Accumulation of microfilaments in a colonial mutant of Neurospora crassa

Show preview

Hide preview

Printed in Sweden Copyright © 1974 by Academic Press, Inc. All rights of reproduction in any form reserved
J. ULTRASTRUCTURE RESEARCH 48, 455-464 (1974) 455
Accumulat ion of Microf i laments in a Colonial Mutant
of Neurospora crassa
EDWARD D. ALLEN, ROBERT J. LOWRY, and ALFRED S. SUSSMAN
Veteran's Administration Hospital, Ann Arbor, Michigan 48105, and Department o.f Botany, University of Michigan, Ann Arbor, Michigan 48104
Received November 14, 1973, and in revised form February 21, 1974
A morphological mutant of Neurospora erassa, snowflake, is shown to contain filaments which are about 70 A in diameter, and up to several microns long, and which usually bunch in groups of a few to several hundred. They may be found longitudinally or transversely arranged with respect to the long axis of the cell and, in many cases, they run up to the plasma membrane, but not through it. The filaments often are arranged in crystalline arrays but may also be found as separate filaments. Sometimes the filaments are closely appressed to nuclei and may be found inside them. It is likely that the filaments are not the result of the dissociation of microtubules and are most likely microfilaments like those found in other organisms. Their relationship to the origin of certain morphological mutants in Neurospora is discussed.
Cytoplasmic filaments ranging in size from 40-120 A in diameter and up to several microns long have been described in many organisms. It has been suggested that filaments in the smaller size range (50-70 ,~), generally referred to as microfilaments, are actinlike and involved in contractile processes.
Therefore, it has been hypothesized that microfilaments are involved in the fol- lowing processes: cytoplasmic streaming in Acanthamoeba (19), Physarum (7, 13), Nitella (14), and higher plants (15, 16); cell elongation in Acytostelium (8), epidermal contraction in Amaroricium (4); and cleavage in jelly fish and polychaete worms (24), and squid cells (1). Other aspects of morphogenesis in animal cells also are believed to be under the control of microfilaments because cytochalasin B, a drug that may disrupt these, can arrest development when added to some systems (26). However, conflicting results have been obtained with this drug and its action on microfila- ments has been questioned (9).
Despite the vast amount of literature on cytoplasmic filaments their fine structure has not been delineated in thin sections. Hatano and Oosawa (7) showed by nega- tive staining that the F-actin they isolated from Physarum is composed of a two- stranded helix. Similarly, negative staining of actinlike filaments, isolated from

456 ALLEN, LOWRY AND SUSSMAN
other organisms, also suggests such an arrangement (13, 18, 19). When thin sections :are used, the helical nature of the microfilaments has been derived from the beaded appearance of the microfilaments when viewed longitudinally. On the other hand, the filaments in transverse sections generally have been described as electron-dense dots, giving the impression that the filaments are solid cylinders.
In general, morphological mutants of Neurospora crassa are characterized by a very slow growth rate (0.5 mm/hour) and a higher frequency of branching as com- pared with wild-type strains. This results in colonies which have very compact my- celia and a mycelial front which advances very slowly. Some morphological mutants may not be as compact as others, but all differ appreciably from wild-type in linear growth rate of the colony, amount of aerial mycelium, compactness of aerial myce- lium, and frequency of branching (12). The various mutants have been described mainly on the basis of the growth rates, a necessity imposed by lack of knowledge of the biochemical lesions of the mutants (6).
During the investigation of several morphological mutants of Neurospora crassa we found one which possesses large numbers of cytoplasmic filaments whose struc- ture could be analyzed. This paper describes the appearance of these filaments in thin sections of fixed material.
MATERIALS AND METHODS
Organism. The morphological mutant of Neurospora crassa used for this study, strain 507 (snowflake), and the wild-type (standard strains), 4121a, and the Lindegren strain, were obtained through the courtesy of the Fungal Genetics Stock Center, Humboldt State College, Arcata, California.
Culture conditions. Cultures of Neurospora were grown in 500-ml Erlenmeyer flasks con- taining 200 ml of distilled water with 2 % sucrose and 2 % of 50 X Vogel's standard salts (25). Growth was initiated in flasks either by hyphal or conidial transfer.
Preparation for electron microscopy. Hyphae from liquid cultures were placed in vials con- taining fixatives at 0°C or room temperature. The fixative used most regularly was 3 % glutaraldehyde in 0.1 M sodium cacodylate (pH 7.4) at 0°C for 2 hours. Other fixatives included (a) 3 % glutaraldehyde in 0.05 M potassium phosphate (pH 7.2) or (b) 1.5 % of both glutaraldehyde and paraformaldehyde in 0.1 M sodium cacodylate (pH 7.4) or 0.05 M potassium phosphate (pH 7.2), or (c) 3 % acrolein in either of the above buffers. After fixa- tion, the hyphae were washed several times and left in buffer overnight.
FIG. 1. Longitudinally oriented microfilaments in three cells. Arrows indicate groups of microfila- ments, x 8 603. FIG. 2. Transverse section of hyphae showing several groups of microfilaments in transverse and longitudinal view. Arrow at ribosomes (R) indicates the small relative size of the microfilaments. × 49 800. Inset: higher magnification of region indicated by outline. Arrow indicates tubular appear- ance of one mierofilament in transverse view. x 164 000. FIG. 3. Three masses of microfilaments, one of which is perpendicular to septum (S). × 16 000.

MICROFILAMENTS IN Ngurospora 457

458 ALLEN, LOWRY AND SUSSMAN
Postfixation (1 hour at 0°C) was carried out with the buffer used for fixation diluted I : 1 with aqueous 2 % osmium tetroxide.
Dehydration in ethanol was followed by transfer to propylene oxide and embeddment in Epon [25 DDSA: 10 Epon 812:0.8 DMP-30 (23)].
Sections were cut with diamond or glass knives and stained with 2% aqueous uranyl acetate followed by lead citrate (21). The sections were viewed with either Philips 300, RCA-EMU4, or Hitachi 11 microscopes.
The microscope was calibrated with a carbon grating replica (28 800 lines per inch) ob- tained from E. F. Fullam Inc., Schenectady, New York. Microfilaments were measured on micrographs enlarged 4 times from negatives taken at 54 000 × at the time of calibration.
RESULTS
The growth pattern of snowflake (strain 507) is characteristic of many of the morphological mutants of Neurospora. Its hyphae radiate outward from a common center after hyphal or conidial transfer and grow at a rate of approximately 0.3 mm per hour compared to about 3.6 mm per hour for wild-type strains. The length of the cells of snowflake is about 100 #m while that of the standard strains used is about 180-200 #m. The width of the cells of snowflake is about 20 ,urn which is approxi- mately one-half that of standard strains.
At the ultrastructural level, the main thing which distinguishes sections of snowflake f rom those of standard strains is the presence of large masses of a fi lamentous ma- terial. Unlike the wild-type strains, in which we have seen a few groups of filaments only in young vegetative hyphae, snowflake commonly contains masses of filaments in vegetative cells.
Figure 1 shows a section of three cells of snowflake, all of which possess the fila- mentous masses.
Filaments appear in groups ranging from a few to several hundred, and nearly all are found dispersed in the cytoplasm. They are not oriented in any particular direction with respect to the long axis of the cell; that is, they may be found in longitudinal or cross-sectional view when the cell is oriented longitudinally (Figs. 2 and 3). However, when extremely large masses of filaments are seen, they are usually oriented parallel to the longitudinal axis of the cell. The filaments are as common in apical cells as in those farther away from the growing tip. However, the extremely
FIG. 4. Group of microfilaments perpendicular to, but not passing through, the plasma membrane at cell wail (W). The microfilaments appear granular near the plasma membrane, x 71 000. FIG. 5. Appearance of microfilaments in hexagonal arrangement (arrows). The presence of longi- tudinally aligned filaments near the hexagonal ones suggests that some of the groups of microfila- ments may be braided arrays, x 61 600. FIG. 6. Longitudinal view of microfilaments in cytoplasm as well as microfilament-like structures in nucleus (arrow). Those found ~n nuclei appear the same as those observed in the cytoplasm, x 37 000. FIG. 7. Nucleus possessing microfilament-like structures (arrows) in transverse view. × 65 000.

