Copyright 0 1991 by the Genetics Society of America

The Genetic Structureof Natural Populations of Drosophila rnelanogaster. XXII. Comparative Studyof DNA Polymorphisms in Northern and Southern Natural Populations Toshiyuki S. Takano, Shinichi Kusakabe' and Terumi Mukai' Department of Biology, Kyushu Uniuersity 33, Fukuoka 812, Japan Manuscript received August 28, 1989 Accepted for publication June 25, 1991 ABSTRACT Restriction map variation in four generegions (Adh, Amy, Pu and Gpdh) was surveyed for 86 second chromosomes from northern (Aomori) and southern (Ogasawara)Japanese populations of Drosophila melanogaster (43 chromosomes from each population). The regions examined cover a total of 62 kilobases. Estimates of nucleotide diversity (P)were approximately constant across the gene regions and populations examined. The distribution of restriction site polymorphisms was compatible with the expectation from the neutral mutation-random genetic drift hypothesis, but insertion/deletion polymorphisms were not consistent with it. While the two populations shared a majority of restriction site polymorphisms, frequencies of individual restriction site variants were significantly different between the two populations at 7 out of 35 segregating sites. In addition, an insertion in the Amy region was found in 15 chromosomes from the Ogasawara sample but absent in the Aomori sample. A considerable difference was observed in the number of rare insertions and deletions between the two populations. The numbers of aberrations uniquely represented were 16 in the Ogasawara sample and only 3 in the Aomori sample. These findings suggest that the two populations were differentiated from each other to some degree by means of random genetic drift and/or otherfactors.

a

gration is presumably restricted. In a series of studies UANTITATIVE analysisof genetic variation on viability, MUKAIand his collaborators have found in natural populations provides information that there is a north-to-south cline of additive genetic essetial for an understanding ofitsmechanismof variance in viability(MUKAI1990). So far, the molecmaintenance and of evolutionary processes. Recent ular basis of mutations responsible for excessive admolecular advances have permitted the study of geditive genetic variancein the southern populations netic variation at the DNA level. By DNA sequencing still remains to be clarified. A difference in the genetic and restriction map analysis, DNA polymorphism in variance of viability between the two populations used natural populations of Drosophilamelanogaster has in the present study haspreviously been reported been reported in several gene regions (e.g., KREITMAN 1983; LANGLEY, MONTGOMERYand QUATTLEBAUM(TAKANO, KUSAKABE and MUKAI1987). Analyses were 1982; LEIGH BROWN 1983) but further study is made independently for four gene regions: alcohol needed. Analysis of DNA polymorphismis also useful dehydrogenase (Adh), a-amylase (Amy), Punch (Pu)and for studying population structure (SLATKIN1987; sn-glycerol-3-phosphate dehydrogenase (Gpdh). For each STROBECK 1987). In particular, it becomes an imporgene region, a restriction map of more than 10 kilotant indicator of interpopulational divergence when bases (kb) in length was constructed, resulting in a it is compared with geographical patterns of other total length of about 62 kb. genetic variation (SIMMONS et al. 1989). The purpose of the present study is to determine MATERIALS AND METHODS the level ofdifferentiation between populations at the Drosophila stocks: More than 400 second chromosomes molecular level. Restriction map variation inD. melawere extracted using the Cy-Pm method from isofemale lines nogaster was examined for 43 second chromosomes collected on Chichijima Island, Ogasawara (southern Japan) from a southern Japanese population [Chichijima Isin 1982 and intwowineries near Hirosaki city, Aomori land, Ogasawara (Bonin)] and the same number of prefecture (northern Japan) in 1983 (TAKANO, KUSAKABE chromosomes from anorthern population of the and MUKAI 1987). These isogenic lines were cytologically examined for thepresence of inversions. Since the estimated mainland ofJapan (Aomori). The island population is age of Zn(2L)t in one of the southern Japanese populations isolated from other populations and, therefore, mi-

' Present address: Department of Biology, Faculty of Integrated Arts and

Sciences, Hiroshima University, Hiroshima 730, Japan. Deceased on April 19, 1990.

'

Genetics 1 4 9 753-761 (November, 1991)

(Ishigakijima) was about 1000 generations (MUKAI,TACHIDA and ICHINOSE 1980), the degree of DNA polymorphism for inversion-carrying chromosomes might be lower than for inversion-free chromosomes in Japanese populations and

754

T. S. Takano, S. Kusakabe and T. Mukai

the frequencies of inversion-carrying chromosomes might have a large effectwhen the variation is compared between the two populations. Thus, the present restriction map analysis was carried out for 43 different second chromosomes with the standard karyotype, randomly chosen from each population, although the frequencies of chromosomes carrying Zn(2L)t and/or Zn(2R)NS were very low (0.05 for the Ogasawara population and 0.03 for the Aomori population). Four gene regions studied and probe preparation: Restriction map variation was surveyed in the following four gene regions: Adh (2-50. l),Amy (2-80),Pu (2-97) and Gpdh (2-20.5). The details of the probes are as follows. Adh: Several probes were developedby A. KOCA (unpublished results).Two X phage clonescontaining the Adh gene were isolated from a laboratory strain (C160) with a probe; a 2.7-kb ClaISalI Adh genomic clone (KREITMAN 1983). The followingthree fragments were subclonedinto pUCl3: Hind111 (-7.O)-BamHI (O.O), Hind111 (-0.7)-HindIII (+4.1) and Hind111 (-7.O)-HindIII(-3.0) fragments. An ll-kb region around the Adh gene was probed in the present survey (see Figure 1A). Amy: An approximately 14-kb region between positions -5.9 and +8.0 (Figure 1B) was probed with four subclones of EcoRIfragments from an a-amylase structural gene clone, XDm65 (GEMMILL, LEVYand DOANE1985), although not included is about 0.3 kb from -3.1 to -2.8. A flanking sequence of the above region was also cloned and an approximately8.1-kb SalI-Hind111 fragment from position +6.3 to +14.4 was used in the analysis. Pu: Construction of plasmid subclones wasmade by EcoRI and HindIII digestion of anapproximately 18-kb phage clone, X525, isolated by LEVYet al. (1982). Initially, southern analysis was performed by using three nonoverlapping plasmid probescovering the sequences betweenthe Hind111 sites at -5.7 and +7.9 (13.6-kb region). For complete construction of maps, further analysis was done using several other subclones. Gpdh: A rough picture of restriction map variationof the Gpdh region has already been reported by TAKANO et al. (1989). The probes described there cover a 23-kb region from position -12.8 to +10.5. For fine analysis, additional plasmid subclones were prepared from the X phage clones et al. (1989): pG8Xb4.8 (HindIII site isolated by TAKANO -15.6-XbaIsite-10.8),~G8Xb3.8(XbaIsites-10.8to-7.0), pG8Xb5.9 (XbaI sites -7.0to - 1.1) and pG9Xb6.9 (XbaI site +5.8-HindIII site +12.7). Restriction map analysis: Smallscale preparations of genomic DNA from adult flies werecarried out as described ~ Restriction maps were by KREITMANand A C U A D(1986). constructed using eight enzymes: EcoRI, HindIII, BamHI, Pstl, SalI, Sad, XbaI and Xhol. The procedures for digestion, electrophoresis, blotting and labeling of probes were essentially the sameasdescribed earlier (TAKANO et al. 1989). With a few exceptions, all variation within the regions probed by the above clones could be ascribed to either a polymorphism of a restriction site or to the presence of an insertion or deletion. In general, two or more adjacent restriction sites not distinguishableby the routine restriction mapping method were scored as a single site. Also, as a general rule, when and onlywhen consistent patterns of restriction fragments were observed from more than three independent enzymes digestions,was a variable site scored as an insertionor deletion.

