Copyright 0 1994 by the Genetics Society of America

The Relationship of Homologous Synapsis and Crossing Over in a Maize Inversion

M. P. Maguire and R. W. Riess

Zoology Department, University of Texas, Austin, Texas 78712 Manuscript received August 30, 1993

Accepted for publication January 18, 1994

ABSTRACT Frequency of homologous synapsis at pachytene for a relatively short heterozygous inversion was com-

pared to the frequency of crossover occurrence within the inversion and to the frequency of the presence of a recombination nodule within the homologously synapsed inverted region. Crossover frequencies were estimated from bridge-fragment frequencies at anaphase I and anaphase 11. Recombination nodules (RNs) were observed in electron micrographs. Results show very similar frequencies of homologous synapsis and the occurrence of reciprocal recombination within the inverted region, consistent with the interpretation that establishment of homologous synapsis in this case is related to at least commitment to the form of resolution of crossover intermediates which gives rise to reciprocal recombination, not conversion only, events. An RN was generally found at pachytene in homologously synapsed inverted regions.

F OR many years conventional wisdom dictated that reciprocal crossing over followed the completion of

synapsis at meiosis I, despite some early warning that this might not be the case. Early warning took the form of studies which indicated a strong correlation in heterozy- gotes between successful homologous synapsis of rear- ranged chromosome segments, and the occurrence of reciprocal exchange between them and the correspond- ing normal sequence segments (MAGUIRE 1965, 1966, 1972,1977). On the basis of these studies itwas proposed that the stage of commitment to crossing over might indeed precede or accompany the full formation of the tripartite synaptonemal complex (SC) structure. Now it appears that, while there was insight in this interpreta- tion, the understanding of the true complexity of the process now emerging should provide many answers.

Studies of meiosis in yeast have been rewarding. Re- cent reports suggest that genome-wide homology search, which leads to initial stages of homolog align- ment, begins very early, immediately following bulk DNA replication or even premeiotically (SCHERTHAN et al. 1992; KLECKNER and WEINER 1993). From synchro- nized cultures (KLECKNER et al. 1991; also reviewed by HAwyand ARBEL 1993), the timing with respect to mei- otic landmarks of appearance and disappearance of DNA configurations which are expected to be recom- bination intermediates has suggested the following se- quence and insights. Homology is probably recognized in permissive chromosome structure at sites for prospec- tive double strand breaks (DSBs) where an initial semi- stable physical connection is made. Then more intimate interaction is thought to occur where homology has been identified, and this probably requires 3’ tails of DSBs with neither 3’ nor 5‘ tails isolatable after this point. Appearance of DSBs with tightly coupled resec-

Genetics 137: 281-288 (May, 1994)

tion of 5’ ends and first appearance of tripartite SC struc- ture seem to be nearly contemporaneous, with hetero- duplex formation relatively late in prophase. Resolution of heteroduplex structures seems to closely follow SC disintegration. It is guessed that bulk SC components may inhibit recombination “maturation” until SCs have been disassembled. Researchers interpret the pattern to suggest that the SC serves some function other than pro- vision for meiotic crossover events which it may in fact inhibit or stall. Additional support for this idea is pro- vided by behavior of certain meiotic mutants in yeast, such as redl , mer l , hop1 and z i p l , and d s y l in maize which are defective for or totally lacking in important SC components but nevertheless show at least moderate lev- els of crossing over (ROCKMILL and ROEDER 1990; ENCEBRECHT et al. 1990; HOLLINGSWORTH et al. 1990; SW et al. 1993; MAGUIRE et al. 1993). On the basis of obser- vations of the timing of appearance and distribution of recombination nodules and synapsis in tomato, STACK and ANDERSON (1986) have also proposed that recom- bination may be initiated early and that the distribution of reciprocal recombination events is somehow depend- ent on some distance function of synaptic extension in accordance with interference. EGEL (1978) and MACUIRE

(1968, 1988) have inferred that the presence of tripar- tite SC may be inhibitory to reciprocal recombinafion on the basis of crossover and interference information.

