Isolation of a novel retina-specific clone (MEKA cDNA) encoding a photoreceptor soluble protein

Molecular Brain Research, 6 (1989) 1-10 Elsevier BRM 70138

1

Research Reports

Isolation of a novel retina-specific clone (MEKA cDNA) encoding a photoreceptor soluble protein Che-Hui Kuo, Mariko Akiyama and Naomasa Miki Department of Pharmacology, Cancer Research Institute, Kanazawa University, Kanazawa (Japan) (Accepted 31 January 1989) Key words: MEKA protein; MEKA cDNA; Retina; Pbotoreceptor cell; Immunohistochemistry

We have reported the isolation of clones which are candidates for retina-specific cDNAs. One of the cDNA clones, pCR-470, was further characterized. We found that mRNA corresponding to the pCR-470 was expressed only in the retina and encodes an unknown soluble protein whose molecular weight and pI are calculated to be 26,935 and 5.35, respectively. We designated it as a MEKA protein, because its amino acid sequence starts from M-E-K-A. It was found by in situ hybridization that MEKA mRNA was transcribed only in the photoreceptor cells and accumulated in the inner segments just like opsin mRNA. The MEKA cDNA was ligated with expression vector PEX 1, and a MEKA-fusion protein synthesized in E. coli was purified and used as an antigen. By the Western blot analysis anti-MEKA protein serum reacted with a soluble 32 kDa protein from bovine retina and 33 kDa for chick, but not with proteins from other tissues. Immunohistochemical study showed that anti-MEKA stained only the photoreceptor cells in bovine, chick, rat and mouse retinas. INTRODUCTION Conversion of a signal of light to an electrical signal (photo-transduction) is one of the most important retinal functions. The photo-transduction is initiated by conformational changes of an opsin molecule by light and then sequential activations of GTP-binding protein and cyclic GMP-dependent phosphodiesterase (PDE) occur 1'5. Finally, the photoreceptor cells show hyperpolarization by closing sodium channels which are associated with a decrease in cyclic G M P concentration 14. We have attempted to isolate retina-specific clones to search for new proteins involved in the retinal phototransduction process. By the differential colony hybridization to retinal and cerebral single stranded c D N A s as probes, we have selected 3 clones (pCR-307,394 and 470) of the possible candidates for retina-specific c D N A s 7. By the determination of nucleotide sequences, we found that the pCR-307 and pCR-394 encode the transducin-7 subunit (T7) and opsin, respectively.

This evidence suggests that differential colony hybridization is a useful method for isolating retinaspecific c D N A clones 7'13. In the present report, we have characterized a remaining clone, pCR-470, and demonstrate that it encodes a novel photoreceptor-specific polypeptide designated as a M E K A protein. The M E K A protein was synthesized as a fusion protein in E. coli by using an expression vector P E X 1 and used as an antigen to produce antibody. We found that the M E K A protein was localized only in the photoreceptor cells of bovine, chick, mouse and rat retinas. MATERIALS AND METHODS Preparation of recombinant D N A and D N A sequencing All procedures for c D N A cloning were carried out according to Maniatis et al. 1o and a previous report 7. (A) A ScaI-PstI fragment (encoding amino acids 29-348 of opsin containing 33 bp of the 3" noncoding region) was ol~tained from LR-8 c D N A 7 and

Correspondence: C.-H. Kuo, Department of Pharmacology, Cancer Research Institute, Kanazawa University, 13-1 Takaramachi, Kanazawa 920, Japan. 0169-328X/89/$03.50 t~) 1989 Elsevier Science Publishers B.V. (Biomedical Division)

ligated with SP64 vector at the HinclI site. (B) a StuI-Pvull fragment (nucleotide numbers 50-621 which encode amino acid 18-207 of the M E K A protein in Fig. 2 A , B ) was ligated with the SP64 vector. The D N A fragment was also ligated with P E X 1 expression vector at the Smal site ( P E X 1 - M E K A plasmid). The ligated SP64 and P E X 1 vectors were transferred into E. coli HB101 and K12AH1AtrplacZamANam7Nam53CI857AH1, respectively. Recombinant plasmids were isolated and digested with a p p r o p r i a t e endonucleases to estimate their orientation of inserted c D N A . Nucleotide sequences were d e t e r m i n e d by the dideoxynucleotide chain-termination m e t h o d 12.