RESULTS Figures1 and 2 summarizetherestrictionmap variation inthe Adh, Amy, Pu andCpdh regions among 86 second chromosomesof D.melanoguster. A total of 72 polymorphismsrepresenting 35 restriction sites and 37 insertions/deletions were scored. Analyses weremadeseparately for restrictionsitepolymorphisms and insertions/deletions, except for analyses of linkage disequilibrium. There is no difference in sample size between the Ogasawara and the Aomori population. Restriction site polymorphism: A total of 146 restriction sites were scored over the four gene regions studied, of which 35 sites were polymorphic. The average number of restriction sites per chromosome was 129. A summary of the number of polymorphic sites in the two populations is shown in Table 1, in which “common”indicatesthatthesame site was polymorphic in both populations and “private” means the sites were polymorphic only in one population sample. The majority of the restriction site polymorphisms werelocatedoutsidethestructural loci. In addition, the four polymorphicsites in theGpdh gene occurinintrons (KUSAKABEet al.1990). The Pu region has been reported to be very complex and to encode many transcripts (O’DONNELL, MCLEANand REYNOLDS1989). It is unclear whether each of the polymorphic sites is located in exons or not. The two populations shared a majorityof restriction site polymorphisms, that is, 23 out of the 35 sites were polymorphic in both populations.The observed number of “private” polymorphic sites was 5 in the Ogasawara population and 7 in the Aomori population. Thus, we have no evidenceof a significant difference in the number of polymorphic sites between the two populations. The polymorphic sites common to both thepopulationsseemto be distributedratheruniformly from high tolow frequency (Figure3). On the otherhand, the averagefrequency of theprivate polymorphic sites was 0.07, which implies that there is considerable migration between local populations and/or the two populations were recently established from a common ancestral population. However, frewere quencies of individual restriction site variants significantly different between thetwo populations at seven sites among 35 segregating sites: BamHI (-7.7) and Xhol (+1.8) in the Adh region, XbaI (-8.5) in the Amy region, Sal1 (-5.6) in the Pu region and Hind111 (-12.2), PstI (-5.3) and HindIII (0.0) in the Gpdh region. It should be mentioned here that all of the tests were not statistically independent since significant linkage disequilibria were observed between the BamHI (-7.7) and the Xhol (+1.8) in the Adh region (in both populations), and the PstI (-5.3) and Hind111 (0.0) in the Gpdh region (in the Ogasawara population). At any rate, this result suggests that the two

DNA Polymorphism

I

I

c - P u

I

I

I

0

"B

il

I

I

I

+5

+IO

i

kI

+I5

i

i

1

I

I

755

populations are differentiated from each other to some degree. One of the measures of DNA polymorphismsis the average number of nucleotide differences per site between two sequences, that is, nucleotide diversity [?r in NEI and TAJIMA (19Sl)l. This value was estimated separately for each population and then for whole chromosomes from the two populations. The estimates and their sampling standard errors are presented in Table 2. From this table, one can see the approximate constancy of the degree ofpolymorphism over the gene regions and populations examined. One exception is the high degree of polymorphism at the Gpdh region, but the estimate in the Cpdh region was not significantly different from that in the remaining three regions when stochastic error TAJIMA 1983; NEI 1987). was taken into account (CJ: HUDSON'S(1982) B is similar to nucleotide diversity, but it is independent of the frequency of variants at segregating sites. This provides a better estimate of 4 X (effective population size) X (mutation rate) than nucleotide diversity under the neutral mutation model. The estimates of B were 0.0050 for the Ogasawara population, 0.0053 for the Aomori population and 0.0051 for thepooled data. These estimates were very close to that of nucleotide diversity. The extent ofDNA polymorphismscanbealso studied by the followingtwomeasures. One is the number ( S ) of segregating sites in the sample and the other is the mean number (k) of differences between two randomly chosen DNA sequences. TAJIMA (1989a) has developed a new method for testing whether the relationship between these two measures is consistent with its prediction based on the neutral mutation model. The expectations of S and k are given as follows: E[S] = M U , E[6] = M where M = 4N,v and a l = [ l

1

I I

L

W I

E

L

I FIGURE1 ."Summaries of the restriction map variation observed among 86 second chromosomes representing the two natural populations (Ogasawara and Aomori). Monomorphic restriction sites in the whole sample are shown along the top line, while restriction sites which could not be detected in the present experiment are ignored (E = EcoRI, H = HindIII, B = BamHI, P = Pstl, SI = SalI, Sc = Sad, Xh = Xhol and Xb = XbaI). Excluding the above sites,

+ 1/2 + 1/3 + . . . +

monomorphic and polymorphic restriction sites for each population are presented above and below the maps, respectively (the second line for the Ogasawara and the third line for the Aomori population). Insertions and deletions are indicated by downward and upward triangles, respectively, and approximate the actual lengths. Insertions and deletions are only known to be located within the indicated regions. Regions probed by the plasmid subclones are shown by double-headed arrows below the maps(see the text). Transcriptional units are indicated by an open box. The direction of transcription is from left to right for the Adh and Gpdh genes. The left and right genes of the two Amy genes correspond to the proximal and distal genes designated by GEMMILL, SCHWARTZ and DOANE(1986), respectively. The Punch region has been reported to encode several transcripts by Northern analysis. The number of transcripts in three regions (double-headed arrows) are shown below the top bar (O'DONNELL, MCLEANand REYNOLDS 1989). (A) Adh, (B) Amy, ( C ) Pu,(D) Gpdh.