MICROFILAMENTS IN Nellrospora 459

460 ALLEN~ LOWRY AND SUSSMAN
large masses appear more commonly in older cells. In many cases, longitudinal groups are seen to run up to the plasma membrane at the septum (Fig. 3) or wall (Fig. 4), but none have been seen to pass through it.
In longitudinal view, the individual filaments are approximately 70 A wide and up to several microns long. Their mean diameter is 68.8 •+ 15 ~ and the range is from 40 to slightly above 100 N. These values, and the standard deviation, were obtained from measurements of 111 filaments. Since the size of these filaments falls into the range of microfilaments, they will be described as such hereafter. Micro- filaments are in close apposition and appear straight or slightly curved, and when in the former condition they present a somewhat crystalline appearance (Fig. 5). How- ever, at other times they may appear loose and separated (Fig. 3).
In cross-sectional views the microfilaments appear as electron-dense dots approxi- mately 70 N in diameter with up to several hundred grouped together in any one area. Occasionally, the microfilaments appear packed into hexagonal arrays (Fig. 5).
The only organelle in the cell with which the filaments seem to be associated is the nucleus. Several nuclei have been found to possess filamentous strands which are very similar in appearance to the microfilaments found in the cytoplasm (Fig. 6). The numbers seen in nuclei never approach those found in the cytoplasm. A single nucleus has been observed to possess about twenty of these strands in cross-sectional view (Fig. 7), which are similar in size to the microfilaments observed in the cyto- plasm. We have not observed filaments in the nuclei of wild-type strains, but Beck et al. (2) have.
At higher magnifications, the usual appearance of cross sections of microfilaments is somewhat irregular and not distinctly circular. A few of the microfilaments at higher magnification appear to possess subunits (Figs. 8 and 9). The range of number of subunits in those microfilaments which show them is from two to four. The diameter of the subunits is approximately 25-35 A.
A few of the microfilaments in cross section give the impression of being hollow tubes (Fig. 2, inset). The number of microfilaments which are tubular in cross-sec- tional view is very low compared to the number which show subunits, although the actual number of either is small when one considers the total number of microfila- ments in any area. It should be mentioned that only extremely thin sections (dark- gray to black) permitted the visualization of this subunit structure.
On viewing longitudinally oriented microfilaments at higher magnification, it ap- pears that they are actually composed of more than one strand (Figs. 10 and 11).
As described in the section Materials and Methods, several different fixatives were used. Although they varied greatly in the effectiveness of preservation of back- ground cytoplasm and various organelles, all showed the microfilaments to be the same size. Microtubules in the range 180-240 N were never observed in the cyto-

MICROFILAMENTS IN Neurospora
~:~ ....
461
Fr~s. 8 and 9. Transversely oriented microfilaments which show subunit structure of individual members. Microfilaments indicated at arrows show 2 or 3 subunits. Fig. 8, x 185 000; Fig. 9, × 385 C00. Ft~s. 10 and 11. Higher magnification of longitudinally oriented microfilaments. Arrows indicate strand subunits of individual microfilaments. Fig. 10, x 105 000; Fig. 11, × 185 000.
plasm, including the regions where the masses of microfi laments were found. How-
ever, microtubules were found in a few nuclei after fixation in 3 % glutaraldehyde
in phosphate buffer. These were best observed in cross sections since it was difficult
to determine whether longitudinal components in the nuclei were fi laments or micro-
tubules.

462 ALLEN, LOWRY AND SUSSMAN
The presence of paraformaldehyde in the fixative containing glutaraldehyde pre- serves more information than the latter alone. This is evident in the retention of an electron dense outer cell wall (Fig. 1) and a denser mitochondrial matrix, neither of which are seen in cells only in glutaraldehyde.
DISCUSSION
The microfilaments just described have infrequently been observed by us in wild- type strains of Neurospora grown and prepared for electron microscopy under the same conditions used for snowflake. But several investigators have observed fila- ments in standard strains of Neurospora, including Beck et al. (2), who have shown that cultures obtained 4 hours after inoculation on media containing 2 % sucrose or 15% glucose possess filaments ("striated inclusions"). These microfilaments, whose appearance is identical to those in snowflake, are 62 N in diameter and appear in the nucleus as well as the cytoplasm. The authors refer to the microfilaments as "rods," and they believe that individual elements are linked by fine filamentous strands and that the rods may be tubular.
As mentioned above, the individual microfilaments in snowflake do not appear to be rods, or solid cylinders. Instead, they appear to be composed of strands when observed in longitudinal sections, a conclusion that is supported by the subunits seen in cross-sectional view. We have not observed fine filamentous strands con- necting microfilaments and feel that the tubular appearance of some microfilaments in cross-sectional view is related to their orientation. That is, osmium tetroxide used as a postfixative could contribute to the electron density of the strands, especially through thicker sections, thereby giving the impression of a continuous cylinder or a tubule, depending on whether the microfilaments are oriented in an exact transverse plane when sectioned.
Filaments in standard strains of Neurospora also have been reported by Wood and Luck (28), who found "paracrystalline inclusions" in a mitochondrial mutant (abnormal - 1). These inclusions could be induced in standard strains by the addition of ethidium bromide or euflavine to the culture medium. The authors describe the filaments as being approximately 70 A wide and dotlike in cross section. After isola- tion and negative staining this material showed a banded pattern formed by units 86 N in length, arranged end-to-end. Wood and Luck found that the protein from paracrystalline inclusions is not a product of mitochondrial genes, and believe that it crystallizes as a result of a mitochondrial defect.
The variation in the diameter of Neurospora microfilaments (approximately 40- 100 A) could be caused by any of several factors related to fixation, orientation dur- ing sectioning, staining, etc. However, such variability could also be explained if

