Novel ENU-induced eye mutations in the mouse - Oxford University ...

165 downloads 27 Views 298KB Size Report
There is also ectropion uveae and a distorted pupil. (F) Raw. The retinal arterioles give a 'shiny' reflex not seen in normal mice. (G) Svc. Numerous vacuoles are ...
© 2002 Oxford University Press

Human Molecular Genetics, 2002, Vol. 11, No. 7

755–767

Novel ENU-induced eye mutations in the mouse: models for human eye disease Caroline Thaung1,2, Katrine West1, Brian J. Clark3, Lisa McKie1, Joanne E. Morgan1, Karen Arnold1,2, Patrick M. Nolan2, Jo Peters2, A. Jackie Hunter4, Steve D. M. Brown2, Ian J. Jackson1 and Sally H. Cross1,* 1Comparative

and Developmental Genetics Section, MRC Human Genetics Unit, Edinburgh EH4 2XU, UK, Mammalian Genetics Unit and UK Mouse Genome Centre, Harwell OX11 0RD, UK, 3Department of Pathology, Institute of Ophthalmology, University College, London EC1V 9EL, UK and 4Neurology CEDD, GlaxoSmithKline, Harlow CM19 5AW, UK 2MRC

Received November 20, 2001; Revised and Accepted February 7, 2002

We have carried out a genome-wide screen for novel N-ethyl-N-nitrosourea-induced mutations that give rise to eye and vision abnormalities in the mouse and have identified 25 inherited phenotypes that affect all parts of the eye. A combination of genetic mapping, complementation and molecular analysis revealed that 14 of these are mutations in genes previously identified to play a role in eye pathophysiology, namely Pax6, Mitf, Egfr and Pde6b. Many of the others are located in genomic regions lacking candidate genes and these define new loci. Four of the mutants display a similar phenotype of dilated pupils but do not appear to be allelic, and at least two of these are embryonic lethal when homozygous. This collection of eye mutations will be valuable for understanding gene function, for dissecting protein function and as models of human eye disease.

INTRODUCTION Sequencing studies indicate that the human genome contains between 30 000 and 40 000 genes (1–3). While gene sequence and expression pattern information provide some indication of gene function, additional evidence is needed to determine the role a gene plays. One way this can be achieved is by carrying out functional studies in the mouse. Ideally, there should be an allelic series of mouse mutants available for every gene so that function could be deduced by examining the mutant phenotype and correlating it with the genetic lesion. Large-scale mouse mutagenesis efforts have been instigated to begin to generate large numbers of mutants, using chemical mutagens (4–6). These take a phenotype-driven rather than genotype-driven approach by examining mice that harbour mutations located throughout the genome for altered phenotypes. This has the advantage that the screens are not biased by pre-conceived ideas about gene function. The mutagen of choice is N-ethylN-nitrosourea (ENU) which mainly causes point mutations (reviewed in 7 and 8). It is extremely efficient and in the male mouse germ-line has a per locus mutation rate of up to 1.3 × 10–3. Point mutations in genes induced by ENU can display a range of mutant effects from complete or partial loss-of-function to gain-of-function. A genome-wide screen for dominantly inherited mutations has been carried out at MRC Harwell (4). Mutant phenotypes were identified by using the SHIRPA [SmithKline Beecham Pharmaceuticals, Harwell MRC Mouse Genome Centre and Mammalian Genetics Unit, Imperial College School of Medicine

(St Mary’s), Royal London Hospital, St Bartholomew’s and the Royal London School of Medicine Phenotype Assessment] protocol that identifies a wide range of phenotypes and is a sensitive screen for mouse neurological, neuromuscular and behavioural mutants. However, this protocol does not allow for detection of vision and eye abnormalities. In order to find mice harbouring mutations that result in such defects, we devised screening protocols and used these to examine mice generated by the MRC Harwell mutagenesis programme for novel eye and vision mutants. Human eye disease is common. Approximately 50 million people worldwide are estimated to be blind and three times that number have significant visual impairment (http://www. fightforsight.org). With increasing longevity in developed countries the impact of late-onset ocular diseases including cataract, glaucoma and age-related macular degeneration is becoming more significant (9) and a high proportion of cases of blindness or visual impairment is due to genetic factors (10). Visual impairment does not generally adversely affect the fitness of laboratory mice and animals with eye mutations are easily studied. The mouse eye has many similarities to the human eye and so mice carrying mutations causing eye defects provide useful models of human eye disease, and may help identify the disease genes involved as well as aiding assessment of therapeutic treatments. We aimed to find new mouse mutants that could be used for such studies and report here the identification of new mouse eye mutants, many of which are in novel genes.

*To whom correspondence should be addressed. Tel: +44 131 332 2471; Fax: +44 131 343 2620; Email: [email protected]

756

Human Molecular Genetics, 2002, Vol. 11, No. 7

RESULTS Screen for novel eye and vision mutants ENU-mutagenized BALB/cAnN (hereafter BALB/c) male mice were mated to wild-type C3H/HeN females. We screened their F1 progeny (designated with a MUTN identifier) for eye and vision defects to identify mutant mice using two different approaches. In the first, 6000 were screened for visual function. We did this by monitoring for a head tracking response to a moving black and white grating using a visual tracking drum (see Materials and Methods). This test is based on one that was developed for rats and utilizes the fact that sighted mice, like other vertebrates, display an unconditioned, involuntary following response of the eyes and head to a moving stimulus, and that this response can be monitored for by observing head movements (11,12). Eleven animals did not respond to a 2° grating on three separate occasions. On clinical examination four of these were found to have eye abnormalities and the remaining seven had clinically normal eyes. However, on inheritance testing none of the offspring of these seven mice demonstrated abnormal vision (see below). Hence we found no inherited mutations that lead to substandard vision in the absence of a clinically detectable lesion of the eye. In the second approach, the eyes of 6500 mice were examined for defects using a slit-lamp biomicroscope to detect anterior segment defects and an indirect ophthalmoscope to detect retinal defects. 2500 of these mice were also vision tested (and are included in the figure of 6000 above) and the remaining 4000 were clinically examined but not vision tested. Forty-four mice with abnormal phenotypes were detected. Of these, nine had already been found to have other abnormalities on inspection or at SHIRPA and four had been found to have abnormal vision on testing in the visual tracking drum, as described above. We found abnormalities affecting all parts of the eye (Table 1).

affected, 0/18 male offspring affected). Three of the founder mice did not breed and so inheritance could not be assessed. Included in the remaining 30 phenotypes that were not dominantly inherited were all eight retinal degeneration phenotypes. We expected to find recessive mutations of Pde6b in this screen because all MUTN animals are already heterozygous for the null allele Pde6brd1 and thus the screen acts as a specific-locus test for new mutations at this locus. The founder animals with retinal degeneration were mated to C3H. In seven cases all the progeny of these test crosses had retinal degeneration (Table 1). These seven are therefore new recessive mutations of Pde6b rather than a mutation in a modifier gene. In one case 11/22 of the progeny had retinal degeneration and the remainder had normal eyes, indicating that the abnormality seen in the founder animal was not due to a recessive mutation in the Pde6b gene (nor was it dominantly inherited because all 24 offspring of a cross to C57BL/6 had normal eyes). The remaining 22 phenotypes were classed as not dominantly inherited. Eighteen were classed as such because the mutant phenotype was not found on examination of between 20 and 30 offspring of the founder animal. In four cases the founders produced fewer than 20 offspring (19, 17, 14, 12, respectively). All these offspring were normal suggesting that the phenotypes were not heritable, although it is possible that any of them may be heritable with low penetrance. In summary we found 25 inherited phenotypes, 18 dominant and seven recessive and the recovery rate of inherited eye mutations was 0.38% (0.28% for dominant mutations). Given that this screen was focused on the visual system and that in the same mutagenesis programme the recovery rate of all detected dominant mutations was 2% (4) ours represents an excellent rate. The per locus mutation rate for Pde6b of 1.08 × 10–3 is at the high end of the range found for other loci in ENU mutagenesis experiments (7). Both these findings indicate that an efficient mutagenesis regime and an effective screen were employed.