756

T. S. Takano, S. Kusakabe and T. Mukai

(A) Osasawara

-

amv

Adh

Line "

cc

(D

..

-- -_ -

h

m

.

I I mnOIu+crmC--

"-- *v

"V

y)

v

a, a

* *a

nccc cccccccc mX-0nT-x"""

1,

29 35

( 6 48

65 53 55 5, 59 1 3 8 1

88

1)s

" "V

" "y1

*"""_

C C O ~ O O a c e xnzw--annnwan--w-

+-+""""""+-

++""""""

"+-"""+""+-

+"""++""-

"+""-"+"-++-

++---+-----"-

++""""""

++"""-"""+-

"+"+"+"+""f"

191

++"""""" ++""""""

213

245 253 2 6 0 262 2,s

++"""""" -+-----+-----"+"""+""" -+""-+"""

284

-+""-+"""

2 9 0 29,

++"""""" ++""""""

299 301 30, 315

++""""""

"+"""-+""+++"+""""+""+++------+------"+-

-+-+-+-+""" ++""+"+"""

"+"""-+"-++"

-+""-+"""

3 4 2

35, 388

++"""""" ++""-+""" +------++-----

+-+"""-+""+++-+"""-+"-++""""+"""++-+"""-+"-++--+-------+---++-

310

"+""+-+"""

+-+""""+"""

++""""""

++---

c c c c cc o c oc c c c c

++ v

-Taanxwnv,-WWa-----axn-ar-----z~W=

-++-++~+"+-+"~"++"++"+"-+++

-++~~+-+"-++~""++~-+""""++

~+-+-+++-~+""+"++"++"+"-+

-++"+-+-~~-+""-++"++"""+++

++-+-

-++-~+++""+"~"++"+"---f ~++-++"+"+~+""-++~~++"+"-+++

+"+" +"++"+-"++--"-+"" ~~++"-

-+-+~+++"+"-+~-~++"++"f"-+

""t""

~++"+-+"-++"~-~++"+""-+ -++"+~+""+""-++-++"--"""+++ ~ + + - - + - + - - ~ ~ + ~ - ~ - ~ + + - - + - - ~ + - + + +

-+"-+-+-~"+"""n+-+"""-+++

~+~+~+++--+---~~---+--++"-++ -++"+-+"~~+~""++"++"--+

-+-+-+++-~+""""+-++"--"""-++

""+""

-++"+~+""+""-++-+"-"-++~++~~+~+"-++""-++~~+""""-+ ~+-~-+-+"-++""-++~~++"""+++

"-+"" +"++"-

~+~~~+-+""+""""+-+"""-+++

+"-+++-+-

325 339

-+""-+"""

*"

~+"-+-+"~~+~""++-~++"""+++

"~

3 2 1

++""""""

O F

z - ~ F ~ + + " .

"

m"_

yI

~+"-+-+-~"+"""n+~+"""-+++ -++~-+-+"-++-~~"++"++"----+

~"

+-+""""""+"+-"""+""" +-+"""-+""+"+"+""+""" +"+""""+"""+ +"+""""""++-

-0-9

* v1 *

"-" * * * * *""

I -.-I

"VVV

-++"+~+-~-++""~++"+"""--+

++"---+----

+-+"""++""++"+""""+"""

* " V

~++"+-+-~"+""-++"+"-+~~-+++ - + + - ~ + ~ + " - + + ~ ~ ~ " + + " + - ~ ~ - ~ " - + +

~"

+"+"""""-+-

I I WUULCI

v

"-+""

"_

+-+"""-+""+-

++"""""" ++""""""