MICROFILAMENTS IN Neurospora 463
microfilaments possessed different numbers of subunit strands. Pollard and/to (17) noted that the smaller filaments in Amoeba range from 50 to 92 A in diameter. This size range might also indicate that variations in the number of subunit strands exist in Amoeba and might also imply that not all small filaments are like actin in forming a two-stranded helix. Variations in the number of strands in individual filaments could explain the large size variations which exist among many small fila- ments and could provide a mechanism to explain how larger filaments are formed, i.e., by increased numbers of subunit strands. Such a mechanism does not, however, explain the short (0.5 #m) thick filaments in Amoeba (17) or the reason why heavy meromyosin binds only to the smaller filaments.
There is evidence that some of the microfilaments seen in electron micrographs are subunits of microtubules. Schultz and Case (22) have shown that neuronal microtubules are replaced by microfilaments in bicarbonate-buffered fixatives. On the other hand, fixatives containing cacodylate or phosphate buffers give normal microtubules. Microfilaments also have been seen after the use of mitotic inhibitors, such as colchicine, vinblastine sulfate, and podophyllotoxin, which are known to disrupt microtubules of the spindle apparatus (27). Consequently, we considered the possibility that the microfilaments observed in snowflake are artifacts of the kind discussed above. Glutaraldehyde when used in combination with osmium tetroxide has been shown to preserve microtubules. After using this fixation procedure as well as the others described in the Materials and Methods section, large numbers of microfilaments still were observed in snowflake and were never found assembled into microtubules.
Furthermore, where microtubules (250 A in diameter) have been observed in nuclei of snowflake, microfilaments of the usual kind still were found in the cytoplasm. Since nuclear microtubules were preserved it is unlikely that those occurring in the cytoplasm of the same cell would not be. Additionally, since microtubules have not been seen in the cytoplasm of Neurospora under fixation regimes that fix other organelles well, it seems unlikely that microfilaments are masses of microtubule sub- units resulting from poor fixation.
This study points to another hypothesis for the origin of certain morphological mutants of Neurospora. Such mutants are characterized, in comparison with standard strains, by the formation of dense colonies with a reduced rate of forward extension and an increased frequency of branching. One hypothesis is based upon the data of Tatum and collaborators (5, 10), which suggest that the walls of mutants contain significantly less glucose and more glucosamine than those of standard strains. In other mutants the primary effect has been traced to the formation of defective en- zymes (3), or higher activites of some (11). That a mucopolysaccharide polymer may regulate the rate of growth by effects upon the plasma membrane has been proposed 30-74•823 J. Ultrastructure Research

464 ALLEN, LOWRY AND SUSSMAN
in another case (20). Our finding of masses of microfilaments in snowflake suggests the additional possibility that the altered growth habit of this mutant is related to the accumulation of this material, perhaps because of effects upon one or more of the physiological processes in which microfilaments have been implicated.
REFERENCES
1. ARNOLD, J. M., Y. Cell Biol. 41, 894 (1969). 2. BECK, D. P., DECKER, G. L. and GREENWALT, J. W., J. Ultrastruct. Res. 33, 245 (1970). 3. BRODY, S. and TATUM, E. L., Proc. Nat. Acad. Sci. U.S. 58, 923 (1967). 4. CLONEY, R. A., J. Ultrastruct. Res. 14, 300 (1966). 5. DETERRA, N. and TATUM, E. L., Science 134, 1066 (1963). 6. GARNJOBST, L. and TATUM, E. L., Genetics 57, 579 (1967). 7. HATANO, S. and OOSAWA, F., J. Cell Physiol. 68, 197 (1966). 8. HOWL, H. R., HAMAMOTO, S. T. and HEMMES, D. E., Amer. Y. Bot. 55, 783 (1968). 9. HOLTZER, H. and SANGER, J. W., Develop. Biol. 27, 444 (1972).
10. MAnADEVAN, P. R. and TATUM, E. L., J. Bacteriol. 90, 1073 (1965). 11. MAHADEVAN, P. R. and MA~ADKAR, U. R., Y. Baeteriol. 101, 941 (1970). 12. MURRAY, J. C. and SRB, A. M., Can. J. Bot. 40, 351 (1962). 13. NACHMIAS, V. T., HUXLEY, H. E. and KESSLER, D., J. Mol. Biol. 50, 83 (1970). 14. NAGM, R. and REBHUN, L. I., J. Ultrastruct. Res. 14, 571 (1966). 15. O'BRIEN, T. P. and THIMANN, K. V., Pro& Nat. Acad. Sci. U.S. 56, 888 (1966). 16. PARTHASARATHY, M. V. and MOHLETHALER, K., J. Ultrastruct. Res. 38, 46 (1972). 17. POLLARD, T. D. and ITO, S., J. Cell Biol. 46, 267 (1970). 18. POLLARD, T. D. and KORN, E. D., J. CellBiol. 48, 216 (1971). 19. POLLARD, T. D., SHELTON, E., WEIHING, R. R. and KORN, E. D., J. MoI. Biol. 50, 91
(1970). 20. RHSSIG, J. L. and GLASGOW, J. E., J. Bacteriol. 106, 882 (1971). 21. REYNOLDS, E. S., J. Cell Biol 17, 208 (1963). 22. SCHULTZ, R. L. and CASE, N. M., J. Cell Biol. 38, 633 (1968). 23. S~eORN, M. B., WANKO, T. and DINGMAN, W., J. Cell Biol. ]15, 109 (1962). 24. SZOLLOSL D., J. Cell Biol. 44, 192 (1970). 25. VOGEL, H., Amer. Natur. 98, 435 (1964). 26. WESS~LLS, N. K., SPOONER, B. S., As~, J. F., BRADLEY, M. O., LUDUENA, M. A., TAYLOR,
E. L., WRENN, J. T. and YAMADA, K. M., Science 171, 135 (1971). 27. WISNIEWSKL H., SHELANSKL M. L. and TERRY, R. D., Y. Cell Biol. 38, 224 (1968). 28. WooD, D. D. and LUCK, D. J. L., J. Cell Biol. 51, 249 (1971).Printed in Sweden Copyright © 1974 by Academic Press, Inc. All rights of reproduction in any form reserved
J. ULTRASTRUCTURE RESEARCH 48, 455-464 (1974) 455
Accumulat ion of Microf i laments in a Colonial Mutant
of Neurospora crassa
EDWARD D. ALLEN, ROBERT J. LOWRY, and ALFRED S. SUSSMAN
Veteran's Administration Hospital, Ann Arbor, Michigan 48105, and Department o.f Botany, University of Michigan, Ann Arbor, Michigan 48104
Received November 14, 1973, and in revised form February 21, 1974
A morphological mutant of Neurospora erassa, snowflake, is shown to contain filaments which are about 70 A in diameter, and up to several microns long, and which usually bunch in groups of a few to several hundred. They may be found longitudinally or transversely arranged with respect to the long axis of the cell and, in many cases, they run up to the plasma membrane, but not through it. The filaments often are arranged in crystalline arrays but may also be found as separate filaments. Sometimes the filaments are closely appressed to nuclei and may be found inside them. It is likely that the filaments are not the result of the dissociation of microtubules and are most likely microfilaments like those found in other organisms. Their relationship to the origin of certain morphological mutants in Neurospora is discussed.
Cytoplasmic filaments ranging in size from 40-120 A in diameter and up to several microns long have been described in many organisms. It has been suggested that filaments in the smaller size range (50-70 ,~), generally referred to as microfilaments, are actinlike and involved in contractile processes.
Therefore, it has been hypothesized that microfilaments are involved in the fol- lowing processes: cytoplasmic streaming in Acanthamoeba (19), Physarum (7, 13), Nitella (14), and higher plants (15, 16); cell elongation in Acytostelium (8), epidermal contraction in Amaroricium (4); and cleavage in jelly fish and polychaete worms (24), and squid cells (1). Other aspects of morphogenesis in animal cells also are believed to be under the control of microfilaments because cytochalasin B, a drug that may disrupt these, can arrest development when added to some systems (26). However, conflicting results have been obtained with this drug and its action on microfila- ments has been questioned (9).
Despite the vast amount of literature on cytoplasmic filaments their fine structure has not been delineated in thin sections. Hatano and Oosawa (7) showed by nega- tive staining that the F-actin they isolated from Physarum is composed of a two- stranded helix. Similarly, negative staining of actinlike filaments, isolated from

