Modulation of cholecystokinin (CCK) gene-expression in a human neuroblastoma cell line: Effects of serum on enhanced CCK and c-FOS mRNA expression

Modulation of cholecystokinin (CCK) gene-expression in a human neuroblastoma cell line: Effects of serum on enhanced CCK and c-FOS mRNA expression

Neuropeptides (1992) 23,107-113 0 Longman Group UK Ltd 1992 Modulation of Cholecystokinin (CCK) Geneexpression in a Human Neuroblastoma Cell Line: Ef...

743KB Sizes 0 Downloads 39 Views

Neuropeptides (1992) 23,107-113 0 Longman Group UK Ltd 1992

Modulation of Cholecystokinin (CCK) Geneexpression in a Human Neuroblastoma Cell Line: Effects of Serum on Enhanced CCK and c-FOS mRNA Expression H. J. MONSTEIN*, K. PEDERSEN*, and P. M. HAAHRt *Department of Clinical Biochemistry (KK 3013) and tMedical Immunology, State University Hospital (Rigshospitalet), DK-2100 Copenhagen, Denmark (Reprint requests to HJM)

Abstract -The effect of fetal-calf serum (FCS) and Forskolin (FKN) on cholecystokinin (CCK) and proto-oncogene c-fos mRNA expression in the human neuroblastoma cell line SK-N-MC, cultured in serum free medium was studied by Northern blot analysis and nuclear run-off transcription analysis. Addition of FCS or FKN gradually increased the basal CCK mRNA level -four to six-fold after 2-4 h. In contrast, a strong and transient increase of the c-fos mRNAlevel was observed, -forty to fifty-fold after !50-60 min over unetimulated cells. Nuclear run-off transcription analysis indicates that c-fos mRNA is constitutivefy expressed and transcription may be further stimulated by FCS and FKN. Moreover, in SK-N-MC nuclei, transcription of the c-fos gene clearly precedes stimulated CCK-mRNA expression. This suggests that FOS, which is known to form a AP-1 heterodimer transcription complex with the proto-oncogen product, Jun, may bind to the tentative AP-1 binding site, found within the human CCK promoter and thereby modulates basal and enhanced CCK-mRNA expression in SK-N-MC cells. Introduction Control of cholecystokinin (CCK) and the protooncogene c-fos gene expression involves the induction of transcription factors which bind to specific cis-acting DNA elements, located in the 5-flanking promoter region of each of these genes. DNA binding analysis and deletion analysis within the promoter regions of the rat CCK- and human c-fos genes, fused to the bacterial Cat gene revealed the

Date received 23 March 1992 Date accepted 2.5 May 1992

presence of DNA elements responsive to CAMP and phorbolesterregulatedexpression (l-5). Expression of c-fos is induced by a variety of agents such as CAMP, phorbolester, and serum. A serum and phorbolester inducible element (SRE, TRE) is found within a dyad symmetry element @SE) located at -292 to -3 17 distal to the mRNA initiation site. This element has been shown to be essential for the induction and control of c-fos gene expression by serum, phorbolester, and EGF (3, 4, 5). Moreover, it has recently been demonstrated that an AU-rich element (ARE), located in the 3’-noncoding region of the cfos gene is responsible for c-fos mRNA decay (6-9). 107

108 In previous studies we showed that the human neuroblastoma cell line SK-N-MC expresses CCK as well as c-fos mRNA and that the mRNA level of these two genes is differently modulated by CAMP and phorbolester induced pathways between 4-24 h of drug treatment (10-12). These and similar studies, however, were carried out in SK-N-MC cells cultured in fetal calfserum(FCS) supplementedmedium(l2,13). As it is well established that FCS contains factors that can stimulate or inhibit cell growth and transcriptional processes (3,4), we performed experiments to investigate the effect of FCS and Forskolin (FKN) on CCKand c-fos mRNA expression in the human neuroblastoma cell line SK-N-MC within the first few hours after FCS and FKN addition.