Inheritance testing All phenodeviant mice were put into inheritance testing by mating the founder animal with C57BL/6 or, for some of the mutants with gross or anterior segment defects, with C3H/HeH (hereafter C3H). The C57BL/6 strain was used in most cases for two main reasons. Both parental strains carry recessive mutations that would confound assessment of eye phenotype when homozygous. C3H carries the retinal degeneration 1 (rd1) mutation of the retinal specific β sub-unit of cGMP phosphodiesterase gene (Pde6b). BALB/c is mutant both at the tyrosinase gene (Tyr), resulting in unpigmented retinal pigment epithelium (RPE), and at the tyrosinase related protein gene (Tyrp1). The Tyrp1 mutation, in the absence of the Tyr mutation, results in the choroid appearing less pigmented than normal. There is also evidence that this mutation predisposes mice to glaucoma (13). C57BL/6 is wild-type for all these loci. In addition, it is genetically distant from both the C3H and BALB/c inbred strains and so it is easier to find variants that can be used for linkage analysis (14). Eighteen phenotypes were inherited in a dominant fashion (Table 1). In a few cases the ratio of mutant carriers to wildtype differed significantly from the expected 1:1 ratio for a dominant mutation indicating penetrance or viability effects. One mutation, in line GENA379, appeared to be X-linked based upon an observed male lethality (13/28 female offspring

Map locations of the mutations As a first step to assign a chromosomal location to the mutants we backcrossed the dominant mutations to C57BL/6. In a few cases a backcross was made to C3H; these were mutations that conferred an anterior segment defect and the segregation of Pde6brd1 would not affect phenotyping. To assign chromosomal location we used a rapid genotyping approach in which, for each line, we pooled DNA samples from mutant backcross progeny and typed the pool using a set of 50 fluorescently tagged simple sequence length polymorphism (SSLP) markers distributed across the mouse genome (15). For those mutations that appeared to be fully penetrant, wild-type backcross animals were typed in the same way. The phenotype of some mutants suggested strong candidate genes and for these the mapping was restricted to the appropriate genomic regions. For each marker the signal ratio of the alleles was compared in order to identify likely map locations. The expected signal ratio for alleles of an unlinked marker is 1 (BALB/c+C3H): 3 C57BL/6. For linked markers this ratio becomes skewed towards 1:1 in the mutant pool and 0:1 in the wild-type pool. To confirm regional chromosomal assignments, and to refine the interval containing the mutation, all animals were genotyped individually using additional SSLP markers from that region. Analysis of the haplotypes of individual animals

Human Molecular Genetics, 2002, Vol. 11, No. 7

757

Table 1. Inheritance testing GENAa

Mutant name (symbol)

Phenotype

No. mutant offspring/total no. offspring C3Hb

C57BL/6b ND

163

microphthalmia Harwell (Mi-H)

Pale patches on coat and iris transillumination

30/61

191

iris-corneal strands (Icst)

Corneal opacity

ND

9/22

232c

lens-corneal adhesion 2 (Leca2)

Anterior cataract and central corneal opacities

ND

25/109

238d

lens-corneal adhesion 1 (Leca1)

Lens-corneal adhesion, small eyes

ND

20/48

239

waved 5 (Wa5)

Eyes open at birth, wavy coat and curly whiskers

39/68

ND

245

atypical retinal degeneration 1 (atrd1)

Retinal degeneration

84/84

246

retinal white spots (Rwhs)

White spots on the retina

ND

20/52

0/8

257

retinal arteriolar wiring (Raw)

Shiny retinal arterioles

ND

29/75

265

lens-corneal adhesion 3 (Leca3)

Lens-corneal adhesion, corneal opacity

32/80

ND

269

dilated pupils 1 (Dilp1)

Dilated pupils

ND

34/53

291c

small with vacuolar cataract (Svc)

Small-sized mouse with vacuolar cataract

ND

8/45

308c

retinal degeneration 2 Harwell (rd1-2H)

Retinal degeneration

46/46

0/20

309

atypical retinal degeneration 2 (atrd2)

Retinal degeneration

21/21

333

retinal vascular mass (Rvm)

Retinal vascular mass and optic disc coloboma

ND

335c

atypical retinal degeneration 3 (atrd3)

Retinal degeneration

67/67

336

retinal orange patches (Rorp)

Oranges patches on the retina

ND

338

retinal degeneration 1 Harwell (rd1-1H)

Retinal degeneration

20/20

358

dilated pupils 3 (Dilp3)

Dilated pupils

362

retinal degeneration 3 Harwell (rd1-3H)

Retinal degeneration

366

retinal degeneration 4 Harwell (rd1-4H)

Retinal degeneration

368

lens-corneal adhesion 4 (Leca4)

Corneal opacity, anterior cataract, small eyes

372

dilated pupils 4 (Dilp4)

Dilated pupils

6/21

ND

379

dilated pupils 2 (Dilp2)

Dilated pupils

13/46

ND

optic disc coloboma (Opdc)

Optic disc coloboma

ND

24/52

lens cloudy (Lcl)e

Cataract

11/22

11/20

380

1/21 32/32 7/7 27/50

0/29 16/52 0/15 35/73 0/22 ND 0/21 0/17 ND

ND, not done. aA GENA number was assigned when a founder mouse was put into inheritance testing (see http://www.mgu.har.mrc.ac.uk/mutabase/). bMouse strain that the founder animal was crossed to for inheritance testing. cMutants found using visual tracking drum. dIn this case the founder was the offspring of an ENU-mutagenized BALB/c male mated with a female Myo7ash1 mouse. eA GENA number was not assigned in this case. The mutation was named and the inheritance testing was carried out by Dr Mary Lyon (personal communication).

enabled each mutation to be placed within a genetic interval. Using this approach we could precisely delineate the genetic location of each of the mutations from analysis of a relatively small number of animals. The mapping information and map intervals are shown in Table 2. For seven mutants this mapping information, together with phenotype data, indicated that they might be new mutations in genes with existing alleles possessing eye phenotypes. These were all confirmed by molecular analysis or complementation tests (detailed below) and the genes affected are shown in Table 2. For the other mutants the regions of conserved synteny in the human genome are given (Table 2). Seven new alleles of Pde6b We identified seven new recessive alleles of the Pde6b gene in the mutant screen by non-complementation (Table 3). Compound heterozygotes of three of them with Pde6brd1 had a

slower onset of retinal degeneration and so were named atypical retinal degeneration 1–3 (atrd1–atrd3). With the other four mutants, compound heterozygous mice developed a very rapid onset retinal degeneration, indistinguishable from Pde6brd1 homozygotes, and these were named retinal degeneration 1–4-Harwell (rd1-1H–4H). We generated homozygotes for Pde6batrd1, Pde6batrd2 and Pde6batrd3 by intercrossing. As the Pde6brd1 mutation has a point mutation in exon 7 that results in a restriction site variant (16), it was possible to distinguish mice of all three expected genotypes (see Materials and Methods). In all cases, mice homozygous for atrd1, atrd2 or atrd3 show milder disease than compound heterozygotes on clinical and histological examination. Preliminary data using the visual tracking drum has indicated that atrd1, atrd2 and atrd3 homozygotes and compound heterozygotes show persistence of the visual tracking response until several weeks of age, in contrast to rd1

758

Human Molecular Genetics, 2002, Vol. 11, No. 7

Table 2. Mutant loci mapped by backcross Mutant name (symbol)

Chromosomal location

Marker symbola

Distance (cM)b

No. of mutant backcross mice tested Affected gene or likely (recombination fraction of BALB/c region of conserved synteny derived allele) in the human genomec

lens-corneal adhesion 1 (Leca1)

Mid Chr 2

D2Mit365

21.9

29 (0.35)

D2Mit327

40.4

21 (0.14)

D2Mit442

50.3

22 (0.0)

D2Mit395

55.7

21 (0.19)

D2Mit493

72.1

24 (0.29)

lens-corneal adhesion 2 (Leca2)

lens-corneal adhesion 3 (Leca3)

lens-corneal adhesion 4 (Leca4)

microphthalmia-Harwell (Mi-H)

retinal orange patches (Rorp)

waved 5 (Wa5)

optic disc coloboma (Opdc)

retinal white spots (Rwhs)

Mid Chr 2

Mid Chr 2

Mid Chr 2

Mid Chr 6

Mid Chr 6

Proximal Chr 11

Mid Chr 19

Mid Chr 11

D2Mit365

21.9

32 (0.34)

D2Mit327

40.4

21 (0.14)

D2Mit442

50.3

33 (0.0)

D2Mit483

52.5

21 (0.0)

D2Mit395

55.7

22 (0.09)

D2Mit224

61.2

22 (0.18)

D2Mit493

72.1

33 (0.36)

D2Mit37

44.8

23 (0.17)

D2Mit442

50.3

23 (0.13)

D2Mit483

52.5

15 (0.13)

D2Mit395

55.7

16 (0.19)

D2Mit372

27.3

19 (0.11)

D2Mit37

44.8

17 (0.0)

D2Mit442

50.3

25 (0.0)

D2Mit395

55.7

19 (0.05)

D2Mit311

83.1

21 (0.29)

D6Mit316

16.4

40 (0.05)

D6Mit230

36.1

40 (0.0)

D6Mit59

56.8

40 (0.125)

D6Mit316

16.4

21 (0.33)

D6Mit230

36.1

23 (0.0)

D6Mit59

56.8

23 (0.13)

D11Mit2

4.4

23 (0.0)

D11Mit306*

8.7

23 (0.39)

D11Mit231

14.2

23 (0.30)

D19Mit128

10.9

42 (0.33)

D19Mit46

24.0

40 (0.00)

D19Mit89

28.4

41 (0.00)

D19Mit53

32.8

41 (0.00)

D19Mit10

35.0

42 (0.00)

D19Mit8

35.0

41 (0.00)

D19Mit103

40.4

42 (0.12)

D19Mit1

43.7

41 (0.12)

D11Mit208

29.5

18 (0.11)

