Role of transcription, translation, and protein turnover in controlling

Role of Transcription, Translation, and Protein Turnoverin Controlling the Distribution of 3-Hydroxy-3-Methylglutaryl Coenzyme A Reductase in the Lens

Richard J. Cenedella

Purpose. To determine ihe principle site (epithelium or .superficial cortex) of gene transcrip-tion and mRNA translation for the regulatory enzyme of lens cholesterol biosynthesis, 3-hydroxy-3-methylglutaryl coenzyme A reductase (HMGR). To evaluate the contribution ofwaning enzyme synthesis versus en/.yme turnover by proteolysis in accounting for the disap-pearance of HMGR protein from elongated fiber cells.

Methods. Young rats were treated with lovastatin, a drug that increases transcripts of the HMGRgene and translation of HMGR mRNA in lens secondary to inhibiting cholesterol biosynthesis.The relative concentration of HMGR mRNA in lens epithelium and superficial cortex wasestimated by a competitive reverse transcriptase-polymerase chain reaction system. RelativeHMGR protein levels were estimated by Western blot analysis. Because lovastatin is clearedrapidly from the lens, the half-life of HMGR protein in epithelium and cortex was estimatedby following the disappearance of the increased pool of en/.yme protein from each compart-ment with time after halting drug treatment.

Results. Between 75% and 90% of the total content of HMGR mRNA and protein in theepithelium and the superficial cortex of control rat lens was located in the cortex. Treatmentwith lovastatin increased the content of the mRNA in epithelium and cortex by approximately0.4-fold and HMGR protein content approximately 5-fold. Although the concentration ofHMGR mRNA and protein was similarly increased in epithelium and superficial cortex, ap-proximately 85% to 90% of the total increase in mRNA and protein content was located inthe cortex because of that area's greater mass. The half-life for the disappearance of theincreased pool of HMGR protein from epithelium and cortex was similar at approximately14 to 17 hours.

Conclusions. The bulk of HMGR gene transcription and mRNA translation apparently is con-fined to elongating fiber cells. The 10-fold greater increase in enzyme protein than mRNAlevels after lovastatin treatment indicates that enzyme concentration in lens is controlledmainly by effects on HMGR mRNA translation or rates of HMGR proteolysis. The observedrapid turnover of en/yme protein in the epithelium and the superficial cortex, if applicableto the deeper cortex and the homeostaiic slate (absence of drug exposure), suggests that thegradual disappearance of HMGR protein from the lens could be caused by waning of enzymesynthesis rather than to proteolysis in the absence of continuing enzyme synthesis. InvestOphthalmol Vis Sci. 1995; 36:2133-2141.

.Describing the regulation of cholesterol metabolismHorn the Depaitmenl nj liiodinm.Mn, Knhsville College o/ (hleopathu Mediant: \X\ the OCular icilS is of interest for at least tWOK,,W/, .v/m«.m T h c iife_iong grOwth of the lens is dependent on on-Suppoiled //> National Institutes nj Health trianl I.YO256X .

Submitted l<» Imbluatwn l-elmia,? 17, 1995, mined Afml 21, 1995. a«(fted Afml s irC CllOlCStCrol biosynthesis tO SUppl)' the Sterol for24• / y y 5 sustained membrane format ion 1 ' and inhibitors of'^nTZprRutnd] Ledeiia. Depa.tmen, ,,jli.o.henustn; Knk.viiie l c n s cholesterol biosynthesis can produce calaracLs inCollegei>/(htmpaiiui Meduinr, HOO w ii-fjnsoti Stieet, Knkwiiir, MO63501 animals*'1 and humans.'1' The hypocholesterolemic

Invcsli^ilivc O|>IHIMIIIIDI<>K\ & Visual Science, Sep lcmbei l!)()f>, Vol lM), No II)Co|>yii)r|M © Asso( i.Hioii loi K C S C I K I I MI Vision and ()|>lilli.ilmnlo^\ 2133

Downloaded From: http://iovs.arvojournals.org/pdfaccess.ashx?url=/data/journals/iovs/933182/ on 02/18/2018

2134 Investigative Ophthalmology & Visual Science, September 1995, Vol. 36, No. 10

drugs lovastalin and simvastatin readily penetrate therat lens, decrease cholesterol synthesis, and can blockthe age-related accumulation of lens cholesterol."/S

These drugs act by inhibiting the rate-limiting enzymein cholesterol biosynthesis," 3-hydroxy-3-methylglu-taryl cocnzyme A reductase (HMGR).10 Lovastatin isthe third most prescribed drug in the United States."

Biosynthesis of lens cholesterol is tightly coordi-nated with synthesis of the other major componentsof the fiber cell plasma membrane, main intrinsic pro-tein-26, and phospholipids.12 Synthesis of all are con-fined to the outer 10% of the lens radius and coordi-nately peak in an arc of radius corresponding to theouter 3% to 6% of the radius.12 We have been inter-ested in describing the factors responsible for control-ling the distribution of cholesterol synthesis and thusmembrane formation.

