NEUROSPORA. I. PRELIMINARY OBSERVATIONS OF THE CHROMOSOMES OF NEUROSPORA CRASSA * Barbara McClintock TE PRESENT report on the chromosomes of Neu- rospora crassa represents the results of observations which were confined to a period of ten weeks in the biological laboratories of Stanford University. The purpose of this study was to obtain some knowledge of nuclear and chromosome behavior in normal and mutant strains. The author realizes that no single phase of these investigations could be adequately studied in so short a period of time. Because of the interest in Neurospora as genetic material, a sum- mary of some of these observations will be given at this time. The observations were confined to the nuclei and chromosomes in the ascus, from fertilization to spore formation. Union of two haploid nuclei occurs in the young ascus. This is followed by a simultaneous enlargement of the ascus and fusion nucleus. During this growth period, the chromosomes in the fusion nucleus enter into meiotic prophase activities, in- cluding homologous association of chromosomes, elongation of chromosomes, chiasmata formation and contraction until typical metaphase I bivalents are produced. Although the consequences of this meiotic prophase activity are essentially similar to those observed in many other organisms, the timing of chromosome synapsis and elongation is dissimilar and is of some theoretical interest. The two meiotic mitoses follow in rapid succession leading to the formation of four haploid nuclei. In essential details and accomplishments, the chromosome and nuclear behavior in these two divisions is typical of meiosis - in general. Particular details, however, are of inter- est to the cytologist. Each of the four haploid nuclei in the now greatly enlarged ascus undergoes a typi- cal equational mitosis resulting in a row of eight haploid nuclei. Associated with each nucleus is a centriole which has become greatly enlarged during the meiotic and first post-meiotic mitoses. Fibers emerging from each centriole extend and encircle the cytoplasm surrounding each nucleus. This proc- ess initiates wall formation and the cutting out of 1 Received for publication August 28, 1945. The author wishes to express appreciation to Dr. G. W. Beadle for presenting the opportunity for these studies and for the many kindnesses he has shown. Much credit should be given to Mrs. Mary B. Houlahan, Dr. Herschel K. Mitchell and Dr. Lotti Steinitz for their interest and collaboration in the chromosome studies, in selecting and supplying the mutant and wild-type strains and for vari- ous helpful suggestions regarding techniques. eight independent ascospores. Shortly after the spore walls are differentiated, the nucleus in each spore undergoes an equational mitosis. The asco- spore continues to maturity with the two resulting nuclei. Mertuops.—Approximately seven days (at 25 °C.) after inoculation of an agar slant with the two sex strains, A and a, perithecia were present containing numerous asci in various stages of pre-fertilization, fertilization, meiosis and spore formation and de- velopment. These perithecia were removed from the slant and placed in a drop of staining solution. With the bent end of a needle, pressure was applied to the perithecial wall. When this pressure was properly exerted, the asci within the perithecium were forced out through the ostiole. They usually emerged as a single mass. The perithecial wall was removed and a cover slip placed over the drop. The slide was then gently heated. Several methods of staining were attempted such as aceto-orcein, aceto-carmine, pro- pionic-orcein, lacto-orcein and acetic-lactic-orcein combinations. After many trials, it was realized that the genetic strain being utilized had much to do with the success of the staining procedure. A cross of two particular wild-type strains always gave excellent results, whereas other strains gave moderate or con- sistently poor results. In general, aceto-orcein was the most reliable chromosome stain but the nucleoli were not differentiated. When, in any particular aceto-orcein preparation, it was necessary to observe the nucleoli, aceto-carmine was subsequently run under the cover slip. The nucleoli, taking up the carmine stain, were then clearly visible. CHromosome NumMBER.—The haploid chromosome number in all the examined strains of N. crassa was seven. This number does not agree with that given by Lindegren and Rumann (1938) for N. crassa (six to nine chromosomes) nor that given by Colson (1984) for N. tetrasperma (six chromosomes ). Seven haploid chromosomes had previously been observed (Dr. E. A. Weaver and author, unpub- lished) in a strain of N. tetrasperma supplied by Dr. B. O. Dodge. The author is indebted to Dr. G. W. Bohn, a former graduate student of the Uni- versity of Missouri, for calling her attention to the Neurospora chromosomes. He observed seven hap- loid chromosomes in his excellent aceto-carmine preparations of Neurospora sp. CuromosomeE sizE.