Materials and Methods Cell cultureand drug treatment SK-N-MC cells were originally obtained from the

NEUROPEPTIDES

Dept. of Pathology, Uppsala University (Uppsala, Sweden) and cultured as described earlier (11, 12). SK-N-MC cells from passage 120 were cultured in Nunc-petri dishes (10 cm) at 37°C in 10 ml Ham’s F12 and Dulbecos modified Eagles medium (1: 1), supplemented with 15% fetal calf serum (FCS), 10 mg/l nonessential aminoacids (GIBCO, 1000 X stock solution), 190 mg/l L-glutamine, and 10 ml/l Penicillin/Streptomycin (10 000 U and 10 000 &ml) in humidified air containing 10% COz. For drug treatment experiments the cells were seeded in petri dishes and cultured in serum free medium under the conditions described above (Fig. 1). After 16-24 h the medium was changed and 15% fetal calf serum or 10 pM Forskolin (FKN) was added. Cell cultures were grown to a density of -10’ cells/dish and harvested in phosphate-buffeted saline (137 mM NaCl, 2.7 mM KCl, and 1.5 mMNa2HP0&H~P04; pH 7.4) supplemented with 0.5 mM EDTA at the time points as indicated.

Fig. 1 Phase contrast photomicrograph of SK-N-MC cells, cultured in FCS-free medium as described in Material and Methods. Magnification was 200 X.

CCK GENE EXPRESSION IN HUMAN NEUROBLASTOMA CELL LINES

Preparation of total RNA and Northern blot analysis Total RNA from 5 x 1O6cells was extracted according to the method of Chomczynski and Sacchi (14), and quantitated by ultraviolet absorption at 260 mm. Electrophoretic separation of total RNA was performed in a 1.2% agarose-0.7% formaldehyde gel followed by transfer onto Hybond-N membrane (Amersham) in 25 mM Na-phosphate buffer @H 6.4). Prior to transfer, RNA preparations were checked for degradation by ethidium bromide staining. Filters were fixed by crosslinking in a W-box (Stratagene). Hybridization probes

and generation

of antisense RNA

Hybridization was performed as described earlier (12, 15). Following hybridization, bands were visualized by exposure for l-3 d to Fuji RXO-1G or Amersham MP films at -80°C using Du Pont intensifier screens: Densitometric scanning of the Northern blots was carried out on a LKB-Pharmacia XL laser densitometer. The level of CCK and c-fos n-RNA signals was estimated in arbitrary units, which was determined relative to the mouse p-actin hybridization signal. The construction and in vitro transcription of the mouse p-actin DNA template was recently described (16). The c-fos DNA template was purchased from Amersham. For each hybridization probe, -1 pg plasmid DNAwas in vitro transcribed either by RNA polymerase SP6 or T7 using Promegas Riboprobe Kit and 50 uCi [32P]a-UTP (400 Ci/mmol). PCR-cloning of a human CCK cDNA probe The human CCK probe phCCK-42 1 was prepared as follows; total RNA from the human cell line SK-NMC was extracted and poly A’ was purified by magnetic oligod(T) Dynabeads (Dynal, Norway). cDNA was prepared by the RiboClone cDNA synthesis kit from Promega and used as template for polymerase chain reactions (PCR), using a sense up-stream Bam Hl linker-primer (5’-CATCGGATCCAAACGCCAAGCCAGCTGCCGTCCTAAT-3’) located in exon one, and an antisense down-stream EcoRl linker-primer (5’-ACGGAATTCTTGGGAGGTTGCTTCCGTTGGGCT-3’), located in the third exon of the CCK-gene according to the sequence in Takahashi et al (17). The PCR product was a frag-

ment of 42 1 base pairs containing part of exon of exon 2, and the coding region of exon 3. The ment was cloned into the polylinker of the scription vector pGEM-1 1Zf (-) (Promega) sequenced.

