- Email: [email protected]
A quantitative estimation of the association of mitochondria to septate desmosomes in the sea urchin larva
Printed in Sweden Copyright @ 1974 by ilcademic Press, Inc. Al! rights of reproduction in any.fom reseraed
Experimental Cell Research 85 (1974) 429436
NTITATIVE ESTIMATION OF T F MITOCHONDRIA TO SEPTA IN THE SEA URCHIN B. LUNDGREN Wenner-GwenInstitute, University of Stockholm, S-Ii3 45 Stockholm, Sweden
SUMMARY Associations of mitochondria with the septate desmosomes are often found in sea urchin embryos. In this investigation a semi-quantitative method of measuring the abundance of these associations is presented and applied to different developmental stages of the sea urchin larva. An increase of the mitochondrion-desmosome association is found before the primary mesenchyme formation. The association is also more common in the ectoderm than in the entoderm. The possible role of such associations in development is briefly discussed.
Associations between mitochondria and desmosomes have been reported in differentiating as well as adult tissues and seem to be common throughout the whole animal kingdom. Different functional roles have been ascribed to such associations, e.g. that mitochondria are necessary for desmosome formation c.5, 61, that they participate in active transport of water and solutes [ZO], and impulse propagation from cell to cell 251. Such associations have frequently been seen in the seaurchin larva (Gustafson, pers. comm.; [19]) in which the desmosomesare of the septate type [I, 2, 41. In view of the important function ascribed to cell adhesion, and hence also to desmosomes,in sea urchin development [13, 271 it was considered of interest to make a more detailed analysis of this phenomenon. A semi-quantitative method was developed to permit a comparison of the frequency of associations in different regions of the embryo and also in different develop-
mental stages. The results showed a marked increase in the frequency of these associations between 6 and 12 h after fertilization and a greater abundance in the ectoderm as compared with the entoderm.
Handling of speckmensfor electron microscopy Larvae of Psammechinus miliaris obtained by artificial fertilization were reared in round flasks with continuous stirring. Samples were taken for fixation at 6 h (unhatched blastula), 12 h (hatched blastula prior to formation of the primary mesenchyme), 22 h (gastrulae with fully extended archenteron) and 72 h (plutei). The material was fixed for 1 h in ! ‘% 0~0, in filtered sea water. After rinsing in sea water rhe larvae were dehydrated in ethanol and embedded in Epon 812 [17]. The embeddings were sectioned with an LKB Ultrotome (ultramicrotome). ‘Ihe sections (approx. 600 A thick) were picked up on one-hole grids and stained with uranyl acetate and lead citrate [22]. After drying, the sections were examined in. a Siemens Elmiskop I electron microscope at x 5 OBQ. The negatives were enlarged to a final ma of 20 000.
430 B. Lundgren
Fig. 1. Septate desmosome with associated mitochondria from the ectoderm of a 22 h old sea urchin larva (gastrula). x 45 000. Fig. 2. A desmosome site from a 6 h old unhatched blastula. The septa are not formed yet, the dark intercellular material is probably polysaccharides. The larva has been treated with ruthenium red in order to reveal extra-cellular polysaccharides. x 45 000.
Sampling The larvae were not oriented in any specific manner in the blocks. Pyramids containing one larva were prepared; from at least 5 different larvae 3 sets of sections were cut from each. Between the individual sets of sections 3-5 pm was removed. Sampling on the grid was performed by photographing the first usable section found on each grid. Micrographs were taken at the first point where the larval wall was hit by the beam, and then at intervals of two screen openings along the larval periphery. The total number of pictures per larva was 3-6 and per developmental stage 45-90. In the pluteus and in the archenteron of the gastrula the sampling was more restricted as only 3, respectively 4, different larvae were used.
Measurements and statistics Estimation of the mitochondrion-desmosome association was performed by comparing the distances between a point in the middle of the desmosome profile to the periphery of the nearest mitochondria profile (d.-m.) and from a randomly chosen point on the cell membrane in the cell junction to the nearest mitochondrial profile (r.-m.). The results were classified into 9 classes with intervals of 0.1 pm; all measurements exceeding 0.9 pm were pooled to make a tenth class. Mitochondrial fractional volume in the cytoplasm was determined by point counting [26]. A priori, the frequency with which mitochondria are associated with desmosomes may depend both on a specific association mechanism and also on the influence of unspecific factors such as the density of packing of mitochondria in the peripheral regions of the cell. Changes in the average mitochondrial packing in the cell were revealed by the figures for mitochondrial fractional volume. Regional changes in the mitochondrial packing in the cell were expected to be Exptl Cell Res 85 (1974)
revealed by an uneven distribution of frequencies in the r.-p. histogram. These unspecific factors will of course be equally registered in the d.-m. measurements; therefore to obtain a measure of the specific association of mitochondria to desmosomes, the frequency of random associations in the class o-O.1 were subtracted from the corresponding d.-m. frequency. The decision to use only this particular class will be considered in the Discussion.