D11Mit242

31.7

37 (0.05)

D11Mit30

37.2

37 (0.05)

D11Mit40

42.6

37 (0.05)

D11Mit38

44.8

37 (0.08)

D11Mit289

55.7

37 (0.14)

Pax6

Pax6

Pax6

Pax6

Mitf

Mitf

Egfr

9p24 or 10q23–26

17p11–13 or 17q21–24

Human Molecular Genetics, 2002, Vol. 11, No. 7

759

Table 2. Continued. Mutant name (symbol)

Chromosomal location

Marker symbola

Distance (cM)b

No. of mutant backcross mice tested Affected gene or likely (recombination fraction of BALB/c region of conserved synteny derived allele) in the human genomec

iris-corneal strands (Icst)

Proximal Chr 2

D2Mit365

21.9

42 (0.00)

D2Mit372*

27.3

42 (0.02)

retinal arteriolar wiring (Raw)

Proximal Chr 8

D2Mit327

40.4

41 (0.15)

D8Mit155

0.0

39 (0.10)

D8Mit124

4.4

39 (0.05)

D8Mit289

8.7

39 (0.17)

12.0

39 (0.15)

D8Mit155

0.0

23 (0.08)

D8Mit124

4.4

23 (0.04)

12.0

23 (0.22)

D8Mit4 small with vacuolar cataract (Svc) Proximal Chr 8

D8Mit4

retinal vascular mass (Rvm)

dilated pupils 1 (Dilp1)

dilated pupils 2 (Dilp2)

Mid Chr 14

Mid Chr 5

Mid X Chr

D8Mit312

44.8

23 (0.26)

D14Mit133

16.4

43 (0.12)

D14Mit174

17.5

43 (0.09)

D14Mit62

19.0

43 (0.19)

D14Mit102

37.2

43 (0.37)

D5Mit387

10.9

46 (0.22)

D5Mit81

18.6

46 (0.07)

D5Mit301

25.1

46 (0.04)

D5Mit303

26.2

45 (0.04)

D5Mit197

26.2

45 (0.04)

D5Mit15

26.2

45 (0.04)

D5Mit356

27.3

45 (0.04)

D5Mit95

57.9

46 (0.17)

DXMit55

1.1

11 (0.27)

DXMit81

12

11 (0.27)

DXMit38

43.7

11 (0.09)

DXMit174

43.7

11 (0.09)

DXMit67

48.1

11 (0.10)

DXMit121

53.6

11 (0.27)

9q33–34

19q13 or 13q32–34

19q13 or 13q32–34

3p14–21 or 10q11–23 or 14q11–21

4p11–15

Xp21–22 or Xq13–28

aA marker shown in bold defines a boundary of the region containing a mutation deduced from examining individual haplotypes, including those of wild-type backcross progeny when informative. For two mutations only the distal boundary has been defined and this is denoted by an asterisk. bDistance (cM) from the centromere on the MIT F2 intercross map (http://www-genome.wi.mit.edu). cWhen shown the identity of the gene affected is known either from molecular analysis or from the result of an allelism test.

homozygous mice which have no tracking response at any age tested (data not shown). Four new alleles of Pax6 Four mutants that we named lens-corneal adhesion 1–4 (Leca1–Leca4) had a range of similar but variable eye phenotypes, with slightly smaller eyes and occasional adhesions between the lens and cornea (Fig. 1). Similar phenotypes have been observed in mice that are haploinsufficient for the paired-homeobox transcription factor Pax6 (17) and these four mutations map close to Pax6 on chromosome

(Chr) 2 (Table 2). Sequencing of the Pax6 gene in Leca1, Leca2, Leca3 and Leca4 revealed single base substitutions that cause mutational changes compared to the sequence from the parental BALB/c strain (Table 4). In each case the sequence change found was confirmed by sequencing at least five mutant animals. Pax6 Leca1, Pax6Leca2 and Pax6Leca4 are missense mutations causing amino acid substitutions in either the homeobox or paired box domain of the protein and Pax6 Leca3 introduces a premature stop codon in the proline-serine-theronine (PST)-rich domain that truncates the Pax6 protein by 69 amino acids (Table 4).

760

Human Molecular Genetics, 2002, Vol. 11, No. 7

Table 3. Mutations mapped by allelism Mutant name (symbol)

Map position

Allele

atypical retinal degeneration 1 (atrd1)

Chr 5

Pde6b atrd1

atypical retinal degeneration 2 (atrd2)

Chr 5

Pde6b atrd2

atypical retinal degeneration 3 (atrd3)

Chr 5

Pde6b atrd3

retinal degeneration 1 Harwell (rd1-1H)

Chr 5

Pde6b rd1-1H

retinal degeneration 2 Harwell (rd1-2H)

Chr 5

Pde6b rd1-2H

retinal degeneration 3 Harwell (rd1-3H)

Chr 5

Pde6b rd1-3H

retinal degeneration 4 Harwell (rd1-4H)

Chr 5

Pde6b rd1-1H

retinal orange patches (Rorp)

Chr 6

MitfRorp

waved 5 (Wa5)

Chr 11

Egfr Wa5

Two new mutations of Mitf Two mutants map in the region of Mitf gene on Chr 6 and resemble known mutants at this locus (18). The first, named microphthalmia Harwell (Mi-H) was identified as a coat colour variant. Heterozygotes have pale patches on their coats, usually a head spot and a belly spot as well as pale ears, feet and tails. Eye examination showed mild iris transillumination, but no other eye abnormality. Homozygotes were generated and showed a more severe phenotype. They were completely unpigmented and lacked eyes. The eyelids were closed and histology showed there was a small amount of eye tissue in the orbit but nothing resembling intact or even misshapen eyes. RT–PCR experiments showed that an abnormal Mitf transcript is produced in MitfMi–H homozygotes indicating that the mutation causes a splicing defect in the Mitf gene (data not shown). A second mutant, named retinal orange patches (Rorp), maps to the same region of Chr 6 (Table 2). This was identified by fundoscopy and has a different phenotype to MitfMi–H. Although the founder has a slightly paler coat, ears and tail, the striking feature of this mutant was the appearance of orange patches on the retina. However, histology of the eye was normal. It may be that since the mutation causes dilute pigment, the orange appearance is due to a difference in light reflex on fundoscopy rather than an anatomical abnormality. Rorp homozygotes lack coat pigment and are white. However, they have pigmented irides although the pupils are often dilated. On fundoscopy, the optic discs are often colobomatous and the retina has reduced pigment with scattered areas of pigmentation (not shown). Compound heterozygotes of Rorp and MitfMi–H have white coats and the eyes have dilated pupils, brown irides, marked iris transillumination, depigmented fundi (albinoid) and optic disc colobomas. The two mutants do not complement and hence Rorp is a new mild allele of Mitf (MitfRorp). Epidermal growth factor receptor (Egfr) Mice heterozygous for the waved 5 (Wa5) mutation are born with open eyelids and curly whiskers. The first coat is wavy and subsequent coats are scruffy in appearance. Mutant neonates were examined and the eyes are present within the orbit and appear normal. A few days following birth the eyelids close, although not fully in some cases, and keratinized tissue fills the interpalpebral aperture. At weaning the eyes of

Figure 1. Phenotypes of potential Pax6 mutants. Representative photographs are shown for each phenotype Leca1–4. (A) Leca1. This mutant has a normal-sized eye with a central corneal dimple (arrowed) and an irregular pupil. (B) Leca2. In this example there is central corneal and lens opacity. There are iris strands leading from the pupillary margin to the central cornea and/or lens. (C) Leca3. Corneal opacity is present and a dimple (arrowed) is present where the lens and cornea have not separated or have separated at the wrong time. (D) Leca4. The pupil is normal but a cataract can be seen (arrowed).

Table 4. Molecular characterization of the Pax6 gene mutations Mutant allele

Mutationa

Exon

Codona

Pax6Leca1

T971A

10

Val270Glu

Homeobox

Pax6Leca2

C586T

7

Arg142Cys

Paired box

Pax6Leca3

C1266A

12

Pax6Leca4

C354A

6

Domain

Tyr368Stop

PST

Asn64Lys

Paired box

aThe numbering of nucleotides and codons follows that in GenBank entry NM_013627.

some mutants remain closed. In others the eyes open again but are often small with marked corneal scarring. However, after rederiving the strain into a specific pathogen free facility, mice born with open eyelids generally had normal-sized eyes at weaning. This suggests that the open eyelids at birth expose the eyes to environmental damage such as infection or dust. Anomalies seen at later stages are therefore unlikely to be due to defects in eye development. This dominant phenotype is similar to the recessive phenotypes waved 1 and waved 2 which are mutations in the transforming growth factor alpha (Tgfa) gene and Egfr gene, respectively (19–21). We tested markers in the vicinity of these two loci for genetic linkage to Wa5 and found that it is located at or close to the Egfr gene. Wa5 fails to complement a null allele of Egfr (Egfrtm1Mag) indicating that it is a new mutant allele of Egfr (D.Threadgill, personal communication). In intercross litters of heterozygous Wa5 mice only wild-type (n = 32) and heterozygous mutant (n = 48) offspring were found showing that Wa5 is homozygous lethal.