The nature of the genetic map construct determines that crossovers shall be randomly distributed along it. However, wherever comparisons have been feasible, they seem to demonstrate that physical cytological maps of chromosomes do not correlate well with their genetic maps. In general, there appear to be hotspots and also coldspots for crossing over along the physical entities, where the genetic map is respectively expanded or

282 M. P. Maguire and R. W. Riess

contracted. Attention to localized recombination fre- quencies in the presence of translocations (HAWLEY 1980) and duplicating fragments (ROSE et al. 1984) has led to the interpretation that centers specialized for ho- mologous pairing indeed exist along chromosomes. Re- cently, various reports have suggested that hotspots for recombination may represent sites for pairing initiation [see GOLDWAY et al. (1993) and references therein].

Late recombination nodules (RNs) which are found at mid to late pachytene with electron micrograph (EM) observation following appropriate fixation and staining have been generally shown to correlate for frequency and position with chiasmata or other crossover mani- festations observed later (VON WETTSTEIN et al. 1984; ALBINI and JONES 1988; CARPENTER 1988; STACK and ANDERSON 1989; ZICKLER et al. 1992; STACK et al . 1993; HERICKHOFF et al. 1993). They therefore provide a tool at pachytene for the observation of probable crossover locations. At zygotene in material with large chromo- somes, initiation of meiotic pairing and synapsis are ac- companied by the appearance of SC-associated nodules which are generally smaller and much more numerous but otherwise appear similar to pachytene RNs (ALBINI andJoms 1987; Anderson and STACK 1988). Some stud- ies have suggested correlation between pairing initiation sites and positions of RNs, chiasmata and recombination (ZICKLER et al. 1992; STACK et al. 1993). ZICKLER et al . (1992) found correlations between frequency and loca- tion of reciprocal exchanges and frequency and location of late RNs in wild-type Sordaria but disparity between location of nodules and exchanges in those mutants which cause reduction in recombination but do not cause SC arrest or major abnormalities. They suggested that possible effects in the mutants, such as premature loss of nodules, or late synapsis might account for the disparities. Information for interrelationships of conver- sion, recombination and synapsis was more limited and less clear. STACK and SOULLIERE (1984) found a 1:l re- lationship between synapsis of distal regions of chromo- somes involved in complex translocation heterozygote configurations in Rhoea spathacea. But, in the absence of rearrangement, with long chromosomes the number of synaptic initiations may greatly exceed the number of chiasmata (HASENKAMPF 1984; GILLIES 1985), and synap tic initiation in heterozygotes for structural change has been found in the absence of evidence for crossing over, e.g., in tomatoes heterozygous for translocations (HERICKHOFF et al. 1993). On the one hand it is possible that in cases where there is a 1:l relationship between crossing over and synaptic initiation with change of pair- ing partner that the genetic map of the region involved is so large that this is the conventional expectation. On the other hand it is also possible that, where there is synaptic initiation in the absence of crossing over, this is due to late synapsis of unpaired ends, too late for crossing over. Other conceivable complicating factors

may exist such as the existence of potential crossover sites with differing probabilities for initiations of cross- overs and change in the use of these sites with chromo- some structural change, as well as possible change in synaptic potential with chromosome rearrangement (PARKER et al. 1982). MCKIM et al. (1993) have reported two types of sites required for meiotic chromosome pair- ing in Caenorhabditis elegans with one localized near one end of each chromosome.

The present study takes advantage of the fact that, for homologous pairing to occur in a heterozygously in- verted region, pairing must be initiated within it rather than extended from an initiation elsewhere. It utilizes an inversion with an extent of about 30 map units so that conventional expectation would call for a maximum crossover frequency within it of about 60%. Maize chro- mosomes are sufficiently large to allow study of their synapsis at pachytene, and with current technology RNs can be reliably stained and identified in EM. The silver- staining procedure with EM viewing allows not only RN visualization but also much more accurate visualization of synaptic configurations than is possible with the light microscope observation of conventional acetocarmine smears used in the prior studies of maize rearrange- ments. In addition, a much smaller proportion of stained cells is unclassifiable. Bridge and fragment analy- sis at anaphase I and I1 provides information on the frequency of crossovers within the inverted region. Re- sults suggest that homologous synapsis of the inverted region is associated with the presence of a crossover and an RN.