The spatial distribution of HMGR enzyme activityin the lens parallels cholesterol biosynthesis, and thedisappearance of enzyme activity over the outer 5%to 10% of the radius appeared due to loss of enzymeprotein mass.'1 We recently observed that the mRNAfor HMGR was detected surprisingly throughout theyoung rat lens (cortex and nucleus), but significantlevels appeared confined to the outer 0% to 5% ofthe radius.1'1 Transcription of the HMGR gene couldbe confined to this outer region because nuclei areclaimed to be present to a depth of only approximately50 fiber cells (~100 fi) below the surface of the rodentlens,1"' a point corresponding to approximately theouter 6% of the radius of the rat lenses used in thisstudy. Because the maximum mRNA levels amountedto only a few copies per cortical fiber cell, we hypothe-sized that translation of HMGR mRNA was also con-fined to this superficial cortex and that the loss ofenzyme protein over the next 5% of the radius (outer5% to 10%) was principally caused by protcolysis, per-haps by general proteolysis of the endoplasmic re Lie u-lum-lhe subcellular location of the enzyme in tissues,"'including ihe lens.1* As described in the Discussion,this possibility would require that the half-life ofHMGR protein in the lens be much longer than inother cells, such as hepatocytes1' and cultured UT-1cells,1" in which it ranges from 2 to 12 hours, respec-tively.

The current study challenges this hypothesis byaddressing the following questions. In view of the lowcopy number of HMGR mRNA in the superficial fibercells, could these mRNA molecules have been im-ported into the lens with differentiating epithelial cellsand then activated for translation? What is the princi-pal site in lens of HMGR gene transcription? Is thehalf-life of HMGR protein sufficiently long to permita gradual decline of enzyme activity over the outer5% to 10% of the lens radius if protein synthesis was

confined only to the outer 0% to 5%? In other words,to what extent is the distribution of HMGR proteindetermined by HMGR mRNA levels and translationversus HMGR protein turnover by proteolysis? To ad-dress these questions, we treated rats with lovastatinand then estimated changes in the levels of HMGRmRNA and protein mass in the epithelium and cortex.Exposure of liver to lovastatin results in increased lev-els of HMGR mRNA transcripts and enzyme proteinsecondary' to inhibition of the cholesterol synthesispathway.11"" The lens site (epithelium or superficialcortex) showing the greater increase in HMGR mRNAand protein levels after treatment could identify theprimary site for mRNA transcription and translation.Because lovastatin is quickly cleared from the lensafter slopping treatment,H measuring disappearanceof the increased pool of HMGR protein from epithe-lium and cortex should permit estimation of the half-life of the enzyme in these two cell populations. Toour knowledge, the half-life of another lens enzymehas not been reported.

MATERIALS AND METHODS

Animals and Lens Fractionation

All procedures involving animals conform to theARVO Statement for the Use of Animals in Ophthal-mic and Vision Research. Sprague-Dawley rats (maleand female from Hilltop Lab Animals, Inc., Scottdale,PA) were placed on a reversed light cycle (10 hoursdark-14 hours light) at least 6 days before the startof treatment with lovastatin. Between 24 and 29 daysof age, rats were fed plain chow (controls) or groundchow supplemented with 0.05% lovastatin, (added asground Mevacor tablets [Merck, Rahway, NJ]) for 5days. Animals were killed by carbon dioxide inhalationusually between hours 3 and 4 of the dark cycle. Lenses(8 to 28) of control and treated rats were decapsu-lated, and the capsules were saved. The encapsulatedlenses were divided into superficial cortex and remain-der of the lens by dissolution for 4 to 6 minutes in 3to 4 ml of buffer (1 mM ethylenediaminetetraaceticacid, 5 mM beta-mercaptoethanol, 0.05 M Tris, pH 8)containing 0.2% (wt/vol) of sodium dodecylsulfatc(SDS). The "undissolved" lens fraclion was homoge-nized in the SDS buffer with a Dounce homogenizerand saved for protein assay, which was performed bya modified Lowry method."1 Based on the protein con-tent of the two fractions and published informationon the radial distribution of protein in the young ratlens, the collected cortical fraction could be equatedto a specific percentage of the lens radius.1" Cortexaccounted for between 3% and 8% of the outer radiusin these studies—that is, 100% to 97% to 100% to


Control of Lens Cholesterol Biosynthesis 2135

92% of the radius (0% radius was the lens center).The pooled capsules were Dounce homogenized in 3ml of the SDS-containing buffer. In some cases, sam-ples of liver were collected and homogenized in twovolumes of the SDS buffer using a motor-driven,smooth glass-teflon homogenizer. The capsule, cor-tex, and liver samples were flash frozen and lyophi-lized for subsequent extraction of RNA or used forestimation of HMGR protein mass.