I

-+-+-+++"+"""~-+"++""++

++-"

~~~

+-+"""-+"+"-

F

I I 0 1

v

-++-~+-+""+~~""++"+-+"-+"""+++

"-

--+-----"--"-+-

++-----------++""""""

I I I

*""""

"-+"" +"++"" "-+""

++-+"" ~~+"+-

+"""""+"""

-+-"+-+"""

v

++""++-~+~+"-+"+~~~+"""++

+"++""

"+"""-+""++-

++""""""

0"m

-NN-_o~~_-________NN--oo-----=~oN

- + + ~ ~ + - + " " + ~ ~ ~ " + + - ~ + + " " + " +

++"+-

"_

+-+""""+""++"+""""+""++-

" "

NO

. . . .

+-+"""-+""

+-+"""-+"""

"

N-.

- + ~ + " + + - ~ + ~ ~ ~ " - + - + ~ ~ + + " " " - + +

"_

120

"

~++-++-+-~+-+""~++"++"+-+++

""+"""""+-

--+""""""++"+"""-+"-++-

mu30

~++--+-+---++----~++--+-"----+

+-+"""-+"+""

"+"""-+""++-

GPdh

---+-"-

"+"""-+"-++-

+"""""+""" +-+""""""++"+"""-+"""

226 22, 2 4 0

"_ + + - + "-+"" "_ + + - + -

"-

. . . . . . . . . . . . .

m---n-l

-

o

F

"+"+"+"+""~~

-+""-+""" -+""+"+""" ++""""""

194 2 1 I

-o

-+"""+"""

99 109

rnw."

anzxmnwv,

++""+"+"""

+-+""""""+-

9 1

c

+-+-------+------

-+-+"-+"+"-

NO-

"++

~~-+"+"-+""

-+"-+-+""-++"""""" ++""""""

r m

" " " "

+""""++""+-

++------------

" "

I -+++a"

* ",- *

n e

++""+""""

0 "

"A

O,.~_~-ON

lnoIuah++ms.-+.-

1 - 1

" " V

0,

"

. . . . . .

m-~v----n---,.,.,.m-

-~---~-------n

0

0-

. . . .

.

"

rn

-.am

Pu

- - -" .. -" - -

"

-++"+-+-~~~+""-++"+"t-f++

"-+""

"-+"" -~~ +"+-

+"++"" "- ++"++-++"++-+"- +"++ "-+"" ~~~

+--+--+"-+""

-+-+-+++"+-~~~""+"++"""++ -++-++-+-~+~+""-++"++"+-~-+++ -++-++~+"+-+""-++"++~-+~~-+++ ~++~-+-+""+"""+~-+++"--f++ -+"-+-+-~~-+""-++"++"""+++ ~+~+~+++"+"""-++-~++--+---+

-+-+-+++"+"""-++++"--+""+"-+

-+,"+-+-~~++~~~-~++"+""""++

~++-~+~+-~"+""-++"++-"+++

-+-+-+-+--+-+-----++--+++-+++

-+"-+-+-~~-+-~~-~-n+-+"""-+++

( B ) Aomori Line

Adh

- -. . .. "- -*- *

-

"

cc

0

C D

-8c

.

0

.

.

.

I I -nolu+=-wc-" " "

"-"-

o m

y1

****

I m l u.

"

""*

Ln"""_

+-+---"-"+""---+-------+-+--+-

-+-----+-----++----+-------

--+-"--"+""++"f"""-+"--+-

T3(

-+"""+"""

+"""-"+"""

T42

-++""+"""

+"+"""-+"+"+-

"+-"+-+""+-+""-+""-+ -+""-+"""

+"-+"""""+-

4

9

"""""""_

44 53

-+-+"-+"""

+-+"""-+""++-+"""-+""++-+""""""+-

8 3

"+"+""+"""

+"+"""""-+-

93 9,

"+""-+"""

+-+""""""++-+"""-+""+-

27

108

++"-+-+"""

"+"C"+"f"""

-+"+n-+-+"" ++"-+""""

+-+""""+"""

-+-+"-+"""

+"""+""""+-

141 144 155

++"-+-+"""

+-+-------+---"-

+-+"""-+""+-

+-+""""""++-+"""-+"""

"+"+---+""--

+"--------+"--+-+-------+----+-

++""+"""-

+-+"""-+""+-

1 6 0 1 6 8 ,,E

-+---+-+----+-

189 190 195

"+"""+"""

+-+""""+""+-

-+"+""+"""

+-+"""-+""+-

2 0 2

-+"+n-+-+""

209 215

++""+"""-+"-+-+"""

2 1 6

-+-+"-+""" ++"-+-+""" -+""""""

225 2 3 2 245

"++""+"""

-+"""+"""

2 5 2

++""""""

256 239

-+---+-+-"-+-

2 7 6

+"+"""-+""++"+""""""++-+"""-+""++"+"""""-+-"+""

V +

+

mnrxaanwa -o

F

,a

_ _ _ ++"_ ~ ++"_ +"++"_ _ _ ++"" f f - + + _ ~ + "_+ +-+"

"-

++"-

~~~

""+"" ~~~

"_ _" "-+""

"-+""-~~ +"+-

_ ~ +"+_ +"+_ ~ ++"_ ~~~

++"-

"~

""+""

-+"-+-+""+-

"+""""+"""

-,

I I

""""Y

I

v

*-"

....

"

"V"

aa

v1 vI

" "*"

r-~co++--

"_

++-+++"-

++"-

-+"+-

++-"

"_ + + - + _ ~ ++"_ "-

++"-

++

" " V

* Ia I

" Y "

oc ecccc zeeu ECCCC -Tm~nxwPm-wWa-----Pxn-~T-----TawT F

-oop

-++"+~+""+""-++"++"""+++ ~++"+-+"-++""-++"++""-+"+ -++--+~+----+--~--++--++--------+

~+"-+-+-~"+"""+"+"""-+++ -++-~+~+""+-~~-"+"++"""-++ ~++"+-+-~"+"""+"++"""-++ "-~+++++-+-+~~~"++"++"""-++ ~++"+-+-~"+""-++-~++"""+++ ~++"+-+-~~-+"""+-~+"-+""-+

-++"+-+"+-+""~++"++"-+""+

-++~~+""-~+""-++"++"+"-+-+

_ _ _ ++"_ ~ + -_+ + -

"-+""

"t"+n"+"""

-

_NN--O___-m____-_-NN--~~-----%=c~N m---n-[, I 0 1 I I a m e . m " ~I.--I

-++~~+-+""+~""++"++"""+-+

~

+-+"""-+""" +-+.."+""""+-

0"-

"-+""

+-+--""-+"-"-

+-+"""-+""+-

" "

NO

. . . .

"-+"" ~~++"-

-~~++"-

""+"--

-+"""+"""

"

N-

~++"+-+""+""-++"++""""+ -++-~+~+"+-+-~~"++"++"+""+ -++"+"-~"+""-++"++"+"-+-+

"_

"f"""+"""+

, .

"_ + + - + "-+"" "_ + + " "_ +"+-

"-+"" "-+""

++"-+""""-

om0

".a_

. . . . . . . . . .

-++-~+-+"+-+""-+"-++"""-++

+""-+""""++--------------+-

3 0 5

GMh

~++"+-+"~-+"""+"+~"""+++ "++"+-+"+-+~~"-++"++"""-++

+""""-+""" +"+"-+""""+-

+""""-+""++-+"""-+""+-

NO?-

"-

++"-

"_ "_

-+---+-+-"-+++"-+""""

" "

++"-

+"+-

+-+"""-+""++-+""-"+"""