MATERIALS AND METHODS

Maize seeds heterozygous for Inv 5083 (provided by GREGORY DOYLE) were grown in a growth chamber with a con- trolled environment. Growth chamber culture has been found to reduce the sample to sample variations in synaptic and cross- over frequencies such as those which were found in the field grown plants of the early study (MAGUIRE 1966). The inversion is reported to have breakpoints at chromosome ZL 0.70 and 0.87 of the length of the long arm from its centromere (LONGLEY 1961). On the basis of the most recent tentative cy- tological map of maize chromosome I in the 1993 Maize Genetics Newsletter the inverted region constitutes about 30 map units. At meiotic stages samples were collected and pre- pared by two differing procedures. (1) Cells from fresh anthers at pachytene were spread on plastic coated slides following maceration of anthers and fixation by a technique described elsewhere ( M A U I R E et al. 1993). Slides were then silver-stained by a procedure which allows reliable visualization of SC lateral elements and RNs (SHERMAN et al . 1992). EM grids were PO- sitioned on suitably spread and stained cells, and grids were floated on their plastic rafts on a water surface and picked up and dried on lens paper before examination and photography with a Siemens Elmiskop 1A electron microscope. (2) A sec- ond meiotic sample was collected and immediately fixed in ethanol-acetic 3:l mixture and stored in a freezer until ex- amination with systematic scanning of conventional acetocar- mine squash preparations of anaphase I and also anaphase 11 slides. Anaphase I cells were scored as either no bridge or

SynapsisCrossing Over Relationship 283

FIGURE 1 .Silver-stained surface spread microsporocyte at pachytene showing reverse pairing loop for the inverted re- gion with an RN (arrow) and also an RN (arrow) in the distal region. This is from grid number 5B box 13. Bar = 2 pm.

FIGURE 3.-Silver-stained surface spread microsporocyte at pachytene showing reverse pairing loop for the inverted re- gion with an RN (arrow) and also an RN (arrow) in the distal region. This is from grid number 9E box 13. Bar = 2 pm.

FIGURE 2.4ilver-stained surface spread microsporocyte at pachytene showing reverse pairing loop for the inverted re- gion with an RN (arrow). This is from grid number 4D box 13. Bar = 2 pm.

fragment present, a single bridge with one fragment, fragment only, or double bridge and two fragments. Anaphase I1 cells were scored for the presence of a bridge.

RESULTS

A total of 100 pachytene cells appropriately fixed and silver-stained were examined in EM for the presence of homologous synapsis of the heterozygously inverted seg-

ment. Of these, 29 showed at least a short homologously synapsed segment (with reverse pairing). The remain- der (71) showed no such reverse synapsis and for the most part showed normal appearing, but necessarily nonhomologous, synapsis across the inverted region, a commonplace occurrence in maize inversion heterozy- gotes (MAGUIRE 1972). Of the 29 cells with homologous synapsis of the inverted region, 24 contained a single identifiable RN within the region of reverse pairing (as illustrated in Figures 1-3), three contained no RN and two were unclassifiable for the presence of an RN. Thus it seems likely that usually homologous synapsis of the inverted region was accompanied by a crossover. In 15 of the cells in which reverse synapsis occurred and an RN was present in the homologously paired inverted region, stretching and distortion appeared to be absent, and the entire extent of chromosome I could be traced and measured. Results of measurements and positions of RNs for these cells are indicated graphically in Figure 4 for both the inverted and distal regions. These results are also tabulated in Table 1. In these configurations the distal region of the long arm of chromosome Iwas usu- ally homologously synapsed, and in 9 of the 15 cells in which chromosome I was traceable, an RN was present in the distal region, three contained no RN and three were unclassifiable for the presence of an RN. There- fore, homologous synapsis of the distal region in the

284 M. P. Maguire and R. W. Riess

9D box 13 40 box 13

1OA box 14 58 box 13 9E box 14 2 8Abox13

2 4Abox13 7Dbox 14 5 8Dbox14 6C box 13 9E box 13 6Abox 14 7C box 13 9A box 12

E 6cbox12

Fraction of Total Length Chromosome 1

.oo .05 .10 .15 .20 .25

.072 .165

" Distal region Inverted region

FIGURE 4.-Graphical representation of the positions of homologous synapsis in the in- verted and distal regions of the long arm of chromosome 1. Grid identitiesof these prepa- rations are indicated at the left, and the hori- zontal axis is the scale representation of the fraction of total length which is involved of the entire chromosome from the distal end of the long arm. Lines at the left represent ho- mologous synapsis extent of the distal region; lines to the right represent homologous syn- apsis of the inverted region. In each case, matching unsynapsed chromosome lengths of unsynapsed regions have been averaged for the two homologs. These tend to stretch more readily than synapsed regions, and the aver- age value seems the most fair representation. RNs are represented by Xs for the two ho- mologs so that each RN within the inversion appears here as two Xs in mirror image po- sitions proximally and distally within the re- verse pairing region. For the distal region each X represents an RN.