Human Molecular Genetics, 2002, Vol. 11, No. 7

761

Figure 2. Clinical phenotypes of mutant lines. Representative photographs are shown for each phenotype. (A) Normal retina. (B) Opdc. The optic disc is enlarged and excavated. (C) Rwhs. Scattered white spots are present. (D) Rwhs histology. Three invaginations of the outer nuclear layer can be seen (arrows). (E) Icst. Corneal haze is present. There is also ectropion uveae and a distorted pupil. (F) Raw. The retinal arterioles give a ‘shiny’ reflex not seen in normal mice. (G) Svc. Numerous vacuoles are present in the lens. (H) Rvm. A large blood vessel is present posterior to the lens. (I) Dilp1. The pupil is irregular and moderately dilated. The Dilp1 phenotype shown here is representative for that seen for Dilp2, Dilp3 and Dilp4.

Eye mutants with no assigned gene The gene affected in the remaining nine mutants has not yet been identified and typical examples of the phenotypes for these are shown in Figure 2. In most cases they do not map close to any strong candidate gene and therefore represent novel loci. Optic disc coloboma (Opdc) mutation The Opdc mutation results in mice with bilateral colobomas of the optic disc and it appears to be fully penetrant (Table 1), although the size of the colobomas varies (Fig. 2B). Opdc maps to Chr 19 between the markers D19Mit46 and D19Mit103 (Table 2). One gene within this interval, Pax2, is a strong candidate for the site of the Opdc mutation. A frameshift mutation in Pax2 (Pax21Neu) has been described for which the heterozygous mutant mice have retinal abnormalities including optic disc dysplasia (22). The similarity of the retinal

phenotype seen in Opdc heterozygotes to that of these mice suggests that Opdc may be a new allele of Pax2. In support of this, the phenotype of Krd/+ mice is attributed to the loss of one copy of Pax2 (23). These mice carry a 7 cM deletion of Chr 19 that is contained within the minimal interval we define for Opdc. There are many other genes within the deletion but haploinsufficiency for these appears to have little impact on the phenotype and it is unlikely that Opdc is due to a mutation in any of these. Retinal white spots (Rwhs) In mice heterozygous for the Rwhs mutation the retina is covered with a number of pale spots (Fig. 2C). The phenotype is first detectable between the ages of 6 and 12 weeks, and the number of spots usually increases with time. The retina in mutant mice has discrete, focal invaginations of the photoreceptors and outer nuclear layer that impinge on the more superficial retinal layers (Fig. 2D). These invaginations are

762

Human Molecular Genetics, 2002, Vol. 11, No. 7

presumed to coincide with the positions of the white spots because of their number and position. There is no abnormality of the RPE or choroid. Linkage analysis showed that Rwhs is located between D11Mit40 and D11Mit38 on Chr 11 (Table 2). One potential candidate gene within this interval was investigated, retinal gene 4 (Rtg4). We considered this to be a strong candidate for two reasons. Firstly, expression studies have shown that Rtg4 and the orthologous rat and human genes are highly expressed in photoreceptors and at low levels in other tissues (24,25). Secondly, a patient with late-onset cone–rod dystrophy has been described with a mutation that introduces a stop codon in one copy of the orthologous human gene and transgenic mice expressing the truncated protein from a rhodopsin promoter exhibit a range of retinal abnormalities, including varying numbers of retinal white spots and streaks that are reminiscent of the Rwhs phenotype (26). However, no mutations were found on sequencing the exons of Rtg4 in Rwhs mice (data not shown). Fewer than expected mutants were found in breeding experiments (95/335 of the mice examined were mutants) and four out of 22 phenotypically normal mice appeared to be carriers by genotyping. This suggests that the Rwhs phenotype is either not fully penetrant and/or the mutant phenotype can be late-onset. Homozygotes for Rwhs have been produced by intercrossing and identified by homozygosity for BALB/c alleles at loci flanking the mutation. The phenotype of these mice is indistinguishable from that of the heterozygotes. It therefore seems likely that this mutation is a true dominant, rather than semidominant, although it is possible that a more severe homozygous phenotype may develop with time. Iris-corneal strands (Icst) Icst is an intriguing mutation with a highly variable phenotype. A mutant phenotype was present in 111 out of 378 mice examined, implying reduced penetrance. A variety of the following eye defects were observed: irregular pupils (53%), corneal haze (51%), iris-corneal adhesions (32%), corneal opacity and scarring and iris abnormalities such as membranes, bands or ectropion uveae. An example is shown in Figure 2E. Some eyes showed corneal vascularization and calcification. Older mice often developed large eyes, possibly because of secondary glaucoma. The clinical phenotype appears to be due to an abnormal proliferation of the corneal endothelial cells. The Icst mutation maps to proximal Chr 2 above 27 cM and is likely to be within 7 cM of the SSLP marker D2Mit365 at 21.9 cM (no recombinants out of 42 Icst mutants tested, 95% upper confidence limit = 6.8 cM) (Table 2). There are no obvious candidate genes in this region. Retinal arterial wiring (Raw) and small with vacuolar cataracts (Svc) Two mutants, Raw and Svc, have overlapping phenotypes and map to the same interval on proximal Chr 8 (Table 2) raising the possibility that they are allelic. The abnormality found in the Raw founder was a ‘shiny’ reflex from the retinal arterioles that was reminiscent of silver-wiring in human patients with dyslipidaemias (Fig. 2F). In subsequent generations this appearance was apparent on phenotyping at the age of 2–3 months and did not appear to progress with time. In addition a few Raw mice were bruised at birth. Svc mice had a variety of

eye abnormalities and most were small and bruised at birth. The most consistent eye defect observed was cataract and the commonest type gave a vacuolar appearance to the lens although some mutants showed anterior subcapsular cataracts instead (Fig. 2G). Other abnormalities observed include enlarged eyes (possibly secondary to glaucoma), corneal opacity and ectasia, peripheral iris-corneal adhesion and irregular pupils. The retina was often difficult to see due to media opacity, but in some cases there was a shiny appearance of the retinal arterioles similar to that seen in the Raw mice described above. In a few mice there appeared to be a fibrovascular tuft in the vitreous. The phenotype of Svc is similar to that described for bruised (Bru) mice which carry a small deletion of band A1.3 of Chr 8 [Del(8)44H)] (27,28). Del(8)44H is located within the minimal interval defined by genetic linkage analysis for Raw and Svc and it seems likely that the gene mutated in these lines falls within this deletion. It is possible that haploinsufficiency for an as yet uncharacterized gene leads to the phenotypes observed in Bru and Svc, and Raw may be a partial loss of function of the same gene. Retinal vascular mass (Rvm) The Rvm mutation has a phenotype that is variable and shows poor penetrance. Some mutants have optic disc coloboma in one or both eyes, while others show an abnormal growth of the retinal vasculature into the vitreous with exudation. An example is shown in Figure 2H. Rvm maps to an interval flanked by D14Mit174 and D14Mit62 on proximal Chr 14, which the MIT map defines as only 1.5 cM, but other crosses, including our own, suggest is more likely to be ∼8 cM (Table 2). Rvm represents a novel locus. Lens cloudy (Lcl) Lcl is dominantly inherited and heterozygotes have an early onset lens clouding which progresses with time to a white cataract. Homozygotes have smaller eyes than normal and a white, nonprogressive cataract (M.Lyon, personal communication). Four mutations cause a dilated pupils phenotype Four of the mutants found have a very similar bilateral dilated pupils phenotype (shown in Fig. 2I) and were named dilated pupils 1–4 (Dilp1–Dilp4). Other than having dilated pupils the mice appear normal. However, fewer than expected Dilp1 heterozygotes were found in the backcross (59/155 backcross mice were classed as mutants). Genotyping of the phenotypically wild-type mice showed that none had a heterozygous genotype, excluding reduced penetrance and implying that some of the heterozygotes die before weaning. Genotyping of litters produced from intercrosses found no homozygous Dilp1 progeny indicating homozygous lethality. The X-linked Dilp2 mutation is male-lethal. Dilp1 is located between D5Mit15 and D5Mit356 on Chr 5 (Table 2). Dilp2 is located between DXMit81 and DXMit38 on the X chromosome (Table 2). These are both novel loci and represent new models for iris dysgenesis diseases in humans. At present we do not have a genetic location for Dilp3 and Dilp4 although genotyping of individual Dilp4 animals with markers flanking Dilp1 has shown that Dilp4 is not allelic to Dilp1. Similar analysis with appropriate flanking markers

