ORIGINAL ARTICLE

Eye development in the Cape dune mole rat

Natalya V. Nikitina & Susan H. Kidson

Received: 9 December 2013 /Accepted: 28 January 2014 /Published online: 26 February 2014# Springer-Verlag Berlin Heidelberg 2014

Abstract Studies on mammalian species with naturally re-duced eyes can provide valuable insights into the evolutionarydevelopmental mechanisms underlying the reduction of theeye structures. Because few naturally microphthalmic animalshave been studied and eye reduction must have evolvedindependently in many of the modern groups, novel evolu-tionary developmental models for eye research have to besought. Here, we present a first report on embryonic eyedevelopment in the Cape dune mole rat, Bathyergus suillus.The eyes of these animals contain all the internal structurescharacteristic of the normal eye but exhibit abnormalities inthe anterior chamber structures. The lens is small but developsnormally and exhibits a normal expression of α- and γ-crystallins. One of the interesting features of these animals isan extremely enlarged and highly pigmented ciliary body. Inorder to understand the molecular basis of this unusual feature,the expression pattern of an early marker of the ciliary zone,Ptmb4, was investigated in this animal. Surprisingly, in situhybridization results revealed that Ptmb4 expression was ab-sent from the ciliary body zone of the developing Bathyerguseye.

Keywords Bathyergus suillus . Cape dunemole rat . Eyedevelopment . Lens . Ciliary body . Crystallins

Introduction

Vertebrates that have adapted to living in constant darkness (incaves or burrows) often have eyes that are reduced in size,dysfunctional and disorganized. Evolutionary processes re-sponsible for the “degeneration” of the eye and the molecularand developmental mechanisms underlying this process arenot completely understood at present. In order to gain a betterinsight into the evolutionary mechanisms of eye reduction,new model animals have to be sought.

The process of vertebrate eye development has been ex-tensively reviewed (Chow and Lang 2001; Fuhrmann 2010;Graw 2010; Sinn and Wittbrodt 2013). The eye forms as alateral outgrowth of the diencephalon (forebrain)neuroectoderm, which enlarges to become the optic vesicle.When the optic vesicle contacts the overlying surface ecto-derm, an exchange of inductive signals between these tissuesis thought to take place, which results in their coordinatedinvagination to form the lens vesicle and the optic cup. Asignal from the distal portion of the optic vesicle induces thepart of the overlying epithelium in contact with it to thickenand form the lens placode, which in turn, promotes the invag-ination of the optic vesicle to form the optic cup. The optic cupconsists of two layers of apposing epithelia, the inner one ofwhich will give rise to the retina, while the outer layer differ-entiates to form the retinal pigmented epithelium (RPE). Thelens vesicle formed from the lens placode is spherical andinitially hollow but soon becomes filled by the primary lensfibres. These fibres form by elongation of the epithelial cellslocated at the posterior of the lens vesicle. Differentiation ofthe lens fibres involves denucleation and loss of organelles,formation of a specialized type of gap junction and synthesisand accumulation of a large number of lens-specific proteins,crystallins. Later, the tip of the optic cup differentiates to formthe ciliary body and the iris; both layers of the optic cup areincorporated into these structures.

Communicated by Andreas Kispert

N. V. Nikitina (*) : S. H. KidsonDepartment of Human Biology, Faculty of Health Sciences,University of Cape Town, Observatory, 7925 Cape Town, SouthAfricae-mail: natalya.nikitina@wits.ac.za

N. V. NikitinaSchool of Molecular and Cell Biology, University of theWitwatersrand, Private Bag 3, Wits, 2050 Johannesburg, SouthAfrica

Dev Genes Evol (2014) 224:107–117DOI 10.1007/s00427-014-0468-x

Numerous experimental studies have demonstrated that thelens serves an essential role in the morphogenesis of theanterior eye structures (cornea, ciliary body and iris)(Breitman et al. 1989; Coulombre and Coulombre 1957;Genis-Galvez 1966; Harrington et al. 1991; Klein et al.1992; Yamamoto and Jeffery 2000). The lens is thought tobe the source of a diffusible inductive signal that instructssubterminal optic cup to adopt the ciliary body fate (Beebe1986; Coulombre and Coulombre 1957; Genis-Galvez 1966;Stroeva 1967; Thut et al. 2001). While the molecular identityof this signal(s) is still not known, two possible downstreamgenes (Tgfb1i4 and Ptmb4) have been identified (Thut et al.2001). Thut et al. (2001) demonstrated that a ciliary bodyspecific gene, Ptmb4, is upregulated in the tip of the opticcup in response to a diffusible lens-derived signal. This genecan therefore be used as the earliest marker for the prospectiveciliary body zone.

Despite a number of investigations in recent years (Aviviet al. 2001; Carmona et al. 2008; Hough et al. 2002; Jeffery2001; Yamamoto et al. 2004), not much is known about thedevelopmental mechanisms responsible for eye degenerationin darkness-adapted vertebrate species. In fact, only very fewrecent microanatomical studies on the eye structure of blindmammals exist; most reports available are over 80 years old.In short, a significant gap in our knowledge of this subjectexists, which needs to be addressed.

