981

Rabbit Heart Fatty Acid-Binding ProteinIsolation, Characterization, and Application of a

Monoclonal Antibody

Anne A. Knowlton, Robert E. Burner, and Peter Brecher

A fatty acid-binding protein (FABP) was purified from rabbit heart and characterized withrespect to size, isoelectric point, and tissue distribution. This protein was found in red muscle,diaphragm, and aorta, as well as in the heart. Amino acid composition of rabbit heart FABPdiffered only slightly from the human and rat proteins. Rabbit heart FABP was shown to bindtwo molecules of fatty acid. A monoclonal antibody was developed and used to demonstrate thefeasibility of a one-step purification with affinity chromatography. Cross-reactivity was foundbetween the human protein and the rabbit antibody, and an immunoassay was developed tohuman heart FABP. Levels of human heart FABP in the plasma of patients with acutemyocardial infarction were significantly elevated (83±9 /ig/ml) compared with patients withpulmonary edema (52±7 /ug/ml) and normal volunteers (28±5 ftg/ml; p<O.OS, mean±SEM).(Circulation Research 1989;65:981-988)

Heart fatty acid-binding protein (hFABP) isone member of a multigene family ofproteins characterized by their relatively

small size, cytoplasmic localization, and their abil-ity to bind hydrophobic ligands including long-chainfatty acids, retinol, retinoic acid, and several otherorganic anions. Distinct fatty acid-binding proteins(FABPs) have been isolated from rat liver,1

intestine,2 kidney,3 adipose tissue,4 and heart,5-6

and the complementary DNA for rat hFABP hasbeen obtained.7 Unlike FABPs from other tissues,hFABP, at least in the rat, has a broad tissuedistribution and is most abundant in heart and slowtwitch skeletal muscle; it is present in kidney,testis, adrenal, placenta, and aortic tissue. Althoughthe function of hFABP, like other members of thismultigene family, is presumably to facilitate theuptake and intracellular metabolism of long-chainfatty acids, this function has not been rigorouslyestablished, and the possibility of pleiotropic effectsfor these proteins has been suggested based oneffects on cell growth and differentiation.8

From the Cardiac Muscle Research Laboratory, Cardiovascu-lar Institute, and the Biochemistry Department, Boston Univer-sity School of Medicine, and the Cardiology Section and theThorndike Memorial Laboratory, Boston City Hospital, Boston,Massachusetts.

Supported by National Institutes of Health Research GrantHL-31195 (P.B.). A.A.IC is the recipient of Physician-ScientistAward HL-01853.

Address for correspondence: Anne A. Knowlton, MD, BostonUniversity School of Medicine, R-215, 80 East Concord St.,Boston, MA 02118.

Received December 1, 1988; accepted April 20, 1989.

In recent studies,9 we have described the leakageof hFABP from the intact rat heart after experimen-tally induced ischemia and suggested that loss ofthis protein after ischemia may make the viablecells more vulnerable to cellular damage duringreperfusion. In other studies,10 we identified hFABPin rat aortic tissues and showed that its expressionwas markedly reduced in aorta, but not in othertissues, after experimental hypertension. The pos-sibility that hFABP may have a protective role inthe heart or aorta and that its reduced expressionmay be a causative factor in different forms ofcardiovascular disease prompted us to examine theproperties of this protein in the rabbit, an animalthat has proven useful for studying ischemia,hypoxia, and myocardial metabolism and that is anexcellent experimental model for atherosclerosis.In the present study, we have purified and charac-terized an hFABP from rabbit heart and producedmonoclonal antibodies to this protein. One of theantibodies was used to show the feasibility of puri-fying the protein by affinity chromatography. Inaddition, monoclonal antibodies to rabbit hFABPwere used to develop an enzyme-linked immuno-sorbent assay (ELISA) to human hFABP. With thisassay we measured human hFABP and quantitatedplasma levels in patients with myocardial infarctionand patients with pulmonary edema.