Human Molecular Genetics, 2002, Vol. 11, No. 7

showed that Dilp4 is also not linked to Foxc1, Foxc2 and Pitx2. These were considered as candidate genes because a dilated pupils phenotype has been observed in mice that are heterozygous for null alleles of these genes (29,30). DISCUSSION This mutagenesis screen has generated 25 inherited phenotypes. We have found a number of new alleles of genes known to be important in the eye and we have also defined a number of novel loci. Both classes of mutation provide useful models of human disease, and are useful tools to dissect eye development and visual function. Scope and effectiveness of the screen for eye and vision mutations Did we find mutations in all the genes that might be expected in this screen? For a number of genes it is known that haploinsufficiency results in eye phenotypes and if the screen was effective and saturated we might expect to find examples for all of these. We did find more than one mutation of two such genes, Pax6 and Mitf, suggesting that the screen was effective. However, we did not find mutations that map close to Foxc1, Foxc2 or Pitx2, all of which have been reported to exhibit a haploinsufficient eye phenotype, although this varies according to genetic background (29,30). We did not find any mutations that gave a phenotype only on a Pde6brd1 heterozygous background, and all the retinal degeneration mutants that we found were recessive alleles of Pde6b. We found no dominant retinal degeneration phenotypes. Retinitis pigmentosa (RP), a disease that causes retinal degeneration in humans, can be inherited in an autosomal dominant fashion and to date eight causative genes have been identified for this form of the disease (31). Most of these are not, however, loss of function mutations. Mutations in rhodopsin, for example, account for ∼30% of cases where the mutated gene has been identified, but these are missense mutations in crucial amino acids (32). In contrast null mutations of rhodopsin appear to cause recessive RP (33). Mice heterozygous for null alleles of rhodopsin or for any of three other genes known to cause autosomal dominant RP, Prph2, Crx and Nrl, show no evidence of photoreceptor death in young mice at least (34–37). The chance of detecting specific missense mutations in particular genes is obviously much less than that of finding inactivating mutations, and the absence of dominant retinal degeneration phenotypes in this screen is readily explained. It is interesting to compare the recovery rate and spectrum of inherited phenotypes that we found with those reported in an earlier screen for eye mutations in the mouse. In a study by Favor and Neuhäuser-Klaus (38), which summarizes the results of several screens for eye mutants induced either by ENU or by irradiation, 203 dominantly inherited mutants were found in 456 890 potential mutants screened, a recovery rate of 0.04%. The majority of phenotypes identified were related to eye opacity. The recovery rate of dominantly inherited phenotypes in our screen was far higher at 0.28% and the phenotypes identified affected all parts of the eye. The major difference between the two studies is that we looked for retinal phenotypes using an indirect ophthalmoscope as well as for anterior segment phenotypes using a slit-lamp biomicroscope. By

763

combining both inspection methods we were able to detect a larger variety of eye phenotypes. At a rate of 0.04% we would have expected to find two to three mutants detectable by slitlamp examination. In fact we found 12. This is an indication of the efficacy of the mutagenesis protocol used, which is also attested to by the satisfactory per locus mutation rate we found for the Pde6b locus of 1.08 × 10–3. Nevertheless, we did not detect many cataracts even though cataract mutations represent one of the most abundant classes of known eye mutations and were the largest class found by Favor and Neuhäuser-Klaus (38). This is probably because most of the mice in the current screen did not undergo slit-lamp examination after dilation, with the operator using only indirect ophthalmoscopy. While it is possible to identify cataracts with an indirect ophthalmoscope, small or subtle cataracts are often missed. However, the intention of this project was to identify new mutations and many cataract mutations have already been described in the literature (38–40). We have not thoroughly screened for phenotypes that develop only with age. A proportion of the mice were aged before phenotypic examination, and indeed one of them (Raw) was found to have an inherited eye mutation, but it was not possible to age all of the mice from the programme. Other phenotypes may have been missed because they were difficult to detect or subtle. However, more comprehensive examination methods (such as electrophysiology on all of the mice screened) would have added greatly to the time spent on each mouse and reduced the numbers screened. Further methods to increase the types of eye abnormalities detected, such as using a non-invasive method for measuring intraocular pressure to detect glaucoma, could be incorporated into future mutagenesis screens. An allelic series of mutant Pde6b alleles The allelic series of Pde6b mutants reported here will be of particular value. Mutations in the corresponding human gene have been implicated in cases of autosomal recessive RP (41,42) and so these mice are an appropriate model for the evaluation of the effectiveness and safety of novel therapeutic strategies for the treatment of human retinal degeneration diseases. Such strategies include pharmacological treatments, cell transplantation and gene therapy. A drawback of using the original mutant, Pde6brd1, for this purpose is that the rod photoreceptors have almost completely degenerated by 17 days (43). Three of the new mutants reported here, Pde6batrd1–3, cause a slower onset of degeneration and thus could provide a longer time window for testing of therapies. In addition, characterization of the underlying genetic lesion in these mutants will provide insights into Pde6b protein function. Novel Pax6 mutations that link phenotype to genotype The four Pax6 mutants reported here extend the allelic series for this gene. The three missense mutations are of particular interest because all except two of the existing alleles are protein null or give rise to a truncated protein product. The phenotypic characteristics of these null and truncation alleles are very similar and similar to the phenotypes of Pax6Leca1–4, although extensive phenotypic characterization of the heterozygous and homozygous phenotypes has not yet been carried out.

764

Human Molecular Genetics, 2002, Vol. 11, No. 7

Pax6Leca3 causes a premature termination in the PST domain and the loss of 69 amino acids. The PST domain is a transactivation domain that mediates interactions between Pax6 and other proteins important for transcription (44). A nonsense mutation at the same position in the protein has been found in a human aniridia patient (45). The missense mutation Pax6Leca1 is located in the homeodomain and causes a valine to glutamic acid substitution at position seven of helix III. Helix III is the recognition helix, which inserts into the major groove of the DNA. The affected valine is highly conserved in the paired class of homeodomains and it is a critical amino acid for DNA recognition as it interacts with a thymidine base in the DNA binding site (46,47). The change in Pax6Leca1 substitutes the acidic glutamic acid for a hydrophobic valine and would be expected to disrupt binding to the DNA recognition sequence. Missense mutations in the PAX6 homeodomain are rare in patients with aniridia. Only one has been found but it is thought that this also causes a splicing defect and protein truncation (48). Recently a patient with unilateral partial aniridia was found to have a missense mutation that substituted threonine for a conserved arginine in the second helix of the homeodomain (49). This is clearly a partially penetrant mutation as the patient’s mother carried the same mutation, but was phenotypically normal. A single mouse missense mutation has been found in the homeodomain, a serine to proline substitution, also in helix III, and this too appears to be a hypomorphic mutation (50). Pax6Leca1 appears to be the only example of a fully penetrant loss-of-function missense mutation in the homeodomain. The other two missense mutations are in the paired box domain. In Pax6Leca2 cysteine is substituted for a highly conserved arginine in the 6th α-helix in the C-terminal subdomain of the paired box that fits directly into the major groove. This arginine is important for DNA binding because it contacts the methyl group of thymidine 19 of the recognition sequence and it also contacts a phosphate in the DNA backbone (51). Interestingly an identical mutation has been found in independent human patients with isolated foveal hypoplasia (52; K.Williamson and V.van Heyningen, personal communication). These patients have a mild phenotype and no anterior segment defects whereas the mouse mutation appears to cause a phenotype similar to loss-of-function alleles (Fig. 1). It seems that the severity of the phenotype caused by this mutation is greater in the mouse than the human. It is also noteworthy that Pax6Leca2 differs from Leca1, Leca3 and Leca4 in that it is incompletely penetrant (Table 1). The final mutation Pax6Leca4 leads to the substitution of lysine for an asparagine that is the first residue of the recognition helix (helix 3) of the N-terminal subdomain of the paired domain. This asparagine recognizes an AT base pair by making van der Waals contacts with the methyl group of thymidine 4 of the DNA recognition sequence and it also makes a water-mediated contact with the phosphate of thymidine 2 (51). No human missense mutation has been reported at this position but the importance of this asparagine for DNA binding shown by the crystal structure suggests that replacement of the asparagine by a lysine would severely disrupt DNA binding. It is worth noting that using ENU we have found three new Pax6 missense mutations, the functional significance of which