RESULTS A close association between septate desmosomes and mitochondria was often found in embryos of Psammechinusmiliaris older than 6 h (fig. 1). In unhatched embryos of 6 h age such associations were rarely seen(fig. 2). The quantitative estimations of the mitochondrion-desmosome complexes for the different stages are summarized in figs 3-7. The distributions of measurements from random points on the plasmalemma to the nearest mitochondrion r-.-m. were relatively constant in all classesfor all stagesof development. However, the desmosome to mitochondria measures (d.-m.) accumulated in the class O-O.1,um in all stages except in the unhatched bastula. The relative frequences in the O-O.1pm class are given in table 1. No differences between d.-m. and r.-m. occurred
Estimation of rviktochon~~ion-~es~~osome~sso~~~t~~~ 43 1 Table 1. Mitochondrion-desmosome Stage (hOXS)
d.-m. in o-o.1
6 Ii 22 ect. t. 72 Zieb.
0.02 0.61 0.75 0.33 0.81
association and corrcation for rQ~dQrn-~o~~td~str~~~ti~~~
S.E.
P
0.016 0.037
0.001
in the 6 h blastula. In more advanced stages on the other hand more than half of the measures of d.-m. were found in this class, while r.-m. was higher than 0.1 in a single case. An interesting feature was the low degree of association in the archenteron of gastrulae as compared with the ectoderm. This difference may explain the lower figure for associations found in the 12 h blastula where no distinction was made between different regions. The figures obtained for the mitochondrion-desmosome association seem to be of high significance. The differences could be due, however, to an alteration of the mitochondrial fractional volume. In table 2 the results of %he morphometric assay of the mitochondrial fractional volume in the cytoplasm are summarized. The mitochondrial fractional volume in developmental stages from 6 to 22 h did not show any drastic variaTable 2. Mttochondrial
fractional
volume in
cytoplasm Stage (hours)
Fract. vol.
SE.
6 12 22 ect. ent. 72 cilieb.a
0.0568 0.0687 0.0690 0.0816 0. 1009a
0.0115Y
P
a The figures for 12 h pluteus ciliary band are not reliable because of the incomplete sampling from that stage. 28-741813
0.02 0.07 0.09 0.04 0.15
0 0.57 0.66 0.29 0.66
tions, but an increase was found between 6 and 12 h. The figure for the ciliate the pluteus is not reliable and is jus in the table for comparison. lit is remarkable that the mitochondrial fractional voltme in the archenteron of gastrula was higher than in the ectoderm. In fig. 8 the mito~bondrio~-~~srnosom~ association and mitochondrial fractional volume are brought together to give a clear, overall summary.
Associations between mitochondria an tate desmosomes ’ sea urchin larvae are wever, it is diffificuh to readily observed. decide on such evidence whether the association is due to true affinity between %he mitochondria and the desmosomes,or merely a passive consequenceof mitochondrial packing in the cytoplasm. ~urthermor~~ when using such a qualitative approach it is difficult to establish differences between varying stages of development and different parts of the larva. %f the mitochondrial associations were due to passive peripheral displacement, the entire lateral plasmalemma would be covered by mitomosomes. iif &is chondria as well were the case a mitochondria was randomly scattered in the cytoplasm, then a comparison belween any random point on the lateral plasmalemma and
432 B. Lundgren
4
l-
o-
5
6
the desmosome would be a measure of the specific association. One problem in this connection is the large nucleus found in the embryonic cell. Still, the method is usable if the nucleus does not represent a too big fraction of the cell. If the distribution of frequences in the r.-m. is dominated by very short and very long measures, it is probably due to a too-big nuclear fraction. An example
of this situation is found in fig. 7. The explanation may in this case also be a high mitochondrial fractional volume, which eliminates a large portion of the expected long measurements. In fig. 6 both these factors contribute to the special distribution of the r.-m. Another question in connection with the evaluation of the method is: “When is a mitochondrion associated with the desmo-
Exptl
Cell Res 85 (1974)
profiles will have larger and 13.5 srnailer values than half the actual rn~~o~bo~~rio~ diameter. This means that 86.4% of the pro-, files of associated mito~hondri~ will be found less than 0.125 pm from the desmosome and the rest 13.6% between 0.125 a pm from the desmosome. However, it will be impossible to ~ete~rni~~ whether a mito~bo~drio~ is associated. not, when using size istributions correlate to the distance unless the mito~bo~~ria are uniform in size and spherical in shape. These conditions are not met. The only usable sections would then be those cut in a plane at absolute right angle to the des plane. Con ring all these lirni~at~o~s~the problem se insoluble with a reasonable Figs 3-7. The histograms consist of two parts; the amount of work and accuracy. part over the zero line represents the distribution of To simplify the estimation, let us define frequences in the d.-m., and the part under the zero line, in the r.