Materials and MethodsIsolation of Rabbit Heart FABP

Rabbit hearts were purchased from Pel-FreezeBiologicals (Gilbertsville, Pennsylvania) and were

by guest on June 22, 2018http://circres.ahajournals.org/

Dow

nloaded from

982 Circulation Research Vol 65, No 4, October 1989

received packed in dry ice. They were stored at-70° C until use. Hearts were minced in a 0.1 Mphosphate buffer, pH 7.4, and then homogenized in2 vol buffer with a Polytron apparatus (BrinkmannInstrs., Westbury, New York). This and all subse-quent steps were carried out at 4° C. The homoge-nate was centrifuged at 18,600g for 20 minutes. Thesupernatant was then centrifuged at 15O,OOQg' in a55.2 Ti rotor (Beckman Instrs., Fullerton, Califor-nia) for 90 minutes. The resulting high speed super-natant was then applied to a Sephadex G-75 column(5x90 cm) preequilibrated with 30 mM Tris, pH 8.3.Elution was performed with 30 mM Tris, pH 8.3,and a low molecular weight fraction, containingpredominantly myoglobin and hFABP, wasobtained. This fraction was dialyzed overnightagainst deionized water; sodium acetate was addedto the sample to a final concentration of 10 mM, pH5.0; and the sample was quickly applied to a CM-52column (Whatman, Clifton, New Jersey). hFABPwas eluted by use of a 10-200 mM NaAc saltgradient, which effectively separated hFABP frommyoglobin. Immediately after elution from the col-umn, 1 M Tris was added to the fractions containinghFABP to neutralize the protein solution. The sam-ples were then dialyzed overnight against 30 mMTris, pH 8.3, and stored in aliquots at -70° C.

Protein samples were analyzed by gel electropho-resis with the method of Laemmli.11 Separating gelscontained 15% acrylamide (National Diagnostics,Manville, New Jersey). Gels were fixed and stainedin either Coomassie brilliant blue (Biorad, Rich-mond, California) or a silver stain (ICN Biochemi-cals, Costa Mesa, California). Amino acid analysisand isoelectric focusing were carried out as describedpreviously.6 Tryptophan content was calculated withthe equation:

A=(5700 n Trp+1300 n Tyr)/MW

where A is the absorbance of a 1 mg/ml solutionthrough a 1 cm path length and n is the number ofamino acids (Trp and Tvr).12 The extinction coeffi-cient was 1.1 for rabbit hFABP. Previously, wehave confirmed the accuracy of this equation bycomparing calculated and measured tryptophan con-tent of the rat heart and kidney FABPs.3 Proteinconcentrations were measured in duplicate usingthe Coomassie assay (Biorad). Bovine serum albu-min was used as a standard.

Liposomal-binding studies were performed withmultilammellar liposomes made with egg yolk leci-thin and oleic acid in a 20:1 ratio as previouslydescribed.13 Radiochemicals were obtained fromNew England Nuclear, Boston, Massachusetts.

Immunologic MethodsSix-week-old Balb/CJ male mice (Jackson Labo-

ratory, Bar Harbor, Maine) were immunized withpurified hFABP in Freund's complete adjuvent.Mice were injected intraperitoneally twice with 2weeks between injections. Ten days after the sec-

ond injection, titers were measured against hFABP,and an intravenous boost of hFABP was given tothe mouse with the highest titer. Three days later,the spleen was removed, and a single cell suspen-sion was made. After lysis of red blood cells withammonium chloride, the donor cells were combinedwith SP2 myeloma cells (American Type CultureCollection), which were hypoxanthine guanine phos-phoribosyl transferase negative.14 The cells weretreated with polyethylene glycol and then gentlypelleted. This pellet was then resuspended in 20%fetal calf serum (FCS) in Dulbecco's Modified EagleMedium (DMEM) and plated the following day onmicrotiter plates at lxlC^/ml in hypoxanthine:

aminopterineUhymidine, 20% FCS, 10% concanava-lin A supernatant DMEM (high glucose) supple-mented with penicillin/streptomycin, L-glutamine,and nonessential amino acids. Concanavalin A super-natant was obtained by stimulating normal spleencells in culture with concanavalin A and then col-lecting the supernatant. The subsequent colonieswere screened at 7-10 days with a noncompetitiveELISA to assess antibody activity. Microtiter plateswere coated overnight at 4° C with 100 ̂ 1 hFABP (1jig/ml) and then washed with phosphate buffer andblocked with 0.2% bovine serum albumin. Dilutionsof media from each colony were incubated in themicrotiter plates. The plates were washed withphosphate buffer, and rat anti-mouse kappa immu-noglobulin G linked to alkaline phosphatase wasadded. After incubation, the plates were developedwith p-nitrophenyl phosphate and read at 410 nmwith a microplate reader (model MR600, DynatechLaboratories, Chantilly, Virginia).

The positive wells were cloned by limiting dilu-tion in 20% FCS DMEM with peritoneal exudatecells from Balb/C mice as a feeder layer. Thepositive walls with one colony were selected andgrown in flasks in 10% FCS DMEM. Media werecollected after the cells became confluent, centri-fuged to remove the cells, and then aliquoted andstored at -20° C until further use.