is supported either by the existence of human patients with identical mutations and/or protein structure information. Prior to the work reported here, there was only one example among the numerous mutant Pax6 alleles known. This finding amply demonstrates the value of ENU as a mutagenic agent for generating subtle mutations in genes that have functional significance. Novel Egfr mutant allele Wa5 is the first dominant mutation found in the Egfr gene. The existing recessive viable mutation, Egfrwa2, is a point mutation in the kinase domain of Egfr that compromises, but does not abolish, receptor activity (21). Gene targeting has generated mice deficient for Egfr and heterozygous animals are normal. The phenotype of homozygotes varies according to genetic background. On some genetic backgrounds homozygotes die in utero, whereas on others the mice are born with open eyes and survive for up to 3 weeks (53,54). Heterozygous null mice are normal. The phenotype of EgfrWa5 heterozygotes most closely resembles that of mice homozygous for the partial lossof-function mutation Egfrwa2. We suggest that the underlying genetic change in EgfrWa5 causes an alteration in the Egfr protein resulting in a dominant-negative effect. This is the first such mutation described for Egfr and will be useful for elucidating receptor function. Dilated pupils phenotypes Four novel mutants, Dilp1–Dilp4, have very similar phenotypes, including dilated pupils or iris dysgenesis, and at least three map to different parts of the genome. Dilp1 is homozygous lethal (and heterozygous semi-lethal) and Dilp2 is an X-linked male lethal. There are several human diseases with phenotypes apparently similar to these mice. Human patients with iridogoniodysgenesis anomaly, Axenfeld–Rieger anomaly (ARA) and a spectrum of glaucoma phenotypes are heterozygous for mutations in the gene encoding the transcription factor FOXC1 (55,56). The related Rieger syndrome can be caused by heterozygosity for mutations in PITX2, another transcription factor gene (57). A dilated pupil phenotype similar to that which we describe here has been reported in mutant mice heterozygous for null mutations of Foxc1 and Pitx2 (orthologues of the human genes described above) and also in Foxc2. These genes encode transcription factors that are important in development, and homozygotes for the mutations die in utero or perinatally (29,30). Although no disease-associated changes were found in the human FOXC2 gene in families with ARA (29), mutations in this gene have been found in lymphedema-distichiasis syndrome (58). The map locations of Dilp1, Dilp2 and Dilp4 indicate that they are not mutations in these previously identified genes. It is possible that iris development is particularly sensitive to dose changes for developmental genes such as transcription factors and that this phenotype is an easily detectable indicator for haploinsufficiency of such factors. The Dilp series of mutations may be a means of identifying important developmental genes. Furthermore, there are still a number of iris dysplasia or hypoplasia diseases in humans whose underlying gene has not yet been discovered, and these mice may aid in their identification.

Human Molecular Genetics, 2002, Vol. 11, No. 7

Models for human eye disease Some of the remaining mouse mutants we describe here are also models for additional human diseases. One of them, Icst, may represent a genetic model of the human disease iridocorneal endothelial (ICE) syndrome that displays the same clinical and histological features. However, there are very few reports of ICE syndrome being familial or bilateral (59–61). Human studies have identified no definite causative factor and Icst may be of value both in the study of ICE syndrome and potentially in therapeutic studies of secondary glaucoma. The Rvm mutation shows poor penetrance and variability of the phenotype. Several human genetic diseases that include neovascular inflammatory vitreoretinopathy have been described (62). The Rvm mutant may be a useful model for these, but also for more common proliferative retinopathies due to factors such as diabetes or ischaemia. Overall, we believe that the series of mouse mutants described here illustrate the power of ENU mutagenesis, when coupled with an effective screen, to provide valuable models of human disease, and to aid in the identification of underlying genetic and molecular defects.

765

was no response to the 2° stripe width, irrespective of whether the mouse responded to either the 4° or 8° widths, the mouse was tested again on a subsequent day. If no response was seen using the 2° stripe width again it was tested for a third time on a different day. If it still did not respond to the 2° stripe it was classified as having substandard vision. The eyes of all mice that did not respond to the 2° stripe on a first test were examined clinically. Clinical examinations

MATERIALS AND METHODS

Most mice were examined at 6–8 weeks of age and all of the mice in the ageing cohort were examined when they reached 6 months of age (females) or 1 year (males). For anterior segment examination and photography, a Nikon FS-3V zoom slit-lamp biomicroscope was used with an attached Kodak DCS420 digital still camera and digital images were saved using Adobe Photoshop 4.0 (Adobe, Inc.). For posterior segment examination, a Heine Sigma 150 spectacle indirect ophthalmoscope with 30D condensing lens was used. Pupil dilatation was achieved using G. Tropicamide 1% Minims (Chauvin Pharmaceuticals). Retinal photography was performed using a Kowa Genesis dedicated handheld retinal camera with a 30D condensing lens.

Animals

Histology

The animal studies described in this paper were carried out under the guidance issued by the Medical Research Council in Responsibility in the Use of Animals for Medical Research (July 1993) and Home Office Project Licence nos 30/1517 and PPL60/2242. Details of the mutagenesis programme are described elsewhere (4).

Mice were killed by cervical dislocation, and eyes removed using fine curved forceps and immediately placed in 10% neutral buffered formalin for 24–48 h followed by incubation in Davidson’s fixative for a further 18–24 h. They were dehydrated by immersion in a series of increasing concentrations of alcohol, embedded in paraffin wax and sectioned. Five micrometre thick sections were stained with haematoxylin and eosin.

Vision testing Mice were vision tested at 6–8 weeks of age, and mice in an ageing cohort were tested when they reached 6 months of age (females) or 1 year (males). A visual tracking drum was used employing a protocol that had been developed to differentiate between mice with normal vision and mice with severely reduced vision owing to retinal degeneration (63). In tests of this protocol, 97% of mice with normal vision responded and all mice with retinal degeneration failed to respond (63). The apparatus used was a scaled-down version of one developed for use with rats (12). The drum is a cylinder 55 cm high by 30 cm in diameter mounted on a motorized base. The drum is lined with stimulus panels consisting of black and white vertical stripes. Stripe frequencies used were 0.25 cycles/ degree (subtending an angle of 2° when viewed from the centre of the drum), 0.125 cycles/degree (4°) or 0.0625 cycles/degree (8°). A mouse is placed on a stationary circular platform 8 cm in diameter in the centre of the drum and 20 cm from the base and allowed to settle for 30 s. The drum was rotated anticlockwise for 30 s, stopped for 15 s and then rotated clockwise for 30 s. Rotation speed was 2 revolutions/min (12°/s). During the rotations the mouse was observed for any head tracking response whereupon it was classified as being able to see and the test was stopped. The 2° stripe width was used first. If the mouse did not respond to this the test was repeated using a 4° stripe width after a rest period of 30 s. If there was no response again the test was repeated using an 8° stripe width. If there

Linkage analysis For linkage analysis of inherited phenotypes genomic DNA was prepared from 1 cm tail snips using Qiagen DNEasy Kits. DNA concentrations were measured using a Hoefer DNA Quant 200 fluorometer as recommended by the manufacturer and each sample was diluted to 50 ng/µl in distilled water. Equimolar amounts of DNA from 37 to 50 mutant backcross mice for each line were pooled. For lines where the phenotype was fully penetrant the same number of wild-type backcross mice were pooled. The DNA pools and DNA from BALB/c, C3H, C57BL/6, MUTN mice, 1:1 and 1:3 mixes of BALB/c:C3H (if a backcross to C3H) or MUTN:C57BL/6 were screened with 50 microsatellite markers spaced at ∼20 cM intervals in the mouse genome that were polymorphic between C57BL/6 and the other two strains. In some cases the allele sizes were different in all three strains. The markers used were Mit markers, and the sequence was obtained from the Whitehead Institute for Biomedical Research/MIT Center for Genome Research (http://www-genome.wi.mit.edu/). The SSLP marker set used was: D1Mit373, D1Mit215, D1Mit285, D1Mit403, D2Mit365, D2Mit442, D2Mit493, D3Mit164, D3Mit49, D3Mit86, D4Mit286, D4Mit27, D4Mit124, D5Mit81, D5Mit259, D5Mit95, D6Mit316, D6Mit230, D6Mit59, D7Mit228, D7Mit238, D8Mit259, D8Mit312, D8Mit200, D8Mit156, D9Mit96, D9Mit11, D10Mit206, D10Mit223,

766

Human Molecular Genetics, 2002, Vol. 11, No. 7

D10Mit10, D11Mit20, D11Mit41, D11Mit104, D12Mit153, D12Mit14, D12Mit263, D13Mit253, D13Mit76, D14Mit62, D14Mit228, D15Mit179, D15Mit70, D16Mit58, D16Mit189, D17Mit202, D17Mit185, D18Mit149, D18Mit8, D19Mit46, D19Mit1. Primers were obtained from Genosys or MWG-Biotech with one primer in each pair being tagged with a fluorescent marker. PCR products were analysed using an ABI 310 Genetic Analyzer (Applied Biosystems) and the results analysed using ABI Genescan (Applied Biosystems). Using Genescan, the peak heights (fluorescent signal) were measured for each allele amplified by each of the markers. The ratios of C57BL/6 peak heights to MUTN (BALB/c plus C3H if the two alleles were separate) peak heights were calculated. To correct for any preferential amplification of the allele from one strain the calculated ratios were divided by the C57BL/6:MUTN peak height ratios obtained from the 3:1 control mix. Assay for the rd1 allele In rd1 there is a C→A transversion in exon 7 of the Pde6b gene that destroys a SnaBI restriction site and creates a DdeI restriction site (14). To distinguish between mice homozygous, heterozygous and not carrying the rd1 allele a 277 bp fragment containing exon 7 of Pde6b was amplified from genomic DNA by PCR using the following flanking primers, 5′-ACCTGAGCTCACAGAAAGGC-3′ and 5′-GCTTCTAGCTGGGCAAAGTG-3′, and tested for digestion with the two restriction enzymes. Mutational analysis of the Pax6 gene Exons 1–13 and the immediate flanking sequences of the Pax6 gene were amplified from Leca1, Leca2, Leca3, Leca4, BALB/c, C3H and C57BL/6 genomic DNA using intronic primers that were also used for subsequent sequence analysis. Exon 13 was not amplified from Leca1, Leca2 and Leca3. PCR products were purified using Millipore Multi-screen PCR 96-well filtration system on a Biomek 2000 robotic platform and sequenced directly using Big DyeTM terminator cycle sequencing. Sequences were analysed using the SequencerTM program. ACKNOWLEDGEMENTS We would like to thank Nick Hastie for support, encouragement and stimulating discussions throughout this project and for critical comments on the manuscript. We thank Veronica van Heyningen for useful discussions, Andrew Carrothers for discussion of statistics and Sandy Bruce for preparing the figures. We thank Mary Lyon, David Threadgill, Veronica van Heyningen and Kathy Willliamson for communicating unpublished results. This work was supported by a grant from the Medical Research Council UK. REFERENCES 1. Lander,E.S., Linton,L.M., Birren,B., Nusbaum,C., Zody,M.C., Baldwin,J., Devon,K., Dewar,K., Doyle,M., FitzHugh,W. et al. (2001) Initial sequencing and analysis of the human genome. International Human Genome Sequencing Consortium. Nature, 409, 860–921. 2. Venter,J.C., Adams,M.D., Myers,E.W., Li,P.W., Mural,R.J., Sutton,G.G., Smith,H.O., Yandell,M., Evans,C.A., Holt,R.A. et al. (2001) The sequence of the human genome. Science, 291, 1304–1351.

