Linkage Mapping and Crossing-over

Independent assortment vs linkage.

Independent assortment occurs when the genes for two different traits or genetic markers are located on different chromosomes. Mendel was lucky enough to have chosen such a configuration. Assuming pea plants have approximately, !,!!! genes and seven chromosomes, then each chromosome might carry around ,"!! genes. #hese genes are not expected to assort independently during meiosis. In other words, there is a roughly "$ chance of the genes %eing linked.

Comparison of independent and linked genetic markers.

Independent

A A

' ' A a A a a a A a a a '

&
' %

a a a a

% %

)
% %

A a

' %

#estcross*

&

% % % % % ' %

+!$ parental phenotypes

+!$ recom%inant phenotypes

Linked genetic markers

A A C C A a C c C c c c c c C c

& &

a a a a

c c c c

A a

C c

#estcross*

Cis-configuration or coupling

A a a a A a a a

,reater than +!$ of progeny with parental phenotype

#rans, or repulsion

less than +!$ of progeny with recom%inant (henotype -resulting from crossing over..

s/ s/

%/ %/ s/ s s/ s

&
%/ % %/ %

s s

% %

)ruit flies s/0 long wings

s 0 short wings %/ 0gray %odies % 0%lack %odies

)
#estcross* In female flies

Long wing, gray %odies

&

s s

% %

Long wing, gray %odies

short wing, %lack %odies

s/ s s s s/ s s s

%/ % % % % % %/ %

" $ " $

Long wing, gray %odies

(arental com%inations

short wing, %lack %odies

1$ 1$

Long wing, %lack %odies 2ecom%inant com%inations short wing, gray %odies

s/ s
/

% % s/ s

&
% %/ s/ % %/

s s

%/ %
/

s/0 long wings

)
Long wing, gray %odies

Long wing, gray %odies

s 0 short wings %/ 0gray %odies % 0%lack %odies

s s/ s s s s/ s s s % % %/ % %/ % % %

&
" $ " $

s s

% %

short wing, %lack %odies

Long wing, %lack %odies

(arental com%inations

short wing, gray %odies

1$ 1$

Long wing, gray %odies 2ecom%inant com%inations short wing, %lack %odies

3hy a maximum of +!$ recom%ination per crossover4 )or a single crossover only 5 strands are involved, the remaining 5 strands are parental. I.e. +!$

'

A A a

' % ' %

If one looks at all the different possi%ilities of dou%le crossovers one arrives at a similar conclusion, again at most +!$ will %e recom%inant. -see next slide.. 6ven if crossovers did not occur %y chance %ut all the time, +!$ would still %e the limit. 7imilarly, if one went through all the possi%ilities of triple crossovers, one would again arrive at a theoretical maximum of +!$.etc. 8nly if nature had 7ome kind of %ias toward four-stranded dou%le crossovers, would the ratio Come out to more than +!$

All even num%er of exchanges %etween two segregating loci will yield parental com%inations and thus go undetected. In general the maximum fre9uency of recom%ination for two genes located 8n the same chromosome is +!$.

A A C A a a c a

C C c c

Two stranded double crossover

:ou%le crossovers occur in the following ways

#wo-stranded dou%le crossover
A B A '

All parental
a A % %

#hree stranded dou%le crossover

r ;p p ; rec r p p r p r ;p ; rec

)our-stranded dou%le crossover

All 2ecom %inant

a '

If one looks at all the different possibilities of double crossovers one arrives at a similar conclusion, again at most 50 will be recombinant. !see ne"t slide#. 6ven if crossovers did not occur %y chance %ut all the time, +!$ would still %e the limit. 7imilarly, if one went through all the possi%ilities of triple crossovers, one would again arrive at a theoretical maximum of +!$.etc. 8nly if nature had some kind of %ias toward four-stranded dou%le crossovers, would the ratio come out to more than +!$

