Differential localization by in situ hybridization of specific crystallin transcripts during mouse lens development

Differentiation (1991) 47: 143-147

Differentiation

Ontogeny, Keoplasia and Differentiation Therapy

0 Springer-Verlag 1991

Differential localization by in situ hybridization of specific crystallin transcripts during mouse lens development Jacques A. TrCton *, Edith Jacquemin, Yves Courtois, and Jean-Claude Jeanny Centre de Girontologie, Association Claude Bernard, INSERM U 1 18, Unite de Recherches Gerontologiques, afflite CNRS, 29 rue Wilheni, F-75016 Paris, France Accepted in revised form May 6, 1991

Abstract. The embryonic development of the mammalian lens is well known at the biochemical and histological level. However few data are available at the molecular level concerning gene expression during the continuous differentiation of the lens. In the present study, we have investigated by in situ hybridization the changes in the distribution of mouse crystallin mRNA as a marker of differentiated lens cells, during development of the lens primordium, when tissue interactions are known to be essential. The transcripts of a and p crystallins are first detected at the early elongation stage of primary fibres ; y-crystallin-transcripts do not appear until the late elongation phase. All areas of the lenses exhibited crystallin mRNA until the beginning of secondary fiber formation at 18 days of development. Hybridization for ct and p crystallin is confined at that time to the equatorial part of the lens. The y crystallin transcripts are no longer found in the equatorial region after 1 post-natal day, but remain in the lens core, decreasing gradually. A possible mechanism is discussed.

Introduction The vertebrate eye lens is characterized by the synthesis and accuniulation of a large amount of structural proteins, the crystallins. In most species three classes of crystallins have been identified: u, p- and y (6 is only found in birds and reptiles; for reviews see [I, 2, 5, 8, 11, 12, 201). The differentiation of the vertebrate eye lens begins with the formation of the lens placode by epidermal ectoderm cells. This layer invaginates into the optic cup, breaks away from the surrounding ectoderm, and closes to form the lens vesicle; the cells from the anterior part of the lens vesicle remain cuboidal and form the epithelial layer. At the equator the epithelial cells of the lens divide and differentiate into fiber cells. As a result of

* To whom offprint requests should be sent

this continuous deposition of new fiber cells, the lens continues to grow throughout life. Early attempts to correlate the specific stage of cellular differentiation at which a particular crystallin is expressed have been principally achieved using the technique of immunofluorescence [7, 9, 11, 18, 21, 22, 231, with antibodies directed toward the various classes of crystallins. Using this method, only the end product (protein) of the gene is determined. Studies at the RNA level, using the Northern-blot technique, have mainly been directed at establishing the program of developmental expression of various crystallin genes. The uA-crystallin gene is active during the whole growth period of the rat lens [7, 191 and the ycrystallin genes are all activated a t the same time in early development but later differentially terminated, in both rat [14, 191 and mouse lenses [4, 10, 171. In the chicken, the S-crystallin gene is expressed from the lenspit stage until 3 months after birth [ 151. A few reports have dealt with a combined approach using immunological and Northern-blot techniques to study molecular differentiation of the mammalian rodent lens, at a cellular level [3, 6, 141. Hereditary defects in lens development resulting in cataract have been reported [21], adding to the interest in studying the appearance of crystallin transcripts during normal development in this species. In the present study, we have used in situ and Northern hybridizations to investigate the cellular distribution of a, p and y crystallin mRNAs, and to quantify the relative amount of crystallin mRNAs in the developing mouse lens.

Methods Tissue preparation. Pregnant OF1 mice were obtained from IFFACREDO, Saint-Germain sur 1’Arbresle (France) and were killed at 9, 10, 11, 12, 14, 16, and 18 days of development/gestation. Embryos were surgically removed, and the age determined by considering the day of the vaginal plug appearance as day 0.