3. Hogenesch,J.B., Ching,K.A., Batalov,S., Su,A.I., Walker,J.R., Zhou,Y., Kay,S.A., Schultz,P.G. and Cooke,M.P. (2001) A comparison of the Celera and Ensembl predicted gene sets reveals little overlap in novel genes. Cell, 106, 413–415. 4. Nolan,P.M., Peters,J., Strivens,M., Rogers,D., Hagan,J., Spurr,N., Gray,I.C., Vizor,L., Brooker,D., Whitehill,E. et al. (2000) A systematic, genome-wide, phenotype-driven mutagenesis programme for gene function studies in the mouse. Nat. Genet., 25, 440–443. 5. Hrabe de Angelis,M.H., Flaswinkel,H., Fuchs,H., Rathkolb,B., Soewarto,D., Marschall,S., Heffner,S., Pargent,W., Wuensch,K., Jung,M. et al. (2000) Genome-wide, large-scale production of mutant mice by ENU mutagenesis. Nat. Genet., 25, 444–447. 6. Nadeau,J.H., Balling,R., Barsh,G., Beier,D., Brown,S.D.M., Bucan,M., Camper,S., Carlson,G., Copeland,N., Eppig,J. et al. (2001) Functional annotation of mouse genome sequences. Science, 291, 1251–1255. 7. Justice,M.J., Noveroske,J.K., Weber,J.S., Zheng,B. and Bradley,A. (1999) Mouse ENU mutagenesis. Hum. Mol. Genet., 8, 1955–1963. 8. Brown,S.D.M. and Balling,R. (2001) Systematic approaches to mouse mutagenesis. Curr. Opin. Genet. Dev., 11, 268–273. 9. Taylor,H.R. (2001) World blindness: a 21st century perspective. Br. J. Ophthalmol., 85, 261–266. 10. Gilbert,C., Rahi,J., Eckstein,M. and Foster,A. (1995) Hereditary disease as a cause of childhood blindness: regional variation. Results of blind school studies undertaken in countries of Latin America, Asia and Africa. Ophthalmic Genet., 16, 1–10. 11. Cowey,A. and Franzini,C. (1979) The retinal origin of uncrossed optic nerve fibres in rats and their role in visual discrimination. Exp. Brain Res., 35, 443–455. 12. Lawrence,J.M., Sauve,Y., Keegan,D.J., Coffey,P.J., Hetherington,L., Girman,S., Whiteley,S.J., Kwan,A.S. and Lund,R.D. (2000) Schwann cell grafting into the retina of the dystrophic RCS rat limits functional deterioration. Royal College of Surgeons. Invest. Ophthalmol. Vis. Sci., 41, 518–528. 13. Chang,B., Smith,R.S., Hawes,N.L., Anderson,M.G., Zabaleta,A., Savinova,O., Roderick,T.H., Heckenlively,J.R., Davisson,M.T. and John,S.W. (1999) Interacting loci cause severe iris atrophy and glaucoma in DBA/2J mice. Nat. Genet., 21, 405–409. 14. Beck,J.A., Lloyd,S., Hafezparast,M., Lennon-Pierce,M., Eppig,J.T., Festing,M.F. and Fisher,E.M. (2000) Genealogies of mouse inbred strains. Nat. Genet., 24, 23–25. 15. Taylor,B.A., Navin,A. and Phillips,S.J. (1994) PCR-amplification of simple sequence repeat variants from pooled DNA samples for rapidly mapping new mutations of the mouse. Genomics, 21, 626–632. 16. Pittler,S.J. and Baehr,W. (1991) Identification of a nonsense mutation in the rod photoreceptor cGMP phosphodiesterase β-subunit gene of the rd mouse. Proc. Natl Acad. Sci. USA, 88, 8322–8326. 17. Hill,R.E., Favor,J., Hogan,B.L.M., Ton,C.C.T., Saunders,G.F., Hanson,I.M., Prosser,J., Jordan,T., Hastie,N.D. and van Heyningen,V. (1991) Mouse small eye results from mutations in a paired-like homeobox-containing gene. Nature, 354, 522–525. 18. Moore,K.J. (1995) Insight into the microphthalmia gene. Trends Genet., 11, 442–448. 19. Mann,G.B., Fowler,K.J., Gabriel,A., Nice,E.C., Williams,R.L. and Dunn,A.R. (1993) Mice with a null mutation of the TGF alpha gene have abnormal skin architecture, wavy hair, and curly whiskers and often develop corneal inflammation. Cell, 73, 249–261. 20. Luetteke,N.C., Qiu,T.H., Peiffer,R.L., Oliver,P., Smithies,O. and Lee,D.C. (1993) TGF alpha deficiency results in hair follicle and eye abnormalities in targeted and waved-1 mice. Cell, 73, 263–278. 21. Luetteke,N.C., Phillips,H.K., Qiu,T.H., Copeland,N.G., Earp,H.S., Jenkins,N.A. and Lee,D.C. (1994) The mouse waved-2 phenotype results from a point mutation in the EGF receptor tyrosine kinase. Genes Dev., 8, 399–413. 22. Favor,J., Sandulache,R., Neuhäuser-Klaus,A., Pretsch,W., Chatterjee,B., Senft,E., Wurst,W., Blanquet,V., Grimes,P., Sporle,R. and Schughart,K. (1996) The mouse Pax2(1Neu) mutation is identical to a human PAX2 mutation in a family with renal-coloboma syndrome and results in developmental defects of the brain, ear, eye, and kidney. Proc. Natl Acad. Sci. USA, 93, 13870–13875. 23. Keller,S.A., Jones,J.M., Boyle,A., Barrow,L.L., Killen,P.D., Green,D.G., Kapousta,N.V., Hitchcock,P.F., Swank,R.T. and Meisler,M.H. (1994) Kidney and retinal defects (Krd), a transgene-induced mutation with a deletion of mouse chromosome 19 that includes the Pax2 locus. Genomics, 23, 309–320.