295

308 307 314 340 351

+-

+-

1 2 1 1 2 6

v

" " " "

~~~

"""""""~

117

-+-+---+-----++""+-+""" -+-+"-+"""

c

-+++a"

I

v

",",- *

-+--+n-+-+"--

e o

ea

0-000-ON

ID~I~o-++wF.-+.-

-+-+---++-----

3 2 7 8

F F F F C ~ C C

"-

"

0 "

. . . . . .

. .

mx-nnz-x-----T T1 TI T2

m

0

c c o o o o )o c c xnzu--mnnnwca--u-

ncoo

PU

- "-

-

"

m-"-hA__-"__-m_

""-N-"-__--

" " "

Amy

-

"

-++"+-+""+"""+"+"-+""-+ -~~"-+++-+~+~""++"-+"""-++

-++"+~+"-++"~"++"++""""+

-++"+-+""+""-++"++"""+++ +-+++-~+-+"""+"++"""-++ """++"+-+""-++"-+"""+++ ~"

-++"+~+"+-++""++"++""+-+-+ _++_~+""-~+""-++"++"+"-+-+

-++"+~+""+""-++"++"""+++

-++"+-+"-++""-++"++"+""-+

-++"+-+""+""-++"++"""+++ ~++"+-+-++-+~""++-+++"""+++ ""-++++-+-+""-++"++"""-++ -++"+-+"-++""-++"++"+""-+ -++"+-+"+-++""++"++""+-+-+ -++"+~+""+""-++"++"""+++ -++"+-+-++-+""-++-+++"""+++ ""+++++-+-+""-++"++"""-++ "-+-+++"+-+"""+"++"""-++ -++"+++"+-+""-++"+"-+""-+ -++"+-+"-++""-++"++"+""-+

-++"+-+""+""-++"++"""+++ _+"_+_+""+"""+"+"""-+++ _+"_+-+""+"""+"+"""-+++ -++"+-+""+""-++"++"""+++

u

757

DNA Polymorphism

TABLE

1

Numbers of variants among86 second chromosomes Restriction site

Insertion/Deletion Polymorphic Private Region

Monomorphic

Common

22 29 24 36

3 5 4 11

111

23

Adh Amy 2

PU

Cpdh Total

Private

Ogasawara

Aomori

Common

1 0

0

2

1

2 (2) 8 (5) 1(1) 8 (7)

6 (0) 1 (0) 1 (0) 4 (1)

19 (15)

12 (1)

2 0

0

2

5

3

65

7

Aomori

Ogasawara

"Common" indicates the same aberration was polymorphic in both populations. "Private" indicates aberrations that were polymorphic only in one population. The figures in parentheses are the numberof uniquely represented variants out of the private insertions and deletions.

4-l-

rate per DNA sequence investigated and n is the number ofchromosomesstudied. TAJIMA (1989a) hasproposed a test statistic as follows:

-

D =d

y1

82

"

/

m

I

v)

5

I

2

3

5

b

6

7

8

9

10 I I

1 2 13 l b 15 I 6 I ? II98

where

2 0 21

d =

Ih I

2

3

5

4

6

7

8

9

I O I I I2 I 3 I 4 1 5 1 6 1 7 l

8

m

Number of occurrences in sample of 43 chromosomes

FIGURE3.-Histograms showing the observed frequency spectra for restriction site polymorphisms. Abscissa: the number of chromosomes with the minor variant among 43 second chromosomes. Ordinate: the number of sites. The dotted and open columns represent common polymorphism and private polymorphism, respectively (see the text). Since four chromosomes out of 43 were not scored for Hind111 site (-2.6) in the Adh region (see Table 2), one was added to the class of size 15.

-

l/(n l)] (WATTERSON1975; TAJIMA 1983). Here Ne is the efective population sire, u is the total mutation

- (S/al)

and f ( d ) is anestimateof the varianceof d . The expectation of D for neutral mutations at equilibrium is nearly zero irrespective of recombination. TAJIMA (1 989a) also suggests that the actual variance of d is smaller than the estimatedvariancewhenrecombination is taken into consideration, and that therefore this testis conservative. Thus, we tested the neutrality of mutationsby using TAJIMA'S D. In Table 3, the values of D are presented with the estimates ofM obtained fromthe above two methods. No significant difference was found between them, although the estimate of D for the Ogasawara population (1.98) wasveryclose to the critical value at a = 0.05 (2.04).

TABLE 2 Estimates of nucleotide diversity( T ) with theirsampling standard errors

Region

Adh Amy

Pu Cpdh Mean

Whole

0.0051 f 0.0004 0.0044 f 0.0004 0.0047 f 0.0005 0.008 1 f 0.0006 0.0056 0.0048 f 0.0002

0.0038 0.0047 0.0033 0.0072

f 0.0005 f 0.0005 f 0.0003 f 0.0007

0.0049 +. 0.0002 0.0046 k 0.0003 0.0041 f 0.0003 0.0078 f 0.0005

0.0053 f 0.0003

f 0.0002

Estimates and their sampling standard errors of nucleotide diversity ( T ) were obtained following N E I and TAJIMA (198 1) separately for each population and then for the pooled population. FIGURE2.--Genetic variation in the Adh, Amy, Pu and Cpdh regions. Polymorphic restriction sites, insertions (Ins) and deletions (Del) are placed left to right along the restriction maps shown in Figure 1. T h e abbreviations for restriction sites are thesame as in Figure 1. + and stand for presence and absence, respectively. Missing observations for restriction sites due to insertion or deletion are shown by n. Since three lines, 48 in the Ogasawara population and 256 and 305in the Aomori population, were lost in the course of experiments, 65 in the Ogasawara population and 239 and 308 in the Aomori population were substituted in subsequent experiments. (A) Ogasawara, (B) Aomori.

T. S . Takano, S. Kusakabe and

758

T. Mukai

TABLE 3

Estimates of M ( W p )obtained fromthe mean number of differences (k)and the number of polymorphisms (S)and D Restriction site

Insertions/Deletions

Estimate of M Region

Adh Amy

Pu Cpdh Sum

Estimate of M

k

S

D

k

S

D

Ogasawara Aomori Ogasawara 1.16 Aomori Ogasawara 1.39 Aomori Ogasawara Aomori

1.47 1.09 1.65 1.79 1.45 1.02 4.41 3.97

0.92 0.69

0.36 1.24 0.93 0.14 0.05 0.09 1.32 0.87

0.92 1.85 2.08 0.46 0.23 0.23 2.54 1.62

Ogasawara Aomori

8.97 7.86

6.47 6.93

1.39 (NS) 1.22 (NS) 1.06 (NS) 0.29 (NS) 0.13 (NS) 0.23 (NS) 1.44 (NS) 0.23 (NS) 1.98 (NS) 0.66 (NS)

2.65 2.34

5.78 4.16

-1.44 (NS) -0.92 (NS) -1.60 (NS) -1.30 (NS) -1.12 (NS) -0.85 (NS) -1.44 (NS) -1.27 (NS) -2.62 (P< 0.001) -1.95 (P< 0.05)

Population

1.62 0.92 3.00 3.70

NS, not significant.