A

1 wlivle

2 I

3 i

4 m

In situ hybridization Anti-sense R N A p r o b e , about 2 x 107 cpm, was synthesized at 40 °C for 40 min from 50 ng of linearized SP64 vector in a total volume of 10 ~1 containing 20 ~tCi a-[35S]UTP. A sense M E K A R N A p r o b e was also used as a control experi-

ment. A n in situ hybridization experiment was perf o r m e d as described by Bloch et al. 2 with minor modifications. Briefly, a rat was anesthetized with p e n t o b a r b i t a l and perfused with 1% formaldehyde in 0.85% NaC1 (pH 7.0). Frozen sections of 10/~m thickness were thaw-mounted on slides which were p r e c o a t e d with 1% g e l a t i n - 0 . 5 % chromium (III) potassium sulfate. They were incubated with x 4 S S C / x 1 D e n h a r d t for 1 h at room t e m p e r a t u r e and prehybridized for 5 h at 45 °C in the same buffer used for Northern blot hybridization, except for the presence of 300 ~g t R N A . The sections were transferred through 70%, 80%, 95% and 100% ethanol and then air-dried. Hybridization was performed at 45 °C for 20 h in the presence of a probe (about 5 x 103 cpm/~l). The slides were rinsed twice in x l SSC and incubated in x0.3 SSC and in x0.1 SSC for 20 min at r o o m temperature. They were washed twice for 1 h at 50 °C in 3 changes of x0.1 SSC/0.1% sarcosyl. The final washing was performed 3 times in the presence of 50% formamide containing x0.1 SSC/0.1% sarcosyl at 35 °C for 30 min, followed by rinsing in x l SSC at room t e m p e r a t u r e . A f t e r the sections were d e h y d r a t e d with ethanol, air-dried and dipped into NTB-3

1

l

8

4

Fig. 1. Blot hybridization of cDNAs. A: poly(A) mRNAs from various tissues of bovine were denatured, separated on 1% agarose gel, and transferred to a nylon filter. Lane 1, 5 ~tg of retina mRNA; lane 2, 10~g of brain mRNA, lane 3, 10/~g of liver mRNA and lane 4, 10/~g of kidney mRNA. The filter was hybridized with nick-translated HaeIII fragment (nucleotide numbers 50-323 in Fig. 2A,B) in 10 mM sodium phosphate buffer (pH 7.0) solution containing 50% formamide, 0.75 M NaCI, 25 mM EDTA, 0.2% sarcosyl, ×1 Denhardt and sheared DNA (200 #g/ml) at 45 °C for 20 h. The filter was washed with ×0.1 SSC and 0.1% sarcosyl at 50 °C. The hybridized band (about 14.5 S) was estimated by using bovine (28 S and 18 S) and E. coli (23 S, 15 S and 5 S) rRNA as molecular size marker. B: 2/~g of the retina, brain, liver and kidney mRNAs (1-4) were dot-blotted on a filter, hybridized with fl-actin cDNA probe and washed as described above.

A 0

0.5

L

i

1.0

l.Skb ,

pc.-470

. - 23 n-27 HO S

A P

B - 60

A A O A G G ~ m ~ CAOACAOTAATTCTTATCCACTGAA CACAATCAAGGACACCAGGCACAGAAAAICTAAACTIATCATATCAAAACTAAATCTCTAAA

1

ATOOAAAAAOCAAAAAGCCAAAOTTTGGAOOAAOACTTTOAAGOACAOGCCTCACATACA

-98

61 121

MetOluLy,AlaLy,SerOlnSerLQuOluOluAspPheOluOlyOlnAXaSerHtsThr

20

GGACCCAAAOGAOTAATAAATOATTGGAQAAAGTTTAAATTGGAGAGTOAAGATAGTGAT GlyProLysGlyV&lIleAsnAspTrpArlLysPheLysLeuOluSerGluAspSerAsp

40

TCAOTTOCACACAGCAAOAAGGAOATTCTCAGACAAATGTCTTCTCCTCAGAGTAGAGAT SerValAlaHisSerLysLysOlulleLeuArgGlnMetSerSerProGlnSerAr|AsP

60

181

OACAAAOACTCAAAAOAAAGATTCAQCAOAAAOATOAGCGTTCAAGAATATGAACTAATC

Z41

eACAAAGACAAAGAAOATGAAAATTGCCTTCGTAAATACCOCAGACAOTOTATGCAGGAT

11isLysAspLysGluAspGluAsnCysLeuAr|LysTyrArgA~|GlnCysNetG1nAsp

I00

301

ATQCACCAGAAOCTGAGTTTTGGGCCTAGATATGGGTTTGTGTATGAGCTGGAATCTQGG HetHllQlnLysLeuSerPheOlyProArlTyrGlyPheValTyrOluLeuGluSerOly

120

361

GAGCAATTCCTOOAAACCATTGAAAAGGAACAGAAAATCACCACTATCGTTGTTCATATT GluOlnPheLeuOluThrlleGluLysOluGlnLyslleThrThrlleValValHisIle