Reverse-Transcriptase-Polymerase ChainReaction Estimation of HMGR mRNACopy Number

The lyophilized lens and liver samples were dissolvedin 0.30 ml of diethylpyrocarbonate (DEPC)-treatedwater and homogenized in 4 ml of TRI REAGENT(Molecular Research Center, Cincinnati, OH), andtotal RNA was extracted as described.1'1 The lens sam-ples consisted of capsules and superficial cortex from8 to 28 lenses, and the liver samples were equivalent to100 ing of tissue. RNA was dissolved in DEPC-treatedwater, and concentrations were estimated from the260/280 nm absorbance ratios. All ratios were be-tween 1.8 and 2.0, indicating RNA of high purity. Be-tween 0.6 to 0.7 /ig of total RNA was recovered perlens epithelium and 5 lo 7 /zg of total RNA per lenssuperficial cortex (outer 3% to 4% of the radius).Liver RNA was separated further into a poly A1 frac-tion using oligo-dT columns from Molecular ResearchCenter according to their instructions. Separation ofpoly A1 RNA from lens samples was not performedbecause of the low levels of RNA recovered. Approxi-mately 6 /ig total RNA/1 mg wet liver was recovered,from which 0.2 to 0.3 /̂ g of poly A1 RNA was isolated.Liver sample size was 100 mg.

HMGR mRNA levels in lens capsules (epithe-lium), superficial cortex, and liver were estimated asrecently described in detail" using a slight modifica-tion of the competitive reverse transcript ion-polymer-asc chain reaction (RT-PCR) method of Powell andKroon."~MThe method basically involves reverse tran-scribing RNA in the presence of a known numberof molecules of the RNA plasmid pAW109 (Perkin-Eliner Cetus, Norwalk, CT). This internal standardcontains 3' and 5' primer recognition sites identicalto those in HMGR mRNA and yields a 303-bp cDNAcompared to a 246-bp segment with the native mRNA.The resultant cDNAs were PCR amplified (Gene AmpRNA PCR kit; Perkin-Elmer Cetus) for 33 cycles inthe presence of 15 to 16 //M digoxigenin-11-dUTP(Bochringcr Mannheim, Indianapolis, IN). After sepa-ration of the amplified products by agarosc gel elcctro-phoresis and transfer to nylon membranes, the mem-branes were exposed to antidigoxigenin-immuno-globulin G-conjugated alkaline phosphatase

(Bochringer Mannheim), and the 246- and 303-bpfragments were located by chemiluminescent expo-sure of x-ray film. The relative densities of the internalstandard and target band signals were estimated bydensitometric scanning of the developed film andmeasuring the relative areas under the resultantcurves. This approach is in essence a titration of agiven mass of RNA with increasing copies of pAW109,and the nearest equivalence point in the titration isused to calculate the copies of HMGR mRNA in theRNA sample."' The current methods were identical tothose used before,1'1 except that the RT-rcaction vol-ume was reduced to 10 /xl (from 20 //I), all of whichwas PCR amplified (50 //I final volume). To use thesmaller volume, aliquots of DEPC-treated water con-taining 400 to 800 ng of lens RNA (stored at -80°C)were lyophilized and then dissolved in 8.5 fj,\ of 1.2XRT master rrrix solution. Between 10 and 50 ng of liverpoly A' RNA was used in the estimation of liver HMGRmRNA content. After the addition of a known numberof copies of pAW109 in 1.5 //I and overlaying with 40ji\ of oil, RT was conducted for 15 minutes at 42°C,and PCR amplification was performed using a PTC-100 programmable thermal controller (MJ Research,Watertown, MA) with cycling times and temperaturesidentical to those used earlier.1'1

Estimation of HMGR Protein Mass

HMGR protein levels can be estimated using Westernblot analysis with a monoclonal antibody coupled tochemilumencent detection.14 Total protein from sam-ples of lens capsules (5 to 50 //g) and superficial cortex(10 to 100 fig) were separated on 12% SDS-pplyacryl-amidc gels, transferred to ProBlott membrane (Ap-plied Biosystems, Foster City, CA), and exposed to a1/100 dilution of monoclonal antibody solution recov-ered from cultures of murinc A9 cells.1 * After expo-sure to the secondary antibody solution (1/80,000 di-lution of peroxidase-conjugated goat anti-mouse im-munoglobulin G, (Jackson Immuno Research, WestGrove, PA), the HMGR band (97 kd) was detected bya nonisotopic enhanced chemiluminescence system(Amersham, Arlington Heights, IL). The relative den-sity of the HMGR bands was estimated by measuringareas under curves generated from densitometricscans of the exposed film. Based on the total proteincontent of each lens fraction examined and the num-ber of lenses that furnished that protein, the concen-tration (band area) of HMGR protein per lens fractionwas estimated relative to that of the epithelium of asingle lens from control rats. The concentration ofHMGR protein in liver of control and lovastatin-treated rats was estimated in microsomal protein (0.1to 5 fj.g), which was isolated from homogenized liverby standard centrifugal methods. The relative concen-



A B Hepatic HMGR mRNA Hepatic Microsomal HMGR

Control ~reated

(3000~)

HMGR 'C .