109 1, all fragtranand

Nuclear run-off transcription analysis The procedure of Greenberg et al (18) with some modifications as described below was used. SK-NMC cultures in 10 cm petri dishes (-0.3-0.4 x lo6 cells/dish) were treated with FCS (5 dishes per experiment; 1.5-2.0 x 10’ cells). Cells were harvested, washed and lysed in a Dounce homogenizer (pestle B, 10 strokes). Nuclei were pelleted by centrifugation at 4°C for 5 min at 500 x g and resuspended in 200 pl ice cold glycerol storage buffer (40% Glycerol, 50 mM Tris-HCl, pH 8.3; 5 mM MgC12and 0.1 mM EDTA) and stored at -80°C until used. For RNA isolation a buffer containing -150 uCi [a-“P] UTP (800 Ci/mmol, Amersham Inc.) was added to the purified nuclei and incubated at 30°C for 30 min. Cellular DNA was digested with 260 U DNase (RNase-free, Boehringer, Manheim) at 30°C for 15 min in the presence of 25 ug tRNA. RNA was purified by proteinase K digestion at 45°C for 30 min, followed by phenolchloroform-isoamylalcohol extraction (25:24:1). The aqueous phase was re-extracted with chloroform and adjusted with Qiagen-QAT-buffer and purified over a pre-equilibrated Qiagen-20 column as recommended by the producer (Qiagen, GmbH, Dtisseldorf). Hybridization filters with 0.5 pg of denatured cDNA equivalents filtered onto Zeta Probe nylon membranes (BioRad) in a dot blot manifold were prepared as follows: mouse S-actin probe; plasmid phCCK-42 1 containing a 42 1-base-pair (bp) human CCK cDNA insert (see PCR cloning); a human proenkephalin A cDNA (11); and a human c-fos cDNA (Amersham Inc., Birkerod, Denmark). Membranes were baked at 80°C for 2 hand pre-equilibrated with hybridization buffer for 4-6 h at 42°C (50% formamide; 0.75 M NaCl; 0.5% SDS; 2 mM EDTA; 20 pg/ml poly A; 500 pg/ml denatured salmon sperm DNA; 50 mM Hepes buffer pH 7.5, and 10 x Denhardts solution). The hybridization was performed with equalized amounts of [ 3ZP]-labelled run-off RNAs at 42°C for 48-60 h followed by 2 washes for 30 min at 25°C in 2 x SSC, 0.1% SDS, and 2 washes in 0.1 x SSC, 0.1% SDS at 60°C for

110

NEUROPEFTIDES

CCK +FCS

‘+ FCS/lOmin

t

20’

t

30’

t

401

C-f08 Actin CCK 50’

*

60’

t

120’

t

180’

t

240’

Fig. 2 Time course of the effects of 15% FCS on CCK and c-fos mRNA expression in SK-N-MC cells. Cultures were treated with FCS for the tunes indicated. Each lane received 10 pg total RNA and autoradiogrsms were derived from successive hybridizations of the Northern blot.

30 min. Membranes were exposed to Amersham MP-films at -80°C for 7-14 d using a DuPont intensifier screen.

Results and Discussion The expression of CCK and c-fos mRNA in the human neuroblastoma cell line SK-N-MC was investigated by Northern blot analysis using human CCK and c-fos antisense hybridization probes. Northern blot filters initially hybridized with CCK and c-fos cRNA probes were re-probed with a mouse

p-actin cRNA probe thus allowing determination of relative amounts of CCK and c-fos mRNA in each slot. The results of these experiments are summarized in Figure 2 and Table. The time course reveals that FCS has a different effect on CCK and c-fos mRNA expression. FCS stimulated c-fos mRNA expression -thirty- to fifty-fold with a maximum after 50-60 min followed by a subsequent decline. Similar results were obtained using 10 @I FKN, which is an activator of the CAMP system. The cfos level rose -ten-fold after 1 h and fell back to -five-fold after 2 h of drug treatment. In contrast we found that the addition of FCS or FKN had differ-

Table Effect of FCS and FKN on CCK and c-fos mKNA levels in SK-N-MC cells. The mRNA abundances are expressed as ratios to the control levels (no drugs) and represent mean densitometric values obtained from Northern blot analysis, *P < 0.01; **P < 0.10 compared with unstimulated cells. (Student’s t test) n.s. = not significant drug treatment FCS, 15%

Forskolin (FKN)

mRNA abundance, relative to controls (k S.E.M., n = 3) minutes CCK c-fos no drug 30 60 120 240