456 ALLEN, LOWRY AND SUSSMAN
other organisms, also suggests such an arrangement (13, 18, 19). When thin sections :are used, the helical nature of the microfilaments has been derived from the beaded appearance of the microfilaments when viewed longitudinally. On the other hand, the filaments in transverse sections generally have been described as electron-dense dots, giving the impression that the filaments are solid cylinders.
In general, morphological mutants of Neurospora crassa are characterized by a very slow growth rate (0.5 mm/hour) and a higher frequency of branching as com- pared with wild-type strains. This results in colonies which have very compact my- celia and a mycelial front which advances very slowly. Some morphological mutants may not be as compact as others, but all differ appreciably from wild-type in linear growth rate of the colony, amount of aerial mycelium, compactness of aerial myce- lium, and frequency of branching (12). The various mutants have been described mainly on the basis of the growth rates, a necessity imposed by lack of knowledge of the biochemical lesions of the mutants (6).
During the investigation of several morphological mutants of Neurospora crassa we found one which possesses large numbers of cytoplasmic filaments whose struc- ture could be analyzed. This paper describes the appearance of these filaments in thin sections of fixed material.
MATERIALS AND METHODS
Organism. The morphological mutant of Neurospora crassa used for this study, strain 507 (snowflake), and the wild-type (standard strains), 4121a, and the Lindegren strain, were obtained through the courtesy of the Fungal Genetics Stock Center, Humboldt State College, Arcata, California.
Culture conditions. Cultures of Neurospora were grown in 500-ml Erlenmeyer flasks con- taining 200 ml of distilled water with 2 % sucrose and 2 % of 50 X Vogel's standard salts (25). Growth was initiated in flasks either by hyphal or conidial transfer.
Preparation for electron microscopy. Hyphae from liquid cultures were placed in vials con- taining fixatives at 0°C or room temperature. The fixative used most regularly was 3 % glutaraldehyde in 0.1 M sodium cacodylate (pH 7.4) at 0°C for 2 hours. Other fixatives included (a) 3 % glutaraldehyde in 0.05 M potassium phosphate (pH 7.2) or (b) 1.5 % of both glutaraldehyde and paraformaldehyde in 0.1 M sodium cacodylate (pH 7.4) or 0.05 M potassium phosphate (pH 7.2), or (c) 3 % acrolein in either of the above buffers. After fixa- tion, the hyphae were washed several times and left in buffer overnight.
FIG. 1. Longitudinally oriented microfilaments in three cells. Arrows indicate groups of microfila- ments, x 8 603. FIG. 2. Transverse section of hyphae showing several groups of microfilaments in transverse and longitudinal view. Arrow at ribosomes (R) indicates the small relative size of the microfilaments. × 49 800. Inset: higher magnification of region indicated by outline. Arrow indicates tubular appear- ance of one mierofilament in transverse view. x 164 000. FIG. 3. Three masses of microfilaments, one of which is perpendicular to septum (S). × 16 000.

MICROFILAMENTS IN Ngurospora 457

458 ALLEN, LOWRY AND SUSSMAN
Postfixation (1 hour at 0°C) was carried out with the buffer used for fixation diluted I : 1 with aqueous 2 % osmium tetroxide.
Dehydration in ethanol was followed by transfer to propylene oxide and embeddment in Epon [25 DDSA: 10 Epon 812:0.8 DMP-30 (23)].
Sections were cut with diamond or glass knives and stained with 2% aqueous uranyl acetate followed by lead citrate (21). The sections were viewed with either Philips 300, RCA-EMU4, or Hitachi 11 microscopes.
The microscope was calibrated with a carbon grating replica (28 800 lines per inch) ob- tained from E. F. Fullam Inc., Schenectady, New York. Microfilaments were measured on micrographs enlarged 4 times from negatives taken at 54 000 × at the time of calibration.
RESULTS
The growth pattern of snowflake (strain 507) is characteristic of many of the morphological mutants of Neurospora. Its hyphae radiate outward from a common center after hyphal or conidial transfer and grow at a rate of approximately 0.3 mm per hour compared to about 3.6 mm per hour for wild-type strains. The length of the cells of snowflake is about 100 #m while that of the standard strains used is about 180-200 #m. The width of the cells of snowflake is about 20 ,urn which is approxi- mately one-half that of standard strains.
At the ultrastructural level, the main thing which distinguishes sections of snowflake f rom those of standard strains is the presence of large masses of a fi lamentous ma- terial. Unlike the wild-type strains, in which we have seen a few groups of filaments only in young vegetative hyphae, snowflake commonly contains masses of filaments in vegetative cells.
Figure 1 shows a section of three cells of snowflake, all of which possess the fila- mentous masses.
Filaments appear in groups ranging from a few to several hundred, and nearly all are found dispersed in the cytoplasm. They are not oriented in any particular direction with respect to the long axis of the cell; that is, they may be found in longitudinal or cross-sectional view when the cell is oriented longitudinally (Figs. 2 and 3). However, when extremely large masses of filaments are seen, they are usually oriented parallel to the longitudinal axis of the cell. The filaments are as common in apical cells as in those farther away from the growing tip. However, the extremely
FIG. 4. Group of microfilaments perpendicular to, but not passing through, the plasma membrane at cell wail (W). The microfilaments appear granular near the plasma membrane, x 71 000. FIG. 5. Appearance of microfilaments in hexagonal arrangement (arrows). The presence of longi- tudinally aligned filaments near the hexagonal ones suggests that some of the groups of microfila- ments may be braided arrays, x 61 600. FIG. 6. Longitudinal view of microfilaments in cytoplasm as well as microfilament-like structures in nucleus (arrow). Those found ~n nuclei appear the same as those observed in the cytoplasm, x 37 000. FIG. 7. Nucleus possessing microfilament-like structures (arrows) in transverse view. × 65 000.

MICROFILAMENTS IN Nellrospora 459

460 ALLEN~ LOWRY AND SUSSMAN
large masses appear more commonly in older cells. In many cases, longitudinal groups are seen to run up to the plasma membrane at the septum (Fig. 3) or wall (Fig. 4), but none have been seen to pass through it.
In longitudinal view, the individual filaments are approximately 70 A wide and up to several microns long. Their mean diameter is 68.8 •+ 15 ~ and the range is from 40 to slightly above 100 N. These values, and the standard deviation, were obtained from measurements of 111 filaments. Since the size of these filaments falls into the range of microfilaments, they will be described as such hereafter. Micro- filaments are in close apposition and appear straight or slightly curved, and when in the former condition they present a somewhat crystalline appearance (Fig. 5). How- ever, at other times they may appear loose and separated (Fig. 3).
In cross-sectional views the microfilaments appear as electron-dense dots approxi- mately 70 N in diameter with up to several hundred grouped together in any one area. Occasionally, the microfilaments appear packed into hexagonal arrays (Fig. 5).
The only organelle in the cell with which the filaments seem to be associated is the nucleus. Several nuclei have been found to possess filamentous strands which are very similar in appearance to the microfilaments found in the cytoplasm (Fig. 6). The numbers seen in nuclei never approach those found in the cytoplasm. A single nucleus has been observed to possess about twenty of these strands in cross-sectional view (Fig. 7), which are similar in size to the microfilaments observed in the cyto- plasm. We have not observed filaments in the nuclei of wild-type strains, but Beck et al. (2) have.
At higher magnifications, the usual appearance of cross sections of microfilaments is somewhat irregular and not distinctly circular. A few of the microfilaments at higher magnification appear to possess subunits (Figs. 8 and 9). The range of number of subunits in those microfilaments which show them is from two to four. The diameter of the subunits is approximately 25-35 A.
A few of the microfilaments in cross section give the impression of being hollow tubes (Fig. 2, inset). The number of microfilaments which are tubular in cross-sec- tional view is very low compared to the number which show subunits, although the actual number of either is small when one considers the total number of microfila- ments in any area. It should be mentioned that only extremely thin sections (dark- gray to black) permitted the visualization of this subunit structure.
On viewing longitudinally oriented microfilaments at higher magnification, it ap- pears that they are actually composed of more than one strand (Figs. 10 and 11).
As described in the section Materials and Methods, several different fixatives were used. Although they varied greatly in the effectiveness of preservation of back- ground cytoplasm and various organelles, all showed the microfilaments to be the same size. Microtubules in the range 180-240 N were never observed in the cyto-

