XeCI laserinduced fluorescence detection ofperoxidized lipoproteins in lipidrich atherosclerotic lesionsAA Oraevsky12 PD Henry3 SL Jacques12 FK Tittel11 Departmentof Electrical and Com ID: 823088
Download Pdf The PPT/PDF document "' R Z Q O R D G H G ) U R P K W W ..." is the property of its rightful owner. Permission is granted to download and print the materials on this web site for personal, non-commercial use only, and to display it on your personal computer provided you do not modify the materials and that you retain all copyright notices contained in the materials. By downloading content from our website, you accept the terms of this agreement.
' R Z Q O R D G H G ) U R P K W W
' R Z Q O R D G H G ) U R P K W W S V Z Z Z V S L H G L J L W D O O L E U D U \ R U J F R Q I H U H Q F H S U R F H H G L Q J V R I V S L H R Q 7 H U P V R I 8 V H K W W S V Z Z Z V S L H G L J L W D O O L E U D U \ R U J W H U P V R I X V HXeCI laser-induced fluorescence detection ofperoxidized lipoproteins in lipid-rich atherosclerotic lesionsA.A. Oraevsky1'2, P.D. Henry3, S.L. Jacques1'2, F.K. Tittel11 .Departmentof Electrical and Computer Engineering,Rice University, Houston, TX 770302. Laser Biology Research Lab, UT/M.D. Anderson CancerCenter, 1515 Holcombe Boulevard, Houston, TX 770303. Cardiology Division, Baylor College of Medicine,One Baylor Plaza, Houston, TX 77030AB STRACTLaser-induced fluorescence spectroscopy of arterial surfaces provides information about thecomposition of atherosclerotic plaques. The aim of the study was to determine whether accumulation ofperoxidized lipoproteins in arterial walls, a process postulated to play a role in initiating atheroscleroticchanges, can be demonstrated by fluorescence spectroscopy.XeC1 excimer laser (= 308nm) induced fluorescence of human aortas containing early lipid-rich,non-collagenous lesions exhibited marked red shifts and broadening of the fluorescence spectra comparedwith spectra from non-atherosclerotic aortas. Similar profiles were observed in spectra obtained fromoxidatively modified LDL, but not native LDL. In hypercholesterolemic rabbits with early foam celllesions, spectral shifts resembled those of oxidized f3-VLDL, the major lipoprotein accumulating in arteriesof rabbits fed cholesterol.XeCl laser-fluorescence spectroscopy of arterial surfaces may be useful for the identification ofarterial plaques indicative of atherosclerosis in its early and probably reversible stages.1. INTRODUCTIONArteries with atherosclerotic changes exhibit altered autofluorescence responses that provideinformation about the fluorophore-chromophore composition of the underlying lesions. Characterization ofarterial tissue by fluorescence spectroscopy has been performed with various lasers including the Ar-ion(476.5nm)1,3?XeCl (308 nm)8'11 ,XeF(351nm)12, He-Cd (325nm)2'9,and N2 (337 nm)13lasers. Advanced human atherosclerotic lesions as seen in unselected autopsy specimens are often highlycollagenous and may contain surprisingly little lipid14. The fluorescence properties of such lesions arestrongly influenced by the matrix proteins forming the fibrous caps that cover advanced plaques9. Incontrast, during early atherogenesis, the subendothelial pace is occupied by lipid-laden macrophages orfoam cells, the major component of fatty streaks15'1°. According to current concets, lipoproteinsaccumulating in foam cell lesions undergo peroxidative and hydrolytic modifications1i-22Ithas beensuggested that these lipoprotein modifications pla3 an important role in attracting circulating monocytesand transforming them into lipid-laden foam cells1In this paper we have tested the hypothesis that lipoprotein peroxidation products in earlyatherosclerotic lesions can be detected by fluorescence spectroscopy 23 Experiments were performedusing a XeCl laser because its wavelength (=308 nm) permits excitation of majority of tissue.fluorophores and because near UV light can be delivered through flexible optical fibers adaptable tocardiovascular
catheterization techniques.0-8194-1 105
catheterization techniques.0-8194-1 105-1/93/$6.00SPIE Vol.1878 (1993)131 ' R Z Q O R D G H G ) U R P K W W S V Z Z Z V S L H G L J L W D O O L E U D U \ R U J F R Q I H U H Q F H S U R F H H G L Q J V R I V S L H R Q 7 H U P V R I 8 V H K W W S V Z Z Z V S L H G L J L W D O O L E U D U \ R U J W H U P V R I X V H2. MATERIALS AND METHODS2. 1 .LaserFluorimeterA XeC1 excimer laser (model LPX-600,Lambda-Physik,Germany) producing 60-ns pulses of 308-nm light was used to elicit fluorescence spectra from arterial surfaces exposed to air or from 1.jpoproteinsolutions contained in high purity quartz cells (Fig. 1). XeC1 laser fluence did not exceed 10 J/cm2 toavoid changes in tissue autofluorescence due to laser irradiation.FIGURE 1.Fluorescence spectroscopic setup.XeC1 excimer laser pulse at 308nm was used to excite arterialsurfaces or isolated lipoproteins insolution. The emitted fluorescencewas collected with an optical fiberand transmitted to a spectrometercoupled to an optical multichannelanalyzer. An interference mirrorreflecting 99 % of radiation at 308nm was placed in front of thespectrometer entrance slit toattenuate back-scattered laserlight. Emitted fluorescence wasevaluated in the spectral regionbetween 330 and 600 nmThe irradiated region of the sample was approximately 1mm in diameter, a compromise betweenspatial selectivity and detectability of fluorescence emission. The emitted fluorescence was collected withfused silica optical fibers (Diaguide, Fort Lee, New Jersey) with a core diameter not exceeding 1/10 that ofthe irradiated area in order to avoid shadowing of the target under irradiation. The end of the optical fiberwas positioned perpendicular to the target surface and its tip