Poly A+ RNA 10 20 50ng 10 20 50ng

cHMGR mRNA/ Relative HMGR Rat mg Liver [Protein]/mg Liver

Control 6.1 94 Treated 14,546

tration of HMGR protein was expressed on the basis of the tnicrosomal protein recovered per milligram of fresh liver.

RESULTS

Effects of Lovastatin on HMGR mRNA and Protein Levels in Liver Feeding young rats 0.05% lovastatin for 5 days resulted in approximately a 2-fold increase in the concentration of HMGR mRNA and approximately a 17-fold increase in immunoreactive HMGR protein in the liver (Fig. 1) . Ness et al"' also reported that hepatic HMGR protein levels increased approximately 10-fold more than mRNA levels after rats were fed lovastatin (0.04% for 3 days).

Effects of Lovastatin on the Concentration and Distribution of HMGR mRNA and Protein in the Lens

Lenses recovered from the same rats that furnished the livers (Fig. 1) were separated into capsule (epithelium) and lens, and the outer 3% to 3.5% of the lens radius (superficial cortex) was removed by dissolution in the SDScontaining buffer. In epithelium and superficial cortex, the cortex accounted for 80% to 90% of the total copies of mRNA in the control lenses (Fig. 2). Treatment with lovastatin increased the concentration of HMGR mRNA in the epithelium and superficial cortex to similar extents. The total content of the mRNA (copies/epithelium and cortex) increased approximately 0.4fold, and the cortex accounted for 80% to 90% of the total increased content because of

Control Treated

97 kD , - <.". -

Pg 5 1 0.5 0.1 5 1 0.5 0.1 Protein

n G m E I. Effects of treatment with lovastatin on hepatic HMGR mRNA and protein concentrations. Rats (29 to 31 days old) were fed plain chow or chow containing 0.05% lovastatin for 5 days. Poly At RNA and microsomal protein was isolated from liver samples. (A) HMGR concentrations were estimated by a competitive reverse transcriptase (RT) - polymerase chain reaction (PCR) method using 3000 copies (c) of the RNA plasmid pAWIO9 as internal standard. (B) Relative HMGR protein levels were estimated by Western blot analysis with chemiluminescent detection. (C) Concen- trations calculated from areas under the scan curves for the samples (*) shown in A and B.

The concentration of HMGR-immunoreactive protein was examined in three separate groups of control and lovastatin-fed rats (Fig. 3). Lens HMGR protein content is expressed relative to the HMGR signal measured per control lens epithelium. On average, the cortex (representing 3% to 8% of the lens outer radius) accounted for approximately 75% of the total content of HMGR protein of control lens epithelium and superficial cortex. Treatment with lovastatin increased the concentration of HMGR protein (relative units per milligram total protein) more in the epithelium than in the cortex (Fig. 3). Although the epithelial cell HMGR protein bands were more intense than those from an equal mass (50 pg) of cortical protein, the aliquot of cortical protein applied to the gels r e p resented a much smaller fraction of the total sample protein. Viewing the epithelium and cortex as a unit, treatment with lovastatin increased the relative HMGR protein content of the lens by approximately 5fold on average, (4.0 + 18.3) - (1.0 + 2.8) + 3.8 (Fig. 3D). As with HMGR mRNA levels, approximately 85% of the total increased content of HMGR protein was accounted for by the cortical fraction because of the cortex's greater mass. Thus, similar to its effects on liver, treatment with lovastatin resulted in approximately a 10-fold greater increase in the relative HMGR protein content of the lens than HMGR mRNA content. Most of the total increased content of mRNA and protein was located in the superficial cortex, the principal site for lens cholesterol synthesis."

Half-Life of Lens HMGR Protein

We attempted to measure the half-life of HMGR protein in the lens to evaluate the importance of mRNA its greater mass.


Control of Lens Cholesterol Biosynthesis

~ d l h d l u m Coltex Control Trntrd Control T m w

HMGR

Asny 2

cHMGR mRNAlhns

, A w W J - A M Y 2

ex En1 Cortex

Control 418 2992 10833 3747

ncuRE 2. Effects of treatment with lovastatin on the copies of HMCR mRNA in lens epithelium and superficial cortex. RNA total wa. extracted from rat lens capsules and superli- cial cortex (outer 3% to 3.5% of the lens radius) of control and treated rats (same animals that supplied the livers in Fig. 1). HMCR mRNA copy numbers were estimated using the competitive reverse transcription-polymerase chain reaction (RT-PCR) method with titration against varying numbers of pAW109 copies (c). * = Nearest equivalency point in the titration. Assay 2 shows the results of a second RT-PCR analysis of the RNA samples run on a separate day. Values in parentheses are percent increases above control levels.

