INFORMATIONAL TRANSFER IN MEIOTIC GENE CONVERSION* BY S. FOGELt AND R. K. MORTIMER - PNAS
←
→
Page content transcription
If your browser does not render page correctly, please read the page content below
INFORMATIONAL TRANSFER IN MEIOTIC GENE CONVERSION*
BY S. FOGELt AND R. K. MORTIMER
DIVISION OF MEDICAL PHYSICS, UNIVERSITY OF CALIFORNIA, BERKELEY
Communicated by Norman H. Giles, November 13, 1968
Abstract and Surmmary.-Aberrant meiotic segregations attributable to in-
tragenic events have been analyzed in an unselected sample of 1611 tetrads from
three heteroallelic diploids of Saccharomyces cerevisiae. Reciprocal recombina-
tion between alleles accounts for only a minor fraction of the total aberrant
tetrads, while the majority component is represented by single- and double-
site conversions. The frequency of double-site conversion is inversely related to
the physical length of the interallelic interval. Since double-site conversions do
not yield prototrophs, their occurrence leads to biased estimates of intragenic dis-
tances. Conversion is viewed as a process of informational transfer distinct from
conventional crossing-over. The implications of the findings for genetic fine
structure mapping and evolutionary theory are discussed briefly.
In diploid organisms that possess a conventional meiotic cycle and are amen-
able to tetrad analysis, it is expected that each ascus will contain two wild-type
(+) and two mutant spores (a) for every heterozygous site. However, aberrant
segregations, mainly 3 +: la and 1+ :3a, not ascribable to conventional genetic
mechanisms such as polyploidy or suppressors, do occur. These exceptional
events have been defined operationally as gene conversions. 1-4 In Saccharomyces
cerevisiae, gene conversion occurs in 1-2 per cent of all segregations.5 But the
respective meiotic conversion frequencies for various heterozygous sites vary
from approximately 0.1 to 10 per cent.5-7 For any given site, 3+: la and 1+:
3a segregations are detected with frequencies that do not differ significantly from
each other.
Analysis of the three mutant spores from 1+:3a tetrads reveals that con-
version operates with complete fidelity. Using a recombinational test, Roman8
was unable to detect differences among the three mutant spores in such an ascus.
Fogel, M\ortimer, and Hawthorne7 have shown that similarity at the mutant site
extends to the single nucleotide level. In addition, nonmutant spores in 3+: la
asci are indistinguishable at the gene product level.9
Gene conversion is also analyzable in diploids carrying in repulsion two inde-
pendent mutants at a single locus (a1+/+a2). Normal segregation yields two
a,+ spores and two +a2 spores in each tetrad. However, prototrophic revert-
ants (+ +) occur among the meiotic products of such heteroallelic diploids.
They result mainly from 3+: la segregations for either allele and to a consider-
ably lesser extent from reciprocal recombination between the input alleles. On
the assumption that these prototrophs arise from recombinational events, their
meiotic frequency has been widely used as a measure of the distance between the
mutated sites. Thus, it is paradoxical that reasonably consistent genetic fine
structure maps can be constructed on the basis of a metric that is essentially an
index of nonreciprocal recombination or conversion. Moreover, it is clear that
96
Downloaded by guest on May 24, 2021VOL. 62, 1969 GENETICS: FOGEL AND MORTIMER 97
selective procedures commonly used in conversion studies do not detect all intra-
genic events.8 Therefore, to resolve this paradox a comprehensive survey of all
intragenic events in unselected tetrads seemed essential.
In this study, gene conversion in unselected tetrads from three heteroallelic
diploids was examined. The emphasis was on a determination of the relationship
between the frequencies of different types of intragenic events and the physical
distance between the mutant sites. The present work leads to the view that
gene conversion involves replacement of the genetic information in the relevant
DNA segment with information that is identical to that carried in the correspond-
ing segment of the homologous non-sister chromatid. This process is designated
as informational transfer.