Insertionsanddeletions: The present study revealed a total of 37 sequence length variants over the four gene regions, 26 insertions and 11 deletions in comparisonwith the most common restriction map (Figures 1 and 2 and Table 1). These variations are scattered over the whole region and rangein sizefrom approximately 100 base pairs to greater than 10 kb. and LANGLEY Previous reports [see CHARLFSWORTH (1 989) for a review] have suggested that the majority of large insertionsare due to insertions of transposable elements. However, three casesin the Gpdh region are of a different nature: two are duplications (insertions m and n ) and one is a triplication (insertion 0 ) (KOGA et al. 1988; TAKANO et al. 1989). Insertion/deletion variation is commonly measured by the number of insertions and deletions per kilobase and LANGLEY per chromosome (e.g., CHARLESWORTH 1989). The present estimates were 0.040 in the Adh, 0.023 in the Amy, 0.002 in the Pu and 0.012 in the Gpdh regions (duplications m, n and o were excluded from the analysis), suggesting an amount of large variation among the regions studied. To see if there issignificantvariation among the four regions, we carried out an analysis of variancetreating each chromosomeas a independent sample. The effectsof region and population-region interaction were highly 6 21.4 and 24.4, respectively). significant ( F S S 9 ~= These results suggest that the number of insertions and deletions per chromosome depends not only on chromosome regionbut also on population. As shown in Figure 4, the majority of the insertions and deletions were found only at low frequencies. This finding is consistent with previous reports [AQUADRO et al. (1986) for Adh; LANGLEYet al. (1988a) for Amy: AQUADRO, LADOand NOON(1988) for rosy; MIYASHITA and LANGLEY (1988) for white]. Indeed, theresults in Table 3 clearly show that the distribution of insertions and deletions is significantly different from the

Number of occurrences in sample 01 43 chromosomes

FIGURE4.-Histograms showing the observed frequency spectra for insertions and deletions. Abscissa and ordinate are the same as in Figure 3.

expectation under theneutral mutation model at equilibrium ( D = -3.1 3, P < 0.001). One exception is insertion i in the Amy region, which is present in 15 chromosomes from the Ogasawara population but absent in the Aomori sample. The mean number of restriction site and insertion/deletion differences in the Amy region was 0.67 f 0.13 for thechromosomes carrying insertion i and 2.59 f 0.24 for the others. This suggests that the origin of insertion i is due toa relatively recent single event and the high frequency of insertion i might represent an example of the founder effect and/or random genetic drift or hitchhiking event, either because of a favorable effect of the insertion itself or of a closely linkedvariant (CHARLFSWORTH and LANGLEY1989). The number of insertions and deletions common to both populations was onlysix. The remaining 31 insertions/deletions were found only in one population sample.The number of “private” polymorphisms detected was 19 in the Ogasawara sample and 12 in the Aomori sample. However, when we compare the

DNA Polymorphism

number of variants that were found only once in a sample (“unique”), this number becomes 15 in the Ogasawara sampleand only one in the Aomori sample (Figure 4). In the Aomorisample, the number of “unique” variants was smaller than that of variants which are present twice. This appears to be attributed to the small number of variants and linkage disequilibria between variants. Indeed, insertions c and j, and g and o in the Gpdh region (all of which werepresent twice) are in complete linkage disequilibrium. At any rate, the number of insertions and deletions tended to be larger in the Ogasawara population than in the Aomori population. Linkage disequilibrium: Nonrandom associations between the polymorphismswere examined using Fisher’s exact test.The tests weremade separately for each population sample andtherefore79 pairwise analyses were made for each sample. The following 13 pairwise comparisons showed significant linkage disequilibria in the same direction in the two samples: BamHI(-7.7) vs. deletion b and BamHI(-7.7) vs. XhoI (+1.8) in the Adh region, Hind111 (-6.1) us. EcoRI (+1.0) andEcoRI (+1.0) us. EcoRI (+8.0) in the Amy region and SalI (-1 2.0) vs. SacI (-1 1.7), SalI ( - 1 2 . 0 ) ~ ~EcoRI . (-9.4), SalI ( - 1 2 . 0 ) ~ ~EmRI . (-8.3), SalI (-12.0) vs. PstI (-2.2), SalI (-12.0) us. insertion m, SacI(-1 1.7) vs. EcoRI (-9.4), EcoRI (-9.4) vs. EcoRI (-8.3), EcoRI (-9.4) vs. SacI (+7.8) and EcoRI (-5.5) vs. SacI (+7.8) in the Gpdh region. These results do not indicate a prevalence of linkage disequilibria in the Gpdh region relative to the otherregions, since most of the tests (66 out of 79 tests) were done for polymorphisms in the Gpdh region. T o compare the level of linkage disequilibrium between the two populations, the polymorphisms which werepresent more than once in each sample were examined. Significant linkagedisequilibriawere found in 59 out of 144 pairwise tests in the Ogasawara sample and 47 out of 257 in the Aomorisample. There is a significant heterogeneity in the number of significant and nonsignificant tests between the twosamples, although many of the tests were not statistically independent. This test also assumes equal power to detect significant linkage disequilibrium in both samples. If the frequencies are different the powers may be different. Thus linkagedisequilibrium was reexamined using only the variants which were present fifth or more in each sample. The average frequencies of the minor variants used for the analysis were 0.30 and 0.26 in the Ogasawara and Aomori populations, respectively. Again, a significant heterogeneity between the two samples was found [55.3% (52/94) and 33.9% (21/ 62) of the pairwise comparisons were significant in the Ogasawara and Aomori populations, respectively]. This mayimply ashort historyof the 0gasawar-a population or recent bottleneck effects.

759