—The lengths of the chromo- 672 somes were measured at various stages from pre- synapsis in the zygote nucleus to the metaphase of the division in the ascospore. The longest chromo- some is approximately 2.7 times the length of the Fig. 1. Diagram illustrating the relative Jengths of the seven chromosomes of Neurospora crassa. The cross lines indicate the positions of the centromeres; these are rea- sonably correct for the two longest chromosomes. The determination of the positions of the centromeres in the other chromosomes needs further confirmation; the as- signed positions (broken cross-lines) should be consid- ered only as tentative. The separation of the minute satel- lite from the main segment of the short arm of chromo- some 2 is indicated by the dashed line. shortest chromosome. Since the relative lengths of all chromosomes are maintained throughout the nu- clear cycles, measurements will be mentioned only * for this longest chromosome. At the end of the pachy- tene, this chromosome may attain a length of ap- proximately 15 microns. At metaphase of the third division in the ascus, it may be approximately 2.5 microns long. At the metaphase of the division in the ascospore, it may be only 1.5 microns long. The chromosomes of the hyphal nuclei were not exam- ined. In contrast to the relatively large size of the nuclei in the ascus and the ascospores, the hyphal nuclei are very minute. It is probable that the meta- phase chromosomes they form are likewise very minute. RELATIVE LENGTHS OF THE CHROMOSOMES.— Meas- urements of the relative lengths of the chromosomes were most satisfactorily obtained from nuclei in late pachytene. The chromosomes are then at their maxi- mum extension (see below). Although all seven chro- mosomes were drawn and measured in only a few AMERICAN JOURNAL OF BOTANY [Vol. 32, meiotic prophase nuclei, the relative lengths of the chromosomes were consistent within each nucleus, Figure 1 illustrates the relative lengths of the seven chromosomes as computed from these measurements, Morruovocy oF THE CHROMOSOMES.—Centromere positions —The centromere position was adequately determined only for the two longest chromosomes. The analysis of centromere positions was suspended temporarily because it was thought that one of the smaller pairs of chromosomes might be heteromor- phic. If this were true, two sets of chromosome mor- phologies with respect to centromere positions. would have to be considered. The presence of a heteromorphic pair was not confirmed in subsequent examinations which were confined mainly to a cross between two particular wild-type strains. Whether a heteromorphic pair is present or could be identified in crosses of other wild-type strains remains to be determined. Due to the pressure for other determina- tions, no time was taken to renew the studies of cen- tromere positions. In order to convey some idea of centromere positions in the complement as a whole. the tentative positions that had been assigned to chromosomes 8 to 7 before this analysis was sus- pended, are included in figure 1. The nucleolus chromosome.—The second longest chromosome (chromosome 2) possesses a nucleolus organizer close to the end of its short arm. Conse- quently, there is a very minute satellite. The nucleo- lus organizer functions in the usual manner and de- velops a nucleolus in each telophase nucleus. Chromomere patterns—At late pachytene, each chromosome shows a distinct chromomere pattern. The pattern for any one chromosome is constant. The chromomeres have various sizes and shapes. They are separated by thinner strands of chromatin but are not spaced equally along the chromosome. The smaller chromosomes have only a few distinct chromomeres (five to six or seven), whereas the longer chromosomes have correspondingly more. No attempt was made in this preliminary study to map the chromomeres of each chromosome. However. these distinctive chromomere patterns could be use- ful in identifying individual chromosomes at pachy- tene. No knobs were recognized in these chromo- somes, Centromeres could not be identified with cer- tainty in the orcein stained preparations of pachy- tene. Heterochromatin.—Heterochromatic segments of chromosomes were not recognized as such in the pachytene chromosomes. However, the presence of heterochromatin was detected in the telophase nu- clei following the second meiotic mitosis and in the resting nuclei of the one- and two-nucleated asco- spores. It could also be observed in the hyphal nu- clei. There are two main segments of heterochro- matin. They are located adjacent to a centromere. It has not been determined whether these two recog” nized segments lie adjacent to the centromere 0” opposite arms of one chromosome or whether they are parts of two separate chromosomes. Congression of the centromeres in late anaphase of division Il. Dec., 1945] and in the spore division, results in the formation of a somewhat pear-shaped resting nucleus. The centromeres of all seven chromosomes lie in the apex of this pear-shaped nucleus. Here, also, are found the two heterochromatic bodies lying so close to- gether that they suggest a single dumb-bell shaped structure. It is believed, however, that they have not fused to form a single chromocenter but are forced close to one another by the intimate spacial association of all seven centromeres. Extensive ob- servations have not been made of these two hetero- chromatic bodies nor has an attempt been made to identify the chromosome or chromosomes involved. NucLeaR FUSION, CHROMOSOME SYNAPSIS AND THE SUBSEQUENT ELONGATION OF THE SYNAPSED CHROMO- somes.