The African mole rats belong to the family Bathyergidae, agroup of exclusively African burrowing rodents, with everymember of the family being adapted to subterranean lifestyle.All members of this family have eyes that are reduced in size;however, the eyes exhibit all structural features typical ofsighted mammals (Nikitina et al. 2004; Cernuda-Cernudaet al. 2003). Studies of the anatomy of the central visualsystem of several bathyergid species suggest that these ani-mals have selectively lost brain structures that mediate formvision while retaining the structures involved in light-darkdiscrimination, suggesting that some level of visual functionis preserved in these animals (Crish et al. 2006; Nemec et al.2004, 2008). Histological and immunocytochemical studiesof the retina in several bathyergid species, as well as behav-ioural studies appear to corroborate this observation (Cernuda-Cernuda et al. 2003; Kott et al. 2010; Peichl et al. 2004;Wegner et al. 2006). Cernuda-Cernuda et al. (2003) demon-strated that functional rods and a small number of functionalcones are present in the Ansell’s mole rat retina. Peichl et al.(2004) demonstrated that the retinas of three bathyergid spe-cies (Cryptomys ancelli , Cryptomys mechowi andHeterocephalus glaber ) are rod-dominated but, surprisingly,possess a small number of functional cones, majority of whichare short-wave sensitive, as opposed to the situation in sightedmammals where the majority of the cones are middle- to long-wave sensitive. Behavioural studies of Heliophobiusargenteocinereus and Fukomys mechowii showed that these

two mole rat species were able to perceive light and avoidlighted areas. It appears that the mole rats use their residuallight discrimination capabilities as an anti-predator response,to detect and repair breaches to their burrow system (Kott et al.2010; Wegner et al. 2006).

While overall retinal morphology and some visual functionare preserved in these animals, there appear to be changes inthe anterior chamber and lens architecture at least in onebathyergid species. A study from this laboratory examinedeye morphology and development in the naked mole rat andfound numerous abnormalities in the shape and cellular ar-rangement of the lens and the structures of the anterior cham-ber. The lens malformation appeared at least in part due todegradation of the major structural lens proteins, gamma-crystallins (Nikitina et al. 2004). No other investigations onthe development of bathyergid eye have been published since.In this work, we provide the first description of the embryonicdevelopment of the eye in the Cape dune mole rat (Bathyergussuillus). In an attempt to understand the molecular basis of theabnormal development of anterior chamber in these animals,we studied some aspects of molecular control of the eyeformation.

Materials and methods

Embryos, histology and eye measurements

Cape dune mole rats (B. suillus) were obtained from CapeTown International Airport, where these animals were cap-tured and culled as part of the airport’s pest control pro-gramme. Animals were sacrificed with chloroform. Embryoswere removed and staged according to Theiler system formouse embryonic development (Theiler 1989). All studieswere carried out in compliance with the guidelines of theAnimal Ethics Committee of the University of Cape Town.

Eyes were dissected and processed using standard histo-logical techniques. Eyes were fixed in 4 % paraformaldehyde(PFA) overnight and then dehydrated through a series ofincreasing ethanol concentrations and embedded in paraffinwax. Sections 4- or 5-μm thick through the vertical meridianof the eye were stained with haematoxylin and eosin usingstandard procedures. The lengths of the Bathyergus retinaand ciliary body were determined using Photoshop 6software package. To ensure consistency in the results,measurements of various structures within the eye weretaken of histological sections through the centre of thepupil. In order to circumvent errors that could be intro-duced into the measurements of the ciliary body andretina length by retinal detachment from the RPE andshrinking of the retina, the lengths of the RPE and thepigmented ciliary epithelium were measured.

108 Dev Genes Evol (2014) 224:107–117

Immunohistochemistry

For anti-α-crystallin and anti-γ-crystallin immunocytochem-istry, 5-μm paraffin wax sections, cut through the centre of thepupil, were placed on APTES-coated slides, dewaxed andrehydrated, rinsed with TBS (Tris-buffered saline) and incu-bated with 0.3 % H2O2. Sections were blocked in 3 % BSA inTBS for 1–1.5 h at room temperature prior to incubation with1/200 dilution of rabbit anti-αA/B-crystallin polyclonal anti-serum (a gift from Prof. H. Kondoh, Osaka University, Japan)or 1/500 dilution of rabbit anti-γ-crystallin antiserum (gener-ously provided by Dr Linlin Ding, Joseph Horwitz Laborato-ry, Jules Stein Eye Institute, University of California) over-night at 4°C. The next day, the sections were washed exten-sively in TBST and incubated with peroxidase-conjugatedswine anti-rabbit antiserum (DAKO) for 2 h. Antigen-antibody complexes were detected by 3,3′-diaminobenzidinetetrahydrochloride (DAB; Sigma) colour reaction. Sectionswere lightly counterstained with haematoxylin, dehydratedthrough ethanol series, cleared in xylol and mounted inEntellan (Merck).

Gel electrophoresis and Western blot analysis

Lenses were dissected from one 2-year-old and one 8-month-old Cape dune mole rat. The lenses were homogenized in20 μl extraction buffer (0.1 M Tris-HCl, 1 % Nonidet P40,0 .01 % SDS, 1 μg /ml ap ro ton in and 0 .1 mMphenylmethylsulfonyl fluoride) at pH 7.2 to extract thewater-soluble proteins and centrifuged at 12,000g for15 min. The water-insoluble proteins from the pellet wereresuspended in 5 μl of 4× sample dye (4 % SDS, 20 %glycerol, 10 % β-mercaptoethanol, 0.125 M Tris, 0.03 %bromophenol blue) and then diluted to 10 μl with extractionbuffer. The supernatant (10 μl) was boiled to denature pro-teins, electrophoresed on a 12 % sodium dodecyl sulphatepolyacrylamide gel (SDS-PAGE), along with 5 and 1 μgsamples of 2-day-old mouse lens protein extract and trans-ferred to a nitrocellulose membrane (Hybond-C, AmershamLife Science) at 8 V, in 25 mM Tris-HCl, 5 % methanol,overnight. The membrane was blocked with 10 % fat-freemilk solution for 1.5 h, and γ-crystallins were detected with1/1,000 dilution of rabbit anti-γ-crystallin antiserum (provid-ed by Dr Linlin Ding). Primary antibodies were detected withhorseradish peroxidase-conjugated swine anti-rabbit poly-clonal antibody, using ECL fluorescent detection system(Amersham Life Science). After transfer, the gel was fixedin 20 % methanol, 7 % acetic acid solution overnight withconstant agitation, washed three times in distilled water andsilver-stained for 15 min with 0.05M solution of AgNO3. Theprotein bands were visualized after addition of the developer(0.005 % citric acid, 0.02 % formaldehyde). The reaction was