Materials and Methods.-The hybrids utilized in this study were synthesized from the
Berkeley yeast culture collection. Their genotypes are as follows:
BZ34 a + arg4-4 + thri + try5-48 + ura3 his5-2 lys1-1 ade2-1
0 -
a p1 + arg4-17 + leul-12 try5-48 metl + his6-2 lysl-l ade2-1
BZ140 a + arg4-2 + + Cul leul-1 + + tryl his5-2 + ade2-1
0 O0
a p1 + arg4-17 thrl cul + leul-12 try5-48 tryl his5-2 lys1-1 ade2-1
X841-1 a + arg4-1 + thrl + + his5-2 +
a pl + arg4-2 + leul tryl his5-2 ade2-1
Hybrids BZ34 and BZ140 were constructed for this study, while X841-1 is a tryptophan-
independent revertant isolated from the diploid X841 previously described.'0
The marker designations are those adopted at the Carbondale and Osaka Yeast Genetics
Conference." The linkage relations do not differ from those previously reported by
Hawthorne and Mortimer.'4 Relevant tetrad data, presented as PD:NPD/T ratios for
gene-gene intervals and as per cent second division segregations (% SDS) for gene-centro-
mere intervals, are given in Table 1. Map distances, X, in centimorgans were calculated
either from the relation X = ((6 NPD + T) X 100)/(2 (PD + NDP + T)) or from X
= % SDS/2." '13
The procedures and media for hybridization, sporulation, tetrad dissection, and scoring
genetic markers have been described.'4 Additionally, to score try5 in the presence of try1,
the tryptophan-deficient medium was supplemented with indole. Each ascosporal clone
was tested for mating type and for the identity of the alleles at the ar g4 and leu, loci by
methods detailed elsewhere."5 In the allele identification tests, mitotic reversion to
prototrophy was enhanced by exposure to a nonlethal dose of ultraviolet light (240 ergs/
mm2). Spore viability for each diploid was greater than 90%.
The mutant sequence within the arg4 cistron was established by two independent meth-
ods. Using the protocols described by Manney and Mortimer," we found the sequence
and distances between mutants consistent with their fine structure map. The ordering of
arg4 alleles relative to the centromere and the outside markers was also determined from
asci in which a reciprocal exchange had occurred between the mutant sites.
With the X-ray fine structure mapping approach, the slopes of linear dose-response
curves are calculated in X-ray map units. These provide an estimate of the nucleotide
interval between mutant sites. An X-ray map unit, defined as one prototroph/108 cells/
roentgen, is equivalent to about 200 nucleotides,'7 a value in agreement with the 150-
nucleotide estimate previously given by Manney anx Mortimer. The mutant sequence
within the arg4 cistron and their linkage relations to other markers on chromosome VIII
are depicted in Figure 1.
Downloaded by guest on May 24, 202198 GENETICS: FOGEL AND MORTIMER PROC. N. A. S.
Results.-Table 2 displays the summed
>e i qaberrant
N 10 segregation frequencies (3+: la
v s Is 55 o > and 1+:3a) for the various heterozygous
> >markers in the three diploids. The values
A range from 5.5 per cent for thr1 to about 0.3
CsE- per cent for the mating type locus. How-
I N ever, values as high as 10 per cent (ad5,7)
< v "4 s :3 and 25 per cent (S6) have been encountered
in our unpublished studies. Thus, meiotic
Ca
t gene conversion is a common event within
Qx X is n , the genome. The average conversion fre-
0va, ° C'i ci 0
quency for all the heterozygous sites in this
O M
study is 1.7 per cent, which is close to the
values reported by Roman5 and Takahashi6
s
C4 e for other sets of heterozygous sites in this
> I yeast. Where values for the identical al-
v s Q leles are available from different hybrids,
S they are remarkably consistent. Thus, the
- N At probability of conversion of a heterozygous
> I osite appears to be insensitive to modifica-
1QL
vZ tion by genetic background, since the three
sets of data were obtained from distantly
Iscv5
S d 6related hybrids. Also, from the data for
the two leua alleles it appears that conver-
t-N
4`4 sion frequency is allele-specific rather than
",a locus-specific.
< °.1k: ~lo ~1 >. > It-From Table2it isalso clearthat conver-
sions in either direction, i.e., 3+: la or
t |;q
I. 1+: 3a, are equally probable. A spurious
> -
=ethe three phenotypically petite spores from
coE °q, °t "- * D 1+: 3p asci, were outcrossed to wild type,
ooQco and their meiotic progenies were analyzed.
A,4:z It was found that most (8/10) 1+:3p con-
*¢ : fl versions were apparent rather than genuine.
. The phenotypic alteration in one of the
X three spores was attributable to a cyto-
0.
bo U plasmic change, since it failed to segregate.
Q Q 0
° . d @e A Some 1600 asci were analyzed in the
Nz N s = < .o - present work. The irregular tetrads for
X arg4 and leua were classifiable (Table 3) in
EH r three major categories, namely, single-
Downloaded by guest on May 24, 2021VOL. 62, 1969 GENETICS: FOGEL AND MORTIMER 99
TABLE 2. Frequency of irregular segregation for miscellaneous markers.