MICROFILAMENTS IN Neurospora
~:~ ....
461
Fr~s. 8 and 9. Transversely oriented microfilaments which show subunit structure of individual members. Microfilaments indicated at arrows show 2 or 3 subunits. Fig. 8, x 185 000; Fig. 9, × 385 C00. Ft~s. 10 and 11. Higher magnification of longitudinally oriented microfilaments. Arrows indicate strand subunits of individual microfilaments. Fig. 10, x 105 000; Fig. 11, × 185 000.
plasm, including the regions where the masses of microfi laments were found. How-
ever, microtubules were found in a few nuclei after fixation in 3 % glutaraldehyde
in phosphate buffer. These were best observed in cross sections since it was difficult
to determine whether longitudinal components in the nuclei were fi laments or micro-
tubules.

462 ALLEN, LOWRY AND SUSSMAN
The presence of paraformaldehyde in the fixative containing glutaraldehyde pre- serves more information than the latter alone. This is evident in the retention of an electron dense outer cell wall (Fig. 1) and a denser mitochondrial matrix, neither of which are seen in cells only in glutaraldehyde.
DISCUSSION
The microfilaments just described have infrequently been observed by us in wild- type strains of Neurospora grown and prepared for electron microscopy under the same conditions used for snowflake. But several investigators have observed fila- ments in standard strains of Neurospora, including Beck et al. (2), who have shown that cultures obtained 4 hours after inoculation on media containing 2 % sucrose or 15% glucose possess filaments ("striated inclusions"). These microfilaments, whose appearance is identical to those in snowflake, are 62 N in diameter and appear in the nucleus as well as the cytoplasm. The authors refer to the microfilaments as "rods," and they believe that individual elements are linked by fine filamentous strands and that the rods may be tubular.
As mentioned above, the individual microfilaments in snowflake do not appear to be rods, or solid cylinders. Instead, they appear to be composed of strands when observed in longitudinal sections, a conclusion that is supported by the subunits seen in cross-sectional view. We have not observed fine filamentous strands con- necting microfilaments and feel that the tubular appearance of some microfilaments in cross-sectional view is related to their orientation. That is, osmium tetroxide used as a postfixative could contribute to the electron density of the strands, especially through thicker sections, thereby giving the impression of a continuous cylinder or a tubule, depending on whether the microfilaments are oriented in an exact transverse plane when sectioned.
Filaments in standard strains of Neurospora also have been reported by Wood and Luck (28), who found "paracrystalline inclusions" in a mitochondrial mutant (abnormal - 1). These inclusions could be induced in standard strains by the addition of ethidium bromide or euflavine to the culture medium. The authors describe the filaments as being approximately 70 A wide and dotlike in cross section. After isola- tion and negative staining this material showed a banded pattern formed by units 86 N in length, arranged end-to-end. Wood and Luck found that the protein from paracrystalline inclusions is not a product of mitochondrial genes, and believe that it crystallizes as a result of a mitochondrial defect.
The variation in the diameter of Neurospora microfilaments (approximately 40- 100 A) could be caused by any of several factors related to fixation, orientation dur- ing sectioning, staining, etc. However, such variability could also be explained if

MICROFILAMENTS IN Neurospora 463
microfilaments possessed different numbers of subunit strands. Pollard and/to (17) noted that the smaller filaments in Amoeba range from 50 to 92 A in diameter. This size range might also indicate that variations in the number of subunit strands exist in Amoeba and might also imply that not all small filaments are like actin in forming a two-stranded helix. Variations in the number of strands in individual filaments could explain the large size variations which exist among many small fila- ments and could provide a mechanism to explain how larger filaments are formed, i.e., by increased numbers of subunit strands. Such a mechanism does not, however, explain the short (0.5 #m) thick filaments in Amoeba (17) or the reason why heavy meromyosin binds only to the smaller filaments.
There is evidence that some of the microfilaments seen in electron micrographs are subunits of microtubules. Schultz and Case (22) have shown that neuronal microtubules are replaced by microfilaments in bicarbonate-buffered fixatives. On the other hand, fixatives containing cacodylate or phosphate buffers give normal microtubules. Microfilaments also have been seen after the use of mitotic inhibitors, such as colchicine, vinblastine sulfate, and podophyllotoxin, which are known to disrupt microtubules of the spindle apparatus (27). Consequently, we considered the possibility that the microfilaments observed in snowflake are artifacts of the kind discussed above. Glutaraldehyde when used in combination with osmium tetroxide has been shown to preserve microtubules. After using this fixation procedure as well as the others described in the Materials and Methods section, large numbers of microfilaments still were observed in snowflake and were never found assembled into microtubules.
Furthermore, where microtubules (250 A in diameter) have been observed in nuclei of snowflake, microfilaments of the usual kind still were found in the cytoplasm. Since nuclear microtubules were preserved it is unlikely that those occurring in the cytoplasm of the same cell would not be. Additionally, since microtubules have not been seen in the cytoplasm of Neurospora under fixation regimes that fix other organelles well, it seems unlikely that microfilaments are masses of microtubule sub- units resulting from poor fixation.
This study points to another hypothesis for the origin of certain morphological mutants of Neurospora. Such mutants are characterized, in comparison with standard strains, by the formation of dense colonies with a reduced rate of forward extension and an increased frequency of branching. One hypothesis is based upon the data of Tatum and collaborators (5, 10), which suggest that the walls of mutants contain significantly less glucose and more glucosamine than those of standard strains. In other mutants the primary effect has been traced to the formation of defective en- zymes (3), or higher activites of some (11). That a mucopolysaccharide polymer may regulate the rate of growth by effects upon the plasma membrane has been proposed 30-74•823 J. Ultrastructure Research