144 Until the 24th day of embryonic development, embryos were decapitated and the bodies discarded. Whole eyes were collected from the 16th day of development to 1 day after birth. All tissues were mounted in O.C.T. Compound, Miles Inc., Elkart, IN (USA), then immediately frozen in liquid nitrogen and kept at -70" C until use. All animal care complied with the "Principles of Laboratory Animal Care" and the guide for the "Care and the Use of Laboratory Animals" (NIH pubkations N"80-23, revised 1978). The tissue was cut by a SLEE Cryo-microtome, London (UK) and sagittal sections of 7-10 pm were collected on slides. The fixation was carried out immediately afterwards in a mixture of methanol and acetic acid (3:1), for 5 min at 4" C and then 0.1 YOglutaraldehyde in 0.1 M cacodylate buffer (pH 7.2) for 20 rnin at 4" C. The slides were dehydrated, air-dried and prehybridized with salmon sperm DNA before in situ hybridization. CI, p and y crystallin cDNAs, pMclA Cr2 (1067 bp), pMp26 Crl (769 bp) and pMy4 Crl (550 bp) were cloned by the GC tailing procedure into the bacterial plasmid pBR322 [4, 13, 16, 171. The cDNA insert was cut out with Pst I and isolated by electrophoresis before hybridization. The probes were labelled [3, 61 with 32P (3000 CijmM) to a specific activity of about 3 x lo8 cpni/pg by random priming and separated from free nucleotides by chromatography on Sephadex G50. The probes did not cross-hybridize.

D N A probes. The murine

Results Histological localization of

c1

crystallin gene expression

At day 9 of development, the outgrowing optic vesicle was found to have made contact with the lateral surface ectoderm. Some prospective lens cells had elongated sufficiently for a lens placode to be recognizable. Its cells formed a very regular columnar epithelium, with most of their nuclei located in a basal position. The autoradiograms with either aAl crystallin and pBR322 probes were negative at this stage (not shown). The level of background was determined as the silver grain density over the area of the tissue section not containing lens. The differential background produced with ctAl crystallin probe was compared with the pBR322 probes labelled to the same specific activity. This comparison may be seen in Figs. 1 and 2. When its invagination occurred at day 10, the lens placode changed into a distinct lens pit. The autoradiograms with all crystallin probes were negative (Fig. 1 a,

In situ hybridization. A drop of labelled probe (10 p l 2.5 ~ ng DNA) was deposited onto the middle of the sections of interest and covered by a coverslip (18-mm diameter) sealed with rubber cement. Hybridization was performed at 42" C and left overnight in 50% formamide, 0.6 M NaC1, 10 m M tris pH 7.4, 1 m M EDTA, 1 x Denhardt's solution, 250 pg/ml single-stranded Salmon sperm DNA, 500 pg/ml yeast tRNA, 10% dextran sulfate. RNAse treatment and heterologous probes (pBR322) constituted the negative controls. After hybridization the slides were washed three times for 10 min in 2 x SSC at room temperature, then twice for 30 min in 0.1 or 0.5 x SSC at 45" C or 50" C or 55" C, and finally three times for 30 rnin each in 2 x SSC at 4" C. The slides were then dehydrated and air dried. The slides were dipped in Ilford K2 emulsion (Lyon, France) diluted 1 : 1 with water and exposed in light-proof boxes with desiccant for 1 01-2 weeks at 4" C. The slides were developed in D19, Kodak (Paris, France), for 2.5 min at 15" C and fixed with fixative (Hypam, Ilford, Lyon, France; dilution 1 :4) for 5 min. The slides were examined under a microscope (Zeiss), and photographs were made using Ilford Pan F (Lyon, France), 50 ASA film. R N A processing und Northern blots. The total RNA was extracted [I71 from mouse lenses at different postnatal stages (15 days, 1, 3, 6 and 12 months). Polyadenylated RNA was purified by oligo (dT) cellulose chromatography and the amount measured by optical density at 260 and 280 nm. Five micrograms mRNA per lane (equivalent to approximately five lenses) was run in 1 % agarose, 2.2 M formaldehyde gels and blotted on nitrocellulose filters [15]. Each Northern-blot filter was paired with a dot blot (not shown) containing an incremental amount of mRNA to provide internal hybridization standards for RNA quantification. The cDNAs were nick-translated to a specific activity of at least lo7 cpm/pg DNA using [ 32P]dCTPand were used for hybridization. Hybridization was performed at 42" C for 16 h in 40% formamide, 4 x SSC, 10% dextran sulfate, 10 m M tris HC1, pH 7.5. Filters were washed for 20 min at room temperature in 2 x SSC and 0.1% SDS, twice at 65" C in 0.1 x SSC and 0.1 SDS and twice at 65" C in 0.1 x SSC [I 51. These conditions result in specific hybridization of each DNA. The dried filters were subjected to autoradiography at -70" C using Kodak xAR5 film. The autoradiograms of hybridized mRNA were scanned at multiple exposures (Transsidyne General Corp, densitometer model 2510, Chicago (USA)), to generate a standard curve relating signal intensity to quantity of hybridizing mRNA.