140

421

TATGAAGATOGTATTAAGGGCTOTGATGCTCTAAACAOTAGCTTGATATGCCTTGCAGCC TyrGluAspGlylleLysGlyCysAspAlaLeuAsnSerSerLeulleCysLeuhlsAla

160

481

OAATACCCTATGGTCAAGTTTTOTAAAATAAAGGCTTCTAATACAGGTGCCGGAGACCGC

AspLysAspSerLysOluAr|PheSarArgLysHetSerV~lGlnGluTyrGluLeulle

GluTyrProMetValLysPheCysLyslleLysAlaSerAsnThrGlyAlsGlyAspArg

80

180

541

TTTTCCTCAGATGTACTCCCCACGCTGCTTOTCTACAAAG6TGGGGAACTCCTAAGCAAT

601

TTCATTAGTGTTACTGAACAGCTOOCTGAAGAATTTTTTACTGGGGATGTGGAGTCTTTC PhelleSerValThrGluGlnLeuAlaOluGluPhePheThrGlyAspValGluSerPhe 220

661

CTAAATOAATATGGGTTATTACCTGAAAAGAGATGCATGTCCTAGAGCAGACGAACATGG LeuAsnGluTyrGlyLeuLeuProGluLysArgCysHetSer***

721 781

AAGAGOATATGGAATAAAGATTCACTATGTCAATQTTTCATATTTCGCATTTCTCCTTTA AGCATTGAACACTGATTTTGGTAGTATTCACATTCTTTTAGGGAATACCAAACATAGCCC TOGCTTTTCTAATTTGOOOAAGAAAAACTCCAGACTGACACTAAAATTATATGATTAGCA TGTCTTAATATTAGTTACTCAAGCTGATATAACACTTTACCTCAAAACATTGTAGTCTTC AOCAATATOTTAGTAGACAAAGAOAATATGAAAAATACTGTGAAAGTTTTCAAAAATCCC TTTCAAGTTATATGTTGGCTTTTTTATTCCATOCTTTTCCTTATTGTTATTACTAGGTTT TTCAAATTATATTOTTAATTATAAATTTGCTTGTGTAATAAGAATAAAGTCTCGTAAGG p o l y ( A ) 75-85

841 901 951 1021 1081

PheSerSerAspValLeuProThrLeuLeuValTyrLysGlyQlyGluLeuLeuS~rAsn

200

234

Fig. 2. Restriction enzyme maps and amino acid sequence of MEKA protein predicted from the nucleotide sequence of MEKA cDNA. A: restriction endonuclease maps of pCR-470, R-23 and R-27 and strategy for the sequencing of pCR-470 and R-27. The vertical lines indicate the restriction enzyme sites and the horizontal arrows show the directions and ranges of the sequences read. The solid boxes and open boxes indicate the coding and non-coding regions, respectively. A, AluI; D, DraI; H, HaelII; P, PvulI; R, RsaI and S, StuI. B: nucleotide sequence of MEKA (R-27) cDNA and a deduced amino acid sequence (MEKA protein). Nucleotide residues are numbered on the left side, beginning with the first residue of the ATG triplet encoding the initial methionine, and the nucleotides on the 5"-side of the residue are indicated by negative numbers. The deduced amino acid sequence of the MEKA protein is shown on the right side beginning with the initiating methionine. The two underlined nucleotide sequences are polyadenylation signal, and *** indicates the stop codon. A boxed sequence and arrows indicate palindromic and symmetric sequences, respectively. nuclear e m u l s i o n ( K o d a k ) , they w e r e e x p o s e d for 2

M E K A fusion protein and anti-MEKA serum

w e e k s at 4 °C, d e v e l o p e d and counter-stained with hematoxylin.

P E X 1-5" M E K A

T h e n u c l e o t i d e s e q u e n c e at the j u n c t i o n of the 3" p l a s m i d was d e t e r m i n e d and the

+5

7~

extracted from the gel as described by W h i t e and Wilcox 19. A b o u t 150/~g of the M E K A fusion protein was recovered from 40 ml of LB culture medium. Purified M E K A fusion protein shows a single band on a 7.5% S D S - P A G E . It was injected into two rabbits to p r o d u c e antiserum.