0.46 0.73 2.39 3.53

1 i 0.22** It 0.05 n.s. f 1.06 n.s. f 0.26’

no drug 30 60 120

1 0.96 f 0.09 n.s. 2.69 f 0.31** 6.06 f 2.07+*

36.67 49.31 31.73 7.08

1 f 8.39* f 8.98; 5 2.05* f 1.25*

1 2.37 f 0.20* 10.93 f 1.74* 5.08 f 0.69*

CCK GENE EXPRESSION

IN HUhtAN NEUROBLASTOhtA

111

CELL LINES

may in part bc due to stimulatedc-fos gene exprcssion. It is well establishedthat Fos together with Jun forms a heterodimer tmnscription complex, AP-1, (29 which interactswith phorbolesteror CAMPresponse elements. A similar sequence corresponding to the AP-1 binding site is also found in the human CCK promoter and may be responsiblefor FCS induction of CCK mRNA expression. This AP-1 binding site is highly conserved and also found in the proenkephalinA (PEA) promoter (Fig, 3), where it is shown to be crucial for CAMP and phorbolester stimulatedPEA transcription(19,20). Second, CCK mRNA turnovermay be modulated by a different mechanism than the c-fos mR.NA Fig. 3 Schematic illustration of the conserved AP-1 element 5’decay and the CCK mRNA may be more stable and CTGCGTCAGC-3’, present within the human CCK (a), rat CCK accumulate in SK-N-MC cells. It is well established (b), and human proenkephalin A (c) promoter. Binding sites of phorbolester and CAMP inducible trans-Acting factors are as that the c-fos mRNA contains an ‘mRNA instabilindicated (2, 19). ity determinant’,an AU-rich element(ARE) located within the c-fos 3’-noncoding region. Recently, ent modulatory effects on CCK mRNA expression Gillis et al (9) clearly has shownthat a cytosolic proin SK-N-MC cells. In this case additionof FCS grad- tein adenosine-uridinebindingfactor (AUBF) comually increased CCK mRNA expression up to 4 h plexes with tandem ARE’s, which subsequently after stimulation(Table), Thus it appeared that c-fos leads to a selective c-fos mRNA degradation.From mRNA expression preceded CCK mRNA expres- our Northernblot analysiswe proposethat FCS may sion (Table). The different c-fos and CCK mRNA induce AUBF like proteins, which complex with expression pattern may be explainedby several dif- ARES in the c-fos mRNA leading to degradationof ferent mechanisms. c-fos mRNA after it has attained a peak level of First, the enhanced CCK mRNA transcription expressionafter -60 min. Examination of the human

a. *

9’

CTGCQTCAGC Q

h B_tntuhmn

-46

QQTCCTCTCTC’

hCCU

-113

GCCGCCCTCTG’

-157

CCCCCTCTCTC2

-1m

GGTGGTCTCTC2

-92

CCATTCCTCTC’

rcxx hCLmh

ANCCTCTCTC’

0B

hGaUrh

5

hCCK

5

ATTCCTCTC -’ CCCCTCTCT$ -167 p-i-

IBH(ANCW~--TATA-~ -

=c?!Ey-m

TATA -

3’

-113 lik4dlenwr ebments

Fig. 4 A) Sequence homology box, containing a silencer-like element present in the human p-interfiin, human and rat CCK, and human gas&in genes. The highly conserved consensus sequence 5’-CCTCTCTC is underlined. DNA sequences are taken From: 1 (23), 2 (2), and 3 (21). B) Structural organization of the silencer-lie elements found in the human CCK and gash-in genes.

112 CCK 3’-noncoding region did not reveal any ARE tandem like structure similar to the one found in the c-fos gene (9,17). This suggests that CCK and c-fos mRNA turnover is controlled by different mechanisms. Third, the increase of CCK mRNA expression after FCS-treatment could also be due to ‘inactivation’ of a silencer mechanism rather than activation of SRE inducible transcription factors. In a recent study it was established that removal of rat CCK 5’flanking sequences between -138 and -102 resulted in a dramatic increase of basal rat CCK mRNA expression, peaking with a six- to eleven-fold increase (2). Surprisingly, we find that both, the human and the rat CCK promoter contain p-interferon-like silencer elements in close proximity to the tentative AP-1 binding site (Fig. 4). In the human CCK gene, two silencer-like sequences containing an essential-CCTCTC-motif (2 1) are present at positions -113 and -157 distal to the mRNA cap site (2). These elements show sequence similarities with an active human gastrin silencer where it was shown that mutation of the 5’-ATTCCTCTC3’ motif into 5’-ATTCCCCC-3’ essentially abolished the activity of the negative element (21). Similar to the gastrin and P-interferon promoters, the tandem array of positive and negative elements in the human CCK promoter (Fig. 4B) may control basal and enhanced CCK mRNA expression at the transcriptional level. To clarify some of these points we performed a set of nuclear run-off transcription assays. SK-NMC cells were treated with FCS for 1, 2, 3, or 4 h, whereafter cell nuclei for run-off transcription analysis were isolated and treated as described in Materials and Methods. The result of the run-off transcription analysis indicates that c-fos gene transcription precedes CCK mRNA expression which is consistent with the results obtained by Northern blot analysis. Moreover, it appears that the c-fos gene is constitutively expressed in SK-N-MC cell nuclei (Fig. 5, lane 1). Northern blot analysis revealed that the basal c-fos mRNA level prior to FCS stimulation in SK-N-MC cells, is very low in this cell line which is in contrast to the high basic CCK mRNA expression (Fig. 2). Therefore, we cannot exclude the possibility that in SK-N-MC cells, FCS may induce a change both in modulating basal c-fos mRNA expression at the transcriptional level and in c-fos mRNA turnover. In addition, the run-off dot