Markers and Segregation Pattern
No. of P thri Cu
asci + - + - + - + - + - + -
Diploid analyzed 3:1 1:3 3:1 1:3 3:1 1:3
BZ34 697 1 1 16 20
BZ140 544 0 1 17 14 4 3
X841 370 1 5 10 12 -
Total 1611 2 7 43 46 4 3
Per cent 0.6 5.5 1.3
leui- 1 leui- 12 trYl.. 48 tryl- 1 meti- 1
+- +- +- +- +- +- +- +- +-_ +-_
Diploid 3:1 1:3 3:1 1:3 3:1 1:3 3:1 1:3 3:1 1:3
BZ34 - 1 0 3 2
BZ140 5 6 0 1 3 5 - 4 5
X841 1 2 - - 2 0
Total 6 8 1 1 3 5 2 0 7 7
Per cent 1.5 0.2 1.5 0.5 1.1
1y8i- 1 ura3 urai ade2 a/a
+- +- +- +- +- +-_ +-_ +-_
Diploid 3:1 1:3 3:1 1:3 3:1 1:3 3:1 1:3 3a:a la:3a
BZ34 - 2 2 - - 1 1
BZ140 13 9 - - - 0 1
X841 7 6 - 5 5 0 3 0 2
Total 20 15 2 2 5 5 0 3 1 4
Per cent 3.8 0.6 2.7 0.8 0.3
site conversion, double-site conversion, and reciprocal recombination. Single-
site conversions represent those asci in which a 3+: la or 1+: 3a segregation
for one of the alleles, either proximal or distal, is accompanied by 2: 2 segregation
of the other allele. Thus, single-site conversional asci may contain the following
spore genotypes: (al+, ++, +a2, +a2); (al+, al+, ala2, +a2); (al+, al+, ++,
+a2); (a,+, ala2, +a2, +a2). Double-site conversional asci, without exception,
were restricted to two subclasses that involve 3+: la, segregation of one allele
accompanied by symmetrical 3a2: 1+ segregation of the other allele. The spore
genotypes in such asci are: a,+, +a2, +a2, +a2, or a,+, a,+, a,+, +a2. The
reciprocal recombination category includes those asci with the spore genotype
distribution a,+, ++, a1a2, +a2 in which the prototrophic and double mutant
spores are reciprocally recombined for the outside markers while the other spores
are of parental genotype.
It is evident that, for all allele pairs tested, reciprocal recombination between
the alleles contributes in only a minor way as a source of irregular tetrads. Of
163 irregular tetrads encountered in this study, only 14 arose by reciprocal re-
combination. For the allele pairs within arg4, an approximately linear relation-
ship is apparent between the frequency of reciprocal recombination and the length
of the intragenic interval as determined by mitotic X-ray mapping. However,
no reciprocal recombination events were observed for the allele pair at leua
though the interallelic distance is comparable to the separation between arg4-4
and arg4-17.
The major source of aberrant tetrads stems from gene conversion. In the case
Downloaded by guest on May 24, 2021100 GENETICS: FOGEL AND MORTIMER PROC. N. A. S.
p7 arg44 thr, CU,
fl
3.7 2.9/ \ 14.2 23.7
/ \
/\
/\
/\
4' 1 2 \17
| -520 -|128
_ 1060
FIG. 1.-Chromosome VIII of Saccharomyces cerevisiae showing the
fine structure of arg4. Intergenic distances are in centimorgans; intragenic
distances are estimated in nucleotides (see text).
of allele pairs that are widely separated, single-site conversions predominate.
For example, in the heteroallelic pair arg4-4/arg4-17, 46 of 49 conversional asci
involved only a single-site event. Similarly, 10 of 12 conversional events for
leul-1/leul-12 involved a single site. In contrast, conversional events involving
the relatively close alleles arg4-2/ar94-17 were mostly double-site conversions
(27/36). Approximately equal numbers of single- and double-site conversion
asci were observed for the allele pair arg4-1/arg4-2, characterized by an inter-
mediate separation. The double-site conversions have also been reported in
Schizosaccharomyces 8 and Neurospora.19
The single exceptional ascus in BZ34 had the constitution +17, +17, 4,17,
4,17. Conceivably, it arose from a single-site conversional event earlier in the
mitotic pedigree or immediately prior to meiotic DNA replication. The second
exceptional ascus was derived from BZ140. In this ascus, two spores were diso-
mic for chromosome VIII which carries the arg4 locus. The arg4 alleles were
heteroallelic in the disomic spores, and these otherwise haploid strains exhibited
mitotic reversion to prototrophy at rates comparable to the parental diploid.