stopped with 1 % acetic acid solution, and the gel wasphotographed using ChemiImager.

Cloning of the mole rat homologue of Ptmb4

Bathyergus-specific probes for Ptmb4was obtained by reversetranscription with polymerase chain reaction (RT-PCR) ontotal RNA isolated either from embryonic heads or eyes.RNA was prepared using TriPure Isolation Reagent (Roche,Germany), according to the manufacturer’s instructions. De-generate primers were designed by aligning the coding se-quences of the mammalian and chick Ptmb4, and selecting themost conserved regions. The primer sequences were GGCTGAGATC/TGAGAAA/GTTGG and AAATAAGAAA/GGCAATGCTC/TGT.

The PCR conditions were as follows: 1 cycle at 94 °C for2 min; 25 cycles of 94 °C for 1 min, 52 °C for 2 min, 72 °C for1–2 min and 1 cycle at 72 °C for 5 min. The amplifiedfragment of the expected size (171 bp) was column-purifiedand cloned into pGem-T Easy Vector System (Promega, Mad-ison, USA) according to the manufacturer’s instructions. Theindividual clones were sequenced by Inqaba Biotech (Preto-ria, South Africa).

In situ hybridization

In situ hybridizations on mouse and mole rat eye and embry-onic head sections were performed as described by Etcheverset al. (2001).

Results

Adult eye morphology of the Cape dune mole rat

The eye of the adult Cape dune mole rat is minute relatively tothe body size of these animals, with the average diameter of3 mm (Fig. 1a). Cape dune mole rats do not appear to usevisual cues for orientation or interaction with conspecifics. Weperformed detailed histological investigation of the internalstructure of the adult eye. The eyes of a sighted species, themouse, were used as a control. We found that all normalintraocular structures were present in the adult Bathyerguseye (Fig. 2). The most striking differences were the extremelylarge, highly pigmented ciliary body and iris (Fig. 2c, f) andirregularly shaped lens, though this may be a processingartefact (compare Fig. 2a to Fig. 3h–l). The lens epitheliumextends about two thirds of the way around the lens, and theequatorial region is irregular and poorly defined (Fig. 2a).Despite the relatively young age of the specimen in Fig. 2,degeneration of pigmented epithelium and pigment dispersionappears to occur in this eye; a large number of pigment-filledmacrophages can be seen in the trabecular meshwork

Dev Genes Evol (2014) 224:107–117 109

(Fig. 2c). All of the retinal layers are visible, except the opticnerve fibre layer, which is very thin and poorly visible onsections. The number of ganglion cells appears to be dramat-ically reduced so that the cells do not form a continuous layer(Fig. 2e).

Embryonic eye development in the Cape dune mole rat

In order to understand how the above ocular abnormalities areformed, we studied the process of Cape dune mole rat embry-onic eye development. The embryos of a range of differentages were obtained from wild-caught pregnant B. suillus fe-males. The crown-to-rump length of every embryo was re-corded, and Theiler staging system for mouse was used toidentify stage-specific features. Embryos up to 30 mm inlength (equivalent to about E16.5 of the mouse) could bereliably staged using such diagnostic features as the presenceand length of the limbs, presence/absence of the fingers andtoes, whether their eyelids were fused or not. Our observationssuggest that early development in these animals proceed at therate comparable to that of the mouse, so that most of themorphogenesis is completed early during gestation. Duringthe rest of the longer gestation period (60 days compared to19 days in the mouse), the embryos greatly increase in sizeand develop a full coat of hair. The largest embryos that were

obtained measured about 83 mm in length, virtually identicalto the size at birth (Bennett and Faulkes 2000).

In the mouse, eye development starts with the lateraloutgrowth of the optic vesicles from the diencephalon,which reach the overlying surface ectoderm, flatten andthen invaginate, forming the optic vesicle by E10.5 (Peiand Rhodin 1970). By E11.0, the lens vesicle is formedby invagination and “pinching-off” from the overlyingectoderm. These initial stages of eye morphogenesis ap-pear to occur similarly in the Cape dune mole rat eye(Fig. 3a, b). The only differences between the mole ratand mouse eye morphogenesis at this stage was thepremature appearance of pigmentation in the prospectiveRPE, and somewhat flattened and misshapen lens primor-dium, though this may be a processing artefact.

As soon as the lens vesicle is formed, the cells at itsposterior hemisphere (closer to the optic cup) begin toelongate to form the lens fibres, and by E13.0, the lumenof the lens vesicle is completely filled in. The lens con-tinues to increase in size by the formation of new fibresfrom the epithelial cells located at the equatorial region ofthe lens. The inner fibres start to undergo terminal differ-entiation, which results in denucleation and loss of otherorganelles. Because the fibre elongation is synchronized,the nuclei of the fibre cells become arranged in very neatand regular pattern. One of the earliest developmentalabnormalities noticeable in the mole rat eye is the de-creased size of the lens primordium and irregular arrange-ment of the nuclei throughout the centre of the lens(Fig. 3d). At later stages, as the central lens fibres beganto differentiate, the bow region of the lens becomes dis-tinguishable. The bow region is clearly located closer tothe posterior pole of the lens (Fig. 3f, h). The lens nucle-us, composed of the primary lens fibres, is flattened, andthe secondary lens fibres are irregular and misshapen.Therefore, abnormalities in the shape of the lens appearearly during embryonic development.