B

Western blot analysis and immunohistochem&try Number of amino acid

234

Fig. 3. Hydropathy analysis of MEKA protein. The hydropathy profile of the MEKA protein was computed according to the method of Kyte and Doolittle9 and plotted with a one-residue interval. The window size is 17 residues.

expected sequence was obtained (Fig. 5). P E X I - M E K A clone was grown in the presence of 50 pg/ml of ampiciilin at 30 °C until 0.2 O . D . at 550 nm and then incubated at 42 °C for 2 h. E. coli was precipitated by centrifugation at 5000 rpm for 10 min and suspended in 50 mM Tris-HCl (pH 7.5) and 1 mM E D T A TM. It was incubated for 30 min at 4 °C in the presence of 1% Triton X-100 and 1 mg/ml lysozyme and stored at - 8 0 °C until use. Insoluble materials containing /3-galactosidase-MEKA fusion protein were analyzed on a 5% S D S - P A G E of 2 mm thickness. A 140 k D a of M E K A fusion protein was

A

.......

B

Bovine retina and various tissues from 2-month chicken were r e m o v e d and h o m o g e n i z e d with 30 m M sodium phosphate buffer (pH 7.4), 0.1 m M E D T A and 0.05 mM DTT. Each of the soluble fractions was obtained by centrifugation at 100,000 g for 2 h. The particulate fractions were washed with 30 rnM sodium p h o s p h a t e buffer (pH 7.4) and 0.05 mM DTT. The soluble and particulate fractions were used for the Western blot analysis as described previously ~. A soluble fraction (about 500 mg protein) from 30 retinas of bovine was applied to a D E A E ion exchanger A-500 column (2.6 × 34 cm) equilibrated with 30 m M sodium p h o s p h a t e buffer (pH 7.4), 0.05 m M D T T (Buffer A). The column was washed with Buffer A and then with 100 m M NaC1 containing Buffer A. The fractions containing M E K A protein (100 mg) were eluted with 250 mM NaC1 in the presence of Buffer A. A n aliquot of 5 mg protein in the fractions was dialyzed against 5

C

Fig. 4. In situ hybridization histochemistry with opsin and MEKA RNA probes. Rat retinal sections were hybridized with anti-sense opsin (A), anti-sense MEKA (B) or sense MEKA (C) RNAs. Both of the cRNA probes showed grains most intensely at inner segments of photoreceptor cells. The following are illustrated: OS, outer segment; IS, inner segment; ONL, outer nuclear layer; OPL, outer plexiform layer; INL, inner nuclear layer; IPL, inner plexiform layer and GCL, ganglion cell layer.

MEKA cDNA ( ME)CA-protein )

PEX l plasmid ( cro-fl-galactosidase )

50 +

CCG TCA GTA TCG GCG GCC C GG GGA - Pro Ser Val Ser Ala Ser

- - CAG G CC TCA CAT - - GAA CAG CTG - I

I ~

SmaI

620 Ser His

StuI

Glu GlnILeu PvuII

Ligation PEX 1 - MEKA plasmid ( c r o - ~ - galactosidase - MEKA fusion protein )

5'

oco Ala -

c

-

3' S'

clcc

I

B

m

- -

Ser ( P r o ) S e r His - - Glu Gln

CGC CGG G GG

/

3'

AGT GTA

\

--

N u c l e o t i d e sequence a t j u n c t i o n o f 3' PEX 1 - 5' MEKA p l a s m i d

0

A

Fig. 5. Construction of PEX I-MEKA plasmid. MEKA cDNA was digested with StuI and PvulI, separated on a 1% agarose gel and then the 570 bp fragment of coding region was recovered. PEX 1 plasmid was digested at the SmaI site from which the stop codon is located at down stream, and ligated with the 570 bp fragment. PEX 1-MEKA clones were isolated by colony hybridization by using the 570 bp as a probe. The direction of the inserted 570 bp was determined by DraI and PstI. The expected nucleotide sequence at the junction of 3" PEX 1-5" MEKA plasmid was obtained as illustrated. mM Tris-HC! (pH 7.0), and 0.05 mM DTI" and was further separated by preparative isoelectric focusing (Rotofor Cell, Bio Rad). Resulting subfractions were dot-blotted on a nitrocellulose filter and reacted with anti-MEKA serum. The fractions containing M E K A protein from the D E A E column was further purified by Sephadex G75 and Mono Q HR5/5 column chromatographies. The partially purified M E K A protein was used for immunohistochemistry (Fig. 9F, G). The procedure for fixation and fluorescent immunohistochemistry with antiopsin (1/1000 dilution) and anti-MEKA fusion protein (1/500 dilution) preabsorbed with proteins of PEX 1 clone was carried out as described previously s.

tained from Amersham. Nitrocellulose and nylon filters were purchased from Schleicher and Schuell, and Pall Ultrafine Filtration Corporation, respectively. M13 sequencing kit and restriction enzymes were obtained from Toyobo. D E A E Cellulofine (A-500) was obtained from Seikagaku Kogyo. FITCand peroxidase-conjugated anti-rabbit IgGs were obtained from Kirkegaard and Perry Labs. BCA protein assay reagent for the determination of protein was from Pierce. PEX expression vectors and mouse fl-actin c D N A containing whole coding region were gifted by Dr. K.K. Stanley and Dr. A.J. Minty, respectively.