464 ALLEN, LOWRY AND SUSSMAN
in another case (20). Our finding of masses of microfilaments in snowflake suggests the additional possibility that the altered growth habit of this mutant is related to the accumulation of this material, perhaps because of effects upon one or more of the physiological processes in which microfilaments have been implicated.
REFERENCES
1. ARNOLD, J. M., Y. Cell Biol. 41, 894 (1969). 2. BECK, D. P., DECKER, G. L. and GREENWALT, J. W., J. Ultrastruct. Res. 33, 245 (1970). 3. BRODY, S. and TATUM, E. L., Proc. Nat. Acad. Sci. U.S. 58, 923 (1967). 4. CLONEY, R. A., J. Ultrastruct. Res. 14, 300 (1966). 5. DETERRA, N. and TATUM, E. L., Science 134, 1066 (1963). 6. GARNJOBST, L. and TATUM, E. L., Genetics 57, 579 (1967). 7. HATANO, S. and OOSAWA, F., J. Cell Physiol. 68, 197 (1966). 8. HOWL, H. R., HAMAMOTO, S. T. and HEMMES, D. E., Amer. Y. Bot. 55, 783 (1968). 9. HOLTZER, H. and SANGER, J. W., Develop. Biol. 27, 444 (1972).
10. MAnADEVAN, P. R. and TATUM, E. L., J. Bacteriol. 90, 1073 (1965). 11. MAHADEVAN, P. R. and MA~ADKAR, U. R., Y. Baeteriol. 101, 941 (1970). 12. MURRAY, J. C. and SRB, A. M., Can. J. Bot. 40, 351 (1962). 13. NACHMIAS, V. T., HUXLEY, H. E. and KESSLER, D., J. Mol. Biol. 50, 83 (1970). 14. NAGM, R. and REBHUN, L. I., J. Ultrastruct. Res. 14, 571 (1966). 15. O'BRIEN, T. P. and THIMANN, K. V., Pro& Nat. Acad. Sci. U.S. 56, 888 (1966). 16. PARTHASARATHY, M. V. and MOHLETHALER, K., J. Ultrastruct. Res. 38, 46 (1972). 17. POLLARD, T. D. and ITO, S., J. Cell Biol. 46, 267 (1970). 18. POLLARD, T. D. and KORN, E. D., J. CellBiol. 48, 216 (1971). 19. POLLARD, T. D., SHELTON, E., WEIHING, R. R. and KORN, E. D., J. MoI. Biol. 50, 91
(1970). 20. RHSSIG, J. L. and GLASGOW, J. E., J. Bacteriol. 106, 882 (1971). 21. REYNOLDS, E. S., J. Cell Biol 17, 208 (1963). 22. SCHULTZ, R. L. and CASE, N. M., J. Cell Biol. 38, 633 (1968). 23. S~eORN, M. B., WANKO, T. and DINGMAN, W., J. Cell Biol. ]15, 109 (1962). 24. SZOLLOSL D., J. Cell Biol. 44, 192 (1970). 25. VOGEL, H., Amer. Natur. 98, 435 (1964). 26. WESS~LLS, N. K., SPOONER, B. S., As~, J. F., BRADLEY, M. O., LUDUENA, M. A., TAYLOR,
E. L., WRENN, J. T. and YAMADA, K. M., Science 171, 135 (1971). 27. WISNIEWSKL H., SHELANSKL M. L. and TERRY, R. D., Y. Cell Biol. 38, 224 (1968). 28. WooD, D. D. and LUCK, D. J. L., J. Cell Biol. 51, 249 (1971).Printed in Sweden Copyright © 1974 by Academic Press, Inc. All rights of reproduction in any form reserved
J. ULTRASTRUCTURE RESEARCH 48, 455-464 (1974) 455
Accumulat ion of Microf i laments in a Colonial Mutant
of Neurospora crassa
EDWARD D. ALLEN, ROBERT J. LOWRY, and ALFRED S. SUSSMAN
Veteran's Administration Hospital, Ann Arbor, Michigan 48105, and Department o.f Botany, University of Michigan, Ann Arbor, Michigan 48104
Received November 14, 1973, and in revised form February 21, 1974
A morphological mutant of Neurospora erassa, snowflake, is shown to contain filaments which are about 70 A in diameter, and up to several microns long, and which usually bunch in groups of a few to several hundred. They may be found longitudinally or transversely arranged with respect to the long axis of the cell and, in many cases, they run up to the plasma membrane, but not through it. The filaments often are arranged in crystalline arrays but may also be found as separate filaments. Sometimes the filaments are closely appressed to nuclei and may be found inside them. It is likely that the filaments are not the result of the dissociation of microtubules and are most likely microfilaments like those found in other organisms. Their relationship to the origin of certain morphological mutants in Neurospora is discussed.
Cytoplasmic filaments ranging in size from 40-120 A in diameter and up to several microns long have been described in many organisms. It has been suggested that filaments in the smaller size range (50-70 ,~), generally referred to as microfilaments, are actinlike and involved in contractile processes.
Therefore, it has been hypothesized that microfilaments are involved in the fol- lowing processes: cytoplasmic streaming in Acanthamoeba (19), Physarum (7, 13), Nitella (14), and higher plants (15, 16); cell elongation in Acytostelium (8), epidermal contraction in Amaroricium (4); and cleavage in jelly fish and polychaete worms (24), and squid cells (1). Other aspects of morphogenesis in animal cells also are believed to be under the control of microfilaments because cytochalasin B, a drug that may disrupt these, can arrest development when added to some systems (26). However, conflicting results have been obtained with this drug and its action on microfila- ments has been questioned (9).
Despite the vast amount of literature on cytoplasmic filaments their fine structure has not been delineated in thin sections. Hatano and Oosawa (7) showed by nega- tive staining that the F-actin they isolated from Physarum is composed of a two- stranded helix. Similarly, negative staining of actinlike filaments, isolated from

456 ALLEN, LOWRY AND SUSSMAN
other organisms, also suggests such an arrangement (13, 18, 19). When thin sections :are used, the helical nature of the microfilaments has been derived from the beaded appearance of the microfilaments when viewed longitudinally. On the other hand, the filaments in transverse sections generally have been described as electron-dense dots, giving the impression that the filaments are solid cylinders.
In general, morphological mutants of Neurospora crassa are characterized by a very slow growth rate (0.5 mm/hour) and a higher frequency of branching as com- pared with wild-type strains. This results in colonies which have very compact my- celia and a mycelial front which advances very slowly. Some morphological mutants may not be as compact as others, but all differ appreciably from wild-type in linear growth rate of the colony, amount of aerial mycelium, compactness of aerial myce- lium, and frequency of branching (12). The various mutants have been described mainly on the basis of the growth rates, a necessity imposed by lack of knowledge of the biochemical lesions of the mutants (6).
During the investigation of several morphological mutants of Neurospora crassa we found one which possesses large numbers of cytoplasmic filaments whose struc- ture could be analyzed. This paper describes the appearance of these filaments in thin sections of fixed material.
MATERIALS AND METHODS
Organism. The morphological mutant of Neurospora crassa used for this study, strain 507 (snowflake), and the wild-type (standard strains), 4121a, and the Lindegren strain, were obtained through the courtesy of the Fungal Genetics Stock Center, Humboldt State College, Arcata, California.
Culture conditions. Cultures of Neurospora were grown in 500-ml Erlenmeyer flasks con- taining 200 ml of distilled water with 2 % sucrose and 2 % of 50 X Vogel's standard salts (25). Growth was initiated in flasks either by hyphal or conidial transfer.
Preparation for electron microscopy. Hyphae from liquid cultures were placed in vials con- taining fixatives at 0°C or room temperature. The fixative used most regularly was 3 % glutaraldehyde in 0.1 M sodium cacodylate (pH 7.4) at 0°C for 2 hours. Other fixatives included (a) 3 % glutaraldehyde in 0.05 M potassium phosphate (pH 7.2) or (b) 1.5 % of both glutaraldehyde and paraformaldehyde in 0.1 M sodium cacodylate (pH 7.4) or 0.05 M potassium phosphate (pH 7.2), or (c) 3 % acrolein in either of the above buffers. After fixa- tion, the hyphae were washed several times and left in buffer overnight.
FIG. 1. Longitudinally oriented microfilaments in three cells. Arrows indicate groups of microfila- ments, x 8 603. FIG. 2. Transverse section of hyphae showing several groups of microfilaments in transverse and longitudinal view. Arrow at ribosomes (R) indicates the small relative size of the microfilaments. × 49 800. Inset: higher magnification of region indicated by outline. Arrow indicates tubular appear- ance of one mierofilament in transverse view. x 164 000. FIG. 3. Three masses of microfilaments, one of which is perpendicular to septum (S). × 16 000.