One of the most prominent features of the mole rat eyeis its very large and highly pigmented ciliary body. In themouse, ciliary body morphogenesis commences at aroundE15.5 with the slight thinning of the non-pigmented tip ofoptic cup and thickening of the corresponding region of thepigmented epithelium. By E16.5, the iris primordium canbe distinguished from the prospective ciliary body, andslight folding of the ciliary processes is evident (Fig. 3i).The morphogenesis of the ciliary body and the outgrowth ofthe iris continue postnatally. In the Cape dune mole rat, amuch larger area of the optic cup becomes specified as theprospective ciliary body (CB)/iris region. This is evidentvery early during development, as soon as the prospectiveCB/iris becomes distinguishable from the rest of the opticcup in 15-mm embryo (Fig. 3f). This large CB primordiumundergoes rapid growth and fold morphogenesis. The outer

Fig. 1 The external appearance and eye size of the Cape dune mole rat. aAdult Cape dune mole rat. Note the extremely small size of the eyeswhich are kept closed most of the time. b External appearance of thedissected Cape dune mole rat eyes. Clear cornea, highly pigmented irisand a layer of fat covering the posterior hemisphere of the eye aredistinguishable. Scale bar equals 1 mm, thus the eye is about 3 mm indiameter

110 Dev Genes Evol (2014) 224:107–117

epithelium thickens and becomes highly pigmented. Inter-estingly, the iris primordium is not identifiable in the eye ofthe Cape dune mole rat until the ciliary fold morphogenesisis well under way (Fig. 3h). Unlike the mouse, the ciliarybody and iris morphogenesis is completed in this animal bythe time of birth.

In the mouse, the first signs of the retinal morphogenesisare apparent by E16.5, when the ganglion cell layer becomesdistinguishable from the uniform non-pigmented neuroblasticlayer (Fig. 3i) (Pei and Rhodin 1970; Theiler 1989). Theretinal morphogenesis is completed postnatally, by P14. Thepigmentation of the RPE layer is first noticeable in the mouseby E11.5 (Fig. 3c) and reaches its full level by E16.5. In theCape dune mole rat, the retinal differentiation occurs muchslower than the development of the anterior chamber struc-tures. By the time the iris muscle starts to develop, the retina isstill double-layered, resembling that of the E16.5mouse (com-pare Fig. 3l to Fig. 3m).

Crystallin gene expression in the Cape dune mole rat lens

Abnormalities in the shape of the lens were observed in theeyes of adult Cape dune mole rats. In a related species, thenaked mole rat, similar but more pronounced abnormalitiesare accompanied by the loss of expression of major structurallens proteins, γ-crystallins (Nikitina et al. 2004). We wantedto find out whether crystallin expression was also abnormal inthis species (Fig. 4). Immunocytochemical study showed thatboth α- and γ-crystallins are expressed normally in the Capedune mole rat lenses, with both the time and the pattern ofexpression being similar to that of the mouse (Fig. 5). Westernblot analysis of the γ-crystallin expression revealed that mul-tiple γ-crystallin bands are present in the water-insolublefraction of the lens extract (Fig. 4; lanes 7, 8), while only asingle band is seen in the water-soluble fraction of the mole ratprotein extract (Fig. 4; lanes 2, 3). The sizes of the larger bands(approximately 44 and 65 kDa) indicated that they may be the

Fig. 2 Histology of adult Cape dune mole rat eye. a Cross sectionthrough the centre of a 2-year-old mole rat eye, showing very largeanterior chamber (ac), abnormally shaped lens (l) and large and highlypigmented iris (i) and ciliary body (cb). bMole rat cornea showing well-defined epithelium (e), stroma (s) and endothelium (ce). c Ciliary body ofthe Cape dune mole rat, showing distinct pigmented outer (oe) and non-pigmented inner epithelial layers (ie) and well-developed ciliary folds.Note large number of circular pigmented granules that appear to fill up the

trabecular meshwork. d Posterior region of the lens, enlarged from thebox in (a). Arrowheads indicate epithelial nuclei, lc lens capsule. e Capedune mole rat retina. All layers of the neural retina are present. RPEretinal pigmented epithelium, pr photoreceptors, onl outer nuclear layer,opl outer plexiform layer, inl inner nuclear layer, ipl inner plexiform layer,gcl ganglion cell layer, nfl nerve fibre layer. f Cape dune mole rat iris.Note the well-developed muscle at the tip of the iris (m). Scale bar=500 μm

Dev Genes Evol (2014) 224:107–117 111

products of covalent cross-linkage of individual γ-crystallinmolecules (Fig. 4; lanes 7, 8). Such polymers have beencharacterized in ageing mouse and human lenses and are oftenassociated with cataract formation. The smaller band (14 kDa)probably represents degraded crystallins. None of these bandswere seen in the mouse lens protein extract (Fig. 4, lane 6).