TABLE 1

Positions of RNs and synapsis of inverted and distal regions as a fraction of total length of chromosome 1

Grid no.

9D box 13 4D box 13 1OA box 14 5B box 13 9E box 14 8A box 13 6C box 12 4A box 13 7D box 14 8D box 14 6C box 13 9E box 13 6A box 14 7C box 13 9A box 12

n Mean SD

Position of reverse pairing

0.120 0.256 0.064 0.205 0.077 0.169 0.048 0.193 0.050 0.169 0.106 0.166 0.075 0.137 0.063 0.150 0.075 0.162 0.097 0.119 0.074 0.159 0.109 0.139 0.128 0.185 0.116 0.140 0.140 0.150

15 15 0.089 0.167 0.029 0.034

Reverse pairing RN position ~~ ~

One homolog

0.127 0.088 0.085 0.092 0.083 0.113 0.086 0.090 0.103 0.097 0.108 0.118 0.152 0.127 0.144

Other Distal region homolog Extent distal region pairing RN position

0.249 0.181 0.161 0.150 0.136 0.159 0.126 0.122 0.133 0.119 0.124 0.130 0.162 0.129 0.146

15 15 0.108 0.148 0.022 0.033

0.000 0.000 0.000 0.000 0.01 1 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000

0.089 0.004 0.032 0.048 0.025 0.050 0.060 0.029 0.054 0.069 0.047 0.039 0.034 0.061 0.082

Unclassifiable None 0.032 0.013 0.018

Unclassifiable 0.039 None 0.000 0.032 None 0.006 0.032

Unclassifiable 0.072

15 9 0.048 0.027 0.022 0.022

presence of inverted region homologous synapsis was probably also usually accompanied by a crossover event.

Findings from bridge and fragment counts in aceto- carmine squash preparations at anaphase I are pre- sented in Table 2. In summary, 74% of 1000 cells at anaphase I showed no bridge or fragment, 20% showed a bridge and a fragment, as expected for cells with either a single crossover within the inversion (Figure 5) (by far the most common case) or for cells with a three-strand double within the inversion (Figure 6), 6% showed a fragment only (see Figure 7 and description below) and 0.40% showed a double bridge and two fragments as expected for cells with four-strand doubles within the inversion (Figure 8). Two-strand doubles within the in-

version are totally invisible. With the assumption of no chromatid interference, three-strand doubles within should constitute 1 the doubles while two- and four- strand doubles should each be : of the total. Thus, use of the double bridge and fragment frequency to esti- mate total frequency of cells with double crossovers within the inversion calls for quadrupling of the 0.40% frequency to 1.60% frequency. An additional source of cells with at least one crossover within the inversion is represented by cells which show a fragment only at an- aphase I and a bridge at anaphase 11. These represent cells in which a single crossover occurred within the in- version, and a three-strand double type second crossover occurred in the proximal region (Figure 7). Anaphase

Synapsis-Crossing Over Relationship

TABLE 2

285

Bridge and fragment frequencies at anaphase I

No bridge or fragment Bridge and fragment Fragment only Double bridge 2 fragments

No. Percent No. Percent No. Percent No. Percent

744/1000 74 196/1000 20 56/1000 6 4/ 1000 0.4

FIGURE 5.-Diagrammatic representation of the inverted re- gion at pachytene with a single crossover within the inversion. One homolog is represented by heavy lines, the other by light lines; boxes at the left represent centromeres; the X intersec- tion represents a crossover. The second and fourth chromatids down from the top are involved in the crossover. Tracing from the left will indicate that these now represent one continuous dicentric chromatid which will form a bridge at anaphase I. Tracing from the rightwill indicate that the other ends of these chromatids now represent a single acentric fragment.

I I

FIGURE 6.-Diagrammatic representation of the inverted re- gion at pachytene with a three-strand double crossover within the inversion. One homolog is represented by heavy lines, the other by light lines; boxes at the left represent centromeres; the X in- tersections represent crossovers. Tracing the topmost chromatid from the left will indicate that it now represents a continuous dicentric chromatid with the fourth chromatid down from the top which will form a bridge at anaphase I. Tracing the second chromatid down from the right will indicate that it now repre- sents a continuous acentric chromatid with the fourth chromatid down from the top which will form a fragment at anaphase I.