translation versus protein loss by proteolysis in accounting for the previously recognized disappearance of HMGR protein after completion of fiber cell elongation.'"ur strategy was to follow the decrease of immunoreactive HMGR protein from the lens epithelium and superficial cortex with time after withdrawal from lovastatin treatment. Lovastatin appears to be cleared rapidly from the rat lens; within approximatley 12 hours of treatment, lens cholesterol synthesis returned to the control level.' We initially assumed that because lens proteins are thought to turn over slowly or not at all,24 turnover of HMGR would be much slower than it is in hepatocytes (2 hours) l 7 or cultured

UT-1 cells (12 hours).'' Thus, lens HMGR protein levels were estimated over a 1-week period after drug treatment was stopped. Surprisingly, within 1 day, epithelial and cortical HMGR protein levels returned to levels similar to those of untreated controls (Fig. 3B). The experiment was repeated, but now the protein was examined at 12-hour intervals after the drug was removed. A stepped decrease in the increased pools of epithelial and cortical HMGR protein was seen clearly over the first 24 hours (Fig. 3C). A plot of changes in the relative protein concentration against time permit- ted estimation of the net half-life of HMGR protein in the epithelium and superficial cortex. The estimated values were approximately 14 hours in the epithelium and 17 hours in the cortex, rates similar to those found in other cells (Fig. 4) . I H These rates of disappearance largely should reflect rates of proteolysis uncon- founded by high rates of protein synthesis because HMGR synthesis is assumed to return quickly to basal levels on removal of lovastatin. If significant HMGR synthesis occurred during the time interval examined, the true half-life of lens HMGR would have been shorter than the one measured. To our knowledge, this is the first reported half-life of any lens enzyme.

DISCUSSION

The rate of cholesterol synthesis in animal cells is thought to be controlled by the activity of the enzyme 3-hydroxy-3methylglutaryl coenzyme A reductase (HMGR)."' We have shown that this also seems to apply to the lens.12'The activity of this enzyme is largely regulated by the cellular concentration of enzyme protein.2"rotein levels are controlled by rates of HMGR gene transcription, mRNA translation, and proteolysis of the en~yme.~"~"" When cellular choles terol levels are lowered, transcription of the HMGR gene is derepressed and mRNA levels in~rease. '~~~' However, accompanying effects on mRNA translation and enzyme degradation seem quantitatively more important in controlling HMGR protein concentra- ti~n.~"onsterol isoprene precursors of cholesterol, possibly farnesyl pyrophosphate, appear to be agents responsible for suppressing HMGR mRNA translation and enhancing enzyme proteolysis."'~'O ~ h u s , when HMGR is inhibited competitively by drugs, such as lovastatin, cellular levels of cholesterol-as well as those of its nonsterol precursors-fall, and this results in increased synthesis of HMGR mRNA and protein and slowing of protein degradation.

Lovastatin was used in the current study as a tool to identify the primary site of HMGR gene transcrip tion in the lens and to permit estimation of the half- life of HMGR protein. One might intuitively expect that enhanced HMGR gene transcription would ac-



Epithelium Cortex Control Treated M - Control Treated M

Epithelium Cortex Control Treated Control Treated

Epithelium Cortex Treated Control Treated Control

97kD - & - - W ax - - ,-----

lHMGRl/lens Control Treated

Epithelium 1.0 4.020.7 Cortex 2.8k0.8 18321.5

FIGURE 3. Relative changes in lens HMGR protein content after treatment with lovastatin. Rats (24 to 29 days old) were fed plain chow or chow containing 0.05% lovastatin for 5 days. Aliquots of total protein from the epithelium and cortex were assayed for immunoreactive protein by Western blot analysis using a monoclonal antibody to HMGR and chemiluminescence detection. (A) Relative HMGR protein mass in 5 to 50 pg of total epithelial cell protein and in 10 to 100 pg of total cortical protein. The cortical fraction constituted approximately 3% of the lens outer radius. M - 0.2 pg of microsomal protein from lovastatin- treated rats. (B,C) Changes in relative HMGR protein mass after treatment with lovastatin (+ hour and + days after stopping treatment). Animals were killed at the third hour of the light cycle in experiment C and between the 3rd and 4th hours of the dark cycle in experiments A and B. Each sample represents the relative HMGR content in 50 pg of epithelial or cortical protein. Cortical fractions constituted 4% to 5% of the lens radius in B and 7% to 8% in <:. (D) Mean 2 SEM concentration of HMGR protein per lens relative to the control lens epithelium (average of experiments A, B, and C).