-m. The first nine classes are made with association as “a distance between desmoa &dtb of 0.1 pm while the tenth class (dotted) consists of all measures longer than 0.9 pm. No. 3, some and mitochondrion less than 0.1 ,uM”, 6 h unha.tched blastula; no. 4, 12 h blastula, just prior a distance still short enough to allow fast, to primary mesenchyme formation; no. 5, ectoderm nt interactions between the of a 22 h gastrula; no. 6, archenteron from the same diff usion-dep stage as no. 5; no. 7, ciliary band and animal-plate may simplify further; ““any components. from a 72 h pluteus. mitochondrion profile found within 0.1 p.m from the desmosome is cut from an associsome?” The most frequently occurring pro- ated mitochondrion”. This is a crude definifiles of mitochondria in sea urchin larvae are tion; associated mitochondria may be lost about 0.5 pm in diameter, and the profiles due to oblique sectioning, but owing to the are almost always circular. It is therefore relatively thick sections and the fact that reasonable to assumeas a first approximation that the mitochondria resemble spheres with a diameter close to 0.5 ,um. If we define the association “when the mitochondrial outer membrane is in contact at any point with the inner surface of the desmosomal membrane” (a case never observed in this investigation) the problem is, on the spherical mitochondrion, that one point alone will show whether an association exists or not. The possibility of finding that point, using 500 A sections, 6 will be 1 out of 10 in an associated mito8. Summary of the result of measurements of chondrion. For such a mitochondrion, accord- Fig. association (dotted) and the mitochondriai fractional ing to Elias & Wenning [7], 86.4% of the volume in the different stages investigated. CL%*- 741813
434 B. Lundgren the mitochondria seemto be flattened on the side facing the desmosome, it does not seem to interfere too much with the results. If there were large errors, they would be most frequently found in the 0.1-0.2 pm class. This is also the case in fig. 4, but the error is small when correcting for the random distribution and considering the relatively high frequency of the classO-01 ,um. Thus, the present method seemsadequate in comparing different developmental stages.It is doubtful, however, if it is usable for comparing different species or different organs in a highly differentiated animal. In such cases differences in the mean mitochondrial diameter must be considered and a more accurate definition of the association be found. In conclusion, the method is no clearcut morphometric assay and the figures are not valid in absolute terms but are still usable as relative figures when comparing developmental stages of the same species. Results In the 6 h old unhatched blastula no association of mitochondria to desmosomes was found. Older stages, 12, 22 h and 72 h larvae, show a well developed mitochondriondesmosome association, at least in the ectoderm. Various theories for the functional significance for such associations have been put forward by different authors. Most investigations are, however, made on vertebrate epithelia and concern a different type of desmosome. For example Deane et al. [5, 61suggestedthat in young and fetal mammals mitochondria were necessary for the formation of the plaque desmosome as a source of energy and metabolites. But, in the larvae of Psammechinus miliaris, the septate desmosomes are formed 1 or 2 h before hatching (about 6 h after fertilization). At this stage the results provide no evidence of an association. Therefore mitochondria do not seem Exptl Cell Res 85 (1974)
to be involved in the formation of the septate desmosomesin Psammechinusmiliaris. It has been known since the turn of the century that a sea urchin embryo transferred into calcium-deficient sea water is dispersed into a cell suspension, which has the ability to recombine into an intact larva when transferred back to normal sea water [8, 141. In other words, calcium-deficient sea water weakens the cell contacts. This calcium dependencemay give a clue to the functional significance of the association between mitochondria and desmosomes.Mitochondria in many tissues are known to be capable of actively accumulating and releasing calcium in response to cellular requirements [3, 231. Therefore it may be hypothesized that the mitochondria might increase the stability of calcium-dependent desmosomesby buffering the intercellular calcium concentrations in case of fluctuations of the exogenous calcium concentrations. This hypothesis has one major weakness; it is not likely that intracellular calcium is affected when larvae are immersed in Ca2+deficient water. The calcium deficiency probably works on ionic bridges in the thin extracellular polysaccharide coating of the plasmalemma [ 181,thereby dispersing the cellcementing capacity of the polysaccharides. However, some support for the theory may be adduced in the fact that no association is found at early stages before hatching. It is possible that the good protection against dispersion provided by the fertilization membrane makes a calcium buffer system unnecessary, whereas in later stages, when no fertilization membrane exists, a mitochondrion-desmosome association is needed. Another function suggestedfor the desmosome is the propagation of intracellular impulses via channels in the desmosome [16]. A possible model for such transmissions could be that of Prince & Berridge [21] in
Estimation of mitochondrion-clesmosowe a$sociattom 435
