Phylogenetic analysis of Leber's hereditary optic neuropathy mitochondrial DNA's indicates multiple independent occurrences of the common mutations

HUMAN MUTATION 6311325 (1995)

RESEARCH ARTlCLE

Phylogenetic Analysis of Leber’s Hereditary Optic Neuropathy Mitochondrial DNA’s Indicates Multiple Independent Occurrences of the Common Mutations Michael D. Brown, Antonio Torroni, Calvin L. Reckord, and Douglas C. Wallace’ Department of Genetics ml Mole& Medicine, Emory University Schod of Medicine, Atlanta, Georgia 30322; Fa: 404-727-3949

Communicclred by R.G.H. Conon

The mitochondria1 DNAs (mtDNA) from 17 Caucasian 11778-positive and 30 Caucasian 11778- negative Leber’s hereditary optic neuropathy (LHON) patients were PCR-amplified and subjected to high resolution restriction endonuclease analysis. Concurrently, all patient mtDNAs were screened for the common primary LHON mtDNA mutations at nucleotide pairs (nps) 3460, 11778, and 14484, the ambiguous intermediate-risk LHON mtDNA mutations at nps 5244 and 15257, and the secondary LHON mtDNA mutations at nps 3394,4216,4917,7444,13708, and 15812. Phylogenetic analysis was performed using mtDNA haplotype data from the 47 LHON patients and 175 non-LHON Caucasian controls. The superimposition of the LHON mutation screening results upon the Caucasian mtDNA phylogeny revealed (1) 35 different LHON haplotypes, (2) that all three common primary mutations have occurred multiple times in Caucasians, (3) that while recurrent mutation is common for the primary mutations, secondary mutations tend to be lineage-specific, (4) that the np 15257 mutation was confined to a single mtDNA lineage but may be etiologically important in some LHON cases since it was found in a LHON pedigree which lacked a common primary mutation; complete sequence analysis of the proband mtDNA revealed only a single other candidate missense mutation (at np 10663 of the ND4L gene) of uncertain pathological significance; and (5) that the np 14484 mutation may be less pathogenic than either the np 3460 or np 11778 mutations, as this mutation most commonly occurred on a single mtDNA lineage and almost always in association with secondary LHON mutations. A phylogenetic approach to this genetically heterogeneous disease has thus provided key genetic data bearing on the relative pathogenicity of the LHON-associated mtDNA mutations. 0 1995 Wdev-Liu, Inc.

KEY WORDS: Leber’s hereditary optic neuropathy, Mitochondrial DNA mutations, Mitochondria1 DNA haplotypes, Phylogenetic analysis, Primary LHON mutations

INTRODUCTION Leber’s hereditary optic neuropathy (LHON) is

a type of maternally inherited blindness clinically recognizable by the rapid, painless, bilateral loss of central vision due to optic nerve atrophy (New- man, 1993). LHON is a genetically heterogeneous disease caused by point mutations in mtDNA genes which encode polypeptide subunits of the mitochondrial electron transfer enzymes (respira- tory complexes I-IV) (Wallace, 1992). To-date, 16 different mtDNA missense mutations in complexes I, 111, and IV have been proposed to be associated with the disease (Brown and Wallace, 1994a,b). Despite progress made in the identifica-

tion of the underlying genetic defects in LHON, it is clear that mtDNA mutations are necessary but, in some cases, not sufficient for disease expression. Penetrance is variable between and within families which share a homoplasmic LHON mutation (in blood) and males are affected much more frequently than are females (Newman et al., 1991). The factors responsible for this variation could be either genetic (nuclear or mitochondrial) or epige- netic.

Recoiwd November 3.1994, accepted November 29,1994 “To whom reprint requWcomspondence should be ad-

dressed.

0 1995 WILEY-US. WC.

312 BROWN ET AL.

LHON base changes are commonly classified as primary or secondary mutations (Table 1). The common primary LHON mutations, which occur at nucleotide pairs (nps) 3460 (ND1 gene; complex I) , 11778 (ND4 gene; complex I) , and 14484 (ND6 gene; complex I), impart a relatively high predisposition for disease expression and together account for approximately 85% of Caucasian LHON patients. Generally, these mutations alter amino acids in evolutionarily constrained regions of mitochondria1 polypeptides, can be either homoplasmic or heteroplasmic, and have not been found in control populations. Moreover, primary mutations are expected to be associated with LHON regardless of mtDNA haplotype. This has been rigorously demonstrated only for the 11778 ND4 mutation (Singh et al., 1989), which is responsible for 50-60% of all Caucasian LHON cases (Newman, 1993), approximately 80% of Jap- anese LHON cases (Mashima et al., 1993) and has most recently been found in an African mtDNA haplotype in association with LHON (Newman et al., 1994). To-date, the three primary mutations have not been found to co-occur in the same mtDNA.

Some LHON mutations appear to be intermediate in risk between primary and secondary mutations. Consequently, their status is currently ambiguous. The np 5244 mutation altered a very highly conserved glycine in the ND2 polypeptide, was not found in greater than 2100 controls, and was heteroplasmic (Table 1). However, it has only been observed in a single patient and in conjunction with the np 15257 mutation. The np 15257 mutation altered a conserved aspartate in the cytochrome b gene product and is commonly found in LHON patients (Table 1). However, it is generally associated with a single mtDNA lineage, with only one additional case proposed to be an independent occurrence in an LHON patient (Howell et al., 1993a). Further, this cytochrome b variant has been found in low frequencies in the control population and can coocur with a common primary LHON mutation (Brown and Wallace, 1994a).

Secondary LHON mutations include base changes at nps 3394, 4216, 4917, 7444, 13708, and 15812 (Table 1). These mutations either impart a relatively lower risk for LHON manifesta- tion or are nonpathogenic. Secondary mutations are often found at increased frequencies in LHON patient cohorts suggesting that some of them may influence LHON expression when associated with a primary or other secondary mutations, possibly

by increasing penetrance when coupled with a primary mutation (Huoponen et al., 1993). For several secondary mutations, however, a direct causal link to LHON has not been demonstrated. To further clarify the pathogenicity of the various LHON mutations, we have employed a phylogenetic approach to determine the number of independent mutational occurrences and the mutational associations of the primary and secondary LHON mutations in Caucasians. We have found that, like the np 11778 mutation, the np 3460 mutation has occurred multiple independent times in Caucasian patients thus confirming pathogenicity. The np 14484 mutation has also occurred more than once in Caucasians, but is more frequently associated with a particular mtDNA haplotype than either the np 11778 or 3460 mutations. The np 15257, by contrast, has only arisen once in our collection of patients. However, complete mtDNA sequence analysis of the proband from a maternally inherited np 15257-positive LHON pedigree which lacked the common primary mutations, revealed only one other candidate missense mutation in the ND4L gene at np 10663. This ND4L mutation was also present in an LHON patient which simultaneously harbored the np 15257 and heteroplasmic np 5244 mutations, but no common primary mutation. Therefore, it is likely that either the np 15257 or the np 10663 mutation is pathogenic in this family, or that both mutations interact to cause the d' isease.

MATERIALS AND METHODS LHON Patients and Controls

Forty-seven LHON patients were analyzed, 17 of which were positive for the np 11 778 mutation and 30 of which were negative for the np 11778 mutation. All patients experienced rapid, painless, bilateral loss of central vision after childhood. The 17 11778-positive patients were randomly chosen for study from a larger group of 11778-positive LHON patients and the potential heteroplasmy of these patients was unknown. Ten of the 11778- negative patients lacked all three of the identified primary LHON mtDNA mutations. Of these, six patients (L12, 19, 28, 29, 38, and 46) experienced rapid-onset, bilateral optic atrophy with central or cecocentral scotoma with a short interval (usually less than six months, not more than one year) between each eye becoming affected. Two of these cases (L12 and L19) had positive family histories. Of the remaining four patients without primary mutations, two (L16 and L22) were diagnosed as LHON and had a positive family history; one

TABL

E 1. M

tDN

A M

utat

ions

Ass

ocia

ted

With

Leb

er's

Her

edita

ry O

ptic

Neu

ropa

thy

App

roxi

mat

e E

stim

ated

0.