Isotyping. This was performed with a kit pur-chased from Amersham (Arlington Heights, Illinois).

Ascites. One week after priming with intraperito-neal pristane (2,6,10,14-tetramethylpentadecane),four 8-week-old male Balb/CJ mice (Jackson Lab-oratory) were injected intraperitoneally with 105

cells in 100 /AI DMEM from the clone for antibody624. Starting at 1 week, the mice were tapped everyother day to remove ascitic fluid. Ascites wascentrifuged at 2,30C& at 4° C to remove all cells andpassed over a glass wool filter to remove lipids.Further purification was carried out with an Affi-gelblue column (Biorad) by following the methodsrecommended by the manufacturer. Antibody wasthen aliquoted and stored at -70° C.

Affinity ColumnBy use of cyanogen bromide-linked Sepharose

4B (Pharmacia, Uppsala, Sweden), an affinity col-

by guest on June 22, 2018http://circres.ahajournals.org/

Dow

nloaded from

Knowhon et al Rabbit Heart FABP 983

umn was made with antibody 624. Activation of thesepharose and binding of the antibody was done asrecommended by Pharmacia. By use of purifiedhFABP, different conditions for application andelution were tested. At 25° C, sample was applied in30 mM Tris buffer and was eluted with either 0.1 Mglycine, pH 2.6, or with 1 M NaCl and 30 mM Tris,pH 7.8. High speed supernatant was then applied tothe column; after washing with four void volumes of100 mM NaCl plus 30 mM Tris at pH 7.8, purifiedFABP was eluted with either the glycine or the 1 MNaCl plus 30 mM Tris buffer.

Immunoblotting was done with the method ofBurnette15 and a Transblot cell apparatus (Biorad).Transfer to nitrocellulose paper was done at roomtemperature over 2.5 hours. The transfer bufferconsisted of 20 mM Tris base, 150 mM glycine, and15% (vol/vol) methanol with a final pH of 7.4. Theblot was incubated for 30 minutes with 3% gelatin toblock nonspecific binding. This was followed byincubating for 30 minutes in blotting buffer (10 mMTris-HCl, pH 8.0,150 mM NaCl, and 0.05% Tween)to which anti-rabbit heart FABP monoclonal anti-body 624 in 1:200 dilution was added. The blot wasthen washed and incubated with anti-mouseimmunoglobulin-linked alkaline phosphatase(Promega, Madison, Wisconsin) for 30 minutes.After washing, the blot was developed with sub-strate, nitroblue tetrazolium and 5-bromo-4-chloro-3-indolyl phosphate in (mM) Tris 100, NaCl100, and MgCl2 5, pH 9.5. Thorough washing withdeionized water was used to stop the reaction.

Development of ELISABy use of one of the rabbit monoclonals (624) a

competitive ELISA for human hFABP was devel-oped with the approach described by Engvall.16

Initially, human hFABP was used for the standardcurves, but because this is in short supply, a com-parison of rabbit/rabbit standard curve competitionwith human/rabbit was done. A correction factorwas calculated by using the curves generated, andrabbit hFABP was used for all standard curves. Thesame technique was used for both the human hFABPassay and for the rabbit hFABP assay. Briefly,microtiter plates (Flow Laboratories, McLean, Vir-ginia) were coated overnight with 0.5 /ug/ml hFABPin 0.1 M bicarbonate buffer, pH 9.6. Plates werewashed with phosphate-buffered saline (PBS).Serial dilutions from 500 to 1 ng/well hFABP weredone in triplicate for a standard curve. Sampleswere done in duplicate at two different dilutions. Alldilutions were done with PBS. Antibody 624 wasdiluted V 1,000 to l:50 depending on its concen-tration and added in 100-jttl aliquots to each well.After a 2-hour incubation with gentle shaking,the plates were washed with PBS and then incu-bated with a anti-mouse immunoglobulin-linkedhorseradish peroxidase (Amersham) for 1 hour.The plates were again washed with PBS followedby deionized water and then incubated with hydro-

gen peroxide in a phosphate and citrate buffer witho-phenylenediamine. After development for 5-15minutes, the reaction was stopped with 4JV H2SO4.Plates were read at 490 nm with a microplate reader(Dynatech Laboratories).