which membrane potential alterations are influenced by LAMP via @a2+release from intracellular deposits. Thus, this theory could also involve the calcium storing capacity of m.itochondria. Some of the results in the present investigation can be taken as support for the impulse propagation hypothesis. The ciliary band of the 72 h pluteus showed a very high mitochondrion-desmosome association. This area is, as concluded in recent investigations [1 1, 12, 151, an excitatory centre, where mechanical or chemical stimulation is rapidly transformed into impulses which are transmitted towards the interior regions of the larva. Furthermore, the archenteron of the 22 h gastrula did not show any marked association, despite the large mitochondrial fractional volume. This is in line with the results of Gustafson et al. [ll]; in larvae irritated by mechanical stimulation, local lobular contractions of the stomach wall occurred at points where neuron-like strands make contact. A good cell-to-cell communication within the stomach wall would probably have led to a general lobulation of the whole stomach. Earlier investigations utilizing vital staining IlO] and particle counting from homogenates 1241were generally not in agreement. Both types of investigation showed, however, an increase in mitochondrial number prior to primary mesenchymeformation. In the work by Gustafson Lenicque [lo], where nileblue sulphate was used as vital stain, the increase was very pronounced. The work by Shaver [24], however, showed a less pronounced increase from about the stage corresponding to the 6 h unhatched Psammechinus miliaris blastula, which is in line with the slight increase in mitochondrial fractional volume found in the present investigation. A priori, changes of mitochondrial fractional volume cannot be exactly compared with the alterations of mitochondrial numbers in the
works of Gustafson & kenicque and of Shaver, but in practice it can as the mean rnit~~~o~~rial profile diameter is not subject to any drastic variations between the stagesin this investigation. The very marked increase in ~ito~~o~~~~a~ number in the IO-12 h blastnla showed by Gustafson & kenicque occurs simultaneous with the increase in mito~ho~drion-desmosome association found in the present work. Furthermore, the low mitochondrial count found by Gustafson & Eenicque IlO] in vegetal parts, in contrast to the relatively high fractional volume found in the present i tion, also reflects the low mitoch desmosome association in this part of the embryo. The results suggest some kind of activation of mitochondria between 10 and 12 h after fertilization, leading to an affinity to the desmosome and also to an increased stainability with nile-blue sulphate. activation may very well reflect an increase in the intercellular communication, wh~.ch may be necessary to coordinate the morphogenetic processes starting at that stage with the formation of the primary rne~e~~~yrne~ The present work therefore supports the conclusions of Gustafson & Lenicque [lo] where a definite change in the physiology of the mitochondrion according to a characfesistic space time pattern occurs; see further Gustafson [9].
The author wishes to thank Professor Bertil Swedmark and his staff at the Kristineberg Zoological Station, Fiskebackskil, Sweden, for arranging working facilities and supplying the animals for the &vestigaI tion. My warmest thanks are also directed to Tudor Barnard for valuable discussions during the experiments and in preparing the manuscript. Professors Bjorn Afzelius and Tryggve Gustafson have also been of great help in the preparation of the manuscript. The skilful technical assistance by Mrs Carina Starkernd has been of greatest importance and is gratefully acknowledged. This investigation was supported by grants from the Swedish Natural Science Research Council.
436 B. Lundgren
REFERENCES 1. 2. 3. 4. 5. 6. 7.
8. 9. 10. 11. 12.
Agrell, I P S, Z Zellforsch 72 (1966) 12. Balinsky, B I, Exptl cell res 16 (1959) 429. Borle, A B, Fed proc 32 (1973) 1944. Dan, K, Int rev cytol 9 (1960) 321. Deane, H W & Wurzelmann, S, Proc 6th int congress EM, Kyoto (1966). Deane, H W, Wurzelmamr, S & Kostellow, A B, Z Zellforsch 75 (1966) 166. Elias, H & Henning, A, Quantitative methods in morphology (ed E R Weibel & H Elias) p. 130. Springer-Verlag, Berlin, Heidelberg, New York (1967). Giudice, G, Dev biol 5 (1962) 402. Gustafson, T, The biochemistry of animal development (ed R Weber) vol. 1, p. 139. Academic Press, New York and London (1965). Gustafson, T & Lenicque, P, Exptl cell res 3 (1952) 251. Gustafson, T, Lundgren, B & Treufeldt, R, Exptl cell res 72 (1972) 115. Gustafson, T, Ryberg, E & Treufeldt, R, Ada embryo1 expt12 (1972) 199.
Exptl Cell Res 85 (1974)
13. Gustafson, T & Wolpert, L, Biol rev Cambridge philosoph sot 42 (1967) 442. 14. Herbst, C, Wilhelm Roux Arch Entwicklungsmech organ 9 (1900) 424. 15. Immers, J & Lundgren, B, Acta embryo1 exptl 2 (1972) 177. 16. Loewenstein, W R, Fed proc 32 (1973) 60. 17. Luft, J H, J biophys biochem cytol9 (1961) 409. 18. Lundgren, B, J submicr cytol 5 (1973) 61. 19. - J ultrastruct res 44 (1973) 440. 20. Oschman, J L & Wall, B J, J morph01 127 (1969) 475. 21. Prince, W T & Berridge, M J, J exptl biol 58 (1973) 367. 22. Reynolds, E S, J cell biol 17 (1963) 208. 23. Schraer, R, Elder, J A & Schraer, H, Fed proc 32 (1973) 1938. 24. Shaver, J R, Exptl cell res 11 (1956) 548. 25. Sternlieb, I, Z Zellforsch 93 (1969) 249. 26. Weibel, E R, Kistler, G S & Scherle, W F, J cell biol30 (1966) 23. 27. Wolpert, L & Gustafson, T, Exptl cell res 25 (1961) 374. Received December 19, 1973
Experimental Cell Research 85 (1974) 429436
NTITATIVE ESTIMATION OF T F MITOCHONDRIA TO SEPTA IN THE SEA URCHIN B. LUNDGREN Wenner-GwenInstitute, University of Stockholm, S-Ii3 45 Stockholm, Sweden
SUMMARY Associations of mitochondria with the septate desmosomes are often found in sea urchin embryos. In this investigation a semi-quantitative method of measuring the abundance of these associations is presented and applied to different developmental stages of the sea urchin larva. An increase of the mitochondrion-desmosome association is found before the primary mesenchyme formation. The association is also more common in the ectoderm than in the entoderm. The possible role of such associations in development is briefly discussed.