Cloning and expression analysis of BsPtmb4

The most striking feature of the Cape dune mole rat eye is theextremely large ciliary body. This feature becomes evident inthe embryonic specimens as soon as the prospective ciliaryzone can be distinguished from the retina. This increasedciliary body size in the adult mole rats could be caused either

by a larger part of the optic cup being specified as CB or by therelatively higher proliferation rate in that area, resulting inincreased growth of CB. In order to distinguish between thesetwo possibilities, the length of the ciliary zone/prospectiveciliary body in a number of adult and embryonic Cape dunemole rat and mouse specimens was measured and expressedas a percentage of the total length of the optic cup (i.e. the sumof the lengths of the retina and the ciliary body). The iris wasnot included in the measurements because it grows dramati-cally during later embryogenesis (Pei and Rhodin 1970), andthis would complicate the interpretation of the results. Forconsistency, all measurements were done on the central sec-tions through the eye, where the optic nerve was visible. Theresults of these measurements (Fig. 6) suggest that indeed the

Fig. 3 Embryonic eye development in the Cape dune mole rat. Mouseeyes of similar stages (c, e, g, i, k, m) are used for morphologicalcomparisons. aNine-millimeter mole rat embryo optic vesicle, beginningto invaginate to form an optic cup (OC). b The eye of a 10-mm mole ratembryo. The lens vesicle has separated from the prospective cornea, andthe elongating fibres are beginning to fill up the lumen of the vesicle (LV).c E11.5 mouse embryo eye. d Sixteen-millimeter mole rat embryo eye.Note prominent RPE pigmentation and irregular arrangement of nuclei inthe lens (L). e E12.5 mouse embryo eye. f Fifteen-millimeter mole rat

embryo eye, with no anterior chamber, very large ciliary body (CB)/irisprimordium and small lens (L). g Eye of E14.5 mouse embryo. h Eye of a29-mm mole rat embryo. i Eye of E16.5 mouse embryo. j Fifty-six-millimeter mole rat embryo eye. k Neonate mouse eye. l Eye of a 65-mm-long mole rat embryo.m Eye of a P5 mouse pup. In h, j, l, note theabsence of the anterior chamber, the extremely large size and highpigmentation of the ciliary body (cb) and iris (i) and the irregular shapeof the lens (L). Scale bars=100 μm (a–h, j), 200 μm (i, l,m) and 500 μm(k)

112 Dev Genes Evol (2014) 224:107–117

region of the optic cup in the mole rat that differentiates as theciliary zone is about three times larger than that of develop-mental age-matched mouse embryos (18.3 % compared to6.5 % in the mouse). This increased to 25 % in late embryosand 27 % in the adults (Fig. 6), which indicates that theproliferation rate in this region is also higher than that of theprospective retina. In the mouse, a comparable dramaticgrowth of the ciliary zone relative to the retina is not observed,and the size of the ciliary zone relative to the retina remainsvirtually the same throughout the embryonic development.

Because molecular studies have never been performed onthe Cape dune mole rat, and no genomic resources exist forthis animal, the degree of sequence conservation between themouse and Cape dune mole rat genes was not known. There-fore, in order to obtain suitable probes, RT-PCR approachusing degenerate primers was adopted. A single prominentband of 170 nt in length was obtained with Ptmb4 primers.Sequence analysis confirmed that the cloned RT-PCR prod-ucts had high sequence similarity but were not identical to the

mouse and human Ptmb4. The nucleotide sequence ofBathyergus Ptmb4 clone (BsPtmb4) was found to be 98 %identical to the homologous gene ofChinchilla lanigera, 95%to the human, mouse and rat Ptmb4 and 87 % to the chickenPtmb4, showing a very high level of conservation. The partialBsPtmb4 sequence was 38-amino acid long, missing six N-terminal amino acids, and showed 100 % identity to themouse, human and chick protein sequences. The nucleotidesequence of BsPtmb4was deposited in the GenBank database(accession number, DQ400347).

Next, we used the Bathyergus-specific probe template de-scribed above to investigate the expression of Ptmb4 in Capedune mole rat eye by in situ hybridization. In the mouse,Ptmb4 expression is first seen at E12.5 at the tip of the opticcup. At later developmental stages, strong signal appears inthe tip of the optic cup, the prospective ciliary zone, whileweaker signal is identifiable in the inner retina and in the cellsat the equator of the lens (Fig. 7a, c, d and data not shown;Thut et al. 2001). This pattern of expression persists until atleast P2. In the Cape dune mole rat, Ptmb4 expression was notseen at earlier developmental stages (data not shown). Lowlevels of expression were apparent in the inner layer of theretina in a 16-mm embryo, which is developmentally equiva-lent to mouse E12.5/13.0 embryo (data not shown). In a laterstage, 25-mm embryo, the retinal and lens equatorial Ptmb4expression was very similar to that of a corresponding mouseembryonic stage, but no Ptmb4 signal in the ciliary bodyregion was found (Fig. 7b, e and f).

Discussion

Vertebrate eye development is a complex, intricately coordi-nated process, which is still not completely understood. Mod-ifications to this process are the “raw material” of the evolu-tion, resulting in diversification of the ocular phenotypes, forexample, in the formation of a microphthalmic eye indarkness-adapted animals. Very little is known about thedevelopmental and molecular mechanisms responsible forthe formation of the microphthalmic eye, but it is clear thateye reduction does not have the same underlying molecularbasis in all subterranean/cave animals (Nemec et al. 2007;Nevo 1979, 1999). In fact, it can be estimated from thephylogenetic relationships between different mammaliangroups that the reduced eye phenotype must have evolvedindependently at least six times in the class Mammalia. Inorder to gain a comprehensive understanding of naturaleye reduction, new mammalian models must be sought.Therefore, the first aim of this project was to conduct adetailed histological and ultrastructural investigation ofthe eye structure and development in the Cape dune molerat, Bathyergus suillus.