DISCUSSION In the present study, the restriction map variation in four gene regions among 86 second chromosomes from two natural populations weresurveyed. The estimates of nucleotide diversity and Hudson’s heterozygosity per nucleotide were roughly constant across the gene regions and populations. They are also comparable with other estimates for the autosomal genes (only the estimates of 6’ are shown) (the Adh region MONTGOMERYand QUATTLE[0.006 in LANGLEY, BAUM (1982)], the 87A heat shock gene region [0.002 in LEIGH BROWN(1983)], the Amy region [0.006 in et al. (1988a)], the rosy region [0.003 in LANGLEY AQUADRO, LADO and NOON (1988)l) and for some Xlinked genes (the white region [0.013 in MIYASHITA andLANGLEY(1988)], the Notch region [0.005 in SCHAEFFER, AQUADRO and LANGLEY (1988)], the zestetho region [0.004 in AGUAD~, MIYASHITAand LANGLEY (1989a)l and the Zw region [0.002-0.003 in MIYASHITA (1990)l). However, the overall estimate in the yellow-achaete-scute region appears to be rather smaller than those for the otherregions, including the , and present results [0.001 in A G U A D ~MIYASHITA LANGLEY (1989b); 0.002 in BEECH and LEIGHBROWN (1989); 0.003 in EANFS, LABATE and AJIOKA(1989); 0.001 in MACPHERSON,WEIR and LEIGHBROWN (1990)], although there might be some differences in heterozygosity betweenthe locations sampledand the enzymesused (EANES,LABATEand AJIOKA 1989; MACPHERSON, WEIR and LEIGHBROWN 1990). We have tested the neutrality of mutations in terms of TAJIMA’S D. The observed Ds for restriction site variation were not significantly different from zero, but they were positive in all ofthe regions studied. In particular, the observed D for the Ogasawara population is nearly statistically significant. The geographical structure of the population might exert an effect on the frequency spectrum of mutations and alter the D values from the expectation based on the neutral mutation model in one panmictic population. TAJIMA (1989b) has studied the effect of population subdivision on the number of segregating sites. From Table 1 in his paper, we can see that d based on the sample from one subpopulation has a positive value when the migration rate is positive but less than one, although the magnitude of the deviation depends on the migration rate. Further molecular studies of population structure are needed to understand possible effects of population structure on the frequency distribution of mutations. In the present study, the number of insertions and deletions per kilobase per chromosome depended not only on regions but also on populations. These estimates have large stochastic variances, and the differencesbetween them maybe explained by chance alone. Otherwise, insertions might bedistributed non-

760

T. S. Takano, S. Kusakabe and T. Mukai

uniformly along the DNA sequence due to a heterogeneity in mutation rate or other factors (e.g., BEECH and LEIGH BROWN 1989). AQUADRO et al. (1986) have reported clustering of large insertions in a particular region 3’ to the Adh gene. The present study showed the similar tendency, that is,most large insertions tended to be found in relativelyshort regions flanking to the coding regions (see Figure 1). These results might support the above possibility. In contrast to restriction site variation, the distribution of insertions and deletions was significantly different from the expectation under the neutral mutation model at equilibrium. Indeed, previous restriction map and in situ hybridization studies [see CHARLESWORTH and LANGLEY(1989) for a review] have reported that individual frequencies of large insertions and deletions are very low. These results have been explained by natural selection againstinsertional mutations or against chromosome rearrangements generated by asymmetric pairing of transposable elements and unequal exchange (LANGLEYet al. 1988b). The absence of a difference in the number of insertions between the autosomal and theX-linked regions has been cited as evidence to reject the generality of CHARLESthe former hypothesis (e.g., MONTGOMERY, WORTH and LANGLEY1987; CHARLESWORTH and LANGLEY1989). But the possibility of natural selection against large insertions cannot finally be ruled out. If large insertions are additive in their effects on fitness, the observed results can also be explained by the simple selection model. The average degree of dominance ofnewly arising, mutant viabilitypolygenes has been estimated to be 0.4 (MUKAI 1990), which is muchgreater than that of lethal mutations in natural populations [0.01-0.02 in MUKAIand YAMAGUCHI (1974)l. The homozygous effects of insertions and deletions such as those observed in the present experiments on fitness, if any, must be much smaller than the average of those of viability polygenes[0.03 or less, in MUKAIet al. (1972)J Thus, the degree of dominance of these mutations might be even closerto 0.5. Moreover, as pointed out by CHARLESWORTH and LANGLEY(1989), it is difficult to fine-tune the selection parameters such as transposition rate and selection intensity to fit the observed results. Further study will be needed to discriminate between the various possibilities. Previous studies on viability polygenes have shown a very high mutation rate of viability polygenes[O. 14 per second chromosome per generation. MUKAIand COCKERHAM (1977) suggested that most mutant viability polygenes must belocated outside the structural loci. Some of the insertions and deletions might be candidates for viability polygenes. We would like to thank C. H. LANGLEY and J. O’DONNELL for providing the phage clones and helpful information, and H.TACH-

IDA and F. TAJIMA for their suggestions and comments. We also thank D. A. JOHNSON and two anonymous reviewers for improving the manuscript. This is paper no. 55 from the Laboratory of Population Genetics, Department of Biology, Kyushu University, Japan. This investigation was supported by research grant from the Ministry of Education, Science and Culture of Japan.

LITERATURECITED A G U A D ~M., , N. MIYASHITA and C. H. LANGLEY, 1989a Restriction-map variation atthe Zeste-tko region in natural populations of Drosophila melanogaster. Mol. Biol. Evol.6 123130. A G U A DM., ~ , N. MIYASHITAand C. H. LANGLEY, 1989bReduced variation in the yellow-achaete-scute region in natural populations of Drosophila melanogaster. Genetics 122 607-615. AQUADRO, C. F., K. M. LADOand W.A. NOON,1988 The rosy region of Drosophila melanogaster and Drosophila simulans. I. Contrasting levels of naturally occurring DNA restriction map variation and divergence. Genetics 119: 875-888. C. F., S. F. DESE, M. M. BLAND, C. H. LANGLEY and C. AQUADRO, C. LAURIE-AHLBERG, 1986 Molecular population genetics of the alcohol dehydrogenase gene region of Drosophila melanogaster. Genetics 114 1165- 1190. BEECH,R. N., and A. J. LEIGHBROWN,1989 Insertion-deletion variation at the yellow-achaete-scute region in two natural populations of Drosophila melanogaster. Genet. Res. 53: 7-15. CHARLESWORTH, B., and C. H. LANGLEY,1989 The population genetics of Drosophila transposable elements. Annu. Rev. Genet. 