I cells of this class were found with a frequency of 6%. An additional estimate of the frequency of anaphase I cells of this last class is provided by the anaphase I1 ob-

FIGURE 7.-Diagrammatic representation of the inverted re- gion at pachytene with a three-strand double crossover with one crossover within the inversion and the other in the region proxi- mal to the inversion. One homolog is represented by heavy lines, the other by light lines; boxes at the left represent centromeres; the X intersections represent crossovers. Tracing the topmost chromatid from the left will indicate that it now represents a loop chromatid with the two left ends attached to sister centromeres of the top homolog. These sister centromeres will move to the same pole at anaphase I and form a bridge at anaphase 11. Tracing the second chromatid down from the top from the right end will indicate that it now represents a continuous acentric chromatid with the fourth chromatid down from the top which will form a fragment at anaphase I.

servations where such cells are represented by a bridge. From scoring of 1074 anaphase I1 cells, 3% showed such a bridge, and since each meiosis I cell produces two an- aphase I1 cells, only one of which will have the telltale bridge, the anaphase I1 cell information also suggests a frequency of 6% anaphase I cells with one crossover within the inversion and a second constituting a three- strand double in the proximal region. The total fre- quency of cells with at least one crossover within the inversion which is indicated by the anaphase data there- fore represents 27.6%. This corresponds to the 29% fre- quency of cells found to have at least some homologous synapsis in the inverted region at pachytene and to the 24% which contained a single identifiable RN within that region, with 2% unclassifiable in this latter respect.

It must be emphasized that there is no test of con- version frequency in this material. Only reciprocal re- combinants are scored, many of which may be accom- panied by conversion, but the frequency of conversion only events is totally unknown.

Findings are therefore consistent with the supposi- tions that in this material, occurrence of homologous synapsis of the inversion at pachytene is closely corre-

286 M. P. Maguire and R. W. Riess

FIGURE 8.-Diagrammatic representation of the inverted re- gion at pachytene with a four-strand double crossover within the inversion. One homolog is represented by heavy lines, the other by light lines; boxes at the left represent centromeres; the X intersections represent crossovers. Tracing the topmost chromatid at the left end will indicate that it now represents a continuous dicentric chromatid with the third chromatid down from the top which will form a bridge at anaphase I. Tracing the second chromatid down from the top at the left will indicate that it now represents a continuous acentric chro- matid with the fourth chromatid down from the top which will also form a bridge at anaphase I. Tracing the topmost chro- matid at the right will indicate that it now represents a con- tinuous acentric chromatid with the third chromatid down from the top which will form a fragment at anaphase I. Tracing the second chromatid down at the right will indicate that it now represents a continuous acentric chromatid with the fourth chromatid down from the top which will form a sec- ond fragment at anaphase I.

lated with the occurrence of at least one crossover within it, and that the existence of an RN at this stage is also closely correlated with the occurrence of a crossover. Unfortunately, there is no test here of crossover fre- quency in the region distal to the inversion. If there are hotspots for crossing over within the inverted region, this study does not resolve them. However, with an es- timate of 30 map units within the inverted region, if this region were homologously synapsed at pachytene throughout its extent in all cells, conventional expec- tation would call for crossing over to occur in this region in 60% of cells. But it is evident that even in those cells where there was homologous synapsis at pachytene of the inverted region, the extent of this synapsis was highly variable and often substantially less than the full inverted region. In addition there was no homologous synapsis of the inverted region in 71% of the cells. It is difficult to escape the conclusion that in this case there is an un- conventional 1:l relationship between occurrence of crossing over and an event of homologous synapsis at pachytene which is assumed to be stable in maize since synaptic adjustment apparently does not occur (ANDERSON et al. 1988).

