and these prevent a decision. Figure 2 is a section through the same region of the dendrite as that shown in Fig. 5 but in a plane at right angles to it. A pair of fibrils is separating about 2 μ above the base of the cilium (upper arrow, left). Immediately to the right is a single fibril with two branches (upper arrow, right). Here the arm extending below is sufficiently long to make it less likely that this is a side view of two fibrils still paired. Note that neither of the upper arms seems to overlie the other. Figure ic is a diagram showing our interpretation of this structure.

Other sense organs besides chemoreceptors provide similar evidence. Figure 7 is a cross-section of a single dendrite from a neuron that innervates a tactile hair. It is enclosed within a cuticular sheath that lies in a vacuole containing granular material and this, in turn, is surrounded by a sheath cell with an irregular inner border and many microtubules in its cytoplasm. Outside the sheath cell is the layered cuticle of the antennal wall. The dendrite contains about 60 microtubules cut in various planes. This suggests that at the level where this section was taken 12 of the original fibrils had divided twice and 6 had divided once to give a total of 60. Whether the 6 would branch again to give a total of 72, we do not know. The tubules cut in cross-section should be compared with those similarly cut in the cytoplasm of the sheath cell. They appear to be identical. A comparison of oblique and longitudinal sections in the two cells leads to the same conclusion.

If, as Behnke & Forer (1967) have shown, ciliary fibrils and cytoplasmic microtubules differ in their reaction to various treatments and,, at the same time, may be continuous structures this would necessitate a change at some point along their length from one type to the other. There is some evidence that such a change does occur in the flagella of rat sperm where the 9 + 2 tubules, or fibrils, were digested by pepsin more readily in certain regions than in others (Behnke & Forer, 1967). Even if the basal bodies, ciliary fibrils and microtubules in the insect sensory dendrite actually are continuous structures, it is certainly not possible to conclude from this that such an association is a general one or, indeed, even common elsewhere. One obvious exception would be the cells of higher plants where centrioles, equivalent to basal bodies, are absent although the fibres of the mitotic spindle (microtubules) are well developed. Here, as Pickett-Heaps & Northcote (1966) point out, some organelle other than the centriole must be involved in the synthesis of the spindle. They suggest that the fibres may originate from the endoplasmic reticulum. In a field where information is accumulating rapidly it is natural to stress first the similarities between various entities. Later, as more data become available and are analysed, differences will be emphasized. The study of microtubules would seem to be in this second stage.

Byers & Porter (1964), Porter (1965), Grimstone & Klug (1966), Sachs & Daems (1966) and others have suggested that microtubules serve as a cytoskeleton. Elements that strengthen and support would seem to be especially necessary in such delicate structures as the dendrites of these sensory neurons. The smallest dendrite branches within the lumen of a thin-walled chemoreceptor may be less than 0·1 μ in diameter and the cell membrane and the microtubules are the only recognizable structures of which they are composed. A fine granular material that surrounds the microtubules (Figs. 5–7) is probably the remains of a fluid present in life. If the microtubules are hollow, as Grimstone & Klug (1966) believe they may be, or if the wall consists of a denser stuff than the central portion, this would indicate their special suitability for increasing the strength of a dendrite. A hollow cylindrical structure has nearly three times the resistance to bending as does a rod of the same material and with the same cross-sectional area (Kennedy, 1927). In an insect, locomotion, respiration, movements of the digestive system etc. all produce internal stresses and strains and alterations in the pressure on different organs. If no tubular elements were present in these excessively thin dendrites and they consisted only of fluid surrounded by a cell membrane, it is difficult to see how they would survive the turbulence, sometimes quite violent, that occurs within the body of an insect.

This work was supported in part by a grant from the National Science Foundation GB-7310.