—-Fusion of two haploid nuclei to form the zygote occurs in the very young ascus. Illustrations of the appearance of the ascus at this stage are given by Colson (1984). At the time of fusion, the chro- mosomes of each nucleus appear to be in a resting stage and a nucleolus is present in each. Following nuclear fusion, the chromosomes contributed by each nucleus undergo what appears to be a typical pro- phase contraction until, in some strains, the chro- mosomes may be almost as short as those of the metaphase of the third division in the ascus. No ob- vious doubleness of the chromosomes was observed, however. During this period, fusion usually occurs between the nucleoli contributed by each nucleus. At the end of the contraction period, the two hap- loid sets of chromosomes lie, roughly, at opposite sides of the zygote nucleus. In this highly con- tracted state, the homologous chromosomes enter into the synaptic phase of the meiotic cycle. In the early synaptic phase, many nuclei were observed with some homologous chromosomes lying adjacent to one another but not in actual physical contact. It is not clear whether this early stage in the associa- tion process is the consequence of a directed migra- tion of homologues toward one another or whether this stage is reached following random movements of the chromosomes within the nucleus. Possibly the movements of the chromosomes could be followed in tissue cultures of the living asci. It is of considerable theoretical interest to determine the range of the force of synaptic attraction. The actual physical association of the chromosomes usually begins at one or both ends and continues along them. In many zygote nuclei, synapsis is completed for some pairs of chromosomes before the members of other pairs have come in contact. Soon, many nuclei show seven short, but completely synapsed, bivalent chromo- somes. (Most of the detailed observations of the synaptic phase were confined to asci resulting from the cross of two wild-type strains (Emerson 52564 > Chilton-a). In crosses of some other strains, synapsis appears to occur when the chromo- somes are less contracted.) Following completion of synapsis and possibly during this period, the chro- mosomes commence their elongation. This is pos- sibly an uncoiling process for in some early post- synaptic nuclei, the elongating chromosomes ap- MC CLINTOCK——CHROMOSOMES OF NEUROSPORA 673 peared to possess compressed gyres. This elonga- tion process continues until the chromosomes have reached their full extension. At this stage, the chro- mosomes are essentially similar in appearance to the pachytene chromosomes of many other organ- isms. The term “pachytene” has therefore been used. Although homologous chromosomes lie side by side at late pachytene, they are often not closely ap- pressed. Often, there is little or no relational coiling of the two homologues around one another. During the period from zygote formation to late pachytene, the volume of the nucleus and nucleolus increases steadily. In all post-synaptic stages, the volume of the nucleus is very much greater than that of the chromosomes. Consequently, the chromosomes’ are widely spaced within the nucleus. During all these stages, the chromosome 2 bivalent remains attached to the nucleolus by the organizer regions. At pachy- tene, the organizer regions of the two homologues may diverge slightly from one another; the satellites may be some distance removed from them. CHROMOSOME BEHAVIOR FROM DIPLOTENE TO THE THIRD DIVISION IN THE ascus.—At diplotene, a wide separation occurs between parts of a bivalent chro- mosome but the individual chromatids were difficult to follow. Coiling commences at diplotene and the contraction of the chromosomes is very rapid. At diakinesis, typical chiasmata may be seen. No at- tempt was made to count chiasmata but it is possible to do so at this stage. The chromosomes continue contraction to form typical metaphase I bivalents with terminal and interstitial chiasmata. Although the nucleolus becomes smaller during the pre-meta- phase stage, it does not disappear. Chromosome 2 remains attached to the nucleolus by its organizer region. Anaphase I separation of the chromosomes appears to be essentially typical except for the nu- cleolus. This may be dragged toward one pole or stretched between the poles because the nucleolus