Chemicals a-[32p]dCTP (spec. act. 3000 Ci/m mol), a[25S]UTP (400 Ci/mol), and SP6 system were ob-

Isolation of a retina-specific cDNA clone (R-27) of full length The pCR-480 c D N A is a clone obtained from a

RESULTS

bovine retinal cDNA library preparared by the G:C tailing method 7. By the Northern hybridization analysis, the pCR-470 PstI fragment (410 bp in Fig. 2A) hybridized to 14.5 S (about 1.3 K nucleotides) bovine retinal mRNA, but not to mRNAs prepared from bovine brain, liver of kidney (data not shown). We concluded that the pCR-470 was another retinaspecific cDNA clone. Therefore, the PstI fragment of the pCR-470 was utilized to select a full length cDNA from another bovine retinal cDNA library constructed by the Okayama-Berg method ~5. An R-27 clone was isolated in the third series of the colony hybridization experiment by a HaeIII fragment (274 bp, nucleotide number 50-323 in Fig. 2A,B) of R-23 clone. The HaeIII fragment, which

A

B

contains a coding region, also hybridized to 14.5 S retina mRNA, but not to mRNAs from the brain, liver or kidney (Fig. 1A). fl-Actin cDNA probe, however, shows the almost similar hybridization signals among the tested mRNAs (Fig. 1B). Nucleotide sequence of R-27 cDNA and M E K A protein Strategy for determining the sequence of the pCR 470 and R-27 cDNAs, and the nucleotide sequence of the R-27 and the deduced polypeptide sequence are illustrated in Fig. 2A and B, respectively. We confirmed that the nucleotide sequence of the pCR-470 was identical and corresponds to nucleotide numbers 678-1087 of the R-27 clone. The R-27 cDNA has almost a full length of about 1.3 kb consisting of a 5" non-coding region (98 bp), a coding region (702 bp) and a 3" non-coding region (437 bp excluding the 75-85 bp of poly(A) tail). In a 5"

~! M KO

1 2 3 4 5 6 7 KO

931 45

22

., 1 4 0 K -.120K 11eK "

Fig. 6. Synthesis of MEKA protein in E. coli. Two of the PEX 1-MEKA (A,B) and PEX 1 clones were grown as described in Materials and Methods. Proteins in each of the clones were separated by 5% SDS-PAGE and visualized with Coomassie brilliant blue. Two arrows, 120 kDa and 140 kDa, indicate cro-fl-galactosidase and MEKA fusion protein, respectively.

Fig. 7. Western blot analysis of MEKA protein. Soluble proteins from bovine retina (lane 1), and the retina (2), cerebrum (3), cerebellum (4), liver (5), heart (6) and intestine (7) of chicken were prepared, and 50/ag protein of each sample except bovine sample (25 /~g) were separated on a 14% SDS-PAGE with marker proteins (M). The proteins were transferred to a nitrocellulose filter and reacted with antiMEKA protein diluted 1/500. Immunoreactive proteins were visualized by the reaction of anti-rabbit IgG-peroxidase and 3,3"-diaminobenzidine. The position of the marker protein on the filter was determined by using a biotin-blot protein detection kit (Bio-Rad). Arrows (32 and 33) indicate the molecular size (kDa) of MEKA protein in bovine and chicken retinal soluble proteins, respectively.

pH 12" 0

2

3

4

.

•

•

e

7

•

*

~o -

~2 ' 1 0 " D ~ 2 8 O n m

I

.1.2

I

I

010-

Q.8

O.4

0 F r a c t t o n number

Fig. 8. Isoelectric focusing of bovine MEKA protein. A fraction (about 5 mg protein) containing MEKA protein eluted from DEAE ion exchanger was mixed with 2.5 ml of BioLyte (3/10, Bio-Rad) in a total volume of 50 ml. Preparative isoelectric focusing was performed at 12 Watt for 3 h and fractionated into 20 fractions. Absorbance at 280 nm and pH in each fraction were measured. The insert shows the result, with aliquots of the sample dot-blotted on a nitrocellulose filter and reacted with anti-MEKA protein (1/500 dilution) as in Fig. 7.