company fiber cell elongation because increased cholesterol synthesis is needed to supply the sterol for membrane formation. However, because superficial fiber cells appeared to contain only a few copies of HMGR mRNA per cell,''' we considered the possibility that HMGR mRNA molecules were imported into the lens with differentiating epithelial cell and then activated for translation." Because lovastatin should in- hibit HMGR activity in both epithelium and cortex, the site (epithelium or cortex) accounting for most of the total increase in mRNA copies after treatment should identify the primary site of gene transcription in the lens. Epithelial cells that possessed increased

levels of HMGR mRNA and protein should have re- mained in the epithelium over the course of the drug treatment period (5 days). Cholesterol synthesis in the lens epithelium likely is confined to mitotically active cells because cholesterol is needed for membrane formation by dividing cells to proceed through the cell cycle.x~ -:u Inhibition of cholesterol synthesis in cul-

tured bovine lens epithelial cells by mevinolin (lovastatin) completely blocked cell proliferation, and the block was overcome by adding cholesterol to the me- dia.":' Because approximately 7 days are required for proliferative epithelial cells to migrate to the lens equator in 5- to &week-old rats,'"' both HMGR mRNA



CORTEX

-TI/2 ^17 5hrs

0 10 20 30 40Time (hrs ) After Stopping Treatment

FIGURE 4. Estimated net half-life: of HMGR protein in lensepithelium and cortex. Concentration values reflect the rela-tive areas under the curves generated hy scanning the bandsshown in experiment C of Figure 2. • = epithelium ofcontrols killed at 0 and +24 hours; • = cortex of controls.

and protein largely could have disappeared from thesecells by the time they began to differentiate and leavethe capsule. We examined changes in HMGR mRNAand protein levels in only the superficial cortex princi-pally for two reasons: Both mRNA and protein levelspeak in the outer 5% of the lens radius1 '•''' and, owingto the great protein concentration of fiber cells, detec-tion of HMGR protein by Western blot analysis wouldbe compromised by including cortex deficient in thisprotein.

Finding that most (85% to 90%) of the total in-creased content of HMGR mRNA and protein in re-sponse to lovastatin treatment was located in the su-perficial cortex rather than epithelium suggests thattranscription of the HMGR gene is confined largelyto elongating fiber cells, and importation of HMGRmRNA molecules with differentiating epithelial cellsis insignificant. It is important to recognize that theincrease in HMGR mRNA transcripts also might beattributed to a lovastatin-related enhancement of thestability of the transcripts, particularly if this mRNApossesses a short half-life in the lens. Because the rela-tive increase in enzyme protein mass greatly exceededthe increase in mRNA copies, modulation of mRNAtranslation could be more important than gene tran-scription in controlling the activity of this enzyme inlens. Also, a decrease in proteolysis (turnover) ofHMGR could contribute to the lovastatin-related in-crease in enzyme concentration. As mentioned above,nonsterol substances formed from mevalonic acid en-

hance degradation of HMGR."1 'w Thus, the need forincreased HMGR activity and cholesterol synthesisthat arises early in fiber cell elongation could be satis-fied by promoting HMGR mRNA translation and slow-ing proteolysis of enzyme protein. Stimulation ofHMGR gene transcription could be quantitatively lessimportant.

The type of HMGR mRNA molecule formed inthe lens also could contribute to regulating lens cho-lesterol biosynthesis. Reynolds cl al1;> identified twoclasses of HMGR mRNA, class I and II.v> Class II con-tains an extremely long 5'-untranslated region thatwas hypothesized to enable regulation at the transla-tional level. Class II messages, therefore, might permitenzyme synthesis in cells in which transcription of theHMGR gene was low.'"' It would be interesting to iden-tity the class of HMGR mRNA in lens epithelial andfiber cells.

One of our main objectives is understanding con-trol of plasma membrane formation in the lens. Wehave focused on cholesterol synthesis because it is anessential component of membranes that must be syn-thesized by the lens,1" and its insertion into the fibercell membrane is coordinated with that of the otheressential components, MP26 and phospholipids.12

Cholesterol synthesis and thus membrane formationdisappeared over an arc corresponding to the outer5% to 10% of the lens radius.1" The waning of choles-terol synthesis in this region appeared to be causedby the disappearance of HMGR protein." Because lowlevels of HMGR mRNA were found over this arc ofradius,1'1 we hypothesized that the loss of enzyme pro-tein was caused by proteolysis of the enzyme ratherthan to decreased enzyme synthesis. Showing that thehalf-life of HMGR protein in the lens epithelium andsuperficial cortex was surprisingly short (14 to 17hours) stiggested revision of this hypothesis.