c-tos

-

PEA

_

CCK Actin

-

Fig. 5

Nuclear Run-off transcription assay of CCK and C-fos mRNA expression. SK-N-MC cells were treated with 15% FCS for 60 (lane 2), 120 (lane 3) and 240 (lane 4) min respectively, and compared to cells without FCS addition; control (lane 1). Nuclear run-off tmnscription analysis were preformed as described in Materials and Methods. Autoradiogram of p-actin, CCK, PEA and c-fos hybridization is shown.

blot analysis indicates that CCK mRNA expression is markedly stimulated between 2-4 h (Fig. 5), which is in agreement with the results obtained by Northern blot analysis (Table). The novel findings in this study are that basal CCK and c-fos mRNA expression takes place in SK-NMC cells growing in the absence of serum, and that the mRNA expression of both genes was stimulated by FCS and FKN (Table). The data presented here are the first demonstration of c-fos mRNA expression in a human neuroblastoma cell line growing in the absence of serum. Furthermore, we conclude that the basal c-fos gene expression, FCS, and FKN accounts for the described changes in CCK mRNA expression. A recent study from Zeytin et al described that in the multipeptide secreting rat cell line 44-2C, calcitonin and c-fos mRNA expression is induced when grown in the absence of serum. The calcitonin and c-fos mRNA expression pattern in rat 44-2C cells was similar to the pattern observed for basal CCK and c-fos mRNA expression in SK-NMC cells, cultured in serum-free medium. In conclusion we find that CCK and c-fos mRNA expression is differently modulated by FCS and FKN. It appears that FCS modulates both CCK and c-fos mRNA expression at the transcriptional level. Moreover, we conclude that modulation of c-fos mRNA expression may also be due to changes in mRNA stability. Experiments to establish whether proto-oncogenes such as c-fos and c-jun are responsible for basal and enhanced CCK mRNA tran-

CCK GENE EXPRESSION

IN HUhtAN NEUROBLASTOMA

CELL LINES

scription are under present investigation in our iaboratory, We fkd that the SK-N-MC cell line, as used in this and previous investigations, provides an excellent model system to study the regulation of proto-oncogene activated neuropeptide gene expression.

Acknowledgements We thank Susanne Hummelgaard and Charlotte Vienberg for providing excellent technical and secretarial assistance and T. Geijer, Stockholm, for the statistical analysis. This study was supported by the Danish MRC, the Danish Cancer Society, and the Danish Biotechnology Center for Neuropeptide Research.