..

Oth

er

/Q

Nuc

leot

ide

neur

olog

ical

A

min

o ac

ids

Eur

oDea

n %

B

ackm

ound

Het

ero

%

- I

Mut

atio

n/ge

nea

chan

ge

Cla

ss

Ass

ocia

tionb

di

seas

e'

Cha

nge

Con

sd

patie

nts

Con

trol

s ha

dohm

es'

Dla

smv'

re

cove

rv

1.14

459m

D69

2.

117

78/N

D4

3.34

60N

Dl

4.14

484/

ND

6 5.

9804

/C03

6.

137

30/N

D5

7. 1

5257

lCY

TB

8.

5244

/ND

2 9.

4160

AD

l 10

.744

4/C

Ol

11.3

394W

Dl

12. 1

3708

WD

5 13

. 943

8tC

03

14.1

5812

iCY

TB

15

.491

7AD

2 16

.421

6ND

l

~ ~

Rar

e 0

M

+ U

N

G-A

Pr

imar

y -

++

A72

V

M

G-A

G

-A

T-C

G

-A

G-A

G

-A

G-A

T

-C

G-A

T

-C

G-A

G

-A

G-A

A

-G

T-C

Pri

maj

-

Prim

ary

-

Prim

ary

7.9-

12,1

4,16

Pr

imar

y -

Prim

ary

-

Inte

rmed

iate

2,

4,8,

12,1

4,16

In

term

edia

te

7,12

,14,

16

Seco

ndar

y 4

Seco

ndar

y 3,

4 Se

cond

ary

4 Se

cond

ary

2-4,

7,8.

11,1

4,16

Se

cond

ary

-

Seco

ndar

y 4,

7,8,

12,1

6 Se

cond

ary

2,3,

16

Seco

ndar

y 2,

3,4,

5,10

, 12

,13,

14,1

5

R34

0H

A52

T M

64V

A

200T

G

465E

D

171N

G

259S

L

285P

T

erm

-tK

Y

30H

A

458T

G

78S

V35

7M

D15

0N

Y30

4H

H

M

M

H

M

H

H

H

NA

H

M

H

M

H

L

50

15

15 1.5

Rar

e 9 R

are

Rar

e 5 R

are

30 2.5

4 3 -4

0

0 M

+/-

0

M*

+ 1-

0 F"

+

/-

0 U

N

-

1*

+ 0 0.

4 F"

1*

+

0 0 1*

F#

1

0.9

1"

6 1"

0.1

1

3 1#

13

F#

-

-

-

-

-

4-6

UN

-

-

-

-

4 22

37

UN

Y

ES 28

U

N 0

UN

-3

0 NA

N

A

NA

N

A

NA

"All

nucl

eotid

e su

bstit

utio

ns ar

e re

porte

d as

L-s

trand

chan

ges.

Mut

atio

ns li

sted

in e

stim

ated

ord

er of

sev

erity

. bG

enot

ypic

asso

ciat

ions

with

oth

er L

HO

N m

utat

ions

. Num

bers

cor

resp

ond

to L

HO

N m

utat

ions

in th

is ta

ble.

dH, h

igh

cons

erva

tion;

M, m

oder

ate

cons

erva

tion;

L, lo

w c

onse

rvat

ion;

NA

, not

ava

ilabl

e.

'M,

mul

tiple

; F, f

ew; U

N, u

nkno

wn;

*on

e re

porte

d ca

se o

r fam

ily; #

est

imat

e fro

m li

tera

ture

. '+

, onl

y fo

und

as h

eter

opla

smic

mut

atio

n: +

I-,

occa

sion

ally

foun

d as

het

erop

lasm

ic m

utat

ion.

gR

efer

ence

s: (1

) Jun

et a

l., 1

994

Bro

wn

et a

l.. 1

994a

; (2)

Wal

lace

et a

l., 1

988 N

ewrn

an e

t al.,

199

1; (3

) Huo

pone

n et al.,

1991

; How

ell et

al.,

1991

a; (4

) Joh

ns

et a

l., 1

992b

; Mac

key

and

How

ell,

1992

Bro

wn

et al

., 19

92b;

(5) J

ohns

and

Neu

feld

, 199

3; (6

) How

ell et a

l., 1

993b

; (7)

Bro

wn et a

l., 1

992a

; How

ell et a

l., 1

993a

; Jo

hns

and

Neu

feld

, 199

1; (8

) Bro

wn et a

l., 1

992a

; (9)

How

ell e

t al.,

199

1b; (

10) B

row

n et al

., 19

92c;

(11)

Bro

wn et al.,

1992

b, Jo

hns et a

l., 1

992b

; (12

) Bro

wn

et a

l., 1

992a

.b; J

ohns

and

Ber

man

, 199

1; Jo

hns

and

Neu

feld

, 199

1; (1

3) Jo

hns

and

Neu

feld

. 199

3 B

row

n et al.,

1994

b; (1

4) B

row

n et

al.,

199

2a; J

ohns

and

N

eufe

ld, 1

991;

(15)

John

s an

d B

erm

an, 1

991;

(16)

John

s an

d B

erm

an, 1

991.

+ + , a

ssoc

iate

d w

ith a

dditi

onal

neu

rolo

gica

l sym

ptom

s; +

I-,

occa

sion

ally

asso

ciat

ed w

ith a

dditi

onal

neu

rolo

gica

l sym

ptom

s.

314 BROWN ET AL.

(L43) had rapid onset bilateral optic atrophy, a short interval between each eye becoming affected, and telangiactasias upon funduscopic examination; and one (L34) suffered from rapid onset bilateral optic atrophy and was diagnosed with LHON by two independent physicians. Some of the 11 778-negative patients described in this study have been previously reported: L6, L l l , L12, and L14 (Brown et al., 1992a), L l and L2 (Brown et al., 1992b), and L40 and L41 (Brown et al., 1992~). The mtDNA from patient L12 was subjected to complete mtDNA sequence analysis. Pa- tient L12 suffered rapid onset, painless, bilateral optic atrophy at the age of 15 years. For each eye, visual loss developed over roughly five weeks and the interval between the two eyes becoming affected was one month. Cecocentral scotomas were noted in both eyes and funduscopic examination revealed optic disk pallor in both eyes, preceded by telangiactasia/tortuous vessels. A positive family history included three maternal uncles and the maternal grandmother being affected.

The 175 non-LHON Caucasian controls used in this report have been previously described (Tor- roni et al., 1994) and represent a random sampling of Caucasians collected in the United States and Canada.

DNA Isolation and LHON Mutation Analysis

Total DNAs were extracted from lymphoblasts by using standard methods. For each sample, eight mtDNA segments were amplified using the poly- merase chain reaction (PCR). These mtDNA PCR fragments were screened for the presence of the Leber’s-associated missense mutations at nucleotide positions (nps) 11778, 3460, 14484, 15257, 13708, 7444, 4216, and 4917 using appropriate restriction endonucleases. LHON mutations at nps 3394, 5244, and 15812 alter HaeIII, Mspl, and RsaI restriction sites, respectively, and thus were tested for during mtDNA haplotype analysis (see below).

For the 11778 mutation, the forward primer extended from np 11141 to 11158, and the reverse primer from np 11851 to 11868 (3’ to 5’). This mutation creates a novel Mae111 restriction site. For the 3460 mutation, the forward primer extended from np 3108 to 3127, and the reverse primer from np 3701 to 3717 (3’ to 5’). This mutation eliminates a BsaHI site. For the 15257 mutation, the forward primer extended from np 15161 to 15180, and the reverse primer from np 15256 to 15234 (3’ to 5’). This primer has a mismatched thymine at np 15254 which creates a diagnostic

MseI restriction site following PCR. This mutation also eliminates a naturally existing AccI restriction site. For the 13708 mutation, the forward primer extended from np 13001 to 13020 and the reverse primer from np 13932 to 13950 (3’ to 5’). This mutation eliminates a BstNI site. For the 7444 mutation, the forward primer extended from np 7240 to 7260 and the reverse primer from np 7610 to 7628 (3’ to 5‘), This mutation eliminates an XbaI site. For the 4216 mutation, the forward primer extended from np 3598 to 3615 and the reverse primer from 4351 to 4370 (3’ to 5‘). This mutation creates an NlaIII restriction site. For the 4917 mutation, the forward primer extended from np 483 1 to 4847 and the reverse primer from np 5174 to 5193 (3’ to 5‘). This mutation creates an Mae111 restriction site. For the np 14484 mutation, the forward primer extended from np 14260 to 14279 and the reverse primer from np 14485 to 14510 (3‘ to 5’). The reverse primer had two mismatched cytosines at nucleotide positions 14487 and 14488. These mismatches create an MvaI site in the presence of the C to T base change at np 14484. Because of sample limitations, we were unable to evaluate patient L6 for the np 14484 mutation.