Discussion.-The major finding in the present work concerns the significance of
symmetrical double-site conversions. Clearly, the magnitude of this component
in an unselected sample of asci is a function of the physical separation between
the mutant alleles. If two alleles are widely separated, they behave as essentially
independent entities in conversion. But when the alleles are in close proximity,
conversion of one allele is characteristically associated with conversion of the
second allele on the same parental strand. The increase in double-site conver-
sions occurs at the expense of single-site events. It is unlikely that double-site
conversions represent simultaneous though independent events. If this were
true, then asci exhibiting six additional genotypic arrays would be encountered.
Instead, all double-site conversion asci were included within the two categories
described earlier. Thus the present data, for both single- and double-site con-
versions, are most simply explained in terms of single events that involve replace-
ment of an informational segment in one homologue with information identical to
Downloaded by guest on May 24, 2021VOL. 62, 1969 GENETICS: FOGEL AND MORTIMER 101
TABLE 3. Analysis of conversional events at arg4 and leul.
Diploid: BZ34 X841 BZ140 BZ140
X901*
Locus: arg4 arg4 arg4 leul
Allele pair al/a2: 4/17 1/2 2/17 1/12
Diagnosis Nucleotide distance: 1060 520 128 1060
Single-site conversions
Proximal allele 3+:la a,+, +a2, +a2, ++ 3 3 1 4
Proximal allele 1+:3a a,+, a,+, +a2, aja2 5 3 3 6
Distal allele 3+:la a,+, a,+, +a2, ++ 18 10 3 0
Distal allele 1+:3a a,+, +a2, +a2, a, a2 20 11 2 0
Double-site conversions
Prox. 3+: la, distal
1+:3a a,+, +a2, +a2, +a2 2 13 14 1
Prox. 1 +: 3a, distal
3+:la ai+, a,+, a,+, +a2 1 10 13 1
Reciprocal recombinants + +, ata2, a,+, +a2 9 5 0 0
Exceptional tetrads 1 1 0 0
Total aberrant segregations 59 56 36 12
Total tetrads analyzed 697 502 544 544
* Pooled data from two closely related hybrids.
that contained in the corresponding section of a non-sister homologue. The modal
length of the segment is in the order of hundreds of nucleotides. Our earlier gene
conversion study on ochre and amber nonsense mutants leads to the further con-
clusion that this informational transfer operates with complete fidelity.7
The present work also bears on the paradoxical situation inherent in most
genetic fine structure mapping studies in which map distances are based on esti-
mates of gross prototroph frequencies. From the data in Table 3 and in an
earlier work,20 it is apparent that prototrophic spores, the frequency of which is
often utilized as a measure of the distance between mutant sites in a heteroallelic
diploid, arise mainly from single site 3+: la conversions and to a lesser extent
from reciprocal recombination between the parental alleles. For example,
among 1081 asci selected to contain a prototrophic spore only 101 arose as a
consequence of reciprocal recombination, while the remainder represented 3+: la
conversions.20 Except in the present study, it was not clear why reasonably
additive maps should result from a metric that primarily estimates nonreciprocal
events. The essential clue is provided by the category of event designated as
symmetrical double-site conversion. As the distance between mutant alleles is
decreased, this ascal class, which contains only auxotrophic spores, increases
linearly at the expense of the anticipated single-site events. Furthermore, the
frequency of reciprocal recombination between alleles appears to be proportional
to the physical separation of mutants as determined by X-ray mapping. Thus,
for three heteroallelic combinations at arg4 representing separations of 1100, 500,
and 150 nucleotides, the respective prototrophic spore frequencies are 30/2788.
18/2008, and 4/2176.
An equation expressing these considerations can be formulated as follows:
f = 1/ c + c2 c1.2 + r,2)
Downloaded by guest on May 24, 2021102 GENETICS: FOGEL AND MORTIMER PROC. N. A. S.
where f = frequency of prototrophic spores, cl = frequency of conversion at al,
C2 = frequency of conversion at a2, c1, 2 = frequency of symmetrical double con-
version at aja2, and rl,2 = frequency of reciprocal recombination between a, and
a2.
In deriving the above, it is assumed that conversion frequencies in either direc-
tion are equal. Estimates for each parameter are directly obtainable from the
experimental data.