C$%C&'(I$%) If over 50 of the gametes produced b* a +, cross contain parental combinations of genetic markers, this is an indication that genes are linked. 3hen a large num%er of genes is analy<ed in that way they are found to assort in linkage groups, which correspond to one set of homologous chromosomes. In :rosophila, the genes fall into four linkage groups, corresponding to the " pairs of chromosomes. &inkage maps #he rationale -logic. %ehind linkage maps* the fre9uency of crossing-over %etween to loci is a function of the length of the interval separating the two loci. #he longer the interval , the more likely it is that a crossing-over occurs. Assuming proportionality* The number of cross-overs . /"! distance#. 'ased on this logic, '= 7turtevant suggested that the fre9uency of recom%ination %e used as an index of distance %etween to loci. $ne map unit is defined as the map distance that produces , of recombinant chromosomes per testcross. #his logic holds true for shorter distances, however, as the distances get longer the correlation %etween recom%ination fre9uency and distance %ecomes less relia%le, as a result of an increase in even num%ered multiple cross-overs that are not counted as recom%inant events -%ecause the two cross-over events cancel each other. and also an increase in odd num%ered crossovers. 8ver longer distances the two are assumed to %e e9ual, then map units can %e defined as map distance -in cM. 0 >5 average num%er of all cross-overs per interval in a meiotic cell.

If one assumes proportionality %etween the distance %etween two loci and the average num%er of crossovers per chromatid then* %umber of crossovers . / !distance#, 3here ? is a proportionality constant. In that case one would also predict that the map distances would also %e additive.

&$ recom %ination

@$ recom%ination

-&/@.$ recom%inationA
3hen distances are large -B ! - 5! map units. the results are less then the additivity would predict. In that case dou%le -or even num%ered. crossovers occur almost as fre9uently as single or uneven num%ered crossover. 6ven num%ered crossovers are not phenotypically detected as recom%inant events, as the second event apparently cancels the first.

If p is the fre9uency for one crossover in interval I and 9 the fre9uency )or interval two, the fre9uency for a dou%le crossover is p9 C >" or - ; x. ;., 7ince p or 9 are C ; each. #he a%ility to identify the parental and the two reciprocal dou%le crossover classes also allows one to determine the order of the D genetic markers.

Three-factor crosses . If one uses three markers to map a chromosome it %ecomes possi%le to detect and 0uantif* double crossovers, and it %ecomes possible to order the markers relative to each other. In diploid organisms three factor crosses are used in analogous fashion to two-factor crosses. #hat is, homo<ygous individuals are crossed to produce triple hetero<ygotes or trihy%rids, and a testcross to recessive homo<ygotes is performed su%se9uently. If the three If traits were independent, the eight resulting gametes would occur in a * * * * * * * ratio. =owever, when the genes are linked, again the ratios of the progen* will depend on the degree of linkage. In a three factor testcross the parental combination will alwa*s occur with the highest fre0uenc*. 'oth parental progeny classes should occur with similar fre9uency. #he parental classes will also give information a%out the linkage relationship in the parental configuration, that is whether the genes were cis or trans-configuration on the homologous chromosomes of the ) generation. #he two reciprocal double crossover progeny classes can %e identified %ecause they will alwa*s e"hibit the lowest fre0uencies

In a three-factor cross for three linked genes* how are the gametes formed4
Eo crossovers, the most likely scenario 7ingle crossover, intermediate fre9uency

1 x 1 x

2 y 2 y

3 < 3 <

1 x 1 x

2 y 2 y

3 < < 3

7ingle crossover, intermediate fre9uency

dou%le crossover, lowest fre9uency

1 x 1 x

2 y y 2

3 < < 3

1 x 1 x y

2 y 3 <

3 <

:etermining the order of three linked markers. #he dou%le crossover results in an interchange of the center marker. #hus, #he recom%inant class with the lowest fre9uency indicates the identity of central marker. ,enotype A'C>a%c a%c>a%c A%c>a%c a'C>a%c A'c>a%c a%C>a%c A%C>a%c a'c>a%c #otal Eum%er of progeny DF! DG+ "+ +! 5 D F+ F! !!!