The np 10663 mutation found in patient L12 eliminates an SpeI restriction endonuclease recog nition site at np 10658. Thus, to screen for the 10663 mutation, the forward primer extended from np 10597 to 10616 and the reverse primer from np 10911 to 10930 (3’ to 5’). The resulting 333 base pair PCR fragment contains the diagnostic SpeI site as well as a second SpeI site which served as an internal control for digestion efficacy.

Following restriction digestion, fragments were separated using 2.5% NuSieve plus 0.9% SeaKem agarose (FMC BioProducts) gels and visualized by ultraviolet-induced fluorescence. Diagnostic restriction site losses were confirmed by DNA sequence analysis.

Haplotype and Phylogenetic Analysis

The entire mtDNA of each patient sample was amplified in nine overlapping fragments by PCR using the primer pairs and amplification conditions previously described (Torroni et al., 1992, 1994). Each PCR segment was digested with 14 restriction endonucleases (AluI, AuaII, BamHI, DdeI, HaeII, HaeIII, HhaI, HincII, Hinfl, HpaI, MspI, MboI, RsaI, TaqI). In addition, all subjects (patients and controls) were screened for the presence of the BstNI site at np 13704 and for the loss of the AccI site at np 15254. This restriction analysis screens

PHYLOGENETIC ANALYSIS OF LHON PATIENTS 315

about 20% of the mtDNA nucleotides and defines the mtDNA haplotype.

The phylogenetic relationships between the haplotypes observed in LHON patients and those previously observed in Caucasians from the United States and Canada (Torroni et al., 1994) were inferred by parsimony analysis. The dendrograms were rooted using a Senegalese mtDN A haplotype (“African outgroup,” Torroni et al., 1993). Max- imum parsimony (MP) trees were generated through random addition of sequences using the Tree Bisection and Reconnection (TBR) algorithm (PAUP 3.1.1s) (Swofford, 1992).

mtDNA Sequence Analysis

MtDNA sequence analysis was performed using an Applied Biosystems 373A automated DNA sequencer and f lourescent dye-labeled ddNTP chain-terminators. For the complete genome sequencing of patient L12, roughly three different sequencing primers were used to ascertain any given region of patient mtDNA. Ambiguities in the sequence were resolved by additional sequencing or more commonly, restriction endonuclease digestion.

RESULTS Haplotype and Phylogenetic Analysis of LHON Patients

Twenty-nine different mtDNA haplotypes, de- fined by 58 restriction fragment length polymorphisms (RFLPs; see Appendix), were observed among the 47 (17 11778-positive and 30 11778- negative) LHON patients. The 17 11778-positive patients were randomly chosen from a larger cohort of 11778-positive patients and thus these patients cannot be used for certain frequency deter- minations. A phylogenetic tree was generated by parsimony analysis using the high resolution RFLP data from all LHON patients and 175 non-LHON Caucasian controls (Fig. 1). LHON patients (designated with an L in Fig. 1) were generally dispersed throughout the Caucasian phylogeny, which represents all major and several minor Caucasian mtDNA lineages (Torroni et al., 1994).

All LHON patients were also screened for the common primary LHON mtDNA mutations at nps 3460, 11 778, and 14484, the ambiguous mutations at nps 5244 and 15257, and the secondary LHON mutations at nps 3394, 4216, 4917, 7444, 13708, and 15812 (Table 2). When the mutation analysis results were combined with the RFLP screening data, an additional six LHON haplotypes were ob-

served, bringing the total number of LHON haplotypes to 35.

The Primary LHON Mutations at np 3460, 11778, and 14484 Have Arisen Multiple Times

The phylogenetic relationships among Cauca- sian LHON patients and controls are shown in Figure 1. The primary mutations at nps 3460, 11778, and 14484 were each found to have occurred more that once in Caucasians. The np 11778 mutation appears to have occurred 12 independent times in this sample of Caucasian patients studied (Fig. 1). Fifty-nine percent (10/17) of the 11 778-positive patients lacked other LHON-associated mutations and the 11778 mutation was not found in the same mtDNA with either the np 3460 or the 14484 mutation. O f the seven 11778-positive patient mtDNAs which harbored additional LHON mutations, all contained the np 4216 mutation in conjunction with either the 4917 mutation (n = 3) or the 13708 mutation (n = 4). One of the np 11778-positive patients (L7) also harbored the np 15257 mutation while the other (L9) had the np 3394 mutation.

The np 3460 mutation appears to have occurred nine independent times in this Caucasian sample (Fig. 1). This mutation was found in 33% (10/30) of 11778-negative LHON patients, was heteroplasmic in one (L45) patient, and was not found with either the np 11 778 or 14484 mutations. Seventy percent (7/10) of the 3460-positive mtDNAs did not contain any other LHON-associated mutations (Table 2). When the np 3460 mutation did occur with other LHON mutations, it was found in association with both the 4216 + 4917 genotype (L30) and with the 4216 + 13708 genotype (Ll). In one case (L40), the 3460 mutation was linked to the 7444 COI mutation.

The np 14484 mutation occurred at least four times in this Caucasian sample with all of our patients being homoplasmic. Twenty-seven percent (8/30) of the 11778-negative LHON patients harbored the np 14484 mutation, and this mutation was not found in mtDNAs with the np 3460 or 11778 mutations. However, unlike the 3460 and 11778 mutations, the np 14484 mutation was frequently associated with the np 15257 mutation and certain secondary LHON mutations. Seventy- five percent (6/8) of the 14484-positive patients harbored the np 4216 and 13708 mutations. Three of these six (L2, L4, and L5) also had the 3394 mutation, while the other three (L8, L10, and L l l ) had the 15257 mutation. Two of the np 14484 + 15257 patients (L10 and L l l ) also had

1 11778 mutation

3460 mutation

A 14484 mutation

* Heteroplasmic

~

94

34 74

1 1 1 21 56

I . . 4

L181* L25. 76 L19 116 112

L20 0 -f 8 20 24

~ 27

~ 38

40 ~ 44

17 19 7 35

37 L21 I __I 55

48 39 I 47 I L271* 1 105

L38

+ Lii L23 I

93 85

110 96 87

I

I

91

16519

4216.13708 103 L1 0

I

79 339 L3 I L5 *

L2 * - 81 106

107

I 57 421 6.491 7

90 92

82

95

32 L26 I

14

101

I 16 12 i 49 10

LO

1 70 L44 I L45

L46

L40 0 (+ 7444) 51 73 29

58

66 72 L47 I EEF 61 L42 L41 A (+ 7444) I

L43

African outgroup

FIGURE 1.

69


the 15812 mutation. One of the 14484-positive patients (L41) had the 7444 mutation and one 14484-positive patient (L39) did not harbor any other LHON variants. Thus, a total of 88% (718) 14484-positive patients harbored additional LHON-associated mtC)NA mutations.