Associations between mitochondria and desmosomes have been reported in differentiating as well as adult tissues and seem to be common throughout the whole animal kingdom. Different functional roles have been ascribed to such associations, e.g. that mitochondria are necessary for desmosome formation c.5, 61, that they participate in active transport of water and solutes [ZO], and impulse propagation from cell to cell 251. Such associations have frequently been seen in the seaurchin larva (Gustafson, pers. comm.; [19]) in which the desmosomesare of the septate type [I, 2, 41. In view of the important function ascribed to cell adhesion, and hence also to desmosomes,in sea urchin development [13, 271 it was considered of interest to make a more detailed analysis of this phenomenon. A semi-quantitative method was developed to permit a comparison of the frequency of associations in different regions of the embryo and also in different develop-
mental stages. The results showed a marked increase in the frequency of these associations between 6 and 12 h after fertilization and a greater abundance in the ectoderm as compared with the entoderm.
Handling of speckmensfor electron microscopy Larvae of Psammechinus miliaris obtained by artificial fertilization were reared in round flasks with continuous stirring. Samples were taken for fixation at 6 h (unhatched blastula), 12 h (hatched blastula prior to formation of the primary mesenchyme), 22 h (gastrulae with fully extended archenteron) and 72 h (plutei). The material was fixed for 1 h in ! ‘% 0~0, in filtered sea water. After rinsing in sea water rhe larvae were dehydrated in ethanol and embedded in Epon 812 [17]. The embeddings were sectioned with an LKB Ultrotome (ultramicrotome). ‘Ihe sections (approx. 600 A thick) were picked up on one-hole grids and stained with uranyl acetate and lead citrate [22]. After drying, the sections were examined in. a Siemens Elmiskop I electron microscope at x 5 OBQ. The negatives were enlarged to a final ma of 20 000.
430 B. Lundgren
Fig. 1. Septate desmosome with associated mitochondria from the ectoderm of a 22 h old sea urchin larva (gastrula). x 45 000. Fig. 2. A desmosome site from a 6 h old unhatched blastula. The septa are not formed yet, the dark intercellular material is probably polysaccharides. The larva has been treated with ruthenium red in order to reveal extra-cellular polysaccharides. x 45 000.
Sampling The larvae were not oriented in any specific manner in the blocks. Pyramids containing one larva were prepared; from at least 5 different larvae 3 sets of sections were cut from each. Between the individual sets of sections 3-5 pm was removed. Sampling on the grid was performed by photographing the first usable section found on each grid. Micrographs were taken at the first point where the larval wall was hit by the beam, and then at intervals of two screen openings along the larval periphery. The total number of pictures per larva was 3-6 and per developmental stage 45-90. In the pluteus and in the archenteron of the gastrula the sampling was more restricted as only 3, respectively 4, different larvae were used.
Measurements and statistics Estimation of the mitochondrion-desmosome association was performed by comparing the distances between a point in the middle of the desmosome profile to the periphery of the nearest mitochondria profile (d.-m.) and from a randomly chosen point on the cell membrane in the cell junction to the nearest mitochondrial profile (r.-m.). The results were classified into 9 classes with intervals of 0.1 pm; all measurements exceeding 0.9 pm were pooled to make a tenth class. Mitochondrial fractional volume in the cytoplasm was determined by point counting [26]. A priori, the frequency with which mitochondria are associated with desmosomes may depend both on a specific association mechanism and also on the influence of unspecific factors such as the density of packing of mitochondria in the peripheral regions of the cell. Changes in the average mitochondrial packing in the cell were revealed by the figures for mitochondrial fractional volume. Regional changes in the mitochondrial packing in the cell were expected to be Exptl Cell Res 85 (1974)
revealed by an uneven distribution of frequencies in the r.-p. histogram. These unspecific factors will of course be equally registered in the d.-m. measurements; therefore to obtain a measure of the specific association of mitochondria to desmosomes, the frequency of random associations in the class o-O.1 were subtracted from the corresponding d.-m. frequency. The decision to use only this particular class will be considered in the Discussion.