The time required for a fiber cell located at the95% point (outer 5%) of the radius of 30- to 35-day-old rat lenses to be displaced inward by growth to 90%(outer 10%) of the radius should be approximately12 days (see Appendix). Thus, if the disappearance ofHMGR protein over this arc of radius was solely causedby proteolysis, the half-life of the enzyme proteinshould have been approximately 4 days (three half-lives). If the half-life of the protein in the 95% to 90%arc of radius was similar to that measured in the moresuperficial cortex (100% to 95%), sustained synthesisof HMGR protein must occur in the deeper region.The observed disappearance of enzyme protein fromfiber cells would have been much too slow to haveoccurred by proteolysis alone without continuing en-zyme synthesis. Thus, a gradual waning in the rate ofHMGR mRNA translation over the 95% to 90% arcof lens radius would be the primary factor accounting



for the disappearance of cholesterol synthesis fromlens fiber cells. On the other hand, if the half-life ofHMGR protein was much longer in this slightly deepercortex, proteolysis could yet be important in account-ing for loss of the enzyme protein. Furthermore, wecannot exclude the possibility that the half-life ofHMGR protein under homeostatic conditions wouldbe much longer than that measured for the increasedpools of enzyme induced by treatment with lovastatin.

Key Words

cholesterol, 3-hydroxy-3-melhylglutaiyl coenzyme A rcduc-tase, lens, lovasialin, proteolysis, transcription, translation

Acknowledgments

The author thanks Dan Hollo for technical assistance andDavid Welsh and Debbie Kramer for assistance with prepara-tion of the figures and manuscript.

References

1. Cenedella RJ. Sterol synthesis by the ocular lens ofthe rat during postnatal development. / Lipid Res.1982:23:619-626.

2. DcVries AC, Vermeer MA, Breclman JJ, Bar PR, CohenLH. Cholesterol content of the rat lens is lowered byadministration of simvastatin, but not by pravastatin.Exp Eye Res. 1993;56:393-399.

3. Cenedella RJ, Bicrkamper GG. Mechanism of cataractproduction by 3-/3(2-diethylaminoethoxy)androst-5-en-17-one hydrochloride, U18666A: An inhibitor ofcholesterol biosynthesis. Exf> Eye Res. 1979; 28:673-688.

4. Gerson RJ, MacDonald JS, Alberts AW, et al. On theetiology of subcapsular lenticular opacities producedin dogs receiving HMG-CoA reductase inhibitors. ExpEye Res. 1990;50:65-78.

5. Kirby TJ, Achor RWP, Perry HO, Winkclmann RK.Cataract formation after triparanol therapy. Arch Oph-thalmol. 1962; 68:486-489.

6. Laughlin RC, Carey TF. CataracLs in patients treatedwith triparanol. JAMA. 1962; 181:389-340.

7. Mosley ST, Kalinowski SS, Schafer BL, Tanaka RD.Tissue-selective acute effects of inhibitors of 3-hy-droxy-3-methylgluiaryl coenzyme A reductase on cho-lesterol biosynthesis in the lens. / Lipid Res.1989; 30:1411-1420.

8. Kalinowski SS, Tanaka RD, Mosley ST. Effects of long-term administration of HMG-CoA reduclase inhibi-tors on cholesterol synthesis in lens. Exp Eye Res.1991; 53:179-186.

9. Alberts AW, Chen J, Kuron G, et al. Mevinolin: Ahighly potent competitive inhibitor of hydroxymethyl-glutaryl-coenzyme A reductase and a cholesterol low-ering agent, Proc Natl Acad Sri USA. 1980; 77:3957-3961.

10. Goldstein JL, Brown MS. Progress in understandingthe LDL receptor and HMG-CoA reduclase, two mem-

brane proteins that regulate the plasma cholesterol. /Lipid Res. 1984; 25:1450- 1461.

11. Gretton C. Like falling off a cliff. Med Ad Neivs.1994; 5:3-25.

12. Cenedella RJ. Apparent coordination of plasma mem-brane component synthesis in the lens. Invest Ophthal-rnol Vts Sa. 1993;34:2186-2194.

13. Shi H, Cenedella RJ. Regional distribution of 3-hy-droxy-3-methylglularyl coenzyme A reductase activityand protein mass in the ocular lens. J Lipid Res.1993; 34:2177-2182.

14. Cenedella RJ, Shi H. Spatial distribution of 3-hydroxy-3-methylglutaryl coenzyme A reductase messengerRNA in the ocular lens: Relationship to cholesterolo-genesis. / Lipid lies. 1994;35:2232-2240.

15. Rafferty NS. Lens morphology. In: Maisel H, ed. TheOcular1jens: Structure, Function and Pathology. New York:Marcel Dekker; 1985:1-60.

16. Liscum L, Finer-Moore J, Stroud RM, et al. Domainstructure of 3-hydroxy-3-methylglutaryl coenzyme Areductase, a glycoprotein of the endoplasmic reticu-lum. JBiol Chern. 1985;260:522-530.

17. Edwards PA, Lan S-F, Tanaka RD, Folegman AM.Mevalonolactone inhibits the rate of synthesis and en-hances the rate of degradation of 3-hydroxy-3-methyl-glutaryl coenzyme A reduclase in rat hepatocytes. /BiolChem. 1983; 258:7272-7275.