organizers of one or both of the dyad chromatids of chromosome 2 have not been released from their attachment to the nucleolus. The nucleolus becomes detached before telophase sets in and may subse- quently be seen in the cytoplasm of the ascus. At telophase I (and likewise telophase II and III) the centromere regions of all the chromosomes form an aggregate that lies at the apex of a distinct protru- sion of the nucleus (the beak: Dodge, 1927). No true resting nucleus is formed. Instead, the chromo- somes uncoil and the individual arms of each chro- mosome extend into an elongated nucleus. A new nucleolus is produced by and remains attached to the nucleolus organizers of chromosome 2. Prophase II proceeds by contraction of these elongated chro- mosomes until the two dyads of each chromosome form very short, parallel rods, each showing a con- spicuous centromere region. Metaphase and ana- phase II proceed normally. At telophase IT, the chro- mosomes, whose centromere regions are again aggre- gated at the apex of the beaked nucleus, uncoil and the two arms of each chromosome extend into the nucleus as individual strands and remain in this con- 674 dition until the following prophase. The extent of elongation of the chromosomes appears to be similar to that of late pachytene. In each nucleus, a new nucleolus is formed at the position of the nucleolus organizer of chromosome 2. Prophase IIT proceeds by contraction of the arms of the chromosomes. Be- cause the chromosomes maintain their previous telo- phase orientation (Js and Vs) during this contrac- tion, the prophase of division III is a satisfactory stage for observing the relative lengths of the arms of a chromosome. Metaphase and anaphase of the Fig. 2. Outline drawing of the synapsed chromosomes in an ascus heterozygous for T 4637. There are five biva- lents and a synaptic configuration of four chromosomes. The nucleolus is outlined by a dashed line. The minute satellites of the pair of nucleolus chromosomes were not detected in this figure. third division proceed as a typical equational mi- tosis. The telophase of this division is followed by a condition of the nucleus resembling a resting stage. Shortly after spore delimitation, a mitosis occurs in each ascospore. This is likewise a typical equational mitosis. In essential details, divisions I and II are typically meiotic. Division III is essentially a so- matic mitosis except that the chromosomes retain their identity from the telophase of division II to the prophase of division III. It would be of interest to determine the time of effective splitting of the chromosomes for this division. RecipRocaL TRANSLOCATIONS.—In the Stanford laboratory, many mutants have been obtained fol- lowing x-ray and ultra-violet irradiations. Chromo- somal abnormalities could likewise be expected to occur from such treatments. Three irradiation-in- duced mutants (4687, 44105 and 45502) whose genetic behavior suggested the presence of some chromosomal abnormality, were selected and crossed to normal wild-type strains. The chromosomes were examined in the asci developing from these crosses. In all three cases, the ascus nuclei were heterozygous for a translocation between two non-homologous chromosomes. In the limited time available, it was not possible to make an intensive study of each trans- location. Nevertheless, some observations and inter- pretations based on these studies will be mentioned. Translocation 4637.—Figure 2 represents an out- line drawing of late pachytene chromosomes in an ascus nucleus developing from the cross of the albino mutant strain 4637 by a wild-type strain. There are five normal bivalents and a synaptic configuration AMERICAN JOURNAL OF BOTANY [Vol. 32, of four chromosomes (right). In these nuclei, ho- mologous associations of all parts of the four chro- mosomes were not always accomplished. Unsynapsed segments, as illustrated in figure 2, were frequently observed. Sometimes, at pachytene, the four chro- mosomes were present as two “bivalents” with synaptic associations only between their respective homologous parts. At diakinesis and metaphase I, either a ring of four chromosomes, a chain of four chromosomes or two “bivalent’’ chromosomes were observed. . Translocation 44105.—Relatively few observa- tions were made of the translocation introduced by mutant strain 44105. These were limited to a few figures of diakinesis and metaphase I. A ring of four chromosomes was observed in one metaphase I fig- ure. In several others, one or more of the chromo- somes were present as univalents. In two figures, all four chromosomes were present as univalents. No pachytene configurations were observed. Translocation 45502.—The reciprocal transloca- tion introduced by mutant strain 45502 involved a very unequal exchange of segments of two non- homologous chromosomes. The breaks appear to have occurred close to the end of the long arm of chromosome J and close to the centromere in the long arm of one of the chromosomes with a sub-terminal centromere. This translocation could serve several purposes which will be outlined below. Estimates of the types of disjunction of chromo- somes in asci heterozygous for T 45502.