MICROFILAMENTS IN Ngurospora 457

458 ALLEN, LOWRY AND SUSSMAN
Postfixation (1 hour at 0°C) was carried out with the buffer used for fixation diluted I : 1 with aqueous 2 % osmium tetroxide.
Dehydration in ethanol was followed by transfer to propylene oxide and embeddment in Epon [25 DDSA: 10 Epon 812:0.8 DMP-30 (23)].
Sections were cut with diamond or glass knives and stained with 2% aqueous uranyl acetate followed by lead citrate (21). The sections were viewed with either Philips 300, RCA-EMU4, or Hitachi 11 microscopes.
The microscope was calibrated with a carbon grating replica (28 800 lines per inch) ob- tained from E. F. Fullam Inc., Schenectady, New York. Microfilaments were measured on micrographs enlarged 4 times from negatives taken at 54 000 × at the time of calibration.
RESULTS
The growth pattern of snowflake (strain 507) is characteristic of many of the morphological mutants of Neurospora. Its hyphae radiate outward from a common center after hyphal or conidial transfer and grow at a rate of approximately 0.3 mm per hour compared to about 3.6 mm per hour for wild-type strains. The length of the cells of snowflake is about 100 #m while that of the standard strains used is about 180-200 #m. The width of the cells of snowflake is about 20 ,urn which is approxi- mately one-half that of standard strains.
At the ultrastructural level, the main thing which distinguishes sections of snowflake f rom those of standard strains is the presence of large masses of a fi lamentous ma- terial. Unlike the wild-type strains, in which we have seen a few groups of filaments only in young vegetative hyphae, snowflake commonly contains masses of filaments in vegetative cells.
Figure 1 shows a section of three cells of snowflake, all of which possess the fila- mentous masses.
Filaments appear in groups ranging from a few to several hundred, and nearly all are found dispersed in the cytoplasm. They are not oriented in any particular direction with respect to the long axis of the cell; that is, they may be found in longitudinal or cross-sectional view when the cell is oriented longitudinally (Figs. 2 and 3). However, when extremely large masses of filaments are seen, they are usually oriented parallel to the longitudinal axis of the cell. The filaments are as common in apical cells as in those farther away from the growing tip. However, the extremely
FIG. 4. Group of microfilaments perpendicular to, but not passing through, the plasma membrane at cell wail (W). The microfilaments appear granular near the plasma membrane, x 71 000. FIG. 5. Appearance of microfilaments in hexagonal arrangement (arrows). The presence of longi- tudinally aligned filaments near the hexagonal ones suggests that some of the groups of microfila- ments may be braided arrays, x 61 600. FIG. 6. Longitudinal view of microfilaments in cytoplasm as well as microfilament-like structures in nucleus (arrow). Those found ~n nuclei appear the same as those observed in the cytoplasm, x 37 000. FIG. 7. Nucleus possessing microfilament-like structures (arrows) in transverse view. × 65 000.

MICROFILAMENTS IN Nellrospora 459

460 ALLEN~ LOWRY AND SUSSMAN
large masses appear more commonly in older cells. In many cases, longitudinal groups are seen to run up to the plasma membrane at the septum (Fig. 3) or wall (Fig. 4), but none have been seen to pass through it.
In longitudinal view, the individual filaments are approximately 70 A wide and up to several microns long. Their mean diameter is 68.8 •+ 15 ~ and the range is from 40 to slightly above 100 N. These values, and the standard deviation, were obtained from measurements of 111 filaments. Since the size of these filaments falls into the range of microfilaments, they will be described as such hereafter. Micro- filaments are in close apposition and appear straight or slightly curved, and when in the former condition they present a somewhat crystalline appearance (Fig. 5). How- ever, at other times they may appear loose and separated (Fig. 3).
In cross-sectional views the microfilaments appear as electron-dense dots approxi- mately 70 N in diameter with up to several hundred grouped together in any one area. Occasionally, the microfilaments appear packed into hexagonal arrays (Fig. 5).
The only organelle in the cell with which the filaments seem to be associated is the nucleus. Several nuclei have been found to possess filamentous strands which are very similar in appearance to the microfilaments found in the cytoplasm (Fig. 6). The numbers seen in nuclei never approach those found in the cytoplasm. A single nucleus has been observed to possess about twenty of these strands in cross-sectional view (Fig. 7), which are similar in size to the microfilaments observed in the cyto- plasm. We have not observed filaments in the nuclei of wild-type strains, but Beck et al. (2) have.
At higher magnifications, the usual appearance of cross sections of microfilaments is somewhat irregular and not distinctly circular. A few of the microfilaments at higher magnification appear to possess subunits (Figs. 8 and 9). The range of number of subunits in those microfilaments which show them is from two to four. The diameter of the subunits is approximately 25-35 A.
A few of the microfilaments in cross section give the impression of being hollow tubes (Fig. 2, inset). The number of microfilaments which are tubular in cross-sec- tional view is very low compared to the number which show subunits, although the actual number of either is small when one considers the total number of microfila- ments in any area. It should be mentioned that only extremely thin sections (dark- gray to black) permitted the visualization of this subunit structure.
On viewing longitudinally oriented microfilaments at higher magnification, it ap- pears that they are actually composed of more than one strand (Figs. 10 and 11).
As described in the section Materials and Methods, several different fixatives were used. Although they varied greatly in the effectiveness of preservation of back- ground cytoplasm and various organelles, all showed the microfilaments to be the same size. Microtubules in the range 180-240 N were never observed in the cyto-

MICROFILAMENTS IN Neurospora
~:~ ....
461
Fr~s. 8 and 9. Transversely oriented microfilaments which show subunit structure of individual members. Microfilaments indicated at arrows show 2 or 3 subunits. Fig. 8, x 185 000; Fig. 9, × 385 C00. Ft~s. 10 and 11. Higher magnification of longitudinally oriented microfilaments. Arrows indicate strand subunits of individual microfilaments. Fig. 10, x 105 000; Fig. 11, × 185 000.
plasm, including the regions where the masses of microfi laments were found. How-
ever, microtubules were found in a few nuclei after fixation in 3 % glutaraldehyde
in phosphate buffer. These were best observed in cross sections since it was difficult
to determine whether longitudinal components in the nuclei were fi laments or micro-
tubules.