Fig. la-d, A-D. In situ hybridization on embryonic mouse lens (pBR322 probe, lower case; CIAIcrystallin probe, upper case). a, A At 10 days of development. b, B At 11 days of development. c, C At 12 days of development. d, D At 14 days of development

145

Fig. 2a+, A-C. In situ hybridization on later embryonic and postnatal mouse lens (pBR322 probe, lower case; CI A1 crystallin probe, upper cusp). a, A At 16 days of development. b, B At 18 days of development; c, C At 1 postnatal day

Fig. 4a-c. In situ hybridization with the 74 crystallin probe on embryonic and postnatal mouse lens at 10 (a) and 1 1 (b) days of development and 1 postnatal day (c)

Fig. 3a, b. In situ hybridization with the p 26 crystallin probe on embryonic mouse lens at 10 (a) and 11 (b) days of development

A). The lens rudiment closed off, forming a vesicle at day 11. At this time virtually all the posterior wall cells showed silver grains with the a crystallin cDNA probe, in comparison with that of pBR322 (Fig. 1 b, B). At

day 12 of development the lumen became occluded by continued growth of the primary lens fibers. The autoradiogram for the %A1crystallin cDNA was strongly labelled for most of the lens fibers but remained weak for the epithelium (Fig. 1 c, C). At day 14, the silver grains produced with the aAl crystallin cDNA probe (Fig. 1d, D) covered both equatorial and fiber cells. At the 16th day of development, the concentration of silver grains produced with the a crystallin probe were now very dense, both in the equatorial region and in the posterior core, and less dense in the anterior core, but still more than the background (Fig. 2a, A). The autoradiograms of the 18-day embryos showed the majority of hybridization located on the equatorial edge of the lens for aAl crystallin cDNA (Fig. 2b, B), mostly in the newly deposited, fully mature secondary fibers. One day after birth, the autoradiogram revealed that %A1 crystallin transcripts were located mostly in the secondary fibers formed at the equatorial region (Fig. 2c, C). An increasing density of silver grains appeared in the lens core either with the aAl crystallin probe or with the pBR322 probe (Fig. 2c, C ) .

146

F

1

z

6

3

6

12

I

3

a

3

50

I

i

1

W

12

> F

a -

E 'O01\

1

3

6

12

MONTHS

Fip. lagiiification of Fig. 4c, in situ hybridization on l-postnatarday ( x i 6 ) mouse lens with y4 cryslal1;n probe

The autoradiograms with the p26 crystallin probe were negative for hybridization at day 10 and became positive at day 11 (Fig. 3 ) . For the other stages, the pattern observed for the 826 crystallin probe was the same as that for the a probe. With the y4 crystallin probe, hybridization also became positive at day l l (Fig. 4a, b). The hybridization pattern from that stage to day 18 was the same as that seen with the aAl and 826 crystallin probes. Then at 1 day after birth, (Fig. 4c) the y4 crystallin cDNA autoradiogram revealed that the silver grain density was very high between the limit of the primary fibers and the secondary fibers, while in the newly formed secondary fibers in the outer cortex, no hybridization was detectable (Fig. 5). After that age, the lens increased in size and in hardness, making it difficult to get good histological sections, e.g. at 7 and 30 days after birth. Thus we studied crystalIin mRNAs, using the Northern-blot technique to measure the relative amount of crystallin mRNA at late postnatal stages. Relative amount of the crystallin mRNA with age