:etermining the order of three linked markers. #he dou%le crossover results in an interchange of the center marker. #hus, #he recom%inant class with the lowest fre9uency indicates the identity of central marker. ,enotype A'C>a%c a%c>a%c A%c>a%c a'C>a%c A'c>a%c a%C>a%c A%C>a%c a'c>a%c #otal Eum%er of progeny DF! DG+ "+ +! 5 D F+ F! !!! (arental configuration -all cis. 7ingle crossover in interval I :ou%le crossover* one in interval I and one in interval II 7ingle crossover in interval II

7ince the recom%ination class A'c, a%C occurs with the lowest fre9uency it must %e marker C>c that is located in the center, since marker C is recom%inant in that class. #hat is the order of the markers must %e AC'. #he linkage distances can now %e calculated as follows* :istance %etween A and C* add fre9uencies of single and dou%le crossover in interval AC* -A%c, a'C,. is -"+/+!/ 5/D.> !!! 0 !. or !$. Linkage distance in interval II, i.e. 'C* -F+/F!/5/D.> !!! 0 !. + or +$. #he dou%le crossovers are included in the calculation %ecause one of the 5 crossovers occurred in the interval. A,a C,c ',%

! map units

+ map units

Interference* In a three factor testcross the o%served fre9uency of a dou%le crossover was + out of a thousand or !.!!+. If the two crossovers in a dou%le crossover had %een completely independent, the expected fre9uency would have %een !. + & !. 0 !.! +. #he o%servation that actually fewer dou%le crossovers occur than expected was called chromosome interference or chiasma interference. #his is o%served in most dou%le crossovers. -#his is not to %e confused with chromatid interference.. #he degree of interference is measured %y the coefficient of coincidence* Coefficient of coincidence 0 o%served dou%le crossover fre9uency expected dou%le crossover fre9uency Coefficient of interference 0 - coefficient of coincidence. 6xample one gives a coefficient of coincidence of !.!!+>!.! +0.DDD Coeff. of interference 0 H !.DDD 0 !.IIF A positive coefficient of interference -%etween ! and . indicates that the first crossover interferes with a second. In %acteria the coefficient of coincidence can %e greater than one -negative interference. indicating that the occurrence of one crossover increases the

-r. recom%ination fre9uency #he genetic mapping function. In general interference increases as the distance %etween two loci decreases. In :rosophila no more dou%le cross-overs are o%served %elow or at ! map units. At small map distances the map distance e9uals the recom%ination fre9uency. #his is true up to + cM. 8range line* interference0 . Ie. At the origin up to a%out +cM all curves have an almost linear section. Interference decreases with increasing map distances. #he other extreme* i0!, no interference. #he corresponding curve is called =aldaneJs mapping function. r 0 !. +x - -e-d>+!. where d is the map distance. ?osam%iJs mapping function assumes that interference decreases with increasing, distance hence -i0 -5r..

4ore reasons wh* recombination fre0uencies are not linear and additive over all distances* 5ecombination fre0uencies differs between male and female individuals, the reason is not clear. (e"es can have different map distances for the same chromosomes with the same primary :EA se9uence. In :rosophila males there is practically no crossing over occuring. #he correlation %etween physical distance and genetic map distance can %reak down within a chromosome as a result of changing chromatin structure, from euchromatin to hetero chromatin. ,enetic distances appears much shorter in heterochromatin than in euchromatin. -2emem%er, the closer the genes are the less recom%ination.. Ksually there is more heterochromatin in the vicinity of the centromere. )or example in euchromatin the map distance may %e !" map units, the same stretch of :EA will give rise to a recom%ination fre9uency of D map units in the heterochromatin state. =owever, generally in euchromatin the correlation %etween physical distance and genetic map distance is relatively good.

6xample II :. melanogaster, curled> straight wings -cu>cu/., e%ony>gray %ody color -e>e/. and scarlet> red eyes -st>st/. ,enotype Eum%er of progeny

e cu e cu e st wt cu st e st cu st

" DII 5" D! !+ DG! G1 5 !!!

6xample II :. melanogaster, curled> straight wings -cu>cu/., e%ony>gray %ody color -e>e/. and scarlet> red eyes -st>st/. ,enotype Eum%er of progeny DII DG! 5" D! G1 !+ 5 " !!!

cu e st/> cu e st cu/ e/ st> cu e st cu e st> cu e st cu/ e/ st/> cu e st cu/ e st> cu e st cu e/ st/> cu e st cu e/ st> cu e st cu/ e st/> cu e st

6xample II :. melanogaster, curled> straight wings -cu>cu/., e%ony>gray %ody color -e>e/. and scarlet> red eyes -st>st/. ,enotype Eum%er of progeny DII DG! 5" D! G1 !+ 5 " !!! parental

cu e st/> cu e st cu/ e/ st> cu e st cu e st> cu e st cu/ e/ st/> cu e st cu/ e st> cu e st cu e/ st/> cu e st cu e/ st> cu e st cu/ e st/> cu e st