Fig. 4 SDS-PAGE and Western blot analysis of gamma-crystallin ex-pression in the Cape dune mole rat. a Ponceau S-stained blotted mem-brane. Lane 1 water-soluble proteins from adult mouse lens; soluble lensproteins from 2-year-old (lane 2) and 8-month-old (lane 3) Cape dunemole rat; lanes 4 and 5 molecular weight markers; ovalbumin (45 kDa),carbonic anhydrase (29 kDa) and egg lysozyme (14.3 kDa); lane 6water-insoluble proteins from adult mouse lens; water-insoluble lens proteinsfrom 2-year-old (lane 7) and 8-month-old (lane 8) Cape dune mole rat. bWestern blot of the above gel, detected with anti-gamma-crystallinantibody

Dev Genes Evol (2014) 224:107–117 113

It was established that the eyes of these mole rats,while being very small relative to the body size, stillretained the general internal organization typical of verte-brate eye. The most prominent features typical of theocular structure of B. suillus were the irregular shapeand abnormal persistence of the nucleated cells in thelens, the extremely large and highly pigmented ciliarybody and the delayed formation of the anterior chamber.The preservation of ocular organization in the mole rats isquite unlike the phenotype seen in other established blindvertebrate models (e.g. Astyanax, Spalax, Talpa), wherethe lens degenerates and the anterior chamber structuresfail to develop (Quilliam 1966; Sanyal et al. 1990;

Yamamoto and Jeffery 2000; Yamamoto et al. 2004).These features make the mole rats suitable models forinvestigations into the evolutionary and developmentalbasis of the abnormal patterning of the anterior eye.Because the development of the ciliary body is dependenton the signalling from the lens, and both of these struc-tures exhibit pronounced abnormalities in the mole rats, itseems likely that the initial defect in the lens formationunderlies the subsequent developmental abnormalities inthe structures of the anterior chamber. Thus, Cape dunemole rats are also potentially very useful alternativemodels for addressing the questions pertaining to the roleof the lens in the ciliary body morphogenesis.

Fig. 5 Expression of γ-crystallins (a–f) and α-crystallins(g–j) in the mouse and the Capedune mole rat lenses, detected byimmunostaining. Sectionscounterstained withhaematoxylin. P0 mouse lens: aγ-crystallin expression and bnegative control (no primaryantibody added). c Fifteen-millimeter-long Cape dune molerat embryo lens. Gamma-crystallin expression has justbegun and is restricted to thecentral fibres. d Fifty-six-millimeter-long mole rat embryolens. e Section through a 65-mm-long mole rat embryo lens. fAdult (2 years old) mole rat lens,showing uniform distribution ofγ-crystallins throughout the fibremass, but not the lens epithelium.P7 mouse lens: g α-crystallinexpression and h negative control(no primary antibody added). iEighty-three-millimeter Capedune mole rat lens. j Thirty-eight-millimeter Cape dune mole ratlens

114 Dev Genes Evol (2014) 224:107–117

The ciliary body and iris of the mole rats aredisproportionately large

Both the iris and the ciliary body of the naked and the Capedune mole rats are highly pigmented, and the ciliary body islarge relative to the size of the eye. Extreme pigmentation andenlargement of these anterior chamber structures has beenreported in other mammals adapted to a fossorial lifestyle.

This phenomenon is most noted in Spalax ehrenbergi, wherethe pupil is completely obliterated by a mass of pigmentedtissue (Sanyal et al. 1990). The presence of an enlarged ciliarybody could be of adaptive value to these animals, as it prob-ably protects the retina from sudden exposure to bright lightwhen the animal emerges from its burrow during the day(Sanyal et al. 1990). Alternatively, the increase in the relativeciliary body size might not, in itself, confer any evolutionaryadvantage, but result from the selective pressure to decreasethe size of the retina. As has been pointed out previously,neuronal tissue uses larger amounts of energy than most othertissues, and therefore, maintenance of excess neurons is anevolutionary luxury that is strongly selected against (Cooperet al. 1993).

In the mouse, the iris and ciliary body epithelia are formedfrom the tip of the optic cup, while the iris stroma and theciliary muscle are derived largely from the cephalic neuralcrest (Beebe 1986). A part of the anterior optic cup is specifiedas the ciliary epithelium at around E12 by as yet unidentifiedsignals from the lens (Beebe 1986; Genis-Galvez 1966;Stroeva 1967; Thut et al. 2001). The morphologically distinctciliary body is first formed in the mouse around E14.5, and theformation of the ciliary processes and its functional maturationare completed postnatally (Napier and Kidson 2005; Theiler1989). In the neonate mole rat, the size of the ciliary body andiris, relative to the eye circumference, is three times greater

Fig. 6 Comparison of the relative ciliary body zone size and growthdynamics during embryonic development of the mouse (black/greycolumns) and the Cape dune mole rat (blue columns). Each columnrepresents an average length percentage of the optic cup (retina+CB)that is taken up by the CB. Three individual eyes were measured for eachdata point

Fig. 7 Thymosin β4 (Ptmb4)expression in the mouse (a, c, d)and the Cape dune mole rat eye(b, e, f), detected by in situhybridization. a E16.5 mouseembryo showing Ptmb4expression in the non-pigmentedepithelium of the ciliary body(arrowheads) and lower levels ofexpression in the inner layer of theretina (arrows) and the equatoriallens fibres. b In a 25-mmmole ratembryonic eye Ptmb4 expressionis observed in the inner retina(arrows) and the equatorial lensfibres, but not in the ciliary body.c Higher magnification of themouse E16.5 ciliary body in a. dHigher magnification of themouse retina in a. e Highermagnification of the 25-mm molerat ciliary body and iris, showingthe absence of Ptmb4 expression.f Higher magnification of the 25-mm mole rat retina