23, 251-287. EANES,W. F., J. LABATEandJ.W.AJIOKA, 1989 Restriction-map variation with the yellow-achaete-scute region in five populations of Drosophila melanogaster. Mol. Biol. Evol. 6 492-502. GEMMILL, R. M., J. N. LEVYand W.W. DOANE,1985 Molecular cloning o f a-amylase genes from Drosophilamelanogaster. I. Clone isolation by use of a mouse probe. Genetics 110: 299312. GEMMILL,R. M., P. E. SCHWARTZ and W. W. DOANE,1986 Structural organization of the Amy locus in seven strains of Drosophila melanogaster. Nucleic Acids Res. 14 5337-5352. HUDSON, R.R., 1982 Estimating genetic variabilitywith restriction endonucleases. Genetics 100: 71 1-719. KOGA,A., S. KUSAKABE,F. TAJIMA,K. HARADA, G . C. BEWLEY and T . MUKAI,1988 Wide-spread polymorphism ofa tandem duplication in the region of the glycerol-3-phosphate dehydrogenase gene in Drosophila melanogaster. Proc. Jpn. Acad. 64: 9-12. KREITMAN, M., 1983 Nucleotide polymorphism at the alcohol dehydrogenase locus of Drosophila melanogaster. Nature 304: 412-417. KREITMAN, M., and M. A G U A D1986 ~, Genetic uniformity in two populations of Drosophilamelanogaster as revealed by filter hybridization of four-nucleotide-recognizing restriction enzyme digests. Proc. Natl. Acad. Sci. USA 83: 3562-3566. KUSAKABE, S., H. BABA,A. KOGA,G . C. BEWLEY and T. MUKAI, 1990 Gene duplication and concerted evolution of the GPDH locus in natural populations of Drosophila melanogaster. Proc. R. SOC.Lond. B 242: 157-162. LANGLEY, C. H., E. MONTGOMERYand W. F. QUATTLEBAUM, 1982 Restriction map variation in the Adh region of Drosophila. Proc. Natl. Acad. Sci. USA 7 9 5631-5635. LANGLEY, C. H., A. E. SHRIMPTON, T. YAMAZAKI, N. MIYASHITA, Y. MATSUO and C. F. AQUADRO, 1988a Naturally occurring variation in the restriction map of the Amy region of Drosophila melanogaster. Genetics 119 619-629. LANGLEY, C. H., E. MONTGOMERY,R. HUDSON, N. KAPLANand B. CHARLESWORTH, 1988b On the role of unequal exchange in

DNA Polymorphism the containment of transposable element copy number. Genet. Res. 52: 223-235. LEIGH-BROWN, A. J., 1983 Variation at the 87A heat shock locus in Drosophilamelanogaster. Proc. Natl. Acad. Sci. USA 80: 5350-5354. LEVY,L. S., R. GANGULY, N. GANGULYandJ. E. MANNING, 1982 The selection, expression and organization of a set of head-specific genes in Drosophila. Dev. Biol. 94: 451-464. MACPHERSON, J. N., B. S. WEIR and A. J. LEIGH BROWN, 1990 Extensive linkage disequilibrium in the achaete-scute complex of Drosophila melanogaster. Genetics 1 2 6 12 1 129. MIYASHITA, N. T., 1990 Molecular and phenotypic variation of the Zw locus region in Drosophila melanogaster. Genetics 1 2 5 407-419. MIYASHITA, N., and C. H. LANGLEY, 1988 Molecular and phenotypic variation of the white locus region in Drosophila melanogaster. Genetics 120: 199-212. MONTGOMERY,E., B. CHARLESWORTH and C. H. LANGLEY, 1987 A test for the role of natural selection in the stabilization of transposable element copy number in a population of Drosophila melanogaster. Genet. Res. 49: 3 1-4 1. MUKAI,T., 1990 Viability polygenes in populations of Drosophila melanogaster, pp. 199-217 in Population Biology of Genes and and J. F. CROW.Baifukan, Molecules, edited by N. TAKAHATA Tokyo. MUKAI,T., and C.C. COCKERHAM, 1977 Spontaneous mutation rates at enzyme lociin Drosophilamelanogaster. Proc. Natl. Acad. Sci. USA 7 4 25 14-25 17. MUKAI,T., and 0.YAMAGUCHI, 1974 The genetic structure of natural populations of Drosophilamelanogaster. XI. Genetic variability in a local population. Genetics 7 6 339-366. MUKAI,T., H. TACHIDA and M. ICHINOSE,1980 Selection for viability at loci controlling protein polymorphisms in Drosophila melanogaster is very weak at most. Proc. Natl. Acad. Sci. USA 77: 4857-4860. MUKAI,T., S. I. CHIGUSA,L. E. METTLER and J. F. CROW, 1972 Mutation rate and dominance of genes affecting viability in Drosophila melanogaster. Genetics 72: 335-355. NEI, M., 1987 MolecularEvolutionaryGenetics. Columbia University Press, New York.

76 1

NEI, M., and F. TAJIMA, 1981 DNA polymorphism detectable by restriction endonucleases. Genetics 97: 145-1 63. O'DONNELL,J. M., J. R. MCLEANand E. R. REYNOLDS, 1989 Molecular and developmental genetics of the Punch locus, a pterin biosynthesis gene in Drosophilamelanogaster. Dev. Genet. 1 0 273-286. SCHAEFFER,S. W., C. F. AQUADROand C. H. LANGLEY, 1988 Restriction-map variation in the Notch region of Drosophila melanogaster. Mol. Biol. Evol. 5 30-40. SIMMONS, G.M.,M.E. KREITMAN,W. F. QUATTLEBAUM and N. MIYASHITA, 1989 Molecular analysis of the alleles of alcohol dehydrogenase along a cline in Drosophilamelanogaster. I. Maine, North Carolina, and Florida. Evolution 43: 393-409. SLATKIN, M., 1987 The average number of sites separating DNA sequences drawn from a subdivided population. Theor. Popul. Biol. 32: 42-49. STROBECK, C., 1987 Average number of nucleotide differences in a sample from a single subpopulation: a test for population subdivision. Genetics 117: 149-153. TAJIMA,F., 1983 Evolutionary relationship of DNA sequences in finite populations. Genetics 105: 437-460. TAJIMA, F., 1989a Statistical method for testing the neutral mutation hypothesis by DNA polymorphism. Genetics 123: 585595. TAJIMA,F., 1989b DNA polymorphism in a subdivided population: the expected number of segregating sites in the twosubpopulation model. Genetics 123: 229-240. TAKANO, T., S. KUSAKABEand T. MUKAI, 1987 The genetic structure of natural populations of Drosophilamelanogaster. X X . Comparison of genotype-environment interaction in viability between a northern and a southern population. Genetics 117: 245-254. TAKANO,T., S. KUSAKABE, A. KOGA and T. MUKAI, 1989 Polymorphism for the number of tandemly multiplicated glycerol-3-phosphate dehydrogenase genes in Drosophilamelanogaster. Proc. Natl. Acad. Sci. USA 86: 5000-5004. WATTERSON, G. A,, 1975 On the number of segregating sites in genetical models without recombination. Theor. Popul. Biol. 7: 256-276. Communicating editor: C. C. LAURIE