Findings overall are similar to those of the first such study with this inversion (MAGUIRE 1966), but frequen- cies differ both within the early study and between that

study and the present results. In that experiment, three plants were studied with pachytene inversion homolo- gous synapsis of 36.0,30.1 and 34.9% compared to com- bined bridge-fragment and fragment only frequencies of 35.8, 29.6 and 34.2%, respectively. In addition, the frequency of the fragment only classes comprised a higher proportion of the total. Since the early study was done with field-grown plants in the presence of much environmental variation, perhaps differences among those plants and between those and the plants of the present study of growth chamber grown plants (with a constant climate) are not surprising. The coordinate variations of synaptic and crossover frequencies throughout may serve to demonstrate a strength of this relationship, however.

DISCUSSION

Findings of this report from studies of a eukaryote with large, complex chromosomes correspond well with expectations from research with yeast in that they are consistent with the interpretation that initiation of cross- over events precedes or accompanies tripartite SC for- mation. Estimates of crossover frequencies within the inversion from bridge and fragment frequencies are ex- pected to be accurate with only small margins of prob- able error, as a result of the fact that crossover frequen- cies may vary among samples from different parts of a tassel. Care was taken, however, to use samples for ace- tocarmine smears from the same general region of the tassel as those used for the pachytene spread prepara- tions. A smaller number of cells at pachytene (100) were examined, however, than at anaphase I (1000) and an- aphase I1 (1074), so that the frequencies observed may be somewhat less reliable at pachytene. The largest er- rors in this study are probably expected to exist in the measurements used to estimate positions of RNs within the inversion and distal regions. Breakpoints previously listed for Inv 5083 were estimated from empirical mea- surements of acetocarmine smear preparations where loop configurations were found and reported to be on the average at 0.70-0.87 of the long arm of chromosome Ifrom the centromere (LONGLEY 1961). From measure- ments in this study the average position of breaks, taking 1.32 as the ratio of long arm to short arm of chromosome I (NEUFFER et nl. 1968) differed somewhat from the pre- viously reported values of 0.072-0.165 from the distal end of the long arm of the chromosome. Here they were 0.089-0.167. Within this study the length of the segment which was homologously synapsed was highly variable, but this is believed to be due not to variation with stage ad- vancement, which (as stated above) seems not to occur in maize (ANDERSON et nL 1988) but to variation in stable ~ y " - aptic extent, some of which may be nonhomologous.

It must be emphasized that, unlike noteworthy cases in yeast (ENGEBRECHT et al. 1990) and Sordaria (ZICKLER et al. 1992) which have been studied, the close rela-

Synapsis-Crossing Over Relationship 287

tionship of homologous synapsis here to the occurrence of crossing over is strictly a relationship to reciprocal recombination, not conversion, although many of the crossovers may also be associated with conversion. How- ever, strong relationship of existence of overall homolo- gous synapsis to conversion only events has been found in the mer1 mutant of yeast in cases where some of the usual defects are corrected by presence of the MER2 gene in high copy number. Synapsis and gene conver- sion are restored to normal by this condition, but re- ciprocal recombination frequency is still depressed, al- though some is found as was the case without the extra copies of MER2 (ENGEBRECHT et al. 1990). Exclusion of noncrossover or conversion only events from the rela- tionship also exists for the many reports of close corre- spondence of pachytene RNs in frequency and position to chiasmata. Where this relationship is also related to synaptic initiation with change of pairing partner, im- portant inferences can be drawn. Perhaps there is com- mitment at some kinds of synaptic initiation to resolu- tion of crossover intermediates to the form which gives rise to reciprocal recombination. At each crossover event the DSB model requires opposite resolutions of two Hollidayjunctions (ORR-WEAVER and SZOSTAK 1985). On the basis of studies with synthetic Hollidayjunctions in vitrowith purified bacterial enzymes, WEST (1990) has proposed that resolution may be unrelated to isomer- ization status (which would be difficult to envision when two DNA molecules are intertwined with a protein fila- ment). Instead, the DNA helices may be held by RecA protein in such a way that resolution can occur in either of the two ways. WEST also suggests that these protein- DNA structures may represent a primitive form of the recombination nodules that occur within synaptonemal complexes of eukaryotes. If some RNs can play a role in commitment to resolution of crossover intermediates, perhaps it is some of those at points of synaptic initiation which serve in this capacity while others may more often relate to events of conversion only. Maybe SC extension at first takes the form of a series of button-up initiations in long chromosomes each followed by zipping up. SC me- tabolism could be more complex than commonly appre- ciated in that it might play a role in chiasma interference (MAGUIRE 1968; EGEL 1978; MAGUIRE 1988). Foss et al. (1993) have noted that chiasma interference seems to de- pend upon genetic distance rather than physical distance and have proposed a model which is consistent with data from Drosophila and Neurospora. This model suggests that the alternative resolutions of randomly distributed re- combination intermediates are somehow constrained so that neighboring reciprocal recombinations must have a certain number of conversion only or noncrossover events between them. The tripartite structural components of the SC may play roles currently not realized.