—Because of the small size of the metaphase and anaphase I chromosomes in Neurospora, it would be very labori- ous to determine by direct observations the modes of disjunction of the four chromosomes involved in translocation configurations. An examination of the eight-spored asci developing from asci whose fusion nuclei were heterozygous for translocation 45502 has suggested a possible method of estimating these disjunctions. In most organisms, a two-by-two dis- junction of the four chromosomes of an interchange complex usually occurs at anaphase I. In organisms having the Oenothera type of disjunction, alternate chromosomes in a ring or chain of four or more chro- mosomes go to the same pole at anaphase I. In maize. _ Pisum, ete., the four chromosomes of a ring usually disjoin so that two members go to one pole and two to the opposite pole. In these forms, alternate dis- junctions occur in some cells. In other cells, how- ever, two adjacent members of the ring or chain of four chromosomes may go to the same pole. When a heterozygous translocation is present in Neurospora, do the chromosomes disjoin according to the Oeno- thera pattern or do disjunctions follow the maize and Pisum pattern? The analysis given below sug- gests that the disjunctions in Neurospora are similar to those observed in maize and Pisum. Although the exact position of breakage in the two chromosomes has not been determined, a diagram illustrating the type of synaptic configuration to be expected in asci heterozygous for T 45502 is given in figure 8. If no crossing over occurs in either re- Dec., 1945] gion a or b, figure 8, alternate disjunctions (1+4 : 2-3, fig. 3) of the four chromosomes at anaphase I when a ring or a chain is present, or the counterpart type of disjunction when two “bivalents” are pres- ent, should produce an ascus with eight normal spores (Type I ascus, fig. 3). In this case every spore would receive a full genomic complement, four with the normal chromosomes (1+4, fig. 3), and four with the translocation chromosomes (2+3, fig. 3). When two adjacent chromosomes of this complex pass to the same pole at anaphase I, all eight of the resulting spores in an ascus would be deficient for some part of the genomic comple- _ment. There are two possible types of adjacent disjunctions, those which result from disjunctions of homologous centromeres (1--2 3+-4) and those which result from non-disjunction of homolo- gous centromeres (1-++3 : 24-4). The former will be called adjacent I disjunction, the latter, adjacent II disjunction. Following adjacent I disjunctions, four of the spores (with 1+-2) would be deficient for 4 . 2 a Fy. . 3 an a 8b ea a ge : of as oe os as ae aa oa oa oe ea as Type T Type I Type 1 Type Fig. 3. Upper. Diagram illustrating the complete synap- tic association of two normal and two translocated chromo- somes in an ascus heterozygous for a very unequal recip- rocal translocation. Chromosomes numbered 1 and 4 repre- sent the normal chromosomes; chromosomes numbered 2 and 3 represent the translocation chromosomes. The cen- tromere in each chromosome is represented by a circle. Lower. Diagrammatic representation of the types of eight- spored asci resulting from several types of disjunctions of the four members of the synaptic complex (see text for ex- planations). The heavily outlined spores are normal in appearance; the lightly outlined spores are visibly defec- tive in appearance. (In the observed material, the type TV ascus is considerably more defective than the diagram suggests. ) MC CLINTOCK—-CHROMOSOMES OF NEUROSPORA 675 nearly all of the long arm of one chromosome. In contrast, the four spores with 3-+4 would be defi- cient only for a small segment of the genomic com- plement. Comparable studies in maize have shown that spores with deficiencies of large segments of the genomic complement are defective in appearance, whereas spores with small deficiencies may be nor- mal in appearance, especially in the early develop- mental stages. If the response in Neurospora is similar, it could be expected that the spores with 1+2 would be defective in appearance, whereas those with 3-+-4 may be normal in appearance, espe- cially in the young eight-spored asci. If this occurs, adjacent I disjunctions would give rise to asci with four adjacent defective spores and four adjacent. spores which appear to be normal (Type III ascus, fig. 8). If adjacent II disjunctions occurred, all eight spores would be deficient for relatively large segments of the genomic complement. All eight spores could be expected to show visible evidence of the deficiencies (Type IV ascus, fig. 3). On the supposition that the spores with 14-2 are defective in appearance and those with 3+4 are normal in appearance, a fourth type of eight-spored