18. Faust JR, Luskey KL, Chin DJ, Goldstein JL, BrownMS. Regulation of synthesis and degradation of 3-hy-droxy-3-methylglutaryl coenzyme A reductase by lowdensity lipoprotein and 25-hydroxycholesterol in UT-1 cells. Proc Natl Acad Sa USA. 1982; 79:5205-5209.

19. Liscum L, Luskey KL, Chin DJ, et al. Regulation of" 3-hydroxy-3-mcthylglutaryl coenzyme A reductase andits rnRNA in rat liver as studied with a monoclonalantibody and a cDNA probe. / Biol Chern.1983; 258:8450-8455.

20. Ness GC, Eales S, Lopez D, Zhao, Z. Regulation of 3-hydroxy-3-methylgluiaryl coenzyme A reductase geneexpression by sterols and nonsterols in rat liver. ArchBiochem Bwphys. 1994; 308:420-425.

21. Lees MB, Paxman S. Modification of the Lowry proce-dure for the analysis of protcolipid protein. Anal Bw-chem. 1972; 47:184-192.

22. Powell EE, Kroon PA. Measurement of mRNA byquantitative PCR with a n on radioactive label. / LipidRes. 1992:33:609-614.

23. Powell EE, Kroon PA. Low density lipoprotein recep-tor and 3-hydroxy-3-methylglutaryl coenzyme A reduc-tase gene expression in human mononuclear leuko-cytes is regulated coordinately and parallels geneexpression in human liver. J Clin Invest. 1994;93:2168-2174.

24. Dejong WW, Lubsen, NH, Kraft HJ. Molecular evolu-tion of the eye lens. Prog Retinal Eye Res. 1994; 13:391-442.

25. Hitchner WR, Cenedella RJ. HMG CoA reductase ac-tivity of lens epithelial cells: Compared with true ratesof sterol synthesis. Curr Eye Res. 1987;6:1045-1049.



26. Goldstein JL, Brown MS. Regulation of"the mevalovatepathway. Nature. 1990; 343:425-430.

27. Gil G, Faust JR, Chin DJ, Goldstein JL, Brown MS.Membrane-bound domain of HMG CoA reductase isrequired for sterol-enhanced degradation of the en-zyme. Cell. 1985; 41:249-258.

28. Nakanishi M, Goldstein JL, Brown MS. Multivalentcontrol of 3-hydroxy-3-methylglutaryl coen/.yme A re-ductase. J Biol Chem. 1988;263:8929-8937.

29. Condi CC, Edwards PA. Mevalonic acid-dependentdegradation of 3-hydroxy-3-melhylglutaryl coen/ymeA reductase in vivo and in vitro. J Biol. Chem.1994;269:633-638.

30. Bradfute DL, Simoni RD. Non-sterol compounds thatregulate cholesterologenesis. J Biol Chem. 1994;269:6645 T 6650.

31. Quesney-Huneeus V, Galick HA, Siperstein MD, etal. The dual role of mevalonale in the cell cycle. J BiolChem. 1983;258:378-385.

32. Roussillon S, Astruc M, DcFay R, et al. DNA and cho-lesterol biosynthesis in synchronized embryonic ratfibroblasts. Biochim Biophys Ada. 1983;763:l-10.

33. El-Sayed GN, Cenedella RJ. Relationship of cholester-ologenesis to DNA synthesis and proliferation by lensepithelial cells in culture. Exp Eye Res. 1987;45:443-451.

34. Cenedella RJ. Aging and rates of lens-cell differentia-tion in vivo, measured by a chemical appioach. InvestOphthalmol Vis Sa. 1989;30:575-579.

35. Reynolds GA, Goldstein JL, Brown MS. MultiplemRNAs for 3-hydroxy-3-merhylglularyl coenzyme A re-ductase determined by multiple transcription initia-tion sites and iniron splicing sites in the 5'-unlrans-lated region. J Biol Chem. 1985; 260:10369-10377.

APPENDIX

Time required for the lens radius to increase by 5% in19- to 29-day-old rats. Because rat lenses are essentiallyspherical and their density is approximately 1 (1 mg= 1 mm1), the lens radius can be estimated from thevolume:

•JVol.mm' (mg)r ~ \ 4.189

The radius of decapsulated lenses from 29- to 39-day-old rats (used in Fig. 3) increased at a rate of approxi-mately 0.00715 mm/day (Fig. 5). At 29 days of age,the lens radius would be approximately 1.649 mm.When the lens grew to a size at which the radius was1.736 mm, 1.649 mm (100% of the radius at 29 days)was at a point corresponding to 95% of the new radius.Approximately 12 days would have been required forthe radius to increase by this 5% (Fig. 5). After anadditional 12 days, 1.649 trim would now correspondto approximately the 90% radius point of the newradius, 1.649 mm +.(1.736 mm + 12 days X 0.00715mm/day).

-1.75

28 30 32 34 36 38 40Age (days)

1.60

FIGURE 5. Changes in rat lens weight (volume) and radiuswith age.


Documents

Role of transcription, translation, and protein turnover in controlling