The association of the np 14484 mutation with the np 4216 and 13708 mutations raised the possibility that the 14484 mutation is a lineage-specific variant that predated the appearance of LHON. To test this possibility, we screened for this mutation in Caucasian controls with the same mtDNA haplotype as the np 14484-positive patients (Fig. l). Five controls representing Cauca- sian haplotype 36 (control haplotype nomencla- ture of Torroni et al., 1994) and 14 controls representing the Caucasian mtDN A branch de- fined by the np 4216 + 13708 variants [haplotypes 78 (n = l), 79 (n = 6), 97 (n = l) , 103 (TI = l), 81 (n = l) , 89 (n = 2) , 106 (n = l) , and 107 (n

FIGURE 1. Phylogenetic tree of Caucasian mtDNAs from 47 LHON patients (Ll-L47, bold) and 175 Caucasian controls (117 haplotypes designated 1-85 and 87-118) obtained from the United States and Canada (Torroni et al., 1994). Haplo- types were determined by screening the entire mtDNA of each individual with 14 restriction enzymes (see Methods and Materials) and also for the 13708 and 15257 mutations with the restriction enzymes BstNl and AccI, respectively. The numbers at the nodes or end of each branch either indicate mtDNA haplotypes observed in individual LHON patients (L designation) or haplotypes observed in Caucasian controls (plain numbers). For the patients, all mtDNA haplotypes are shown and the presence of a common primary mutation or heteroplasmy is indicated (see key), although we were unable to test patient L6 for the presence of the np 14484 mutation. Occassionally, the superimposition of the LHON mutation screening data on the phylogeny would al- low the further subdivision of clustered LHON haplotypes. For example, patients Ll8, L19, and L20 had identical haplotypes until it was noted that L18 harbored the 11778 mutation, L20 harbored the 3460 mutation, and L19 harbored neither of these mutations. Additional mutations of interest are noted in parentheses. Multiple occurrences of the three common primary LHON mutations are indicated by their presence in independent rntDNA lineages and, when occur- ring within the same large mtDNA lineage, by the presence of intermediate haplotypes lacking LHON mutations. For the controls, 19 of the 117 Caucasian haplotypes were observed in more than one individual (Torroni et al., 1994). The number of individuals for each of these haplotypes is indicated in parentheses and is as follows: 4 (23), 13 (4), 28 (2). 36 (5), 41 (2). 42 (2), 43 (2), 49 (21, 57 (2), 59 (2). 68 (21, 79 (7), 82 (6). 88 (3), 89 (2), 94 (2), 96 (3). and 98 (2). The numbers in large, bold characters on branches indicate mutations which define specific groups of mtDNAs but are not necessarily the mutational steps used to create the phylogeny. The horizontal branch lengths are proportional to the number of mutational events that separate the haplotypes. This tree is 221 steps in length, has consistency and retention indices of 0.515 and 0.858, and is one of the 3,000 MP trees generated with the TBR branch-swapping algorithm.

= l)] were tested and found to be negative for this mutation. Thus, the np 14484 mutation appears to be specific to LHON patients.

The Secondary LHON Mutations Are Typically Lineage-Specific and Associated With Various Primary Mutations

The 4216 polymorphism has arisen at least twice in the Caucasians (Fig. l), once in association with the np 4917 mutation and once in association with the np 13708 mutation. In each np 4216-positive lineage, LHON is well-represented and all three common primary mutations have occurred in conjunction with the np 4216 mutation. The np 4216 mutation changes a poorly conserved NDl tyrosine 304 to a histidine, is always homoplasmic, and is present in roughly 13% of random controls reported in the literature (Table 1). We found this variant in 43% (20/47) of LHON patients and in all Caucasian controls (n = 21) which were associated with the two np 4216-positive mtDNA lineages. Thus, the np 4216 mutation is not specific for LHON since all individuals from these two Caucasian lineages were homoplasmic for the mutation irrespective of their pheno- types.

The np 13708 mutation has occurred one time in our Caucasian sample and is always found in association with the np 4216 mutation. The np 13708 mutation changes the moderately conserved ND5 alanine 458 to a threonine and is present in roughly 6% of random controls reported in the literature (Table 1). We found this variant in 30% (14/47) of LHON patients and in 9% (16/175) of our Caucasian controls. All common primary LHON mutations were found in individuals harboring the np 4216 + 13708 genotype, as were all of the np 15257-positive individuals in our sample including the single np 15257-positive patient (L14) which contained the heteroplasmic np 5244 mutation. As mentioned, nearly all (6/8) of our np 14484-positive patients were part of the 4216 + 13708 mtDNA lineage.

The np 3394 mutation was also found exclusively with the np 4216 + 13708 genotype and therefore arose a single time in the Caucasians de- picted in Figure 1. This mutation alters a moderately conserved ND1 tyrosine and has been found in approximately 1% of random controls reported in the literature. In this study, the mutation was found in four LHON patients (8.5%), but also two controls (1.1%). However, all np 3394 containing patients also harbored one of the primary LHON mutations: one patient (L9) had the np 11778 mu-

318 BROWN ET AL.

TABLE 2. MtDNA Mutations Found in 11778-Positive and 11778-Negative LHON Patientsa

Intermediate Primary mutations mutations Secondary mutations

Patient number 3460 11778 14484 15257 5244 3394 4216 4917 7444 13708 15812

3 7 9

13 15 17 18 21 23 26 27 31 32 33 36 44 47

1 2 4 5 6 8

10 11 12 14 16 19 20 22 24 25 28 29 30 34 35 37 38 39 40 41 42 43 45 46

"Patients 6.11, 12, and 14 were reported by Brown et al. (1992a); patients 1 and 2 were reported by Brown et al. (1992b); patients 40 and 41 were reported by Brown et al. (1992~). bMutation was heteroplasmic. 'nd. no data.

tation and three patients (L2, L5, L9) had the np 14484 mutation. Hence, the np 3394 mutation is insufficient by itself to cause LHON. Therefore, it may represent a polymorphism specific to the np 4216 + 13708 lineage, or it might influence the penetrance of individuals which also harbor one of the primary mutations.

The np 4917 mutation occurred once in Cau- casians and in exclusive association with the np 4216 mutation. The np 4917 variant changes a highly conserved ND2 aspartate and has been reported to be present in 4% of random controls. We found this mutation in 13% (6/47) of our Cauca- sian LHON patients. Haplotype-matched controls


[haplotypes 26 (n = I ) , 57 (n = 2), and 98 (n = 2)] were positive for the np 4917 mutation, thus this polymorphism is not specific to LHON. Like the np 4216 + 13708 lineage, LHON is common in the np 4216 + 491 7 lineage as 55% (6/11) of these individuals were LHON patients. All but two of these LHON patients also harbored a primary mutation, with patient L30 containing the np 3460 mutation and patients L31, L32, and L33 containing the np 11 778 mutation. Hence, the np 4917 mutation is not a primary LHON mutation, but could possibly exacerbate a more primary etiological factor.

The np 7444 mutation appears to have occurred at least twice in Caucasians and is also insufficient to cause LHON by itself. This mutation extends the COI polypeptide by three amino acids and was found in two LHON patients. Both patients harbored one of the primary mutations: patient L40 harbored the np 3460 mutation while patient L41 harbored the 14484 mutation. When haplotype- matched controls [haplotypes 36 (n = 4), 61 (n = l), 63 (n = l), and 99 (n = l)] were tested for the presence of this mutation by XbaI digestion, only one (haplotype 45) individual was positive. The consistent presence of primary mutations in np 7444-containing patients and its presence in an unaffected control confirm that the np 7444 mutation is not a primary cause of LHON. However, the novel nature of the mutation and its slightly increased frequency in LHON patients raise the possibility that it might also contribute to the penetrance of the primary mutat ions.

The Intermediate Mutations at np 5244 and np 15257 Have Features of Primary Mutations

The mutation at np 5244 has only been found in patient L14 and is probably a primary LHON mutation. This mutation converts the very highly conserved glycine 259 to a serine in the ND2 polypeptide. The homologous glycine is maintained in nearly all organisms studied including vertebrates, sea urchin, Drosophih, Paramecium, Neurospora crassa, Escherichia coli, Paracoccus denitrificans, and liverwort mitochondria and choloroplasts. Further, this mutation is heteroplasmic in the proband, and has not been observed in greater than 2100 random controls (Brown et al., 1992a). The mutation occurred on a np 4216 + 13708 + 15257 mtDNA which lacked the three common primary LHON mutations (Fig. 1). Although the np 5244 mutation is probably the primary LHON mutation in this patient, a primary status for this mutation re-

mains provisional, since it has only been observed in one singleton pedigree, and this patient was found to harbor a second novel mtDNA variant at np 10663 in this study (see below).