RESULTS A close association between septate desmosomes and mitochondria was often found in embryos of Psammechinusmiliaris older than 6 h (fig. 1). In unhatched embryos of 6 h age such associations were rarely seen(fig. 2). The quantitative estimations of the mitochondrion-desmosome complexes for the different stages are summarized in figs 3-7. The distributions of measurements from random points on the plasmalemma to the nearest mitochondrion r-.-m. were relatively constant in all classesfor all stagesof development. However, the desmosome to mitochondria measures (d.-m.) accumulated in the class O-O.1,um in all stages except in the unhatched bastula. The relative frequences in the O-O.1pm class are given in table 1. No differences between d.-m. and r.-m. occurred
Estimation of rviktochon~~ion-~es~~osome~sso~~~t~~~ 43 1 Table 1. Mitochondrion-desmosome Stage (hOXS)
d.-m. in o-o.1
6 Ii 22 ect. t. 72 Zieb.
0.02 0.61 0.75 0.33 0.81
association and corrcation for rQ~dQrn-~o~~td~str~~~ti~~~
S.E.
P
0.016 0.037
0.001
in the 6 h blastula. In more advanced stages on the other hand more than half of the measures of d.-m. were found in this class, while r.-m. was higher than 0.1 in a single case. An interesting feature was the low degree of association in the archenteron of gastrulae as compared with the ectoderm. This difference may explain the lower figure for associations found in the 12 h blastula where no distinction was made between different regions. The figures obtained for the mitochondrion-desmosome association seem to be of high significance. The differences could be due, however, to an alteration of the mitochondrial fractional volume. In table 2 the results of %he morphometric assay of the mitochondrial fractional volume in the cytoplasm are summarized. The mitochondrial fractional volume in developmental stages from 6 to 22 h did not show any drastic variaTable 2. Mttochondrial
fractional
volume in
cytoplasm Stage (hours)
Fract. vol.
SE.
6 12 22 ect. ent. 72 cilieb.a
0.0568 0.0687 0.0690 0.0816 0. 1009a
0.0115Y
P
a The figures for 12 h pluteus ciliary band are not reliable because of the incomplete sampling from that stage. 28-741813
0.02 0.07 0.09 0.04 0.15
0 0.57 0.66 0.29 0.66
tions, but an increase was found between 6 and 12 h. The figure for the ciliate the pluteus is not reliable and is jus in the table for comparison. lit is remarkable that the mitochondrial fractional voltme in the archenteron of gastrula was higher than in the ectoderm. In fig. 8 the mito~bondrio~-~~srnosom~ association and mitochondrial fractional volume are brought together to give a clear, overall summary.
Associations between mitochondria an tate desmosomes ’ sea urchin larvae are wever, it is diffificuh to readily observed. decide on such evidence whether the association is due to true affinity between %he mitochondria and the desmosomes,or merely a passive consequenceof mitochondrial packing in the cytoplasm. ~urthermor~~ when using such a qualitative approach it is difficult to establish differences between varying stages of development and different parts of the larva. %f the mitochondrial associations were due to passive peripheral displacement, the entire lateral plasmalemma would be covered by mitomosomes. iif &is chondria as well were the case a mitochondria was randomly scattered in the cytoplasm, then a comparison belween any random point on the lateral plasmalemma and
432 B. Lundgren
4
l-
o-
5
6
the desmosome would be a measure of the specific association. One problem in this connection is the large nucleus found in the embryonic cell. Still, the method is usable if the nucleus does not represent a too big fraction of the cell. If the distribution of frequences in the r.-m. is dominated by very short and very long measures, it is probably due to a too-big nuclear fraction. An example
of this situation is found in fig. 7. The explanation may in this case also be a high mitochondrial fractional volume, which eliminates a large portion of the expected long measurements. In fig. 6 both these factors contribute to the special distribution of the r.-m. Another question in connection with the evaluation of the method is: “When is a mitochondrion associated with the desmo-
Exptl
Cell Res 85 (1974)
profiles will have larger and 13.5 srnailer values than half the actual rn~~o~bo~~rio~ diameter. This means that 86.4% of the pro-, files of associated mito~hondri~ will be found less than 0.125 pm from the desmosome and the rest 13.6% between 0.125 a pm from the desmosome. However, it will be impossible to ~ete~rni~~ whether a mito~bo~drio~ is associated. not, when using size istributions correlate to the distance unless the mito~bo~~ria are uniform in size and spherical in shape. These conditions are not met. The only usable sections would then be those cut in a plane at absolute right angle to the des plane. Con ring all these lirni~at~o~s~the problem se insoluble with a reasonable Figs 3-7. The histograms consist of two parts; the amount of work and accuracy. part over the zero line represents the distribution of To simplify the estimation, let us define frequences in the d.-m., and the part under the zero line, in the r.