Dev Genes Evol (2014) 224:107–117 115

than in the newborn mouse (Fig. 6). In the Cape dune mole ratembryos, the ciliary body takes up about 17% of the optic cupcompared to 6 % in mouse embryos of comparable develop-mental stage; the relative size of the ciliary body increasesthroughout embryogenesis. This suggests that a greater pro-portion of the anterior optic cup is instructed to adopt theciliary body fate during embryogenesis. There is abundantexperimental evidence that the lens epithelium is the sourceof the signal that instructs part of the optic cup to form theciliary body. The region of close apposition of the lens and theanterior optic cup is also significantly enlarged in the mole ratsdue to the posterior extension of the lens epithelium. There-fore, it can be hypothesized that the enlarged ciliary zoneobserved in the mole rats could be caused by a larger portionof the optic cup receiving the signal from the lens.

Expanded ciliary zone of the Cape dune mole rat: the roleof the lens

The next question that was addressed in this study concernedthe role of the lens-derived signalling in the determination ofthe ciliary body zone in the Cape dunemole rat. The larger CBzone of the mole rats could be established in one of thefollowing ways: (1) a greater proportion of the optic cup couldbe specified as the prospective ciliary body; (2) the increasedsize of the CB could be due to the differential growth after theCB primordium is specified, but before it is morphologicallydistinct. In the first case, it could be hypothesized that the CBsize is determined by the larger area of the optic cup receivingthe lens-derived inductive signal or a larger portion of theoptic cup being able to respond to it. If the latter possibility istrue, then the area of the optic cup that is specified as the CB inthe mole rat is the same as in the mouse, but the rate of celldivision in this area is increased. To distinguish between thesetwo possibilities, in situ hybridization with CB specificmarkerPtmb4 was performed on the mole rat eye. Ptmb4 was chosenbecause it is one of the earliest and the most specific CBmarkers known and also because it was previously demon-strated that it was upregulated in the subterminal optic cup bythe signals from the lens (Thut et al. 2001). It was expectedthat if the first explanation above is correct, then a larger areaof the optic cup would exhibit Ptmb4 expression at all embry-onic stages. However, if the alternative is correct, a similarzone of expression of these markers in the early embryos ofthe mouse and the mole rat would be seen, while at laterstages, the area of expression should increase.

Surprisingly, in situ hybridization results revealed thatPtmb4 expression was absent from the CB zone altogether.Even though the expression pattern was investigated in a widerange of embryonic ages, no CB-specific expression was seenin any of the specimens. Therefore, it is not possible at presentto resolve the question of the role of the lens in the enlargedCB morphogenesis in the Cape dune mole rat. Both of the

above possibilities may be true. In order address this questionfurther, it is essential to search for additional CB-specificmarkers that are downstream from the lens-derived signals.A more direct experimental approach to investigate whetherthe mole rat lens can induce the formation of a larger CB is totransplant the mole rat lens into the chick optic cup and studythe resulting phenotype. However, this latter approach israther difficult to accomplish due to the very narrow timeframe in which the mole rat embryos of the age appropriatefor such experiments are available (not more than 2 weeks peryear).

The functional significance of the absence of Ptmb4 fromthe prospective ciliary zone of the Cape dune mole rat isunclear. Ptmb4 might facilitate cell shape changes accompa-nying the folding of the epithelial cell layers of the ciliarybody into processes or it might play a role in the formation ofthe capillaries. However, the observation that both the CBfolding and the angiogenesis in the CB stroma occur in theCape dune mole rat despite the absence of Ptmb4 expressioncontradicts this hypothesis. It is possible that another actin-sequestering molecule functionally substitutes for Ptmb4 inthe mole rat eye. Alternatively, it is possible that, due to thelarge extent of the prospective CB that undergoes folding,individual cells do not need to undergo such dramatic shapechanges as, for instance, when mouse CB folding occurs.Therefore, there could be no need for elevated Ptmb4expression.

Acknowledgments We are thankful to our grant sponsors: the SouthAfrican Medical Research Council (MRC) and the University of CapeTown. We are very grateful to Prof Hisato Kondoh (Osaka University,Japan) and Dr Linlin Ding (Joseph Horwitz Laboratory, Jules Stein EyeInstitute, University of California) for their generous gifts of anti-crystallin antibodies.

References

Avivi A, Joel A, Nevo E (2001) The lens protein alpha-B-crystallin of theblind subterranean mole-rat: high homology with sighted mammals.Gene 264:45–49

Beebe DC (1986) Development of the ciliary body: a brief review. TransOphthalmol Soc U K 105(Pt 2):123–130

Bennett NC, Faulkes CG (2000) African mole-rats: ecology andeusociality

Breitman ML, Bryce DM, Giddens E, Clapoff S, Goring D, Tsui LC,Klintworth GK, Bernstein A (1989) Analysis of lens cell fate andeye morphogenesis in transgenic mice ablated for cells of the lenslineage. Development 106:457–463

Carmona FD, Jimenez R, Collinson JM (2008) The molecular basis ofdefective lens development in the Iberian mole. BMC Biol 6:44

Cernuda-Cernuda R, Garcia-Fernandez JM, Gordijn MC, Bovee-GeurtsPH, DeGrip WJ (2003) The eye of the African mole-rat Cryptomysanselli: to see or not to see? Eur J Neurosci 17:709–720