Human Plasma SamplesVenous blood samples were obtained from

patients in the coronary care unit at Boston CityHospital within 24 hours of admission after obtain-ing informed consent. Samples were drawn frompatients admitted with clearcut acute myocardialinfarction or anginal syndrome, from patients withpulmonary edema but without evidence of myocar-dial ischemia, and from normal volunteers. Myocar-dial infarction was defined as S-T segment elevationof at least 1 mm in at least two leads and a typicalpresentation of chest pain. All these patients showedevolution of electrocardiographic changes and hadelevation of creatinine kinase levels with positiveMB fractions. Anginal syndrome was defined asclearcut myocardial ischemia as evidenced by S-Tsegment changes or T-wave inversions with clinicalsymptoms characteristic of myocardial ischemiaand was distinguished from myocardial infarctionby absence of evolution of electrocardiogram andnormal serum creatinine kinase levels as deter-mined by the hospital laboratory. Pulmonary edemawas defined as severe congestive heart failure withrales to the apexes. Many of these patients requiredtransient intubation until diuresis improved theiroxygenation. Patients with both pulmonary edemaand clear-cut myocardial infarction or anginal syn-drome were included in the myocardial infarction/anginal syndrome group. Samples were centrifugedat 2,30Qg for 5 minutes at 4° C, aliquoted, andstored at -70° C until assay. The protocol wasapproved by the Human Investigation Committee atBoston City Hospital.

Statistics were performed by an analysis of vari-ance to measure differences between groups, andthe Bonferroni t test was used to test for signifi-cance. Values are expressed as mean±SEM.

Rabbit tissues used for tissue localization studieswere the gift of Dr. Carl S. Apstein, Boston Uni-versity School of Medicine. Human hFABP was thegift of Dr. Robert Troxler, Boston University Schoolof Medicine. All chemicals were at least reagentgrade and were obtained from Sigma (St. Louis,Missouri) unless otherwise indicated.

ResultsRabbit hFABP was purified with minor modifica-

tions of the strategy used to purify hFABP from ratand human hearts. A low molecular weight fractioncontaining predominantly myoglobin and hFABPwas derived by gel filtration of the soluble fractionobtained by high speed ultracentrifugation of thehomogenate. This fraction was further purified bycation exchange chromatography with CM-52, whichselectively absorbed rabbit myoglobin and permit-

by guest on June 22, 2018http://circres.ahajournals.org/

Dow

nloaded from

984 Circulation Research Vol 65, No 4, October 1989

17—

PI

8.55

6.57

5.13 -

4.85

BFIGURE 1. Sodium dodecyl sulfate-pofyacrylamide gelelectrophoresis (15%) illustrating purification steps forrabbit heart fatty acid-binding protein. Lane A: Highspeed supernatant. Lane B: Law molecular weight frac-tion from G-75 column. Lane C: Pure heart fatty acid-binding protein eluted from CM-52 column.

ted an effective purification of the hFABP. Figure 1shows sodium dodecyl sulfate-polyacrylamide gelelectrophoresis of the high speed supernatant frac-tion; the low molecular weight fraction was obtainedby Sephadex G-75 filtration, and the purified rabbithFABP was eluted off the CM-52 column. Therabbit hFABP had an apparent molecular mass of 14kDa, which is consistent with that found for otherhFABPs.

A comparison of the isoelectric points of rat,human, and rabbit hFABP is shown in Figure 2. Thepi for rabbit hFABP was virtually identical to thatof the rat protein, whereas human hFABP wasslightly more acidic than the FABP from the rodentspecies. Table 1 shows the amino acid compositionof rabbit hFABP compared with human and rathFABP and indicates the overall similarity betweenthe different proteins.17 Rabbit hFABP is distin-guished only by the presence of two prolines butwas similar with respect to the absence of cysteine,relative abundance of glutamine and lysine, and thepresence of two tryptophans. Although threoninewas abundant, it was less so than in the rat andhuman proteins. In contrast, the valine content wassimilar in the rabbit and human protein, but greaterin the rat. All three proteins contain a large amountof leucine; the rabbit hFABP has somewhat morethan the rat and less than the human.