The np 15257 mutation also has features consistent with a primary mutation. It changes a highly conserved amino acid in the cytochrome b gene, has been shown in the literature to have arisen two independent times in association with LHON (Howell et al., 1993a) (although only a single time in our sample), and is present at increased frequency in LHON patients relative to controls. We found this variant in 7/47 (15%) of patients and in 2/175 (1.1%) of controls (haplotypes 78 and 97). However, five of the six infor- mative cases also harbored known or suspected primary mutations: one patient (L7) had the np 11778 mutation, three patients (L8, L10, L11) had the np 14484 mutation, and one patient (L14) had the np 5244 mutation.

One np 15257-positive patient (patient L12), lacked any previously identified common primary mutations. To determine if patient L12 harbored another as yet unidentified primary mutation, which could explain the phenotype, we sequenced this patient’s mtDNA. Ninety-nine percent of the mtDNA was sequenced, revealing 47 nucleotide substitutions relative to the standard Cambridge sequence (Anderson et al., 1981) (Table 3). All of these mutations but one could be excluded as either a known sequence variant or a new polymorphism. The remaining variant was a homoplasmic T to C transition at np 10663 in the ND4L gene. This converts the poorly conserved valine at resi- due 65 to an alanine in the ND4L polypeptide (Table 3). The np 10663 mutation eliminates an SpeI restriction endonuclease recognition site at np 10658 thus providing a rapid screen for the presence of the mutation in LHON patients and controls. To assess the origin of the mutation in the L12 family, we tested three other maternal rela- tives (mother and two siblings), and, like the proband, all were found to be homoplasmic for the np 10663 mutation. To determine if the variant was a neutral polymorphism we screened 96 Cau- casian (random United States subjects), 93 Asian (15 Malay Chinese, 29 Vietnamese, 22 Sabah Ab- originies, 18 Taiwainese, 9 Korean), and 91 Afri- can (60 Mandinka and 31 Eastern and Western Pygmy) controls. All were negative for the np 10663 mutation. Further, we screened all controls (except haplotype 81; n = 1) and LHON patients (except L6, L8, and L13) which harbored the np 4216 and 13708 mutations and hence were on the

320 BROWN ET AL.

TABLE 3. MtDNA Sequence Changes in np 15257-Positive LHON Patient L12”

Nucleotide Amino Amino acid position Nucleotide acid conservation and gene change change HlClMlXlCh/SlD

1. 3191Dloop T t o C - 2. 512lDloop G t o A - 3. 1850116s rRNA T to C - 4. 4216ND1 T to C Y to H YIWHIHIHILII 5. 71991COI C to A syn 6. 77891COII G to A syn 7. 10499ND4L A to G syn 8. 10663ND4L T to C V to A VlVlVlPlAlTlT 9. 13053ND5 C to T syn

10. 13708ND5 G to A A to T AILIAIAIAIAIL 11. 13722ND5 A to G syn 12. 14133ND5 A to G syn 13. 16126lDl00p T to C - 14. 16145lDl00p G to A - 15. 16231IDl00p T to C - 16. 16261lDl00p C to T -

“The missense mutation at np 10663 has not been previously described in the literature. Previously reported polymorphisms or Cambridge sequencing errors found in the proband include mutations at nps 73, 150, 152, 195, 489, 750, 1438, 2706, 3106, 3423,4769,4985,7028,7476,8860,9559,10398,11251,11335, 11719, 12612, 13161, 13702, 14199, 14272, 14365, 14368, 15326. 15452, and 16069. Mutations in table are reported as L-strand base changes. H, human: C, cow: M, mouse; X, Xeno- pus; Ch, chicken: S, sea urchin: D, Drosophiia.

same mtDNA lineage as L12. All patients and controls from this lineage were negative for the ND4L mutation, with the exception of patient L14, who harbors the homoplasmic np 15257 mutation and heteroplasmic np 5244 mutation. Di- rect mtDNA sequence analysis (data not shown) confirmed that patient L14, who is the only other np 15257-positive patient that lacks one of the three primary mutations, was also homoplasmic for the T to C base change at np 10663.

To see if the np 10663 mutation was present in other LHON patients not harboring known common LHON mutations, we screened 20 LHON and 16 putative LHON patients for the presence of the np 10663 variant. All were negative with the exception one 48-year-old Caucasian woman with ophthalmologic clinical signs including monocular vision loss with spared visual acuity, both of which are atypical of LHON. To unambiguously determine the cause of the Spel site loss in this individual, we sequenced a portion of her ND4L gene. This revealed that the SpeI site loss in this individual was due to a silent A to G transition at np 10658, and thus of no relevance to the np 10663 mutation or her clinical phenotype.

DISCUSSION

LHON has been proposed to be genetically heterogeneous, resulting from at least three common

primary LHON mtDNA mutations at nps 3460, 11778, and 14484 which together account for approximately 85% of Caucasian LHON patients (Brown and Wallace, 1994a). Thirteen other mtDNA mutations have also been associated with this disease in Caucasians. With the exception of the most common LHON mutation at 11778, the population frequencies and mutational associations of these mutations have not been rigorously exam- ined. The current paper addresses this deficiency by a phylogenetic analysis of patient and control mtDNAs. This has confirmed the primary nature of the three common mutations, demonstrated the lineal nature of the secondary LHON mutations, provided insight into the pathogenicity of the ambiguous np 5244 and np 15257 mutations, and identified a new ND4L np 10663 mutation of possible relevance to LHON.

The 11778 Mutation

The 11 778 mutation has not been found in over 380 controls throughout the world and has occurred multiple times in Caucasians, each time resulting in disease (Wallace et al., 1988; Holt et al., 1989; Vilkki et al., 1989; Korman et al., 1991; Newman et. al., 1991; Mashima et al., 1993; Newman et al., 1994). The current study has confirmed the repeated occurrence of the 11 778 mutation, having shown that it has arisen at least 12 independent times in a sample of 17 11778-positive Caucasian patients (Fig. 1). Indeed, this variant not only has arisen multiple times in Caucasians, but has been reported in Asian (Jap- anese) (Mashima et al., 1993) and African-de- rived mtDNA haplotypes (Newman et al., 1994), the latter characterized by the African-specific HpaI polymorphism at np 3592 (Denaro et al., 1981).

The 11778 mutation did not co-occur with other primary LHON mutations in our survey. Moreover, in 59% (10/17) of cases, it did not co- occur with any other LHON-associated mutations. However, when it occurred with secondary LHON mutations, it occurred together with the 4216 + 4917, 4216 + 13708, or the 4216 + 13708 + 15257 genotypes. Thus, this base change unequiv- ocally imparts a high risk for blindness, regardless of background mtDNA haplotype.

The 3460 Mutation

The 3460 mutation has not been observed in over 630 controls surveyed (Huoponen et al.,


1991; Howell et al., 1991a; Johns, 1992; Johns et al., 1992a; Newman et al., 1994;). Moreover, in the present study, this mutation was shown to have occurred at least nine independent times, one case of which was heteroplasmic (Fig. 1). So far, this mutation has been limited to Caucasian patients in North America and Europe. This apparent speci- ficity could be explained by differences in genetic background (mitochondria1 or nuclear) or could simple reflect inadequate sampling in other populations.

The 3460 mutation accounted for 33% of our 11778-negative patients. Assuming (from the literature) that the 11 7718 mutation is responsible for 50-60% of all Caucasian LHON cases (Wallace et al., 1988; Newman et al., 1991; Wallace, 1992; Newman, 1993; Brown and Wallace, 1994a) the np 3460 mutation would then account for about 13-17% of all Caucasian LHON patients, a figure in good agreement with recently published estimates of 15% (Johns et al., 1992a; Howell et al., 1991a). In 70% of our 3460-positive patients, this mutation did not co-occur with other LHON-associated mutations and it never occurred with the other primary LHON mutations. Hence, this mutation is also a primary risk factor for LHON.

The 14484 Mutation

The 14484 mutation has not been detected in roughly 430 random, mostly Caucasian, controls (Mackey and Howell, 1992; Johns et al., 199213, 1993a; Newman et al., 1994) nor in controls with a similar mtDNA haplotype (this study). It appears to have occurred four separate times in our collection of Caucasian patients (Fig. I). This mutation has been found primarily in Caucasians, and was not found to co-occur with other common primary LHON mutations. Therefore, this mutation must also be a primary cause of LHON in the population.