7ingle crossover interval I

7ingle crossover interval II dou%le crossover, per interval

#he advantage of working in Neurospora. Neurospora crassa spends a significant part of its life cycle the in haploid state. 2ight after meiosis N.crassa undergoes one more mitotic division, su%se9uently spores can germinate and reproduce asexually %y mitotic division of haploid cells to form mycelia. #hus the phenotype of the spores can %e assessed after meiosis, without any test crosses, since the recessive genes are not masked in the haploid state. )urthermore, in N. crassa the haploid products of meiosis -ascospores. are kept in linear order within the tu%e like structure called ascus. #his order reflects the order in which they were formed during meiosis. In Neurospora the products of meiosis can %e phenotypically analy<ed since they germinate into haploid mycelia. #hus in E. crassa it is possi%le to o%tain and analyse all four product of a meiotic event. 7uch data are called tetrad data. 8ther advantages include the simple growth conditions.

If crossing over occurs %efore chromosome replication*

A % A ' a % a ' A % A % a ' a '

!!$ recom%inant

If crossing over occurs after chromosome replication* 5+$(arental

A ' A ' A ' a % A % a ' a % a %

+! $recom%inant

5+$ parental

)irst divisional segregation pattern Alleles are segregated into different nuclei In first division
Meiosis I A Meiosis II A A a a tedrad a Mitosis

As well as opposite All A

All a

second division segregation pattern Alleles are segregated into different nuclei in 5nd div.
Meiosis I A a A a A a Meiosis II A A

5A ascospores 5a 5A 5a

a tedrad

#hese four second-division segregation ascus patterns are e9uivalent. they all reflect the same event, that is a single crossover relative to the centromere . A a A a a A a A A a a A a A A a

(arental ditype ascus pattern, no crossover

Meiosis I A ' Meiosis II A ' Mitosis

ascospores
A A A A ' ' ' '

All A'

A ' a %

a %

a %

a % a % a % a %

All a%

#etratype ascus pattern* single crossover

Meiosis II Meiosis I A ' A ' A % A ' A A a a % a a ' % a % a % % ' a ' % A '

5A' 5A% 5a' 5a%

If tedrads are ordered in the ascus and recovered> dissected without distur%ing the order, as in E crassa, - %ut not in 7. cerevisiae. it %ecomes possi%le to order the two markers relative to the centromere. #he recombination fre0uenc* from a gene to the centeromere is 6 the fre0uenc* of asci that e"hibit second7division segeragation patterns for the alleles of the gene %ecause only half of the chromatids are involved in the crossover. #he map distance from centromere to gene A 0 !.+x-num%er of asci with second division segregation pattern.& !! total num%er of asci some %ooks say !.+ %ecause only half of the single crossovers are productive.

Eon-parental ditype, four strand dou%le crossover

A A ' A % %

"A%

a %

' a '

"a'

#his ditype will %e rare in the case of linked genes %ut will %e as fre9uent as parental ditype for independently assorting genes. #he pattern for a two stranded dou%le crossover %etween the two loci looks again like the parental ditype. #he relative fre9uencies of parental ditype , tetratype and nonparental ditype asci can %e used to calculate the linkage distance %etween the two loci. In ordered tetrad data, the centromere can %e used as a marker. #his is done %y determining whether each ascus is the result of first or second division segregation.

#ype of ascus pattern 7pore pair - . A' 5 D " #otal num%er of asci* A' a% a% 5 -5. A' A% a' a% "G -D. A' a% A' a% DG -". A' a' A% a% 5 A% A% a' a' !

#ype* (: ## ## E(: :ivision segregation pattern* ):A ):A 7:A 7:A ):' 7:' 7:' ):' (:, parental ditype ##, tetratype, single cross over %etween two of the markers. 7: second division segregation pattern* cross-over %etween centromere and the two markers, ):* first division segregation pattern, no crossover %etween that marker and the centromere. E(:, non parental ditype, four-stranded dou%le cross-over.