asecus could be anticipated. This would be formed whenever a crossover had occurred between the cen- tromere and the position of break (regions a or b, fig. 8). In this Neurospora translocation, such a crossover is probably confined almost entirely. to the long segment of region a. Few crossovers would be expected to occur in the very short b segment. Studies of the disjunction of the four chromosomes involved in a heterozygous translocation in maize have re- vealed that whenever a crossover has taken place between the centromere and the position of breakage, homologous centromeres will pass to opposite poles at anaphase I (McClintock, unpublished). If this crossover-disjunction relationship likewise applies to Neurospora, the resulting eight-spored asci should possess four spores with normal genomic comple- ments (two with 144 and two with 2-+3), two nor- mal appearing spores with the short deficiency (8-+-4) and two adjacent defective spores with the longer deficiency (1+2) (Type II ascus, fig. 3). As table 1 indicates, four main types of asci cor- responding to types I to IV, figure 8, were observed. The eight-spored asci were all relatively young, as the counts were made from slides prepared for chro- mosome studies. In each count, the relative frequen- cies of the types of asci are similar. Observations of the spore appearances in mature asci were made by Mrs. Mary B. Houlahan. She found that the asci with two very defective spores had, in addition, two immature appearing spores plus four normal ap- pearing spores. These should be type IT asci; the spores with the short deficiency, (8+-4), not dis- tinguishable in the young stage from spores having a normal genomic complement, are now detectable because of their slower rate of maturity. It should be stated that in ascus type II, the two adjacent defective spores occupied any one of the four possible positions in the ascus, with approxi- 676 TasLe l. Frequencies of asci with normal and defective spores in six preparations. The zygote nuclei were heterozygous for a translocation associated with mu- tant strain 45502." Typel TypelIIascus TypeIIIascus TypelV ascus 2 defective 4 adjacent ascus sister spores defective spores All8spores and6normal and4normal Al} 8 spores normal spores spores defective 20 50 28 Vw 24 41 34 8 37 85 32 18 37 58 30 17 54 97 50 16 25 50 24 7 Totals 197 381 198 93” * Record was made of 17 asci with normal and defective spore orientations other than types II and III of this table. See text for description. >In making these slides for chromosome studies, many of the asci of types I to III were broken and their spores scattered. Only non-broken asci were scored. Type IV asci were not so readily broken. Thus, the figure for type IV probably is relatively too high. mately equal frequencies. This is to be expected if the orientation of the chromosomes at metaphase I and II is at random with respect to the long axis of the ascus. Likewise, in ascus type III, the four adja- cent defective spores occupied positions either at the base or the tip of the ascus. On the basis of the explanation of the types of eight-spored asci given above, the following conclu- sions may be drawn: (1) When no crossing over occurs between the centromere and the point of in- terchange, alternate and adjacent I disjunctions will occur equally frequently (types I and ITI, table 1). (2) Adjacent II disjunctions are relatively infre- quent (type IV, table 1; see accompanying foot- note). (3) A crossover occurs in the longer chromo- some between the centromere and the position of breakage in approximately half of the ascus nuclei (type II, table 1). It is fully realized that these studies are only preliminary and require further investigation. Nevertheless, the author wishes to emphasize the possible usefulness of this type of analysis as a complement to the cytological ob- servations. . A POSSIBLE METHOD FOR DETERMINING THE FRE- QUENCY OF TRANSPOSITION OF sPoRES.—In many genetic analyses, the order of the spores in an ascus is of prime importance. The eight spores in an ascus are linearly arranged and are assumed to reflect the orientation of the nuclei and spindles in the three preceding divisions in the ascus. Following division I, the two resulting nuclei are some distance apart in the ascus cytoplasm. The spindles they form are par- alle] to the long axis of the ascus. Thus, following the second division, four nuclei are present, the upper two derived from one nucleus, the lower two derived from the second nucleus. Maintaining their respec- AMERICAN JOURNAL OF BOTANY [Vol. 32, tive positions in the ascus cytoplasm, each nucleus again divides and a row of eight free nuclei are formed. It is not until then that walls appear cutting out the eight spores. If no disturbances have occurred in the arrangement of the nuclei and spindles during the free-nucleated stage, the position of each spore reflects its origin with respect to the three preceding divisions. Lack of wall