The 14484 mutation accounted for 27% of our 11 778-negative LHON population, which trans- lates to an overall frequency of 11-13% in Cauca- sian LHON patients, a range in good agreement with previously published estimates (Johns et al., 1992b). Unlike the 3460 and 11778 mutations, however, this variant was found only in the homoplasmic state and co-occurred with secondary LHON mutations in seven out of eight patients (see Fig. 1). Six 14484-positive patients contained the 4216 and 13708 variants. The combined data from this study and the literature indicate that nearly 80% of 14484-positive patients have a very similar mtDNA haplotype (Caucasian haplogroup “J” from Torroni et al., 1994) characterized by variants at

nps 4216, 10398, 13708, and 16069 (Brown et al., .1992a,b; Mackey and Howell, 1992; Johns et al., 1993a). Further, roughly 38% of all np 14484-positive patients simultaneously harbor the np 15257 mutation and a separate 38% simultaneously contain the np 3394 mutation. These strong haplotype associations are unique among the primary LHON variants and may not be fully explained by a common mutational origin (founder), since the 14484 mutation appears to have arisen at least twice within this lineage. Since a number of secondary LHON mutations occur exclusively within this lineage, it is possible that expression of the 14484 mutation is enhanced by the presence of one or more of the mutations within the np 4216 + 13708 lineage. This implies that the np 14484 mutation may be less deleterious than the 11778 or 3460 mutation, a notion supported by the high rate of spontaneous visual recovery associated with the 14484 mutation (37% Johns et al., 1993a) relative to the 3460 (22% Johns et al., 1993a) or 11778 (4%, Stone et al., 1992; Johns et al., 1993a) mutations. The np 14484-positive patient L39 did not contain any other known LHON-associated mutations, which supports the primary pathogenicity of this mutation. Thus, the dispersal of the three primary LHON mutations around the Caucasian phylogenetic tree confirms their recurrent origin, with 71% (12117) of 11778-positive patients, 90% (9/ 10) of 3460-positive patients, and 50% (4/8) of 14484-positive patients apparently representing independent mutational events. This is in marked contrast to most secondary LHON mutations like the 3394, 4917, 13708, and 15812 which are restricted to a single mtDNA lineage and thus arose as a single founder event.

The 15257 Mutation

The 15257 mutation may also contribute to LHON and could thus be of intermediate risk between the primary and secondary LHON mutations. This mutation alters a conserved aspartate in the cytb gene, is relatively common in LHON patients, and has been found in LHON patients with different mtDNA haplotypes (see below). How- ever, unlike the three common primary LHON mutations, the np 15257 mutation has been detected in at least 4/1044 (0.4%) worldwide controls (Brown et al., 1992a; Johns et al., 199313; Torroni et al., 1994; unpublished data, this laboratory) including the 2/175 Caucasian controls represented in Figure 1. Furthermore, in patients, this variant is often found in association with one

322 BROWN ET AL.

of the common primary LHON mutations, most commonly the np 14484 mutation, but also the np 11778 mutation (Huoponen et al., 1993; Johns et al., 1993b). In our Caucasian survey, the np 15257 mutation appears to have arisen once and was found in six 11778-negative patients and in one 11778-positive patient. In the 1 1778-negative patients, the 15257 variant was linked to the 14484 mutation in three cases and with the heteroplasmic 5244 mutation in one case.

One np 15257-positive patient (L12) did not harbor any known common primary LHON mutations, although this patient did harbor the 4216 and 13708 mutations. Patient L12 is the proband from a three generation maternally inherited LHON family with a total of five affected family members (see Methods and Materials). Complete sequence analysis of this patient’s mtDNA revealed only one other missense mutation at np 10663 which could play an etiological role in the disease in this family. This mutation alters a poorly conserved valine codon in the ND4L gene, was not found in a large number of random controls representing all major racial groups, and was not found in haplotype-matched controls or patients with the exception of patient L14. The occurrence of the np 10663 mutation in patients L12 and L14 is intriguing since these are the only two patients harboring the np 4216 + 13708 genotype which do not contain one of the three primary mutations, although patient L14 does harbor the heteroplasmic np 5244 mutation.

A screening of 36 likely or putative LHON patients which lacked the np 3460, 11778, 14484, and 15257 mutations was also negative for the np 10633 mutation. Thus, if this mutation is pathogenic it may require other mtDNA mutations in the np 4216 + 13708 + 15257 lineage to result in a sufficient energetic defect to cause the disease. However, the pathogenicity of the np 10663 mutation must still be tentative because of its low frequency in LHON patients lacking other primary LHON mutations, its homoplasmic occurrence on a single mtDNA haplotype, the alteration of a poorly conserved amino acid, and its presence in a patient (L14) simultaneously harboring the heteroplasmic np 5244 ND2 mutation which is believed to be the primary LHON mutation in this patient.

Pending confirmation that the np 10663 mutation contributes to LHON, the molecular genetic evidence from patient L12 supports the hypothesis that the np 15257 mutation has an etiological role in LHON. This proband comes from a pedigree with classical LHON presentation and clear famil-

ial evidence of maternal transmission. Yet the sequence of the entire mtDNA revealed only the np 4216 + 13708 + 15257 mutations. Hence, with the possible exception of the np 10663 mutation, other pathogenic mtDNA mutations were excluded. Furthermore, Howell et at. (1993a) have reported a singleton np 15257-positive Australian LHON case, in which sequence analysis of most of the mtDNA protein coding genes failed to reveal any additional pathogenic mtDNA mutations, including the np 10663 mutation. The mtDNA haplotype of this np 15257-positive patient was very different from the mtDNA haplotype of our patient L12. This patient must therefore represent a second independent np 15257 mutation also in association with LHON. Thus, the current data con- tinue to support a pathological role for this mutation, though it is clearly less pathogenic than the np 11778, 3460 and 14484 LHON mutations.

Other LHON-Associated Mutations

Certain secondary LHON mutations appear in LHON patients more frequently than controls. This raises the possibility that they also contribute to the expression of LHON. LHON appears to cluster on the np 4216 + 13708 mtDNAs, although the reason for this is not obvious. Both the np 4216 and np 13708 mutation are found at polymorphic frequencies in non-LHON Caucasians, are consistently homoplasmic, and are rarely found in LHON patients in the absence of a common primary LHON mutation (Brown et al., 1992a,b; Johns et al., 1993a; Johns and Berman, 1991). It is possible that the np 4216 + 13708 haplotype partially impairs oxidative phosphorylation such that the occurrence of other primary LHON mutation is more likely to result in phenotypic expression.

The relative pathogenicity of the np 3394 mutation is also unclear. This mutation is restricted to the 4216 + 13708 lineage, was never found in patient mtDNAs independent of the np 14484 mutation, and was found in two control individuals which lacked the np 14484 mutation. These results are concordant with other data concerning the np 3394 mutation (Brown et al., 1992b; Johns et al., 1992b). Thus, this mutation is either a rare polymorphism or a weakly contributing factor in the pathology of this disease.

The pathological significance of the 4917 mutation also remains in question. This mutation changes a conserved aspartate to an asparagine in the ND2 subunit of complex I and has been found in higher frequencies in an 11778-negative LHON patient cohort relative to a control population


(36% versus 4%; Johns and Berman, 1991). In the present study, we found this mutation equally dispersed among 11778-positive (n = 3; 18%) and 11778-negative (n = 3; 10%) patients, and in one individual containing the 3460 mutation. This variant was also found in all haplotype-matched control individuals. Like the np 13708 mutation, the np 4917 mutation generally occurs with the 4216 mutation, partially defining a lineage where LHON is relatively common. Again, the 4216 + 4917 genotype may be mildly deleterious and thus may influence the penetrance of other primary LHON mutations.