#ype of ascus pattern 7pore pair - . A' 5 D " #otal num%er of asci* A' a% a% 5 -5. A' A% a' a% "G -D. A' a% A' a% DG -". A' a' A% a% 5

-I. Eot alll patterns are e9ual in their fre9euncy, so genes A and ' are linked. i.e. on the same chromosome. -II. Ascus pattern t*pe one is most fre9uent and represent the parental ditype, and first division segregation, no crossover occurred. -III. Ascus pattern t*pe four is least fre9uent* therefore a two stranded dou%le crossover gives rise to that pattern. -IL. Mudging %y there intermediate fre9uency, ascus pattern type 5 and D are most likely the result of a single crossover.

#ype of ascus pattern 7pore pair - . A' 5 D " #otal num%er of asci* A' a% a% 5 -5. A' A% a' a% "G -D. A' a% A' a% DG -". A' a' A% a% 5

Map distance 0 !!x!.+x -N of asci w> 5nd div. segr patterns.> total N of asci* =ence, distance A,' 0 !! x !.+-"G/5.>5!!0 5.+ map units :istance cetromere, A* 0 !!x!.+-DG/5.>5!!0 ! map units.

-I. #ype 5 displays a second division segregation patter with respect to allele ', and first division segreg. pattern with respect to A alleles, indicating that a crossover occurred %etween markers A and ', %ut not %etween the markers and the centromere. #hus the fre9uency represents interval A'. -LI. #ype D represents a second division segregation for %oth A and ' loci, indicating a single crossover %etween %oth , A and ' loci, and the centromere, %ut not %etween A and ', since that would have re9uired a dou%le crossover. #hus the centromere has to %e one the side of either A, or '. -LII. Mudging %y the fre9uency, #ype " must result from a two stranded dou%le Crossover. If the centromere were in the middle, the pattern would have to %e parental, having a second division segregation %etween the centromere and A, %ut not %etween the centromere and ', indicates that the dou%le crossover happened in intervals centromere and A, and A'. #hus the order must %e centromere, A,'

89

%89

TT

TT

(9(

TT

3hen taking all the different crossover patterns into account, including multiple crossovers then the formula %ecomes map distance . 0.5 !single crossovers#:total ; < " 0.5 !doubles#:total ; = " 0.5 !triples#:total ; > " 0.5!0uadruples#:total etc. #he factors !.+, , .+ is a matter of how many times does the crossover type affect the given interval in a tetrad. 6g. A single exchange affects half of the strands in that interval , a dou%le exchange affects

, is parental <

Tetrad anal*ses in unordered asci eg (. cerevisiae.

'nlinked genes)

#ypes of unordered asci produced with two genes on two different chromosomes. -A. If no cross-over occurs, only parental ditype and nonparental ditype result. (: and E(: are e9ually likely here. -'. 7ingle cross-overs %etween one of the genes and the centromere yields tetratype tetrads. If %oth gene are linked closely to their centromeres few tetratypes are produced. Eote, even though the tetrads are unordered, there is still some information regarding the relation ship %etween a gene and its centromere.

Tetrads produced with two linked genes) AB1ab

Linkage is indicated %y E(:CC(:

Calculating map distances with unordered tetrads #he relative fre9uencies of different types of tetrads can %e used to determine the map distance %etween two linked genes, as long as the genes are close enough, so that triple crossovers and higher levels of crossing-over can %e neglected. E(: give the fre9uency for dou%le crossovers. All four types of dou%le crossovers are expected in e9ual fre9uency. =ence* #otal N of dou%le crossovers0 " Oo%served E(:P #he N of ## tetrads from dou%le stranded crossover is 5x E(:, hence ## from single cross over 0 Oo%served##P-5Oo%served E(:P. In general the map distance %etween two is >5x fre9uency of single crossover ## tetrads -Oo%served##P-5Oo%served E(:P. plus the fre9uency of dou%le Hcrossover tetrads - "OE(:P. If all of that is com%ined, then Map distance 0 !!x - !.+Oo%served##P /DOo%servedE(:P.>total num%er of tetrads. 6g* if a two factor cross yields 5 (:, "E(:, 5"## tetrads, then 5/"/5".0 F.

the map distance is !!x-!.+O5"P / -DxO"P..> -

The two thirds rule is also true for unordered asci.

#he left hand side pertains to unlinked genes. #he fre9uency of ## allows inferences a%out the distance from a genes centromere