formation following divisions I and II in the ascus is a distinct disadvantage. Irregularities in spindle orientation or transposition of the usual order of two or more of the free nuclei will lead to linear arrangements of spores which do not reflect their origin in the previous spindles. Irregularities of this sort are known to occur and it is important for some investigations to determine their frequencies. : The reciprocal translocation in mutant strain 45502 or a chromosomal abnormality giving similar types of recognizable defective spores, might be use- ful for estimating the frequency of occurrence of aberrant alignments of some of the spores in an ascus. In addition to the ascus types recorded in table 1, there were 17 asci with normal and defective spore orientation other than types II and III. If. after the second meiotic mitosis following an adja- cent I disjunction described above, the two inner nuclei (with 1-2 and 8+-4, respectively) exchanged positions, ‘the spore alignment would not be type III. Instead, two adjacent normal appearing spores (with 3+-4) would be inserted between the two scts of sister defective spores (with 1+2). Seven of the 17 aberrant asci were of this type. If, following division III in an ascus destined to be of type II. two non-sister nuclei exchanged positions, a spore alignment other than type IT could appear. This would occur if one of these nuclei possessed the long deficiency (1-+2) which gives rise to the defective appearing spores. In these asci, the two defective appearing spores would now be separated by a normal appearing spore. Five such asci were ob- served among the 17 aberrant asci mentioned in the footnote to table 1. These observations are not con- sidered adequate for estimating the frequency of nuclear displacements. More study needs to be given to the aberrant asci to determine whether displace- ment of spores may occur after spore delimitation through rough handling, or whether additional dis- turbances, such as aberrant chromosomal behavior. are contributing factors. Because of the significance of aberrant alignment of spores in genetic investiga- tions, it was considered worth while to mention 4 possible rapid method of estimating their frequen cies. Concxusions.—The usefulness of fungi as genetic material has been well demonstrated in recent years. To interpret properly the results of many genetic investigations, it is either advantageous or necessary to know the accompanying chromosomal conditions. On the basis of this brief study of Neurospora chro- mosomes, the author anticipates that some fungi may prove to be adequate and in some respects superior cytogenetic material. A review of the literature sug Dec., 1943] gests that some forms may be distinctly superior to Neurospora for studies of chromosome behavior, particularly of those stages from fertilization to the first meiotic metaphase. Forms with two haploid chromosomes, one of which is associated with the nucleolus, might prove to be very satisfactory in fol- lowing the stages and motions of the chromosomes during synapsis, in studying the consequences of various chromosomal rearrangements and for other studies involving the meiotic prophase periods. In ascomycetes, the ease of isolation of the asci, the abundance of asci and the relation of size to stage in meiosis should recommend this material for tissue cultures when it is desired to observe the chromo- somes during the meiotic stages in living nuclei. The haploid chromosomal complement of Neuro- spora crassa is similar in its organization to that observed in many organisms. Each of the seven chro- mosomes may be identified not only by its relative length, the position of its centromere, but also by the constancy of its internal organization as exhibited by chromomere patterns in the meiotic prophase. One chromosome of the haploid complement pos- sesses a nucleolus organizer which functions just as it does in other organisms. Because of the location of the nucleolus organizer near the end of one arm of this chromosome, there is a minute satellite. Even the coiling and uncoiling processes leading to con- traction and expansion of the chromonema appear to be similar to that observed in many other organ- isms. No distinctively unique features of chromo- somal organization were recognized. The presence of translocations between non-homologous chromo- somes following irradiation treatment and the be- havior of these translocated chromosomes in the meiotic stages of heterozygous asci likewise are in- dicative of the orthodox organization of the Neuro- spora nuclei and chromosomes. It has been observed that the behavior of the chro- mosomes in the first two mitoses in the ascus results in the formation of four haploid nuclei whose chro- mosomes have been subjected to the processes com- mon to meiosis in general: synapsis of homologous chromosomes, chiasma formation, and typical ana- phase I and II disjunctions and segregations of chromatids. The