The lesson is that many factors contribute to syn- apsis and crossover relationships, and superficial ap-

pearances are determined by those which predomi- nate. The task is to learn what these factors are and how they interact.

This work was supported by U.S. Department of Agriculture grant 92-01569. The authors are grateful to GREGORYDO~ZE for supplying the seeds of the Inv 5083 stock.

LITERATURE CITED

ALBINI, S. M., and G. H.]ONES, 1987 Synaptonemal complex spread- ing in Allium cepa and A. f istulosum. I. The initiation and se- quence of pairing. Chromosoma 95: 324-338.

ALBINI, S. M., and G. H. JONES, 1988 Synaptonemal complex spread- ing in Allium cepa and A. fistulosum. 11. Pachytene observations of the SC karyotype and the correspondence of late recombina- tion nodules and chiasmata. Genome 30: 399-340.

ANDERSON, L. IC, and S. M. STACK, 1988 Nodules associated with axial cores and synaptonemal complexes during zygotene in Psilotum. Chromosoma 97: 96-100.

ANDERSON, L. K., S. M. STACK and J. D. SHERMAN, 1988 Spreading synaptonemal complexes from Zea mays. I. No synaptic adjust- ment of inversion loops during pachytene. Chromosoma 96:

CARPENTER, A. T. C., 1988 Thoughts on recombination nodules, mei- otic recombination, and chiasmata, pp. 529-548 in Genetic Recombination, edited by R. KUCHERLAPOTT and G. R SMITH. American Society for Microbiology, Washington, D.C.

EGEL, R., 1978 Synaptonemal complexes and crossing over: struc- tural support or interference? Heredity 41: 233-237.

ENGEBRECHT,J.,]. HIRSCH and G. s. ROEDER, 1990 Meiotic gene con- version and crossing over: their relationship to each other and to chromosome synapsis and segregation. Cell 62: 927-937.

FOS, E., R. LANDE, F. W. STAHL and C. M. STEINBERG, 1993 Chiasma interference as a function of genetic distance. Genetics 133

GILLIES, C. B., 1985 An electron microscopic study of synaptonemal complex formation at zygotene in rye. Chromosoma 92: 165-175.

GOLDWAY, M., T. ARBEL and G. SIMCHEN, 1993 Meiotic nondisjunction and recombination of chromosome I1 and homologous frag- ments in Saccharomyces cerevisiae. Genetics 133: 149-158.

HASENKAMPF, C. A., 1984 Synaptonemal complex formation in pollen mother cells of Tradescantia. Chromosoma 90: 275-284.

HAWLEY, R. S., 1980 Chromosomal sites necessary for normal levels of meiotic recombination in Drosophila melanogaster. I. Evidence for and mapping of the sites. Genetics 9 4 625-646.

a m y , R. S., and T. ARBEL, 1993 Yeast genetics and the fall of the classical view of meiosis. Cell 72: 301-303.

HERICKHOFF, L., S. STACK and]. SHERMAN, 1993 The relationship be- tween synapsis, recombination nodules and chiasmata in tomato translocation heterozygotes. Heredity 71: 373-385.

HOLLINGSWORTH, N. M., L. G~ETSCH and B. BYERS, 1990 The HOP1 gene encodes a meiosis-specific component of yeast chrome somes. Cell 61: 73-84.

KLECKNER, N., and B. WEINER, 1993 Potential advantages of unstable interactions for pairing of chromosomes in meiotic, somatic and premeiotic cells. Cold Spring Harb. Symp. Quant. Biol. 58: (in press).

KLECKNER, N., R. PADMORE and D. K. BISHOP, 1991 Meiotic chrome

295-305.

681-691.

some metabolism: one view. Cold Spring Harbor Symp. Quant. Biol. 5 6 729-743.

LONGLEY, A. E., 1961 Breakage points for four corn translocation series and other corn chromosome aberrations. Crops Res. A.R.S.