462 ALLEN, LOWRY AND SUSSMAN
The presence of paraformaldehyde in the fixative containing glutaraldehyde pre- serves more information than the latter alone. This is evident in the retention of an electron dense outer cell wall (Fig. 1) and a denser mitochondrial matrix, neither of which are seen in cells only in glutaraldehyde.
DISCUSSION
The microfilaments just described have infrequently been observed by us in wild- type strains of Neurospora grown and prepared for electron microscopy under the same conditions used for snowflake. But several investigators have observed fila- ments in standard strains of Neurospora, including Beck et al. (2), who have shown that cultures obtained 4 hours after inoculation on media containing 2 % sucrose or 15% glucose possess filaments ("striated inclusions"). These microfilaments, whose appearance is identical to those in snowflake, are 62 N in diameter and appear in the nucleus as well as the cytoplasm. The authors refer to the microfilaments as "rods," and they believe that individual elements are linked by fine filamentous strands and that the rods may be tubular.
As mentioned above, the individual microfilaments in snowflake do not appear to be rods, or solid cylinders. Instead, they appear to be composed of strands when observed in longitudinal sections, a conclusion that is supported by the subunits seen in cross-sectional view. We have not observed fine filamentous strands con- necting microfilaments and feel that the tubular appearance of some microfilaments in cross-sectional view is related to their orientation. That is, osmium tetroxide used as a postfixative could contribute to the electron density of the strands, especially through thicker sections, thereby giving the impression of a continuous cylinder or a tubule, depending on whether the microfilaments are oriented in an exact transverse plane when sectioned.
Filaments in standard strains of Neurospora also have been reported by Wood and Luck (28), who found "paracrystalline inclusions" in a mitochondrial mutant (abnormal - 1). These inclusions could be induced in standard strains by the addition of ethidium bromide or euflavine to the culture medium. The authors describe the filaments as being approximately 70 A wide and dotlike in cross section. After isola- tion and negative staining this material showed a banded pattern formed by units 86 N in length, arranged end-to-end. Wood and Luck found that the protein from paracrystalline inclusions is not a product of mitochondrial genes, and believe that it crystallizes as a result of a mitochondrial defect.
The variation in the diameter of Neurospora microfilaments (approximately 40- 100 A) could be caused by any of several factors related to fixation, orientation dur- ing sectioning, staining, etc. However, such variability could also be explained if

MICROFILAMENTS IN Neurospora 463
microfilaments possessed different numbers of subunit strands. Pollard and/to (17) noted that the smaller filaments in Amoeba range from 50 to 92 A in diameter. This size range might also indicate that variations in the number of subunit strands exist in Amoeba and might also imply that not all small filaments are like actin in forming a two-stranded helix. Variations in the number of strands in individual filaments could explain the large size variations which exist among many small fila- ments and could provide a mechanism to explain how larger filaments are formed, i.e., by increased numbers of subunit strands. Such a mechanism does not, however, explain the short (0.5 #m) thick filaments in Amoeba (17) or the reason why heavy meromyosin binds only to the smaller filaments.
There is evidence that some of the microfilaments seen in electron micrographs are subunits of microtubules. Schultz and Case (22) have shown that neuronal microtubules are replaced by microfilaments in bicarbonate-buffered fixatives. On the other hand, fixatives containing cacodylate or phosphate buffers give normal microtubules. Microfilaments also have been seen after the use of mitotic inhibitors, such as colchicine, vinblastine sulfate, and podophyllotoxin, which are known to disrupt microtubules of the spindle apparatus (27). Consequently, we considered the possibility that the microfilaments observed in snowflake are artifacts of the kind discussed above. Glutaraldehyde when used in combination with osmium tetroxide has been shown to preserve microtubules. After using this fixation procedure as well as the others described in the Materials and Methods section, large numbers of microfilaments still were observed in snowflake and were never found assembled into microtubules.
Furthermore, where microtubules (250 A in diameter) have been observed in nuclei of snowflake, microfilaments of the usual kind still were found in the cytoplasm. Since nuclear microtubules were preserved it is unlikely that those occurring in the cytoplasm of the same cell would not be. Additionally, since microtubules have not been seen in the cytoplasm of Neurospora under fixation regimes that fix other organelles well, it seems unlikely that microfilaments are masses of microtubule sub- units resulting from poor fixation.
This study points to another hypothesis for the origin of certain morphological mutants of Neurospora. Such mutants are characterized, in comparison with standard strains, by the formation of dense colonies with a reduced rate of forward extension and an increased frequency of branching. One hypothesis is based upon the data of Tatum and collaborators (5, 10), which suggest that the walls of mutants contain significantly less glucose and more glucosamine than those of standard strains. In other mutants the primary effect has been traced to the formation of defective en- zymes (3), or higher activites of some (11). That a mucopolysaccharide polymer may regulate the rate of growth by effects upon the plasma membrane has been proposed 30-74•823 J. Ultrastructure Research

464 ALLEN, LOWRY AND SUSSMAN
in another case (20). Our finding of masses of microfilaments in snowflake suggests the additional possibility that the altered growth habit of this mutant is related to the accumulation of this material, perhaps because of effects upon one or more of the physiological processes in which microfilaments have been implicated.
REFERENCES
1. ARNOLD, J. M., Y. Cell Biol. 41, 894 (1969). 2. BECK, D. P., DECKER, G. L. and GREENWALT, J. W., J. Ultrastruct. Res. 33, 245 (1970). 3. BRODY, S. and TATUM, E. L., Proc. Nat. Acad. Sci. U.S. 58, 923 (1967). 4. CLONEY, R. A., J. Ultrastruct. Res. 14, 300 (1966). 5. DETERRA, N. and TATUM, E. L., Science 134, 1066 (1963). 6. GARNJOBST, L. and TATUM, E. L., Genetics 57, 579 (1967). 7. HATANO, S. and OOSAWA, F., J. Cell Physiol. 68, 197 (1966). 8. HOWL, H. R., HAMAMOTO, S. T. and HEMMES, D. E., Amer. Y. Bot. 55, 783 (1968). 9. HOLTZER, H. and SANGER, J. W., Develop. Biol. 27, 444 (1972).
10. MAnADEVAN, P. R. and TATUM, E. L., J. Bacteriol. 90, 1073 (1965). 11. MAHADEVAN, P. R. and MA~ADKAR, U. R., Y. Baeteriol. 101, 941 (1970). 12. MURRAY, J. C. and SRB, A. M., Can. J. Bot. 40, 351 (1962). 13. NACHMIAS, V. T., HUXLEY, H. E. and KESSLER, D., J. Mol. Biol. 50, 83 (1970). 14. NAGM, R. and REBHUN, L. I., J. Ultrastruct. Res. 14, 571 (1966). 15. O'BRIEN, T. P. and THIMANN, K. V., Pro& Nat. Acad. Sci. U.S. 56, 888 (1966). 16. PARTHASARATHY, M. V. and MOHLETHALER, K., J. Ultrastruct. Res. 38, 46 (1972). 17. POLLARD, T. D. and ITO, S., J. Cell Biol. 46, 267 (1970). 18. POLLARD, T. D. and KORN, E. D., J. CellBiol. 48, 216 (1971). 19. POLLARD, T. D., SHELTON, E., WEIHING, R. R. and KORN, E. D., J. MoI. Biol. 50, 91
(1970). 20. RHSSIG, J. L. and GLASGOW, J. E., J. Bacteriol. 106, 882 (1971). 21. REYNOLDS, E. S., J. Cell Biol 17, 208 (1963). 22. SCHULTZ, R. L. and CASE, N. M., J. Cell Biol. 38, 633 (1968). 23. S~eORN, M. B., WANKO, T. and DINGMAN, W., J. Cell Biol. ]15, 109 (1962). 24. SZOLLOSL D., J. Cell Biol. 44, 192 (1970). 25. VOGEL, H., Amer. Natur. 98, 435 (1964). 26. WESS~LLS, N. K., SPOONER, B. S., As~, J. F., BRADLEY, M. O., LUDUENA, M. A., TAYLOR,
E. L., WRENN, J. T. and YAMADA, K. M., Science 171, 135 (1971). 27. WISNIEWSKL H., SHELANSKL M. L. and TERRY, R. D., Y. Cell Biol. 38, 224 (1968). 28. WooD, D. D. and LUCK, D. J. L., J. Cell Biol. 51, 249 (1971).