non-coding region, a palindromic sequence and a symmetric sequence were found. In a 3" non-coding region, two polyadenylation signals, A A T A A A , at nucleotide numbers 733 and 1123 were found. We designated the protein deduced by R-27 cDNA as a M E K A protein, because its amino acid sequence starts from Met(M)-GIu(E)-Lys(K)AIa(A). By computer homology research, we found that there are no significant homologous proteins or nucleotide sequences with M E K A protein or with the M E K A cDNA. Bovine and human genome D N A fragments digested with various restriction enzymes were separated on an agarose gel, transferred to a nylon filter and analyzed by the Southern hybridization method. Each of the digested genome DNAs showed a single hybridization band, indicating that the M E K A protein gene has a single copy number in both bovine and human genome (data not shown). The molecular weight and pI of the M E K A protein were calculated to be 26,935 (234 amino acid residues) and 5.35, respectively. No transmembrane

sequence was observed by the hydrophobicity plots of amino acid sequences of M E K A protein as shown in Fig. 3. The M E K A protein contains 24 serine and 8 threonine residues, suggesting a possibility that M E K A protein serves as a substrate for and acts through phosphorylation. It has been reported that the consensus sequences of cyclic AMP-dependent serine phosphorylation sites are -Arg-Arg-Y-Serand Lys-Arg-X-Y-Ser4"6. Two possible phosphorylation sites in M E K A protein are found at amino acid residues 73 and 234 in amino acid sequences -Arg-Lys-Met-Ser- and -Lys-Arg-Cys-Met-Ser. A possible glycosylation site was also found at amino acid residue 153 in amino acid sequence -AsnSer-Ser-. Another distinctive feature of M E K A protein is an abundance of methionine (7 residues) and cysteine (6 residues).

In situ hybridization histochemistry of the retina A StuI (nucleotide number 50 in Fig. 2B)-PvulI (621) fragment of M E K A c D N A was ligated with SP64 vector and transferred into HB101 cells. Both directions, determined by digestion of DraI and HindlII, of the inserted D N A were used. Opsin cDNA in SP64 vector was also constructed and the anti-sense R N A was used as a positive control probe. Autoradiograms of sections in rat retina with R N A probes are illustrated in Fig. 4. The anti-sense opsin probe (Fig. 4A) labeled the inner segment most intensely as described by Brann and Young 3. The M E K A cRNA hybridized with the inner segment and outer nuclear layer of photoreceptor cells (Fig. 4B). When a sense R N A probe for M E K A m R N A was used, no regional labeling w.as observed (Fig. 4C). The result in Fig. 4 also suggests that opsin m R N A is more abundant than M E K A m R N A . An anti-sense T7 R N A probe stained the inner segment, but the number of grains was less than that of M E K A cRNA (data not shown). Therefore, we concluded that m R N A contents in rat photoreceptor cells appear to be in the following order: opsin > M E K A > T7.

Expression of MEKA protein in E. coli Construction of PEX 1-MEKA plasmid is illustrated in Fig. 5 and described in Materials and Methods. The StuI-PvulI fragment of the M E K A cDNA encoded about 20 kDa protein was ligated

Fig. 9. Immunohistochemistry with anti-opsin and anti-MEKA. Retinal sections were reacted with anti-opsin (A) and anti-MEKA serum (B-G). Chick retina, A and B; rat retina, C; mouse retina, D; bovine retina, E-G. Anti-MEKA serum was preabsorbed without (E) or with 1 ~g/ml (F) or 10/~g/ml (G) of purified bovine MEKA protein for 30 min at 37 °C. Arrows indicate a position of the outer plexiform layer. Bars = 50/~m.

with P E X 1 at the SmaI site. The nucleotide sequence at the junction of 3" P E X 1-5" M E K A plasmid was determined as shown in Fig. 5. Two r e c o m b i n a n t P E X 1 - M E K A clones (A and B) and the P E X 1 clone were grown at 42 °C, and then their proteins were analyzed by 5% S D S - P A G E (Fig. 6). Both of the A and B clones synthesized 140 kDa

protein, while PEX 1 clone synthesized 120 kDa protein (cro-fl-galactosidase). The 140 kDa ( M E K A fusion) protein was recovered from the gel and used as an antigen.

Characterization of MEKA protein By using a n t i - M E K A serum, Western blot anal-

ysis was carried out with soluble fractions prepared from bovine retina, and the retina, cerebrum, cerebellum, liver, heart and intestine of chicken. The soluble fractions were separated by the 14% SDS-PAGE, transferred to a nitrocellulose filter and reacted with anti-MEKA. The anti-MEKA serum reacted with 32 kDa and 33 kDa proteins of bovine and chick retinal soluble proteins, respectively. However, it did not react with soluble fractions from the other tissues (Fig. 7). The washed particulate fraction of bovine and chick retinas did not show a positive reaction on Western blot analysis (data not shown). We concluded that MEKA protein is a retina-specific soluble protein. A fraction eluted from the D E A E column was separated into 20 fractions by a preparative isoelectric focusing apparatus and each fraction was dotblotted on a nitrocellulose filter. The anti-MEKA serum reacted with fraction number 4 and 5, whose pI corresponded to 4.3 and 5.1, respectively (Fig. 8).