The np 7444 variant is novel among mtDNA mutations in that it converts the COI mRNA ter- mination codon to a lysine codon, thereby extend- ing the COI polypeptide by three amino acids. This structural alteration is reflected in the altered COI mobility upon SDS-PAGE and a reduction in the specific activity of: Complex IV (Brown et al., 1992~). However, this mutation had previously been found in 61545 (1.1%) worldwide controls (Brown et al., 1992c) and the present study has revealed that all LHON patients with the np 7444

variant also harbor a primary LHON mutation, specifically the np 3460 or the np 14484 mutations. An individual with a similar mtDNA haplotype and the 7444-positive LHON patients but no primary mutation had no phenotype. Thus, this mutation is either a novel polymorphism or is only weakly contributory to the disease.

In conclusion, this phylogenetic analysis of Cau- casian LHON patients has further substantiated the pathogenicity of the np 3460, 11778, and 14484 mutations. It has also provided support for a pathogenic role for the np 15257 and np 5244 mutations, perhaps through synergistic interaction with a new ND4L np 10663 variant which might contribute to LHON, but does not so far appear to be a primary mutation. Finally, these data support the hypothesis that certain mtDNA haplotypes (such as the np 4216 + 13708 or the np 4216 + 4917 haplotype) might contribute to the expression of LHON. Thus, it remains possible that multiple genetic factors can play a role in an individual’s predisposition to LHON, although genetic counseling for these patients should currently focus on the less ambiguous common primary LHON mutations.

APPENDLX I. Polvmomhic Restriction Sites”

Sites 9 5 1 j 1 3 0 1 f 1 7 1 5 c 1 7 1 8a 2 8 4 9 k 31 9 2 c 3 3 9 1 e 4 3 3 2 b 4 6 4 3 k 4 7 6 9 a 4 7 9 3 e 5 0 0 3 c 5 2 4 2 1 5 2 5 9 b I 5 2 6 1 e 5 8 2 3 a 5 8 3 7 e 7 0 2 5 a 7 4 7 4 a 7 6 0 7 e 7 8 5 3 0 8 2 4 9 b I 8 2 5 0 e 8 5 1 9 9 8 6 1 6 j 8 7 4 8 c 8 8 5 8 f 8 9 9 4 e 9 0 5 2 n / 9 0 5 3 f 9 1 4 7 c 1 0 3 9 4 c 1 0 7 7 1 j 1 1 3 1 3a 1 1 6 4 1 e

“L” Haplotypes 1 1 1 1 1 1 1 1 1 1 2 2 2 2 2 2 2 2 2 2 3 3 3 3 3 3 3 3 3 3 4 4 4 4 4 4 4 4

1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7

1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 0 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 0 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 0 1 1 1 1 1 1 1 1 1 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 1 0 0 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 0 0 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 0 1 1 1 1 1 0 1 0 1 1 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 0 1 1 1 1 1 1 1 1 1 1 1 1 1 0 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 0 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 0 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 0 0 0 0 0 0 0 0 0 0 0 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 0 1 1 1 1 1 0 0 0 1 0 0 0 1 0 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 0 1 1 1 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 0 1 1 1 1 1 1 1 1 1 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 0 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 0 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0

(continued)

324 BROWN ET AL.

APPENDIX 1. Polymorphic Restriction Sites (Continued)

12345k 00000000000000000000000000000000010000000000000 12629b 11111111111111111111101111111111111111111111111

13604k 00000000000010000000000000000000000000000000000 136351 00000000000000000000000000000000000000000001100 13702er 00000000000000000000000000000000000000000000000 141 9901 00000000000000000000000000000000000000000000000

14304a 11111111111111111111111111111111111111110111111 14347k 00000000000000000000000000000000000000000000001

15606a 00000000000000000000000000011111100000000000000 15776a 11111111111111111111111111111111111111110111111 1581 2k 11111111100110111111111111111111111111111111111 1 5925 i 11111111111111111111111111100000011111111111111 160659 00000000000000111111111111111111111111111111111 16208k 11111111111111111111111111111111111111111101111 16303k 11111111111111111111111000110011111111111111111 1631 O k 11111111111111111111111110111111111111111111111 163899/16390b 00000000000000000000001000000000000000000000000 16398e 00000000000000000000000000000000000000000000010 1651 7e 10000000000000111111111000011111111111000000100 9 - b p deletion 00000000000000000000000000000010000000000000000 3457p 01 1 1 1 1 1 1 1 1 1 1 1 1 1 1 111011 1001 1 1 1011 110101 10101 101 1 421 6 q 11111111111111000000000000011111100000000000000 491 4r 0000000J000000000000000000011111100000000000000 74405 11111111111111111111111111111111111111100111111 1 1 7753 001 0001 01 0001 01 01 1001 01 001 10001 1 1 001 00000001 001 1 3 7 0 4 ~ 00000000000000711111111111111111111111111111111 14484v 01 01 1 ? 0 1 0 1 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 ’ 0 0 0 0 0 0 0 0 0 0 0 0 1 0 1 0 0 0 0 0 0 152542 11111100100010111111111111111111111111111111111

“Table presenting the polymorphic restriction sites which define the 47 mtDNAs (Ll-L47) observed in LHON patients. A “1” indicates the presence of a site and a “0’ indicates the absence of a site except for the COII-tRNALy” 9-bp deletion where “1” indicates the presence of a 9-bp deletion between the COII and tRNALys genes and a “0’ indicates the absence of the deletion. Sites are numbered from the first nucleotide of the recognition sequence according to the published sequence (Anderson et al., 1981). The restriction enzymes used in the analysis are designated by the following single-letter code: a, AluI; b, Auall; c, DdeI; e, HaelII; f, Hhak g. Hinfl; h, Hpal; i, H p a k j, Mbol; k, RsaI; I, Taql; m, BamHI; n, HaeIk 0, Hincll; p, BsaHI; q. NlaIlI; r, MaeI; s, XbaI; t, MaeIlI; u, BstNI; v, Musk z. Accl. Sites separated by a diagonal line indicate either simultaneous site gains or site losses for two different enzymes or a site gain for one enzyme and a site loss for another because of a single inferred nucleotide substitution; these sites are considered to be only one restriction site polymorphism in the parsimony analysis. Underlined sites indicate mutations which have been associated with LHON. The question mark indicates that L6 could not be tested for the 14484 mutation. Sites marked with an asterisk were found to be present or absent in all samples contrary to the published sequence.

13366m/13367b/13367j 00000000000000000000000000011111100000000000000

1426891 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1

1436891 00000000000000000000000000000000000000000000000 1 4869 j 11111111111111111111011111111111111111111111111

ACKNOWLEDGMENTS

The authors would like to thank Ms. Marie T. Lott and Dr. Nancy Newman of Emory University School of Medicine for intellectual and technical input and Dr. John M. Shoffner and the Emory University Molecular Diagnostics Laboratory for providing some of the LHON patient samples. We also thank Dr. Birgit Lorenz for clinical informa- tion on LHON patient L12 and family as well as Stephanie Neil1 and Lori Griffin (Emory Univer- sity General Clinical Research Center) for cell cul- ture work. This work was supported in part by NIH Grants NS21328 and GM46915 (DCW) and MOl- RR-00039 (Emory University General Clinical Re- search Center).

REFERENCES Anderson S, Bankier A, Barrel1 BG, deBruijn MHL, Coulson AR,

Drouin J , Eperon IC, Neirlich DP, Rose BA, Sanger F,

Schreier PH, Smith AJH, Staden R, Young 1G (1981) Se- quence and organization of the human mitochondrial genome. Nature (London) 290:457-465.

Brown MD, Voljavec AS, Lott MT, Torroni A, Yang C-C, Wal- lace DC (1992a) Mitochondria1 DNA complex I and 111 mutations associated with Leber’s hereditary optic neuropathy. Genetics 130:163-173.

Brown MD, Voljavec AS, Lott MT, MacDonald I , Wallace DC (1992b) Leber’s hereditary optic neuropathy; a model for mitochondrial neurodegenerative diseases. FASEB J 6:2791- 2799.

Brown MD, Yang C-C, Trounce I, Torroni A , Lott MT, Wallace DC ( 1 9 9 2 ~ ) A mitochondrial DNA variant, identified in Leber hereditary optic neuropathy patients, which extends the amino acid sequence of cytochrome c oxidase subunit 1. Am J Hum Genet 51:378-385.