Lens mRNA for each age was analysed on gels by the Northern-blot technique. The corresponding Northern blot was scanned and standardized (as explained in

Fig. 6a-c. Five micrograms lens mRNA at different stages was run in a gel and transfered as a Northern blot (see details in Methods). A blot for each was hybridized respectively with xAl, /3 26, and 74 crystallin probes. Relative amounts of hybridization were determined for each probe (ordinate) and at different stages (ahscissa) by densitometric scans of the autoradiograms. a a A l . b /3 26. c y4; 100% was determined as the labelling intensity in the first stage studied (1 5 days)

Methods), and the labelling intensity determined for each band: 100% was determined as the labelling intensity in the first stage studied (15 days). The relative amount of ctAl transcripts increased (four times) with age specifically until 3 months of age (Fig. 6a). However for 826 and y4 transcripts from 2 weeks onwards, the amount of those transcripts progressively decreased (826, twice; y4, four times) until 6 months. They stayed at this same level until 12 months (Fig. 6b, c).

Discussion Our results show that the appearance of crystallin mRNAs during embryonic development was correlated with morphological changes. a A l , p26 and y4 crystallin transcripts could be detected synchronously when the elongation of the primary fibers had begun [23]. Our data can be compared with immunofluorescence studies. Indeed this last technique detects the lens protein present in the tissues, and is different from our technique which localizes the lens mRNA in situ. But our finding is in good accordance with previous immunofluorescence studies on uAl crystallin localization (slightly before

147

11 days [22, 231) but not for p26 crystallin (which was found after 12 days) and y4 crystallin (before 12 days). This last difference could be linked with the nature of the antibody or the fact that we are measuring quite different and distinct entities, i.e., the mRNA could be present but not translated. It is known that eight different P-crystallins and six y-crystallins are synthesised in the mouse lens [12]. Our data are also in accordance with the work of Van Leen et al. who found nonuniformly distributed aAl and p26 crystallin transcripts in the fiber cells of the newborn rat lens [19], concentrated in the co.rtical zone. These authors did not find a sharp decrease in the amount of y4 transcripts in the lens equator, perhaps because of slight differences in species’ gestation times. In addition, they did not study any age older than that of the newborn animal. Our data thus suggest that the y4 crystallin gene is turned on as soon as the primary fibers elongate; the gene then seems to turn off just before birth. The decrease of silver grains in the outer secondary fibers was not due to loss of the fibers from the slide, since Fig. 5 shows that fibers were present. The sharp demarcation on the left of Fig. 5 between the signal over equatorial cells and the lack of signal over the more-mature fiber cells has been found consistently at that age and symmetrically on each side of the lens section. The rapid increase in yCtranscripts before birth has also been observed by other authors [ 101. The accumulation of y4-crystallin mRNA in the elongating cells may be the result of increased transcription or increased stabilization of mRNA. In the lens core after birth, we observe the same pattern of increasing density of silver grains for the crystallin and pBR322 probes, indicating that any hybridization at this time is nonspecific. This could be due to the higher protein content of the central fiber mass, reported to increase nonspecific binding to a nucleic acid probe (from chicken [15]). Zwaan has observed that the internal pH of the mouse lens [21] changes during development: The cytoplasm of the fiber cells is basophilic throughout the lens in the 13-day-old embryo. After that time, a weak acidophilia occurs in the central fibers, and spreads to the entire lens of the 19-day-old embryo. This chemical change of the lens core (“nucleus”) may be responsible for any artefactual hybridization observed. In conclusion, eye lens crystallin mRNA and protein expression appear to be differentially regulated over time and within the lens space. It will be interesting now to extend these studies in mutant mice, whose lens abnormalities may be related to similar molecular events. Acknowledgmeflts. We thank Dr. David McDevitt for reviewing the manuscript and Yveline Maville for expert typing. Laurent Joiiet for his technical assistance and Herve Coet for photographs.