The binding characteristics of rabbit hFABP weretested by an assay using multilamellar liposomes as

RABBIT RAT HUMANFIGURE 2. Isoelectric focusing comparing rat, human,and rabbit heart fatty acid-binding protein.

a source of labeled oleic acid (Figure 3). Bindingwas maximum at a molar ratio of 1.5:1 (oleate toprotein), consistent with the previously establishedfinding of two binding sites per mole of protein forrat and human hFABPs under similar experimentalconditions.6-16

Monoclonal antibodies were developed by use ofthe purified rabbit heart protein (see "Materials andMethods"). Two antibodies were obtained; the more

TABLE 1. Amino Add Composition of Rabbit, Rat, and HumanHeart Fatty Add-BIndlng Proteins

Amino acid

Asx

ThrSerGlu

Pro

Gly

Ala

Val

Metlieu

Leu

Tyr

Phe

LysHisArg

Try

Rabbit

14.1

16.08.6

11.8

2.2

11.0

6.8

10.42.56.4

11.6

2.5

6.4

13.02.74.22.1

Rat

15198

12

1

10

6

1336

10

2

6

13

34

2

Human

14.6

19.26.6

11.5

1.1

10.5

7.0

9.1

2.3

5.8

13.7

2.2

6.0

14.7

3.04,2

2.4

Values for rat were determined by DNA sequence.7 Humanvalues were based on amino acid sequence."

by guest on June 22, 2018http://circres.ahajournals.org/

Dow

nloaded from

Knowlton et al Rabbit Heart FABP 985

0.ID

U.Q

i 1 . 0

ml

TQ,

oHUO •+- •+•

0 0 - i e . 2 e . 3OLEIC ACID ADDED CroM)

FIGURE 3. Graph showing liposomal binding assay forrabbit heart fatty acid-binding protein (FABP) demon-strating maximal uptake of 1.5 mol oleic acid/mol heartFABP.

sensitive was produced in larger amounts by theascites technique and used for further studies. Thisantibody was shown by isotyping to be an immuno-globulin G2a antibody. The specificity of this anti-body was tested with hFABP from several species(Figure 4). The antibody showed a high affinity forrabbit hFABP but did not react with either rathFABP or with myoglobin. However, cross-reactivity toward human hFABP was found, andalthough the affinity was less than for the rabbitprotein, it was sufficient for the development of anELISA for human hFABP.

The monoclonal antibody was used in Westernblot analysis to show the tissue distribution ofrabbit hFABP in the rabbit. Figure 51 shows thatsoluble fractions from heart, diaphragm, soleus,

and gastrocnemius muscles all contained hFABP.In Figure 511, a separate Western blot analysisshowed the presence of hFABP in the aorta. Thehigher molecular weight bands noted on this blotare a result of nonspecific binding by the secondantibody and were noted on an identical blot thatwas developed with second antibody but withoutfirst antibody. No hFABP was detected in brain,liver, or lung. The above findings were similar tothose made for the tissue distribution of rat hFABP.

The usefulness of the monoclonal antibody wasestablished by additional approaches. An immunoaf-finity column was prepared and tested for its utilityin effecting a one-step purification of rabbit hFABP.The high-speed supernatant fraction was applieddirectly to the column, and absorbed proteins wereselectively eluted with various salt solutions. Figure6 shows the supernatant fraction and the materialthat flowed directly through the column (lanes Aand B); this figure indicates that most of the pro-teins were not retained. When the column waseluted with 0.1 M glycine, pH 2.6, a single bandcorresponding in size to hFABP was eluted fromthe column (lane C). The bands seen at the top ofthe gel reflect an artifact derived from the silverstain and were found even when conventionallypurified hFABP was run on the gel.

Because of the cross-reactivity of antibody 624with the human protein, we were able to measurecirculating levels of hFABP in patients with acutemyocardial infarction or anginal syndrome or pul-monary edema as well as in normal volunteers. Thevolunteers were housestaff with a mean age of27.9±0.6 years («=8, four males). Their mean levelwas 28 ±5 pg/ml plasma. Levels were measured in14 patients admitted with myocardial infarction11 oranginal syndrome.3 This group had a mean age of61.3±3.5 years, and 57% of the patients were male.The nine patients with pulmonary edema (without

e

90)

> A

l l*

UJQ

_1

OP

TIC

P

2

1

1 .

8 .

8 .