From the viewpoint that gene conversion is a process of informational transfer,
it is appropriate to argue that approximately 2 per cent of a haploid genome is
replaced in each meiosis by information contained in the homologous non-sister
chromatids. The replacement involves randomly located segments of which
modal length is of the order of some hundreds of nucleotides. In this way,
conversion provides a mechanism for rapid meiotic rearrangement of genetic
variability. Unlike crossing-over which typically involves the physical transfer
of comparatively long DNA segments by breakage-reunion, conversion is not
restricted by positive interference and results in the oscillation of relatively short
informational segments. In effect, since conversion acts to rearrange genetic
variability, it must be viewed as a significant mechanism in the dynamic aspects
of evolution, particularly as it applies to the evolution of coadaptive gene clusters.
What is the molecular mechanism that leads to gene conversion? At present,
this cannot be fully specified. However, we are aware that our consideration of
gene conversion as a process of informational transfer, distinct from recombina-
tion via breakage-reunion of DNA segments, imposes restrictions on a choice of
models. Tentatively, we are inclined to the view that gene conversion entails a
DNA repair mechanism.21-23 We assume that repair is initiated by a cycle that
includes single-strand breaks in DNA helices of opposite polarity and different
parental origin, annealing of strands by complementary base pairing yielding a
heteroduplex with mismatched base pairs at the included heterozygous sites, and
excision of some hundreds of nucleotides from either helical member of the hybrid
DNA molecule followed by or simultaneous with resynthesis of an equivalent
segment along the remaining template. Closure of the single-strand breaks
would complete the cycle. While the present study does not provide any direct
evidence bearing on the molecular events in gene conversion, the genetical data
are consistent with the postulated mechanism.
* This
work was aided by research grant RG 06979 from U.S. Public Health Service, a
John Simon Guggenheim Fellowship (S. F.), and a grant from the Atomic Energy Commis-
sion (R. K. M.). We gratefully acknowledge the skilled technical assistance of Mrs. Ruth
Lerner.
t On leave from Brooklyn College, Department of Biology, Brooklyn, New York.
1 Winkler, H., Die Konversion der Gene (Jena: Verlag Gustav Fischer, 1930).
2 Lindegren, C. C., J. Genet., 51, 625 (1953).
3 Mitchell, M. B., these PROCEEDINGS, 41, 215 (1955).
4 Roman, H., D. C. Hawthorne, and H. C. Douglas, these PROCEEDINGS, 37, 79 (1951).
5 Roman, H., in Methodology in Basic Genetics, ed. W. J. Burdette (San Francisco: Holden
Day, 1963).
6 Takahashi, T., Bull. Brewing Sci., 12, 15 (1966).
7 Fogel, S., R. K. Mortimer, and D. C. Hawthorne, manuscript in preparation.
8 Roman, H., Ann. Genet., 1, 11 (1958).
9 Zimmerman, S., Mol. Gen. Genet., 101, 171 (1968).
Downloaded by guest on May 24, 2021VOL. 62, 1969 GENETICS: FOGEL AND MORTIMER 103
Mortimer, R., T. Brustad, and D. V. Cormack, Radiation Res., 26, 465 (1965).
1' von Borstel, R. C., ed., Carbondale Yeast Genetics Conference, Microb. Gen., Suppl. 19
(1963).
12 Perkins, D. D., Genetics, 34, 607 (1949).
13 Barratt, R. W., D. Newmeyer, D. D. Perkins, and L. Garnjobst, Advan. Genet., 6, 1
(1954).
14 Hawthorne, D. C., and R. K. Mortimer, Genetics, 45, 1085 (1960).
15 Hurst, D. D., and S. Fogel, Genetics, 50, 435 (1964).
16 Manney, T. R., and R. K. Mortimer, Science, 143, 581 (1964).
17 Parker, J. H., Ph.D. thesis, University of Rochester (1967).
18 Gutz, H., Proc. Intern. Congr. Genet. 12th, 1, 6 (1968).
19 Case, M. E., and N. H. Giles, Genetics, 49, 529 (1964).
20 Fogel, S., and D. D. Hurst, Genetics, 57, 455 (1967).
21 Haynes, R. H., Radiation Res., Suppl. 6, 1 (1966).
22 Whitehouse, H. L. K., and P. J. Hastings, Genet. Res., 6, 27 (1965).
23 Holliday, R., Genet. Res., 5, 282 (1964).
Downloaded by guest on May 24, 2021You can also read