Immunohistochemistry of the retina with anti-MEKA protein The anti-opsin as a control and anti-MEKA protein were reacted with retinal sections from various species. We found that the anti-MEKA serum showed the same positive reaction with the photoreceptor cells of various species that anti-opsin had shown (Fig. 9A), in accordance with the data of in situ hybridization study (Fig. 4). Staining of the outer segment, inner segment and the outer plexiform layer was also observed in chick retina (Fig. 9B). However, different staining patterns in the photoreceptor cells were observed among species. In the rat and mouse retinal sections, anti-MEKA serum mainly stained the inner segment (Fig. 9C,D), whereas mainly outer segments show positive reaction in bovine retina (Fig. 9E). When anti-MEKA serum was preincubated without (E) or with 1/~g/ml (F) or 10 /~g/ml (G) of purified bovine MEKA protein (about 30% purity), immunoreaction had completely disappeared at 10/~g/ml of the MEKA protein. DISCUSSION By the differential colony hybridization method, we have already isolated several retina-specific

cDNA clones, suggesting that the method is very useful for isolating cDNA clones unique to the retina. Recently, McGinnis et al. 11 have reported the isolation of photoreceptor-specific cDNA clones by the differentiation of retinal, proteins between normal and degeneration (rd) mice. We have chosen pCR-470 as a candidate for retina-specific clones 7 and cloned almost a full length cDNA by using pCR-470 as a probe, and R-27 clone was isolated. By the Northern hybridization the MEKA mRNA was expressed only in the retina among the bovine tissues tested, and the mRNA size was about 1300 nucleotides (Fig. 1A). From the colony hybridization experiment, about 0.3% of the total colony of the bovine cDNA library exhibited positive signals by the HaelII (nucleotide number 50-323 in Fig. 2A,B) probe. The results of in situ hybridization (Fig. 4A,B) also suggest the abundance of the MEKA mRNA in the retina. The opsin mRNA content in the retina was estimated to be about 1.6% by precipitating with anti-opsin in the labeled peptides conducted by total retina poly(A) mRNA in vitro 7. From the results of colony and in situ hybridization experiments, the content of the MEKA mRNA is roughly estimated to be a range of one tenth percent of the total retina poly(A) mRNA. By the Southern hybridization analysis, we found that the MEKA genome is a single copy number both in bovine and human genome (data not shown). We have synthesized the MEKA protein as a fusion protein in E. coli. This is the first report on the synthesis of retina-specific protein in E. coli. We have also attempted to express opsin protein in E. coli, but the transformed E. coli did not grow well. From one liter culture of the PEX 1-MEKA clone in LB medium, about 4 mg of the MEKA fusion protein could be recovered from 5% SDS gel electrophoresis. The amount of the fusion protein was enough to make anti-MEKA serum in rabbits. By Western blot analysis using the anti-MEKA, it was also proven that MEKA protein is a retinaspecific protein. By computer analysis, no homologous sequences with MEKA protein were found. We think that M E K A protein is a novel photoreceptorspecific protein. The molecular size (32 kDa) and pI (4.3-5.1) of native MEKA protein of bovine retina (Figs. 7 and 8) are reasonable values, compared with