André
,
J.
&
Thiéry
,
J. P.
(
1963
).
Mise en évidence d’une sous-structure fibrillaire dans les filaments axonematiques des flagelles
.
J. Microscopie
2
,
71
80
.
Barnicot
,
N. A.
(
1966
).
A note on the structure of spindle fibres
.
J. Cell Sci
.
1
,
217
222
.
Behnke
,
O.
&
Forer
,
A.
(
1967
).
Evidence for four classes of microtubules in individual cells
.
J. Cell Sci
.
2
,
169
192
.
Byers
,
B.
&
Porter
,
K. R.
(
1964
).
Oriented microtubules in elongating cells of the developing lens rudiment after induction
.
Proc. natn. Acad. Sci. U.S.A
.
52
,
1091
1099
.
Gall
,
J. G.
(
1965
).
Fine structure of microtubules
.
J. Cell Biol
.
27
,
32
?.
Gonatas
,
N. K.
&
Robbins
,
E.
(
1965
).
The homology of spindle tubules and neurotubules in the chick embryo retina
.
Protoplasma
59
,
377
391
.
Grimstone
,
A. V.
&
Klug
,
A.
(
1966
).
Observations on the substructure of flagellar fibres
.
J. Cell Sci
.
1
,
351
363
.
Guthrie
,
D. M.
(
1966
).
The function and fine structure of the cephalic airflow receptor in Schistocerca gregaria
.
J. Cell Sci
.
1
,
463
470
.
Kennedy
,
C. H.
(
1927
).
The exoskeleton as a factor in limiting and directing the evolution of insects
.
J. Morph
.
44
,
267
312
.
Ledbetter
,
M. C.
&
Porter
,
K. R.
(
1963
).
A ‘microtubule’ in plant cell fine structure
.
J. Cell Biol
.
19
,
239
250
.
Ledbetter
,
M. C.
&
Porter
,
K. R.
(
1964
).
Morphology of microtubules of plant cells
.
Science, N. Y
.
144
,
872
874
.
Luft
,
I. H.
(
1961
).
Improvements in epoxy resin embedding methods
.
J. blophys. biochem. Cytol
.
9
,
409
414
.
Pease
,
D. C.
(
1963
).
The ultrastructure of flagellar fibrils
.
J. Cell Biol
.
18
,
313
326
.
Pickett-Heaps
,
J. D.
&
Northcote
,
D. H.
(
1966
).
Organization of microtubules and endoplasmic reticulum during mitosis and cytokinesis in wheat meristems
.
J. Cell Sci
.
1
,
109
120
.
Porter
,
K. R.
(
1965
).
Cytoplasmic microtubules and their functions
.
In Ciba Fdn Symp. on Principles of Biomolecular Organisation
, pp.
308
356
.
London
:
Churchill
.
Renaud
,
F. L.
,
Rowe
,
A. J.
&
Gibbons
,
I. R.
(
1968
).
Some properties of the protein forming the outer fibres of cilia
.
J. Cell Biol
,
36
,
79
90
.
Reynolds
,
E. S.
(
1963
).
The use of lead citrate at high pH as an electron-opaque stain in electron microscopy
.
J. Cell Biol
,
11
,
208
212
.
Ringo
,
D. L.
(
1967
).
The arrangements of subunits in flagellar fibers
.
J. Ultrastruct. Res
.
17
,
266
277
.
Sabatini
,
D. D.
,
Bensch
,
K. G.
&
Barrnett
,
R. J.
(
1963
).
Cytochemistry and electron microscopy. The preservation of cellular ultrastructure and enzymatic activity by aldehyde fixation
.
J. Cell Biol
.
17
,
19
58
.
Sachs
,
E. S.
&
Daems
,
W. T.
(
1966
).
Microtubules in human aortic intima cells
.
Z. Zellforsch. mikrosk. Anat
.
73
,
553
558
.
Slautterback
,
D. B.
(
1963
).
Cytoplasmic microtubules. I. Hydra
.
J. Cell Biol
,
18
,
367
388
.
Slifer
,
E. H.
(
1967a
).
Sense organs on the antennal flagella of earwigs (Dermaptera) with special reference to those of Forficula auricularia
.
J. Morph
.
122
,
63
79
.
Slifer
,
E. H.
(
1967b
).
The thin-walled olfactory sense organs on insect antennae
.
In Insects and Physiology
(ed.
J. W. L.
Beament
&
J. E.
Treherne
), pp.
233
245
.
Edinburgh
:
Oliver and Boyd
.
Slifer
,
E. H.
&
Sekhon
,
S. S.
(
1964
).
The dendrites of the thin-walled olfactory pegs of the grasshopper (Orthoptera, Acrididae)
.
J. Morph
.
114
,
393
410
.
Stuart
,
A. M.
&
Satir
,
P.
(
1968
).
Morphological and structural aspects of an insect epidermal gland
.
J. Cell Biol
.
36
,
527
549
.
Watson
,
M. L.
(
1958
).
Staining of tissue sections for electron microscopy with heavy metals
.
J. biophys. biochem. Cytol
.
4
,
475
478
.