Brown MD, Wallace DC (1994a) The spectrum of mitochondrial DNA mutations in Leber’s hereditary optic neuropathy. Clin Neurosci 2138-145.

Brown MD, Wallace DC (199413) The molecular basis of mitochondrial DNA disease. J Bioenerget Biomembranes 26263- 279.

Brown MD, Jun AS, Wallace DC (1994a) Maternally inherited


Leber’s hereditary optic neuropathy and dystonia associated with a mitochondrial DNA ND6 mutation. Neurology 44 (suppl 2):A285.

Brown MD, Torroni A, Houponen K, Chen, Y-S, Lott MT, Wal- lace DC (1994b) Pathological significance of the mitochondrial DNA COIII mutation at np 9438 in Leber’s hereditary optic neuropathy Am J Hum Genet 55:410.

Denaro M, Blanc H , Johnson MI, Chen KH, Wilmsen E, Cavalli Sforza LL, Wallace DC (1981) Ethnic variation in HpaI endonuclease cleavage patterns of human mitochondrial DNA. Proc Natl Acad Sci USA 785768-5772.

Holt IJ, Miller DH, Harding AE (1989) Genetic heterogeneity and mitochondrial DNA heteroplasmy in Leber’s hereditary optic neuropathy. J Med Genet 26:739-743.

Howell N , Bindoff LA, McCullough DA, Kubacka I, Poulton J , Mackey D, Taylor L, Turnbull DM (1991a) Leber hereditary optic neuropathy: identification of the same mitochondrial NDl mutation in six pedigrees. Am J Hum Genet 49:939-950.

Howell N , Kubacka I, Xu M, McCullough DA (1991b) Leber hereditary optic neuropathy: involvement of the mitochondrial ND1 gene and evidence fix an intragenic suppressor mutation. Am J Hum Genet 48:935-942.

Howell N, Kubacka I, Halvorson S, Mackey D (1993a) Leber’s hereditary optic neuropathy: The etiological role of a mutation in the mitochondrial cytochrome b gene. Genetics 133:133- 136.

Howell N , Halvorson S, Burns J , McCullough DA, Poulton J (199312) When does bilateral optic atrophy become Leber hereditary optic atrophy? Am J Hum Genet 53:959-963.

Huoponen K, Vilkki J, Aula P, Nikoskelainen EK, Savontaus ML (1991) A new mtDNA mutation associated with Leber hereditary optic neuroretinopathy. Am J Hum Genet 48:1147-1153.

Huoponen K, Lamminen T, Juvonen V, Aula P, Nikoskelainen EK, Savontaus ML (1993) The spectrum of mitochondrial DNA mutations in families with Leber hereditary optic neuroretinopathy. Hum Genet 92:379-384.

Johns DR (1992) Mitochondrial ND-1 mutation in Leber hereditary optic neuropathy. Am J Hum Genet 504724474.

Johns DR, Berman J (1991) Alternative, simultaneous complex 1 mitochondrial DNA mutations in Leber’s hereditary optic neuropathy. Biochem Biophys Res Commun 174:244-250.

Johns DR, Neufeld MJ (1991 ) Cytochrome b mutations in Leber hereditary optic neuropathy. Biochem Biophys Res Commun 181: 1358-1364.

Johns DR, Neufeld MJ (1993) Cytochrome c oxidase mutations in Leber hereditary optic neuropathy. Biochem Biophys Res Commun 196310-815.

Johns DR, Smith KH, Miller NR (1992a) Leber’s hereditary optic neuropathy. Clinical manifestations of the 3460 mutation. Arch Ophthalmol 110:1577-1581.

Johns DR, Neufeld MJ, Park RD (1992b) A n ND-6 mitochondrial DNA mutation associated with Leber hereditary optic neuropathy. Biochem Biophys Res Commun 187:1551-1557.

Johns DR, Heher KI, Miller NR, Smith KH (1993a) Leber’s hereditary optic neuropathy. Clinical manifestations of the 14484 mutation. Arch Ophthalmol 11 1 :495-498.

Johns DR, Smith KH, Savino PJ, Miller NR (1993b) Leber’s hereditary optic neuropathy. Clinical manifestations of the 15257 mutation. Arch Ophthalmol 110:981-986.

Jun AS, Brown MD, Wallace DC (1994) A mitochondrial DNA

mutation at np 14459 of the ND6 gene associated with maternally inherited Leber’s hereditary optic neuropathy and dystonia. Proc Natl Acad Sci USA 91:6206-6210.

Korman BA, Shuster H, Berninger TA, Leo-Kohler B (1991) De- tection of the G to A mitochondrial DNA mutation at position 11778 in German families with Leber’s hereditary optic neuropathy. Hum Genet 88:98-100.

Mackey D, Howell N (1992) A variant of Leber hereditary optic neuropathy characterized by recovery of vision and by an un- usual mitochondria genetic etiology. Am J Hum Genet 51:

Mashima Y, Hiida Y, Oguchi Y, Kudoh J , Shimizu N (1993) High frequency of mutations at position 11778 in mitochondrial ND4 gene in Japanese families with Leber’s hereditary optic neuropathy. Hum Genet 92:lOl-102.

Newman NJ (1993) Leber’s hereditary optic neuropathy. Ophthal- mol Chn NA 4:431-447.

Newman NJ, Lott MT, Wallace DC (1991) The clinical charac- teristics of pedigrees of Leber’s hereditary optic neuropathy with the 11778 mutation. Am J Ophthalmol 111:750-762.

Newman NJ, Torroni A , Brown, MD, Lott MT, Marquez Fernan- dez M, Wallace DC, and the Cuban Field Investigation Team (1994) Epidemic neuropathy in Cuba not associated with Leb- er’s mitochondrial DNA mutations. Am J Ophthalmol 118: 158-168.

Singh G, Lott MT, Wallace DC (1989) A mitochondrial DNA mutation as a cause of Leber’s hereditary optic neuropathy. N Engl J Med 320:1300-1305.

Stone EM, Newman NJ, Miller NR, Johns DR, Lott MT, Wallace DC (1992) Visual recovery in patients with Leber’s hereditary optic neuropathy and the 11778 mutation. J Clin Neuro Oph- thalmol 12:lO-14.

Swofford D (1992) Phylogenetic Analysis Using Parsimony Anal- ysis (PAUP) 3.0s. Illinois Natural History Survey, Champaign, Illinois.

Torroni A , Schurr TG, Yang C-C, Szathmary EJE, Williams RC, Schanfield MS, Troup GA, Knowler WC, Lawrence DN, Weiss KM, Wallace DC (1992) Native American mitochondrial DNA analysis indicates that the Amerind and Na Dene populations were founded by two independent migrations. Ge- netics 130:153-162.

Torroni A, Schurr TG, Cabell MF, Brown MD, Nee1 JV, Larsen M, Smith DG, Vullo CM, Wallace DC (1993) Asian affinities and continental radiation of the four founding Native Ameri- can MtDNAs. Am J Hum Genet 53:563-590.

Torroni A, Lott MT, Cabell MF, Chen, Y-S, Lavergne L, Wallace DC (1994) MtDNA and the origin of Caucasians. Identifica- tion of ancient Caucasian-specific haplogroups, one of which is prone to a recurrent somatic duplication in the D-loop region. Am J Hum Genet 55:760-776.

Vilkki J. Savontaus ML, Nikoskelainen E (1989) Genetic heterogeneity in Leber hereditary optic neuropathy revealed by mitochondrial DNA polymorphism. Am J Hum Genet 45:206- 211.

Wallace Dc (1992) Diseases of the mitochondrial DNA. Annu Rev Biochem 61~1175-1212.

Wallace Dc, Singh G, Lott MT, Hodge JA, Schurr TG, Lezza AMS, Elsas LJ, Nikoskelainen E (1988) Mitochondrial DNA mutation associated with Leber’s hereditary optic neuropathy. Science 242:1427-1430.

1218-1228.

Documents

Phylogenetic analysis of Leber's hereditary optic neuropathy mitochondrial DNA's indicates multiple independent occurrences of the common mutations