References 1. Aarts HJM, Lubsen NH, Schoenmakers GG (1989) Crystallin gene expression during rat lens development. Eur J Biochem 183: 31-36

2. Bloemendal H (1985) Lens research: from protein to gene. Exp Eye Res 42 : 429-448 3. Bower DJ, Errington LH, Pollock BJ, Morris S, Clayton RM (1983) The pattern of expression of chick d-crystallin genes in lens differentiation and in trans differentiating cultured tissues. EMBO J 2: 333-338 4. Breitman ML, Lok S, Wistow G, Piatigorsky J, Treton JA, Gold RJM, Tsui LP (1984) y-crystallin family of the mouse lens: structural and evolutionary relationship. Proc Natl Acad Sci USA 81 :7762-7766 5. Harding JJ, Dilley KJ (1976) Structural proteins of the mammalian lens: a review with emphasis on changes in development, aging and cataract. Exp Eye Res 22: 1-74 6. Jeanny JC, Bower DJ, Errington LH, Morris S, Clayton RM (1985) Cellular heterogeneity in the expression of the &crystallin gene in non-lens tissues. Dev Biol 112:9499 7. McAvoy JW (1978) Cell division, cell elongation and distribution of E-, 0- and y-crystallins in the rat lens. J Embryo1 Exp Morphol44: 349-165 8. McAvoy JW (1980) Induction of the eye lens. Differentiation 17~137-149 9. McDevitt DS, Brahma SK (1981) Ontogeny and localization of the CI- p- and y-crystallin in newt lens development. Dev Biol84: 449-454 10. Murer-Orlando M, Paterson RC, Lok S, Paterson LC, Breitman ML (1987) Differential regulation of y-crystallin genes during mouse lens development. Dev Biol 119:260-267 11. Piatigorsky J (1981) Lens differentiation in vertebrates. Differentiation 19:134-153 12. Piatigorsky J (1987) Gene expression and genetic engineering in the lens. Invest Ophthal Visual Sci 28 :9-28 13. Shinohara T, Robinson EA, Appela E, Piatigorsky J (1982) Multiple crystallins of the mouse lens fractionation of mRNAs by cDNA cloning. Proc Natl Acad Sci USA 79:2783-2787 14. Siezen RJ, Wu E, Kaplan ED, Thomson JA. Beneder GB (1988) Rat lens gamma crystallins. Characterization of the six gene products and their spatial and temporal distribution resulting from differential synthesis. J Mol Biol 199:475-490 15. Treton JA, Shinohara T, Piatigorsky J (1982) Degradation of crystallin mRNA in the lens fiber cells of the chicken. Dev Biol92 :60-65 16. Treton JA, Jacquemin E, Jeanny JC, Courtois Y (1986) A critical age during the development and ageing of the mouse lens: lens growth and RNA content. In: Courtois, Faucheux, Forette, Knook, Trtton (eds) Colloque INSERM, Paris Vol 147, pp 323-33 1 17. Treton JA, Jacquemin E, Courtois Y (1988) Variation abundance of y-crystallin gene transcripts during development and ageing. Exp Eye Res 46 :405-41 3 18. Van de Kamp M, Zwaan J (1973) Intracellular localization of lens antigens in the developing mouse embryo. J Exp Zoo1 186123-35 19. Van Leen RW, Breuer ML, Lubsen NH, Schoenmakers JGG (1987) Developmental expression of crystallin genes: in situ hybridization reveals a differential localization of specific mRNAs. Dev Biol 123:338-345 20. Wistow G, Piatigorsky J (1988) Lens crystallins: the evolution and expression of proteins for a highly specialized tissues. Annu Rev Biochem 51 :419-504 21. Zwaan J , Williams RM (1968) Morphogenesis of the eye in a mouse strain with hereditary cataracts. J Exp 2001169:407422 22. Zwaan J (1983) The appearance of 1-crystallin in relation to cell cycle phase in the embryonic mouse lens. Dev Biol 96: 173181 23. Zwaan J, Silver J (1983) Crystallin synthesis in the lens rudiment of a strain of mice with congenital anophthalmia. Exp Eye Res 36:551-558