synaptic period, however, is dis- tinetly atypical. In many organisms, synapsis is in- itiated in the meiotic prophase when the chromo- somes are much extended. In the Neurospora strains most extensively studied, this period occurs when the chromosomes are contracted, short rods simulat- ing late prophase chromosomes. Elongation of the chromosomes to their maximum meiotic prophase extension takes place after the chromosomes have become homologously associated throughout their lengths. If the chromonema within each chromosome at the time of synaptic attraction and association is tightly coiled, the homologous associations along the chromosomes cannot be equally intimate. Other cases of synaptic attraction of condensed chromosomes have been described but Neurospora offers rather unique opportunities for studying this process. MC CLINTOCK——-CHROMOSOMES OF NEUROSPORA 677 The centriole has not been considered in previous sections of this report, but it deserves a brief men- tion because of its steady enlargement during the interphase stages of the divisions in the ascus, its relation to the centromeres during this enlargement, as well as its previously known function in initiating spore wall formation (Harper, 1905; Dodge, 1927; Wilcox, 1928). As mentioned previously, the inter- kinetic nuclei following divisions I, II and III are somewhat pear-shaped because of a decided protru- sion or “beak.’” The centromeres of all chromosomes form a compact aggregate at the apex of this beak. The centriole begins to enlarge into a rod-shaped structure following division I. It functions asa typi- eal centriole in division II. During the following interkineses, the process of enlargement in contact with the centromeres continues. It again functions as a typical centriole during the third mitosis. (For illustrations, see Plates I and II, Dodge, 1927). Following the third division, the greatly elongated centriole, associated with the beak of each nucleus, comes to lie close to the ascus wall. Fibers emerge from it and encompass a mass of cytoplasm about each nucleus thus initiating spore wall formation. That centromeres, centrioles and blepharoplasts are interchangeable cell organelles has been demon- strated in the classic investigations of Pollister and Pollister (1943). In line with these investigations, it is possible to consider that the centromeres of Neurospora may contribute to the substance of the centriole during these periods of enlargement. Cen- tromeres, centrioles and blepharoplasts all have the common function of producing fibers. It is possible that the fibers formed by these three interrelated but morphologically distinct cellular organelles are structurally identical or much alike in that they all possess one particular type of molecular organiza- tion which is responsible for their capacity to con- tract or alternately contract and expand. SUMMARY A summary report is given of the results obtained from a very brief study of chromosome and nuclear behavior in Neurospora crassa. The investigations are admittedly incomplete and possibly some errors have been made. Nevertheless, they have revealed that Neurospora offers adequate and in some re- spects unique opportunities for cytogenetic research. The chromosomes were followed from the nuclear division preceding zygote formation through the division in the ascospore. Chromosome morphology was considered with reference to the absolute and relative sizes of the seven chromosomes in various division cycles, the centromere positions, the nucleo- lus chromosomes, the pachytene chromomere mor- phology and the presence of heterochromatin. Chro- mosome behavior was followed with reference to the atypical timing of chromosome synapsis, the elonga- tion of the chromosomes during a prolonged “‘pachy- tene,” chiasma formation and the general behavior of the chromosomes in the two meiotic mitoses and the two subsequent equational mitoses. Several re- 678 ciprocal translocations were investigated and their usefulness for special studies indicated. DEPARTMENT OF GENETICS, Cannecre [xstiTuTION or WasHINGTON, Cotp Srrtna Harson, New Yorx LITERATURE CITED Corson, B. 1934. The cytology and morphology of Neu- rospora tetrasperma (Shear and Dodge). Ann, Bot. 48: 211-224. . Donor, B. O. 1927. Nuclear phenomena associated with heterothallism and homothallism in the ascomycete Neurospora. Jour. Agric. Res. 34: 289-305. AMERICAN JOURNAL OF BOTANY [Vol. 32, Harper, R. A. 1905. Sexual reproduction and the or- ganization of the nucleus in certain mildews. Car- negie Inst. Wash. Pub. 87: 104 p. LIinperGREN, C. C., anv 8S. Rumann. 1938. The chromo- somes of Neurospora crassa. Jour. Genetics 36: 395- 404. Poutister, A. W., anv P. F. Powrister. 1943. The rela- tion between centriole and centromere in atypical spermatogenesis of viviparid snails. Ann. New York Acad. Sci, 45: 1-48. Witcox, M. S. 1928. The sexuality and arrangement of the spores in the ascus of Neurospora sitophila. My- cologia 20: 3-17.