10 those estimated from the nucleotide sequence of the M E K A cDNA. The size and content of chick M E K A protein (33 kDa) are about 1 kDa larger and about one-third, compared to that of the bovine, respectively. The localization of the M E K A protein was also studied on various retinal sections by using antiM E K A serum. The distribution of M E K A protein differed slightly among species, although the photoreceptor cells were intensively stained with the a n t i - M E K A in all species tested (Fig. 9). AntiM E K A stained both inner and outer segments in chicken retina, the inner segments in rat and mouse, and the outer segments in bovine retina. No positive immunoreaction was observed with sections from mouse cerebrum, cerebellum and hippocampus (data not shown). Philp et al. ~6 have reported that transducin and arrestin move between the outer segment and inner segment by dark and light adaptation. The different distribution of the M E K A protein in photoreceptor cells observed in this study may be due to species differences and the experiREFERENCES 1 Baehr, W. and Applebury, M.L., Exploring visual transduction with recombinant DNA techniques, Trends Neurosci., 9 (1986) 198-203. 2 Bloch, B., Popovici, T., Le Guellec, D., Normand, E., Chouham, S., Guitteny, A.E and Bohlen, P., In situ hybridization histochemistry for the analysis of gene expression in the endocrine and central nervous system tissues: a 3-year experience, Mol. Brain Res., 1 (1986) 251-260. 3 Brann, M.R. and Young iii, S.W., Localization and quantitation of opsin and transducin mRNAs in bovine retina by in situ hybridization histochemistry, FEBS Lett., 200 (19861 275-278. 4 Feramisco, J.R., Glass, D.B. and Krebs, E.G., Optimal spatial requirements for the location of basic residues in peptide substrates for the cyclic AMP-dependent protein kinasc, J. Biol. Chem., 255 (1980) 4240-4245. 5 Hubbel, W.L. and Bownds, M.D., Visual transduction in vertebrate photoreceptors, Annu. Rev. Neurosci., 2 (1979) 17-34. 6 Krebs, E.G. and Beavo, J., Phosphorylation of enzymes, Annu. Rev. Biochem., 48 (1979) 923-959. 7 Kuo, C.-H., Yamagata, K., Moyzis, R.K., Bitensky, M.W. and Miki, N., Multiple opsin mRNA species in bovine retina, Mol. Brain Res., 1 (1986) 251-26(I. 8 Kuo, C.-H., Tamotsu, S., Morita, Y., Sbinozawa, T., Akiyama, M. and Miki, N., Presence of retina-specific proteins in the lamprey pineal complex, Brain Res., 422 (1987) 147-151. 9 Kyte, J. and Doolittle, R.F., A simple method for displaying the hydropathic character of a protein, J. Mol.

mental conditions under which eyes were isolated. We feel that the staining pattern was slightly altered by light and dark, but more precise experiments are necessary. It is also interesting to know whether the M E K A protein is expressed in rod and/or cone cells. Since the M E K A protein appears to exit in both cone and rod cells, electron microscopic study is under investigation. At present, the physiological function of the M E K A protein is u n k n o w n . It is, however, interesting to note that the M E K A protein has two consensus sequences which may be phosphorylated by cyclic AMP-, cyclic G M P - d e p e n d e n t or other protein kinase. Recently, Shinozawa and Yoshizawa 17 have reported the presence of two soluble proteins (pI 5.2), which are phosphorylated by cyclic GMP- or AMP-kinase, in the rod outer segment. The M E K A protein may be different from the two proteins (10.5 and 8.5 kDa) in molecular size. Purification of M E K A protein is in progress and will offer a clue to the physiological function of the M E K A protein in the photoreceptor cells. Biol., 157 (1982) 105-132. 10 Maniatis, T., Fritsch, E.E and Sambrook, J., Molecular Cloning, A Laboratory Manual, Cold Spring Harbor Laboratory, Cold Spring Harbor, NY, 1982. 11 McGinnis, J.F. and Leveille, P.J., A biomolecular approach to the study of the expression of specific genes in the retina, J. Neurosci. Res., 16 (1986) 157-165. 12 Messing, ,1., New M13 vectors for cloning, Methods Enzymol., 101 (19831 20-78. 13 Miki, N., Kuo, C.-H., Nakagawa, Y., Bitensky, M.W., Isihii, K., Shiosaka, S. and Tohyama, M., Cloning and characterization of retina-specific cDNAs. In Y. Tsukada (Ed.), Molecular Genetics in Developmental Neurobiology, Japan Scientific Societies Press, Tokyo, 1986, pp. 167-176. 14 Nicol, G.D. and Miller, W.H., Cyclic GMP injected into retinal rod outer segments increases latency and amplitude of response to illumination, Proc. Natl. Acad. Sci. U.S.A., 75 (1978) 5217-5220. 15 Okayama, H. and Berg, P., High-efficiency cloning of full-length cDNA, Mol. Cell. Biol., 2 (1982) 161-170. 16 Philp, N.J., Chang, W. and Long, K., Light-stimulated protein movement in rod photoreceptor cells of the rat retina, FEBS Lett., 225 (19871 127-132. 17 Shinozawa, T. and Yoshizawa, rE, Cyclic nucleotidedependent pbosphorylation of proteins rod outer segments in frog retina, J. Biol. Chem., 261 (1986) 216-223. 18 Stanley, K.K. and Luzio, .I.P., Construction of a new family of high efficiency bacterial expression vectors: identification of cDNA clones coding for human liver proteins, EMBO J., 3 (19841 1429-1434. 19 White, R.A.H. and Wilcox, M., Protein products of the bithorax complex in Drosophila, Cell, 39 (1984) 163-171.