-m. The first nine classes are made with association as “a distance between desmoa &dtb of 0.1 pm while the tenth class (dotted) consists of all measures longer than 0.9 pm. No. 3, some and mitochondrion less than 0.1 ,uM”, 6 h unha.tched blastula; no. 4, 12 h blastula, just prior a distance still short enough to allow fast, to primary mesenchyme formation; no. 5, ectoderm nt interactions between the of a 22 h gastrula; no. 6, archenteron from the same diff usion-dep stage as no. 5; no. 7, ciliary band and animal-plate may simplify further; ““any components. from a 72 h pluteus. mitochondrion profile found within 0.1 p.m from the desmosome is cut from an associsome?” The most frequently occurring pro- ated mitochondrion”. This is a crude definifiles of mitochondria in sea urchin larvae are tion; associated mitochondria may be lost about 0.5 pm in diameter, and the profiles due to oblique sectioning, but owing to the are almost always circular. It is therefore relatively thick sections and the fact that reasonable to assumeas a first approximation that the mitochondria resemble spheres with a diameter close to 0.5 ,um. If we define the association “when the mitochondrial outer membrane is in contact at any point with the inner surface of the desmosomal membrane” (a case never observed in this investigation) the problem is, on the spherical mitochondrion, that one point alone will show whether an association exists or not. The possibility of finding that point, using 500 A sections, 6 will be 1 out of 10 in an associated mito8. Summary of the result of measurements of chondrion. For such a mitochondrion, accord- Fig. association (dotted) and the mitochondriai fractional ing to Elias & Wenning [7], 86.4% of the volume in the different stages investigated. CL%*- 741813
434 B. Lundgren the mitochondria seemto be flattened on the side facing the desmosome, it does not seem to interfere too much with the results. If there were large errors, they would be most frequently found in the 0.1-0.2 pm class. This is also the case in fig. 4, but the error is small when correcting for the random distribution and considering the relatively high frequency of the classO-01 ,um. Thus, the present method seemsadequate in comparing different developmental stages.It is doubtful, however, if it is usable for comparing different species or different organs in a highly differentiated animal. In such cases differences in the mean mitochondrial diameter must be considered and a more accurate definition of the association be found. In conclusion, the method is no clearcut morphometric assay and the figures are not valid in absolute terms but are still usable as relative figures when comparing developmental stages of the same species. Results In the 6 h old unhatched blastula no association of mitochondria to desmosomes was found. Older stages, 12, 22 h and 72 h larvae, show a well developed mitochondriondesmosome association, at least in the ectoderm. Various theories for the functional significance for such associations have been put forward by different authors. Most investigations are, however, made on vertebrate epithelia and concern a different type of desmosome. For example Deane et al. [5, 61suggestedthat in young and fetal mammals mitochondria were necessary for the formation of the plaque desmosome as a source of energy and metabolites. But, in the larvae of Psammechinus miliaris, the septate desmosomes are formed 1 or 2 h before hatching (about 6 h after fertilization). At this stage the results provide no evidence of an association. Therefore mitochondria do not seem Exptl Cell Res 85 (1974)
to be involved in the formation of the septate desmosomesin Psammechinusmiliaris. It has been known since the turn of the century that a sea urchin embryo transferred into calcium-deficient sea water is dispersed into a cell suspension, which has the ability to recombine into an intact larva when transferred back to normal sea water [8, 141. In other words, calcium-deficient sea water weakens the cell contacts. This calcium dependencemay give a clue to the functional significance of the association between mitochondria and desmosomes.Mitochondria in many tissues are known to be capable of actively accumulating and releasing calcium in response to cellular requirements [3, 231. Therefore it may be hypothesized that the mitochondria might increase the stability of calcium-dependent desmosomesby buffering the intercellular calcium concentrations in case of fluctuations of the exogenous calcium concentrations. This hypothesis has one major weakness; it is not likely that intracellular calcium is affected when larvae are immersed in Ca2+deficient water. The calcium deficiency probably works on ionic bridges in the thin extracellular polysaccharide coating of the plasmalemma [ 181,thereby dispersing the cellcementing capacity of the polysaccharides. However, some support for the theory may be adduced in the fact that no association is found at early stages before hatching. It is possible that the good protection against dispersion provided by the fertilization membrane makes a calcium buffer system unnecessary, whereas in later stages, when no fertilization membrane exists, a mitochondrion-desmosome association is needed. Another function suggestedfor the desmosome is the propagation of intracellular impulses via channels in the desmosome [16]. A possible model for such transmissions could be that of Prince & Berridge [21] in
Estimation of mitochondrion-clesmosowe a$sociattom 435