Human Molecular Genetics, 2002, Vol. 11, No. 7

24. Higashide,T., Murakami,A., McLaren,M.J. and Inana,G. (1996) Cloning of the cDNA for a novel photoreceptor protein. J. Biol. Chem., 271, 1797–1804. 25. Swanson,D.A., Chang,J.T., Campochiaro,P.A., Zack,D.J. and Valle,D. (1998) Mammalian orthologs of C. elegans unc-119 highly expressed in photoreceptors. Invest. Ophthalmol. Vis. Sci., 39, 2085–2094. 26. Kobayashi,A., Higashide,T., Hamasaki,D., Kubota,S., Sakuma,H., An,W., Fujimaki,T., McLaren,M.J., Weleber,R.G. and Inana,G. (2000) HRG4 (UNC119) mutation found in cone-rod dystrophy causes retinal degeneration in a transgenic model. Invest. Ophthalmol. Vis. Sci., 41, 3268–3277. 27. Lyon,M.F., Glenister,P.H. and West,J.D. (1984) Bruised (Bru). Mouse News Lett., 71, 26. 28. Cattanach,B.M., Evans,E.P., Burtenshaw,M., Glenister,P.H., Vizor,L. and Woodward,A.M. (1993) Radiation-induced deletions. Mouse Genome, 91, 853–854. 29. Smith,R.S., Zabaleta,A., Kume,T., Savinova,O.V., Kidson,S.H., Martin,J.E., Nishimura,D.Y., Alward,W.L., Hogan,B.L. and John,S.W. (2000) Haploinsufficiency of the transcription factors FOXC1 and FOXC2 results in aberrant ocular development. Hum. Mol. Genet., 9, 1021–1032. 30. Gage,P.J., Suh,H.Y. and Camper,S.A. (1999) Dosage requirement of Pitx2 for development of multiple organs. Development, 126, 4643–4651. 31. Wang,Q., Chen,Q., Zhao,K., Wang,L., Wang,L. and Traboulsi,E.I. (2001) Update on the molecular genetics of retinitis pigmentosa. Ophthamic. Genet., 22, 133–154. 32. Dryja,T.P. and Li,T. (1995) Molecular genetics of retinitis pigmentosa. Hum. Mol. Genet., 4, 1739–1743. 33. Rosenfeld,P.J., Cowley,G.S., McGee,T.L., Sandberg,M.A., Berson,E.L. and Dryja,T.P. (1992) A null mutation in the rhodopsin gene causes rod photoreceptor dysfunction and autosomal recessive retinitis pigmentosa. Nat. Genet., 3, 209–213. 34. Humphries,M.M., Rancourt,D., Farrar,G.J., Kenna,P., Hazel,M., Bush,R.A., Sieving,P.A., Sheils,D.M., McNally,N., Creighton,P. et al. (1997) Retinopathy induced in mice by targeted disruption of the rhodopsin gene. Nat. Genet., 15, 216–219. 35. Travis,G.H., Brennan,M.B., Danielson,P.E., Kozak,C.A. and Sutcliffe,J.G. (1989 Identification of a photoreceptor-specific mRNA encoded by the gene responsible for retinal degeneration slow (rds). Nature, 338, 70–73. 36. Furukawa,T., Morrow,E.M., Li,T., Davis,F.C. and Cepko,C.L. (1999) Retinopathy and attenuated circadian entrainment in Crx-deficient mice. Nat. Genet., 23, 466–470. 37. Mears,A.J., Kondo,M., Swain,P.K., Takada,Y., Bush,R.A., Saunders,T.L., Sieving,P.A. and Swaroop,A. (2001) Nrl is required for rod photoreceptor development. Nat. Genet., 29, 447–452. 38. Favor,J. and Neuhäuser-Klaus,A. (2000) Saturation mutagenesis for dominant eye morphological defects in the mouse Mus musculus. Mamm. Genome., 11, 520–525. 39. Ehling,U.H., Favor,J., Kratochvilova,J. and Neuhäuser-Klaus,A. (1982) Dominant cataract mutations and specific-locus mutations in mice induced by radiation or ethylnitrosourea. Mutat. Res., 92, 181–192. 40. Ehling,U.H. (1985) Induction and manifestation of hereditary cataracts. Basic Life Sci., 33, 345–367. 41. McLaughlin,M.E., Sandberg,M.A., Berson,E.L. and Dryja,T.P. (1993) Recessive mutations in the gene encoding the beta-subunit of rod phosphodiesterase in patients with retinitis pigmentosa. Nat. Genet., 4, 130–134. 42. McLaughlin,M.E., Ehrhart,T.L., Berson,E.L. and Dryja,T.P. (1995) Mutation spectrum of the gene encoding the beta subunit of rod phosphodiesterase among patients with autosomal recessive retinitis pigmentosa. Proc. Natl Acad. Sci. USA, 92, 3249–3253. 43. Carter-Dawson,L.D., LaVail,M.M. and Sidman,R.L. (1978) Differential effect of the rd mutation on rods and cones in the mouse retina. Invest. Ophthalmol. Vis. Sci., 17, 489–498. 44. Glaser,T., Jepeal,L., Edwards,J.G., Young,S.R., Favor,J. and Maas,R.L. (1994) PAX6 gene dosage effect in a family with congenital cataracts,

767

aniridia, anophthalmia and central nervous system defects. Nat. Genet., 7, 463–471. 45. Love,J., Axton,R., Churchill,A., van Heyningen,V. and Hanson,I. (1998) A new set of primers for mutation analysis of the human PAX6 gene. Hum. Mutat., 12, 128–134. 46. Kissinger,C.R., Liu,B.S., Martin-Blanco,E., Kornberg,T.B. and Pabo,C.O. (1990) Crystal structure of an engrailed homeodomain-DNA complex at 2.8 A resolution: a framework for understanding homeodomain-DNA interactions. Cell, 63, 579–590. 47. Wilson,D.S., Guenther,B., Desplan,C. and Kuriyan,J. (1995 ) High resolution crystal structure of a paired (Pax) class cooperative homeodomain dimer on DNA. Cell, 82, 709–719. 48. Chao,L.Y., Huff,V., Strong,L.C. and Saunders,G.F. (2000) Mutation in the PAX6 gene in twenty patients with aniridia. Hum. Mutat., 15, 332–339. 49. Morrison,D., FitzPatrick,D., Hanson,I., Williamson,K., van Heyningen,V., Fleck,B., Jones,I., Chalmers,J. and Campbell,H. (2002) National study of microphthalmia, anophthalmia, and coloboma (MAC) in Scotland: investigation of genetic aetiology. J. Med. Genet., 39, 16–22. 50. Favor,J., Peters,H., Hermann,T., Schmahl,W., Chatterjee,B., Neuhäuser-Klaus,A. and Sandulache,R. (2001) Molecular characterization of Pax62Neu through Pax610Neu: An extension of the Pax6 allelic series and the identification of two possible hypomorph alleles in the mouse Mus musculus. Genetics, 159, 1689–1700. 51. Xu,H.E., Rould,M.A., Xu,W., Epstein,J.A., Maas,R.L. and Pabo,C.O. (1999) Crystal structure of the human Pax6 paired domain-DNA complex reveals specific roles for the linker region and carboxy-terminal subdomain in DNA binding. Genes Dev., 13, 1263–1275. 52. Azuma,N., Nishina,S., Yanagisawa,H., Okuyama,T. and Yamada,M. (1996) PAX6 missense mutation in isolated foveal hypoplasia. Nat. Genet., 13, 141–142. 53. Threadgill,D.W., Dlugosz,A.A., Hansen,L.A., Tennenbaum,T., Lichti,U., Yee,D., LaMantia,C., Mourton,T., Herrup,K., Harris,R.C. et al. (1995) Targeted disruption of mouse EGF receptor: effect of genetic background on mutant phenotype. Science, 269, 230–234. 54. Sibilia,M. and Wagner,E.F. (1995) Strain-dependent epithelial defects in mice lacking the EGF receptor. Science, 269, 234–238. 55. Mears,A.J., Jordan,T., Mirzayans,F., Dubois,S., Kume,T., Parlee,M., Ritch,R., Koop,B., Kuo,W.L., Collins,C. et al. (1998) Mutations of the forkhead/winged-helix gene, FKHL7, in patients with Axenfeld–Rieger anomaly. Am. J. Hum. Genet. 63, 1316–1328. 56. Nishimura,D.Y., Swiderski,R.E., Alward,W.L., Searby,C.C., Patil,S.R., Bennet,S.R., Kanis,A.B., Gastier,J.M., Stone,E.M. and Sheffield,V.C. (1998) The forkhead transcription factor gene FKHL7 is responsible for glaucoma phenotypes which map to 6p25. Nat. Genet. 19, 140–147. 57. Semina,E.V., Reiter,R., Leysens,N.J., Alward,W.L., Small,K.W., Datson,N.A., Siegel-Bartelt,J., Bierke-Nelson,D., Bitoun,P., Zabel,B.U. et al. (1996) Cloning and characterization of a novel bicoid-related homeobox transcription factor gene, RIEG, involved in Rieger syndrome. Nat. Genet. 14, 392–399. 58. Fang,J., Dagenais,S.L., Erickson,R.P., Arlt,M.F., Glynn,M.W., Gorski,J.L., Seaver,L.H. and Glover,T.W. (2000) Mutations in FOXC2 (MFH-1), a forkhead family transcription factor, are responsible for the hereditary lymphedema-distichiasis syndrome. Am. J. Hum. Genet. 67, 1382–1388. 59. Blum,J.V., Allen,J.H. and Holland,M.G. (1962) Familial bilateral essential iris atrophy (Group 2). Tr. Am. Acad. Ophth. Otol., July–Aug, 493–500. 60. Jampol,L.M., Rosser,M.J. and Sears,M.L. (1974) Unusual aspects of progressive essential iris atrophy. Am. J. Ophthalmol., 77, 353–357. 61. Kaiser-Kupfer,M., Kuwabara,T. and Kupfer,C. (1977) Progressive bilateral essential iris atrophy. Am. J. Ophthalmol., 83, 340–346. 62. Stone,E.M., Kimura,A.E., Folk,J.C., Bennett,S.R.,Nichols,B.E., Streb,L.M. and Sheffield,V.C. (1992) Genetic linkage of autosomal dominant neovascular inflammatory vitreoretinopathy to chromosome 11q13. Hum. Mol. Genet., 1, 685–689. 63. Thaung,C. (2001) New mouse models of inherited eye disease and evaluation of vision testing methods. PhD Thesis, Oxford University, UK.

768

Human Molecular Genetics, 2002, Vol. 11, No. 7