.e

. 6

e

6

e<

i *

# 0 •RflOSIT hFABP

A A^KUfMN hFABP• • «HAT hPABP

' • iT H

• — •

• • — # -f 4 -

—•—

\

N

—4—

—f

\

\ y

FIGURE 4. Graph showingenzyme-linked immunosorbentassay with antibody 624. Sepa-rate curves compare activity ofthe antibody with rabbit heartfatty acid-binding protein(hFABP), rat hFABP, and humanhFABP. The antibody showed noactivity against rat hFABP.

ie i0e ieee

ANTIBODY DILUTION

1 E 4

by guest on June 22, 2018http://circres.ahajournals.org/

Dow

nloaded from

986 Circulation Research Vol 65, No 4, October 1989

B

FIGURE 5. Western blot analysis of tis-sue distribution of heart fatty acid-binding protein in the rabbit. Panel I:Gastrocnemius muscle (A), soleus muscle(B), diaphragm (C), heart (D), and pureheart fatty acid-binding protein (E). Tis-sue samples were matched by tissue wetweight. Panel II: Aorta (A), heart (B), andpure heart fatty acid-binding protein (C).

myocardial infarction or anginal syndrome) had amean age of 65.2±2.6 years, and 56% of thesepatients were male. The patients with acute myo-cardial infarction or anginal syndrome had signifi-cantly higher levels with a mean of 83±9 ^g/tnlplasma compared with normals and with those withpulmonary edema, who had a level of 53±7 tiglrsA.These findings are illustrated in Figure 7. Althoughpatients with pulmonary edema had a higher meanlevel than normal volunteers, this difference wasnot statistically significant.

DiscussionThese studies document several properties of

rabbit hFABP and illustrate a potentially usefulclinical application of a monoclonal antibody againstrabbit hFABP for measuring circulating levels ofthe homologous protein in man. Purification of theprotein was straightforward but differed somewhatfrom the approach used to isolate rat hFABP in thatcation exchange chromatography was required toseparate myoglobin from hFABP. These proteinscould not be resolved adequately by an ion exchangechromatography although several different experi-mental conditions involving pH and salt gradientswere attempted. We minimized the time the proteinwas exposed to acidic conditions during purificationsince we had noticed irreversible aggregation of theliver FABP when exposed to pH conditions below5.13 The protein was obtained in relatively highyields of about 1 mg protein per gram of tissue;these high yields indicate that the protein is abun-dant in rabbit heart as in rat heart.6

The physical and chemical properties of the rab-bit protein were similar to that of rat and humanhFABP, having an acidic isoelectric point, a molec-ular weight of about 14,000, and the capacity to bindmore than one mole of long-chain fatty acid permole of protein. A monoclonal antibody directedagainst the rabbit hFABP cross-reacted with thehuman, but not the rat, protein; thus, a speciesdifference was indicated between rat and rabbit in atleast one epitope. We used this monoclonal anti-body 1) to establish tissue distribution of the pro-

tein, 2) to show the feasibility of purification byaffinity chromatography, and 3) as a tool for mea-surement of the human hFABP in plasma frompatients. Interestingly, there have been no previousreports on the production or use of monoclonalantibodies directed against any of the other FABPs.

The precise function(s) of hFABP remains unclear.The several FABPs show tissue selective expres-sion as observed by us in the rabbit and has beenpreviously been reported in the rat; hFABP isprimarily found in tissues of the heart, aorta, skel-etal muscle, and diaphragm with a predilection for

kDa

68 —

43 —

29 —17 —

BFIGURE 6. Sodium dodecyl sulfate-potyacrylamide gelelectrophoresis (15%) (silver stain) showing one-step puri-fication of rabbit heart fatty acid-binding protein fromhigh-speed supernatant with an affinity column. Lane A:High-speed supernatant. Lane B: Void volume fromcolumn. Lane C: Purified heart fatty acid-binding proteinelated with 0.1 M gfycine. Upper bands represent arti-fact. See text.

by guest on June 22, 2018http://circres.ahajournals.org/

Dow

nloaded from

HUMAN PLASMA LEVELS OF hFABP

Knowlton et at Rabbit Heart FABP 987

158 '

FABP iee-

(ug/ml)

68-

e-

•

o *go.

A

1 ?"

A

1

FIGURE 7. Plot showing human heartfatty acid-binding protein (FABP) plasmalevels. C, controls with no acute cardio-vascular disease (o),- MI, patients withmyocardial infarction (•) or anginal syn-drome (O) (mean value is for entire group);PE, patients with pulmonary edema (A)and pulmonary edema with known coro-nary artery disease (A) (mean value is forentire group). Overlap of symbols occurssecondary to duplicate values. Each of thetop three solid triangles represents twopatients. *Mean±SEM. **p<0.05.