Chow RL, Lang RA (2001) Early eye development in vertebrates. AnnuRev Cell Dev Biol 17:255–296

116 Dev Genes Evol (2014) 224:107–117

Cooper HM, Herbin M, Nevo E (1993) Visual system of a naturallymicrophthalmic mammal: the blind mole rat, Spalax ehrenbergi. JComp Neurol 328:313–350

Coulombre AJ, Coulombre JL (1957) The role of intraocular pressure inthe development of the chick eye: III. Ciliary body. Am JOphthalmol 44:85–93

Crish SD, Dengler-Crish CM, Catania KC (2006) Central visual systemof the naked mole-rat (Heterocephalus glaber). Anat Rec A: DiscovMol Cell Evol Biol 288:205–212

Etchevers HC, Vincent C, Le Douarin NM, Couly GF (2001) Thecephalic neural crest provides pericytes and smooth muscle cells toall blood vessels of the face and forebrain. Development 128:1059–1068

Fuhrmann S (2010) Eye morphogenesis and patterning of the opticvesicle. Curr Top Dev Biol 93:61–84

Genis-Galvez JM (1966) Role of the lens in the morphogenesis of the irisand cornea. Nature 210:209–210

Graw J (2010) Eye development. Curr Top Dev Biol 90:343–386Harrington L, Klintworth GK, Secor TE, Breitman ML (1991)

Developmental analysis of ocular morphogenesis in alpha A-crystallin/diphtheria toxin transgenic mice undergoing ablation ofthe lens. Dev Biol 148:508–516

Hough RB, Avivi A, Davis J, Joel A, Nevo E, Piatigorsky J (2002)Adaptive evolution of small heat shock protein/alpha B-crystallinpromoter activity of the blind subterranean mole rat, Spalaxehrenbergi. Proc Natl Acad Sci U S A 99:8145–8150

Jeffery WR (2001) Cavefish as a model system in evolutionary develop-mental biology. Dev Biol 231:1–12

Klein KL, Klintworth GK, Bernstein A, Breitman ML (1992)Embryology and morphology of microphthalmia in transgenic miceexpressing a gamma F-crystallin/diphtheria toxin A hybrid gene.Lab Invest 67:31–41

Kott O, Sumbera R, Nemec P (2010) Light perception in two strictlysubterranean rodents: life in the dark or blue? PLoS One 5:e11810

Napier HR, Kidson SH (2005) Proliferation and cell shape changesduring ciliary body morphogenesis in the mouse. Dev Dyn 233:213–223

Nemec P, Burda H, Peichl L (2004) Subcortical visual system of theAfrican mole-rat Cryptomys anselli: to see or not to see? Eur JNeurosci 20:757–768

Nemec P, Cvekova P, Benada O, Wielkopolska E, Olkowicz S, TurlejskiK, Burda H, Bennett NC, Peichl L (2008) The visual system in

subterranean African mole-rats (Rodentia, Bathyergidae): retina,subcortical visual nuclei and primary visual cortex. Brain Res Bull75:356–364

Nemec P, Cvekova P, BurdaH, BenadaO, Peichl L (2007)Visual systemsand the role of vision in subterranean rodents: diversity of retinalproperties and visual system designs. In: Begall S, Burda H,Schleich CE (eds) Subterranean rodents: news from underground.Springer, Heidelberg, pp 129–160

Nevo E (1979) Adaptive convergence and divergence of subterraneanmammals. Ann Rev Ecol Syst 10:269–308

Nevo E (1999) Mosaic evolution of subterranean mammals: regression,progression and global convergence. Oxford Uniuversity Press,Oxford

Nikitina NV, Maughan-Brown B, O'Riain MJ, Kidson SH (2004)Postnatal development of the eye in the naked mole rat(Heterocephalus glaber). Anat Rec A: Discov Mol Cell Evol Biol277:317–337

Pei YF, Rhodin JA (1970) The prenatal development of the mouse eye.Anat Rec 168:105–125

Peichl L, Nemec P, Burda H (2004) Unusual cone and rod properties insubterranean African mole-rats (Rodentia, Bathyergidae). Eur JNeurosci 19:1545–1558

Quilliam TA (1966) The problem of vision in the ecology of Talpaeuropaea. Exp Eye Res 5:63–78

Sanyal S, Jansen HG, de Grip WJ, Nevo E, de JongWW (1990) The eyeof the blind mole rat, Spalax ehrenbergi. Rudiment with hiddenfunction? Invest Ophthalmol Vis Sci 31:1398–1404

Sinn R, Wittbrodt J (2013) An eye on eye development. Mech Dev 130:347–358

Stroeva OG (1967) The correlation of the processes of proliferation anddetermination in the morphogenesis of iris and ciliary body in rats. JEmbryol Exp Morphol 18:269–287

Theiler K (1989) The house mouse: atlas of embryonic developmentThut CJ, Rountree RB, HwaM, Kingsley DM (2001) A large-scale in situ

screen provides molecular evidence for the induction of eye anteriorsegment structures by the developing lens. Dev Biol 231:63–76

Wegner RE, Begall S, Burda H (2006) Light perception in "blind"subterranean Zambian mole-rats. Anim Behav 72:1021–1024

Yamamoto Y, Jeffery WR (2000) Central role for the lens in cave fish eyedegeneration. Science 289:631–633

Yamamoto Y, Stock DW, Jeffery WR (2004) Hedgehog signalling con-trols eye degeneration in blind cavefish. Nature 431:844–847

Dev Genes Evol (2014) 224:107–117 117