MAGUIRE, M. P., 1965 The relationship of crossover frequency to syn-

MAGUIRE, M. P., 1966 The relationship of crossinrr over to chre

34-16: 1-40.

aptic extent at pachytene in maize. Genetics 51: 23-40. -

mosome synapsis in a short paracentric inversion. Genetics 53: 1071-1077.

MAGUIRE, M. P., 1968 The effect of synaptic partner change on cross- over frequency in adjacent regions of a trivalent. Genetics 59:

MAGUIRE, M. P., 1972 The temporal sequence of synaptic initiation, 381-390.

crossing over and synaptic completion. Genetics 70: 353-370.

288 M. P. Maguire and R. W. Riess

MAGUIRE, M. P., 1977 Homologous chromosome pairing. Philos. Trans. R. SOC. Lond. Ser. B 227: 245-258.

MAGUIRE, M. P., 1988 Crossover site determination and interference. J. Theor. Biol. 134: 565-570.

MAGUIRE, M. P., R. W. RIES and A. M. PAREDES, 1993 Evidence from a maize desynaptic mutant points to a probable role of synap- tonemal complex central region components in provision for sub sequent chiasma maintenance. Genome 36: 797-807.

MCKIM, K. S., K PETERS and A. M. ROSE, 1993 Two types of sites re- quired for meiotic chromosome pairing in Caenorhabditis el- egans. Genetics 134 749-768.

NEUFFER, M. G., L. JONES and M. S. ZUBER, 1968 The Mutants ofMaize. Crop Science Society of America, Madison, Wisc.

ORR-WEAVER, T. L., and J. W. SZOSTAK, 1985 Fungal recombination. Microbiol. Rev. 49: 33-58.

PARKER, J. S., R. W. PALMER, M. A. F. WH~EHORN and L. A. EDGAR, 1982 Chiasma frequency effects of structural chromosome change. Chromosoma 85: 673-686.

ROCKMILL, B., and G. S. ROEDER, 1990 Meiosis in asynaptic yeast. Ge- netics 126 563-574.

ROSE, A. M., D. L. BAILLIE and J. CURRAN, 1984 Meiotic pairing be- havior of two free duplications of linkage group I in Caenorhab- ditis elegans. Mol. Gen. Genet. 194: 52-56.

SCHERTHAN, H., J. LOIDL, T. SCHUSTER and D. SCHWEIZER, 1992 Meiotic chromosome condensation and pairing in S a c c h u m y c a dae studied by chromosome painting. Chromosoma 101: 590-595.

SHERMAN, J. D., L. D. HERICKHOFF and S. M. STACK, 1992 Silver staining two types of meiotic nodules. Genome 3 5 907-915.

STACK, S. M., and L. ANDERSON, 1986 Two dimensional spreads of synaptonemal complexes from solanaceous plants. 111. Recombi- nation nodules and crossing over in Lycopersicon esculentum (to- mato). Chromosoma 94: 253-258.

STACK, S. M., and L. ANDERSON, 1989 Chiasmata and recombination nodules in Lilium longi,florurn. Genome 3 2 486-498.

STACK, S. M., and D. SOULLIERE, 1984 Rhoea spathacea. I. The rela- tionship between synapsis and chiasma formation. Chromosoma

STACK, S. M., J. D. SHERMAN, L. K. ANDERSON and L. S. HERICKHOFF, 1993 Meiotic nodules in vascular plants, pp. 301-311 in Chro- mosomes Today,VoI. 11, edited by A. T. S U M N E R ~ ~ ~ A . C. CHANDLEY. Chapman & Hall, London.

SYM, M., J. ENGEBRECHT and G. S. ROEDER, 1993 ZZPZis a synaptonemal complex protein required for meiotic chromosome synapsis. Cell 7 2 365-378.

VON WETTSTEIN, D. S., W. RASMUSSEN and P. B. HOLM, 1984 The syn- aptonemal complex in genetic segregation. Annu. Rev. Genet. 18:

WEST, S. C., 1990 Processing of recombination intermediates in vitro. BioEssays 12 151-154.

ZICKLER, D., P. J. E. MOREAIJ, A. D. HUYNH and A. M. SLEZEC, 1992 Cor- relation between pairing initiation sites, recombination nodules and meiotic recombination in Sordaria macrospora. Genetics 132: 135-148.

90: 72-83.

331-413.

Communicating editor: P. J. PUIUULA