C (8)

* Hmmn */- SEM

** p< e.ee

red rather thai? white muscle.6 In previous workfrom our laboratory, we have demonstrated thathFABP in the rat is present in the rat aorta and thatits expression is inhibited at the transcriptional levelby deoxycorticosterone salt hypertension.10 Smallamounts of hFABP have been identified in the ratkidney, along with a second more abundant kidneyFABP.3

Functional studies have focused on liver andintestinal FABP. Stimulation of microsomal enzymesby these proteins as well as inhibition of mitochon-drial enzymes has been reported.18-20 Rat liverFABP binds acyl-CoA and rysophosphatidylcho-line, but rat hFABP does not.21-22 More work isneeded to define the function(s) of hFABP. Theinhibition of its expression with hypertension in therat aorta raises the possibility that it or its absencehas a role in the underlying molecular mechanism ofvascular change during hypertension.

In a previous study,9 we observed that hFABPleaks from the myocardium with ischemia. We nowpresent data demonstrating the elevation of hFABPlevels in the plasma of patients with acute myocar-dial infarction. hFABP is a much smaller proteinthan creatine kinase and may leak much morereadily from the cell. Work by other investigatorshas shown an increase in serum creatine kinaseafter ischemia in the absence of myocardialnecrosis.23 In humans with acute myocardial infarc-tion or an anginal syndrome, we found a clearcutincrease in plasma hFABP. We observed in a lim-ited number of patients with anginal syndromes, butwithout myocardial infarction, a significant increaseof plasma FABP compared with normals. Althoughthis group was small, having only three patients, itprovides further data suggesting that leak of smallproteins may occur from the myocardium withoutclinical evidence of myocardial infarction.24-23

Because all levels were single samples drawn within24 hours of admission, they do not necessarilyrepresent peak levels. Patients with pulmonaryedema and no evidence by EKG or enzymes ofmyocardial ischemia showed a mixture of plasmalevels most likely representing the mixed etiologiesof their underlying cardiovascular disease. Some ofthese patients with known coronary artery disease(as shown in Figure 7) had no evidence of acuteischemia by EKG on presentation and had normalcreatine kinase levels. The unique properties ofhFABP and the deleterious effects of long-chainfatty acids raise the possibility that loss of hFABPcontributes to the myocardial damage from isch-emia rather than just reflecting tissue necrosis.Further work needs to be done to determine theprecise function of this protein in the myocardiumin die experimental animal model and in humans.

AcknowledgmentsThe authors gratefully acknowledge the expert

technical assistance of Rachel C. Ettinger, MA, andthe assistance of the medical housestaff at BostonCity Hospital in collection of blood samples.

References1. Ockner RK, Manning JA, Kane JP: Fatty acid binding

protein: Isolation from rat liver, characterization, and immu-nochemical quantification. J Biol Chem 1982;257:7872-7878

2. Ockner RK, Manning JA: Fatty acid-binding protein in smallintestine. Identification, isolation, and evidence for its role incellular fatty acid transport. / Clin Invest 1974;54:336-338

3. Lam KT, Borkan S, Claffey KP, Schwartz JH, ChobanianAV, Brecher P: Properties and differential regulation of twofatty acid binding proteins in the rat kidney. / Biol Chem1988;263:15762-15768

4. Matarese V, Bernlohr DA: Purification of murine adipocytelipid-binding protein. Characterization as a fatty acid-and retinoic acid-binding protein. / Biol Chem 1988;263:14544-14551

by guest on June 22, 2018http://circres.ahajournals.org/

Dow

nloaded from

988 Circulation Research Vol 65, No 4, October 1989

5. Said B, Schulz H: Fatty acid binding protein from rat heart.JBiolChem 1984;259:1155-1159

6. Offner GD, Troxier RF, Brecher P: Characterization of afatty acid-binding protein from rat heart. J Biol Chem 1986;261:5584-5589

7. Claffcy KP, Herrera VL, Brechcr P, Ruiz-Opazo N: Cloningand tissue distribution of rat heart fatty acid binding proteinmRNA: Identical forms in heart and skeletal muscle. Bio-chemistry 1987;26:7900-7904

8. Sweetser DA, Heuckeroth RO, Gordon JI: The metabolicsignificance of mammalian Fatty-Acid Binding Proteins:Abundant proteins in search of a function. Annu Rev Nutr1987;7:337-359

9. Knowlton AA, Apstein CS, Saouaf R, Brecher P: Leakageof heart fatty acid binding protein with ischemia and reper-fusion. J Mol Cell Cardiol (in press)

10. Sarzani R, Claffey K, Chobanian AV, Brecher P: Hyperten-sion induces tissue-specific gene suppression of a fatty acidbinding protein in rat aorta. Proc Nad Acad Sci USA 1988;85:7777-7781

11. Laemmli UK: Cleavage of structural proteins during assem-bly of the head of bacteriophage T4. Nature 1970-,227:680-685