which membrane potential alterations are influenced by LAMP via @a2+release from intracellular deposits. Thus, this theory could also involve the calcium storing capacity of m.itochondria. Some of the results in the present investigation can be taken as support for the impulse propagation hypothesis. The ciliary band of the 72 h pluteus showed a very high mitochondrion-desmosome association. This area is, as concluded in recent investigations [1 1, 12, 151, an excitatory centre, where mechanical or chemical stimulation is rapidly transformed into impulses which are transmitted towards the interior regions of the larva. Furthermore, the archenteron of the 22 h gastrula did not show any marked association, despite the large mitochondrial fractional volume. This is in line with the results of Gustafson et al. [ll]; in larvae irritated by mechanical stimulation, local lobular contractions of the stomach wall occurred at points where neuron-like strands make contact. A good cell-to-cell communication within the stomach wall would probably have led to a general lobulation of the whole stomach. Earlier investigations utilizing vital staining IlO] and particle counting from homogenates 1241were generally not in agreement. Both types of investigation showed, however, an increase in mitochondrial number prior to primary mesenchymeformation. In the work by Gustafson Lenicque [lo], where nileblue sulphate was used as vital stain, the increase was very pronounced. The work by Shaver [24], however, showed a less pronounced increase from about the stage corresponding to the 6 h unhatched Psammechinus miliaris blastula, which is in line with the slight increase in mitochondrial fractional volume found in the present investigation. A priori, changes of mitochondrial fractional volume cannot be exactly compared with the alterations of mitochondrial numbers in the
works of Gustafson & kenicque and of Shaver, but in practice it can as the mean rnit~~~o~~rial profile diameter is not subject to any drastic variations between the stagesin this investigation. The very marked increase in ~ito~~o~~~~a~ number in the IO-12 h blastnla showed by Gustafson & kenicque occurs simultaneous with the increase in mito~ho~drion-desmosome association found in the present work. Furthermore, the low mitochondrial count found by Gustafson & Eenicque IlO] in vegetal parts, in contrast to the relatively high fractional volume found in the present i tion, also reflects the low mitoch desmosome association in this part of the embryo. The results suggest some kind of activation of mitochondria between 10 and 12 h after fertilization, leading to an affinity to the desmosome and also to an increased stainability with nile-blue sulphate. activation may very well reflect an increase in the intercellular communication, wh~.ch may be necessary to coordinate the morphogenetic processes starting at that stage with the formation of the primary rne~e~~~yrne~ The present work therefore supports the conclusions of Gustafson & Lenicque [lo] where a definite change in the physiology of the mitochondrion according to a characfesistic space time pattern occurs; see further Gustafson [9].
The author wishes to thank Professor Bertil Swedmark and his staff at the Kristineberg Zoological Station, Fiskebackskil, Sweden, for arranging working facilities and supplying the animals for the &vestigaI tion. My warmest thanks are also directed to Tudor Barnard for valuable discussions during the experiments and in preparing the manuscript. Professors Bjorn Afzelius and Tryggve Gustafson have also been of great help in the preparation of the manuscript. The skilful technical assistance by Mrs Carina Starkernd has been of greatest importance and is gratefully acknowledged. This investigation was supported by grants from the Swedish Natural Science Research Council.
436 B. Lundgren
REFERENCES 1. 2. 3. 4. 5. 6. 7.
8. 9. 10. 11. 12.
Agrell, I P S, Z Zellforsch 72 (1966) 12. Balinsky, B I, Exptl cell res 16 (1959) 429. Borle, A B, Fed proc 32 (1973) 1944. Dan, K, Int rev cytol 9 (1960) 321. Deane, H W & Wurzelmann, S, Proc 6th int congress EM, Kyoto (1966). Deane, H W, Wurzelmamr, S & Kostellow, A B, Z Zellforsch 75 (1966) 166. Elias, H & Henning, A, Quantitative methods in morphology (ed E R Weibel & H Elias) p. 130. Springer-Verlag, Berlin, Heidelberg, New York (1967). Giudice, G, Dev biol 5 (1962) 402. Gustafson, T, The biochemistry of animal development (ed R Weber) vol. 1, p. 139. Academic Press, New York and London (1965). Gustafson, T & Lenicque, P, Exptl cell res 3 (1952) 251. Gustafson, T, Lundgren, B & Treufeldt, R, Exptl cell res 72 (1972) 115. Gustafson, T, Ryberg, E & Treufeldt, R, Ada embryo1 expt12 (1972) 199.
Exptl Cell Res 85 (1974)
13. Gustafson, T & Wolpert, L, Biol rev Cambridge philosoph sot 42 (1967) 442. 14. Herbst, C, Wilhelm Roux Arch Entwicklungsmech organ 9 (1900) 424. 15. Immers, J & Lundgren, B, Acta embryo1 exptl 2 (1972) 177. 16. Loewenstein, W R, Fed proc 32 (1973) 60. 17. Luft, J H, J biophys biochem cytol9 (1961) 409. 18. Lundgren, B, J submicr cytol 5 (1973) 61. 19. - J ultrastruct res 44 (1973) 440. 20. Oschman, J L & Wall, B J, J morph01 127 (1969) 475. 21. Prince, W T & Berridge, M J, J exptl biol 58 (1973) 367. 22. Reynolds, E S, J cell biol 17 (1963) 208. 23. Schraer, R, Elder, J A & Schraer, H, Fed proc 32 (1973) 1938. 24. Shaver, J R, Exptl cell res 11 (1956) 548. 25. Sternlieb, I, Z Zellforsch 93 (1969) 249. 26. Weibel, E R, Kistler, G S & Scherle, W F, J cell biol30 (1966) 23. 27. Wolpert, L & Gustafson, T, Exptl cell res 25 (1961) 374. Received December 19, 1973