References 1. Deschenes, R. J., Haun, R. S., Funckes, C. L. and Dixon J. E. (1985). A gene encoding rat cholecystokinin. J. Biol. Chem., 260: 128&1286. 2. Haun, R. S. and Dixon, J. E. (1990). A transcriptional enhancer essential for the expression of the rat cholecystokinin gene contains a sequence identical to the -296 element of the human c-fos gene. J. Biol. Chem., 265: 15455-15463. 3. Verma, 1. M. and Sassone-Corsi, P. (1987). Proto-oncogene fos: Complex but versatile regulation. Cell, 5 1: 5 13-514. 4. Ramsone, L. J. andtrerma, I. M. (1990). Nuclearproto-oncogenes fos and jun. Ann. Rev. Cell Biol., 6: 539-557. 5. Des&s, J., Meijlink, F. and Verma, I. M. (1985). Identification of a tmnscriptional enhancer element upstream from the proto-oncogene fos. Science, 230: 1174-l 177. 6. Piechaczyk, M., Blanchard, J-M. and Bonnieu, A. et al. (1988). Role of RNA structures in c-myc and c-fos gene regulations. Gene, 72: 287-295. 7. Shyu, A-B., Greenberg, M. E. and Belasco, J. G. (1989). The c-fos tmnscript is targeted for rapid decay by two distinct mRNA degradation pathways. Genes & Develop., 3: 60-72. 8. Jackson, R J. and Standart, N. (1990). Do the Poly(A) tail and 3’-untranslated region control mRNA translation? Cell, 62: 15-24. 9. Gillis, P. and Malter, J. S. (1991). The adenosine-u&line binding factor recognizes the AU-rich elements of cytokine, lymphokine, and oncogene mRNAs. J. Biol. Chem., 266: 3172-3177. 10. Folkesson, R., Monstein, H-J., Geijer, T. and Terenius, L.

11.

12.

13.

14.

15.

16.

17.

18.

19. 20.

21.

22.

23.

113 (1989). Modulation of proenkephalm A gene expression by cyclic AMP. Mol. Brain Res., 5: 21 l-217. Monstein, H-J., Folkesson, R and Geijer, T. (1990). Procholecystokinin and proenkephalin A mRNA expression is modulated by cyclic AMP and noradrenaline. J. Mol. Endocrinol., 4: 3741. Monstein, H-J. andFolkesson, R. (1991). Phorbol12-myristate- 13-acetate (PMA) stimulates a differential expression of cholecystokinin (CCK) and c-f0 mRNA in a human neuroblastoma cell line. FEBS, 293:145-148. Verbeeck, M. A. E. and Burbach, P. H. (1990). The cholecystokinin gene is abundantly co-expressed with gastrin releasing peptide, enkephalin and neuropeptide Y genes in a clonal human neuroepithelioma cell line. FEBS, 268: 88-90. Chomczynski, P. and Sac&i, N. (1987). Single-step method of RNA isolation by acid guanidinium thiocyanate-phenolchloroform extraction. Analyt. Biochem., 162: 156159. Monstein,H-J.andGeiierT.(l988).Ahiahlvsensitivenorthem blot assay detects multiple pro&@&& A-like mRNAs in human caudate nucleus and pheochromocytoma. Bioscience Reports, 8: 255-261. Geijer, T., Folkesson, R., Rehfeld, J. F. and Monstein, H-J. (1990). Expression of the cholecystokinin gene in a human (small-cell) lung carcinoma cell-line. FEBS, 270: 3&32. Takahashi, Y., Kato, K, Hayashixaki, Y., Wakabayashi, T., Ohtsuka, E., Matsuki, S.. lkebara. M. and Matsubara. K. (1985). Molecular cloning of the’ human cholecystokmin gene by use of a synthetic probe containing deoxyinosine. Proc. Natl. Acad. Sci., 82: 1931-1935. Greenberg, M. E. and Ziff, E. B. (1984). Stimulation of 3T3 cells induces transcription of the c-fos proto-oncogene. Nature, 3 11: 433438. Goodman R. H. (1990). Regulation of neuropeptide gene expression. Annual Rev. Neurosci., 13: 11 l-127. Kobierski,L. A.,Chu,H. M.,Tan,Y. andCombM. J. (1991). CAMP-dependent regulation of proenkephalin by JunD and JunB: Positive and negative effects of AP-1 proteins. hoc. Natl. Acad. Sci., 88: 10222-10226. Wang, T. C. and Brand, S. J. (1990). Islet Cell-specific Regulatory Domain in the Gastrin Promoter Contains Adjacent Positive and Negative DNA Elements. J. Biol. Chem., 265: 89088914. Zeytin, F. N., Rusk, S. F., Baird, A., Raymond, V., Leff, S. E., Haaparanta, T. and Mandell, A. J. (1988). Induction of cfos, calcitonin gene expression, and acidic fibroblast growth factor production in a multipeptide-secreting neuroendocrine cell line. Endocrinol., 122: 1114-l 127. Ban&mad, A., Muller, M., Steiner, C. and Renkawitz, R. (1987). Activity of two different silencer elements of the chicken lysozyme gene can be compensated by enhancer elements. EMBO J., 8: 2297-2303.