12. Cantor CR, Schimmel PR: Biophysical Chemistry, Part II.San Francisco, WH Freeman and Co., 1980, pp 380-381

13. Brecher P, Saouaf R, Sugarman JM, Eisenberg D, LaRosaK: Fatty acid transfer between multilamellar liposomes andfatty acid binding proteins. JBiol Chem 1984;259:13395-134Ol

14. Marshak-Rothstein A, Fink P, Gridley T, Raulet DH, BevanMJ, Gefter ML: Properties and applications of monoclonalantibodies directed against determinants of the Thy-1 locus.J Immunol 1979;122:2491-2497

15. Bumcttc WN: "Western Blotting": Electrophorctic transferof proteins from sodium dodecyl sulfate-poryacrilamide gelsof unmodified nitrocellulose and radiographic detection withantibody and radioiodinated protein A. Anal Biochem 1981;112:195-203

16. Engvall E: Enzyme immunoassay. ELISA and EMIT. Meth-ods Enzymol 1980;70:419-439

17. Ofifner GD, Brecher P, Sawlivich WB, Costello CE, TroxierRF: Characterization and the amino acid sequence of a fattyacid-binding protein from human heart. Biochem J 1988;252:191-198

18. Ockner RK, Manning JA: Fatty Acid Binding Protein. Rolein esterification of absorbed long chain fatty acid in ratintestine. / Clin Invest 1976;58:632-641

19. Wu-Rideout MYC, Elson C, Shrago E: The role of fatty acidbinding protein on the metabolism of fatty acids in isolatedrat hepatocytes. Biochem Biophys Res Commun 1976;71:809-816

20. Lunzer MA, Manning JA, Ockner RK: Inhibition of rat liveracetyl coenzyme A carboxylase by long chain acyl coen-zymc A and fatty acid. Modulation by fatty acid bindingprotein. J Biol Chem l?77;252:5483-5487

21. Burrier RE, Manson CR, Brecher P: Binding of acyl-CoA toliver fatty acid binding protein: Effect on acyl-CoA synthe-sis. Biochim Biophys Ada 1987;919:221-230

22. Burrier RE, Brecher P: Binding of lysophosphatidylcholineto the rat liver fatty acid binding protein. Biochim BiophysAda 1986;879:229-239

23. Heyndrickx GR, Amano J, Kenna T, Fallon JT, Patrick TA,Manders WT, Rogers GG, Rosendorff C, Vatner SF: Creat-inine kinase release not associated with myocardial necrosisafter short periods of coronary artery occlusion in consciousbaboons. J Am Coll Cardiol 1985;6:1299-1303

24. Ellis AK, Little T, Masud ARZ, Klocke FJ: Patterns ofmyoglobin release after reperfusion of injured myocardium.Circulation 1985;72:639-647

25. Katus, HA, Diederich KW, Hoberg E, Kubler W: Circulat-ing cardiac myosin chains in patients with angina at rest:Identification of a high risk subgroup. / Am Coll Cardiol1988; 11:487-493

KEY WORDS • myocardial metabolism• fatty acid-binding protein

myocardial ischemia

by guest on June 22, 2018http://circres.ahajournals.org/

Dow

nloaded from

A A Knowlton, R E Burrier and P Brechermonoclonal antibody.

Rabbit heart fatty acid-binding protein. Isolation, characterization, and application of a

Print ISSN: 0009-7330. Online ISSN: 1524-4571 Copyright © 1989 American Heart Association, Inc. All rights reserved.is published by the American Heart Association, 7272 Greenville Avenue, Dallas, TX 75231Circulation Research

doi: 10.1161/01.RES.65.4.9811989;65:981-988Circ Res. 

http://circres.ahajournals.org/content/65/4/981World Wide Web at:

The online version of this article, along with updated information and services, is located on the

  http://circres.ahajournals.org//subscriptions/

is online at: Circulation Research Information about subscribing to Subscriptions: 

http://www.lww.com/reprints Information about reprints can be found online at: Reprints:

  document. Permissions and Rights Question and Answer about this process is available in the

located, click Request Permissions in the middle column of the Web page under Services. Further informationEditorial Office. Once the online version of the published article for which permission is being requested is

can be obtained via RightsLink, a service of the Copyright Clearance Center, not theCirculation Research Requests for permissions to reproduce figures, tables, or portions of articles originally published inPermissions:

by guest on June 22, 2018http://circres.ahajournals.org/

Dow

nloaded from