was precisely centered above the target at adistance equal to the length of the excitation beam diameter. These precautions minimize fluorescencedistortion due to scattering and reabsorption within the tissue24.Collected fluorescence was transmitted through a fiber optical system to a spectrometer coupled to agated multichannel spectral analyzer (model OMA ifi, EG&G Princeton Applied Research, Princeton,New Jersey). Fluorescence was induced by a single laser pulse and recorded during a 20-ns gatepositioned at the maximum amplitude of the fluorescence pulse. Before each experiment spectralcalibration of spectrometer was performed.A high reflectance (99 %) mirror (308 FR-iD-FL, ARC Corp., Acton, Massachuseus) was placed infront of the spectrometer entrance slit to suppress 308 nm back scatter from the sample. To improve thesignal-to-noise ratio, at least four (average 7) fluorescence spectra were obtained from each target site.Profiles of the spectra were numerically smoothed using a narrow interval procedure. Smoothed spectrafrom different arteries and different arterial locations were compared by superimposing them. To facilitatecomparison of the spectra, they were normalized with respect to integrated fluorescence intensity.32 ISPIE Vol. 1878 (1993) ' R Z Q O R D G H G ) U R P K W W S V Z Z Z V S L H G L J L W D O O L E U D U \ R U J F R Q I H U H Q F H S U R F H H G L Q J V R I V S L H R Q 7 H U P V R I 8 V H K W W S V Z Z Z V S L H G L J L W D O O L E U D U \ R U J W H U P V R I X V H2.2. Studies wit
h human lipoproteins and arteriesHuman l
h human lipoproteins and arteriesHuman lipoproteins were fractionated by ultracentrifugation of pooled plasma collected in EDTA(1mg/mi) as previously described 25,Thelipoproteins were dialyzed against phosphate-buffered salinecontaining 0.01% EDTA. Protein in the samples was determined by the Lowry method using serumbovine albumin as a standard26 O,uthflve modification of the low density lipoprotein fraction (LDL,density 1.019 -1.063g/ml) was performed by exposing native LDL for 24 h to cultured rabbit endothelialcells equilibrated with serum-free FlO medium25Samplesof native LDL serving as control weresubjected to incubations in FlO medium containing no cells. All LDL samples were re-isolated byultracentrifugation. The extent of lipoprotein modification was assessed by measurements of thiobarbituricacid -reactivesubstances (TBARS) and electrophoretic mobility as previously reportedHuman aorta samples were obtained at post mortem examination from 19 patients within 6 hoursafter death. The wide age range of the donors (10-73 years) insured the collection of specimens with orwithout atherosclerosis of varying severity15ples were examined rapidly after excision to avoidartifacts resulting from prolonged storage. Throughout the experiments, the specimens were superfusedwith saline (21 °C) to avoid tissue desiccation. Six to ten samples, approximately 1cm2 in size, were cutout from each aorta at sites thought to contain no atherosclerotic changes, fatty lesions (fatty streaks), oradvanced fibrous plaques. Areas selected for sectroscopic study were marked with India ink andsubsequently excised for histologic examination0. The photomicrographs of histologic sections wereevaluated by two blinded independent observers and categorized according to the Stary classification ofatherosclerotic plaques15Sections exhibiting no or minima! changes including varying degrees of intimal thickening werecalled nonatheroscierotic (normal zones, Stary type I). Fatty streaks (early lesions) corresponded to lesionscontaining layers of macrophages (foam cells) and extracellular lipid particles (Stary lesion type II). Theselesions contained by defmition no stainable collagen deposits. Samples exhibiting such lesions wereobtained from the aorta of one child and four adults. Advanced atheromas represented lesions collectedfrom middle-aged patients (56-73yearsold) with either fibro-lipid (Stary lesion type IV) or purely fibrous(collagenous) plaques (Stary lesion type V). In our context, the term fibrous plaque (Stary lesion type V)was used, when irradiation sites of the superficial intima (1mm thick subendothelial layer) contained acollagenous cap free of stainable lipid. In contrast, when irradiation sites contained in addition to collagenany stainable lipid, the lesions were called fibro-fatty (Stary type IV).2.3. Studies with lipoproteins and arteries from rabbits with and without hypercholesterolemiaRelationships between spectroscopic characteristics of plasma lipoproteins and alteration in arterialautofluorescence were further evaluated in rabbits. Male albino New Zealand rabbits were placed onstandard chow or on 1% rabbit cholesterol pellets as previously described28. After a 10 week dietaryperiod, plasma total cholesterol measured by the cholesterol oxidase method averaged 1.2 0.2 (SE) and31.19.0 mM in six control and six cholesterol4ed rabbits, respectively.
Plasma was fractionated asdescribed for
Plasma was fractionated asdescribed for human samples. The major cholesterol-carrying lipoprotein fraction in cholesterol-fed rabbitsis recovered at a density of ()VLDLwere oxidized by incubation with endothelial cells25. Spectroscopic evaluation of rabbit aortas andnative or oxidized rabbit lipoproteins (-VLDL) were performed as described above.2.4. Analysis of Spectra and Statistics.Sample groups selected according to their histologic categories yielded in general similarfluorescence spectra (except Stary group IV) as assessed by gross fluorescence band-shapes. To assessSPIEVol. 1878 (1993)133 ' R Z Q O R D G H G ) U R P K W W S V Z Z Z V S L H G L J L W D O O L E U D U \ R U J F R Q I H U H Q F H S U R F H H G L Q J V R I V S L H R Q 7 H U P V R I 8 V H K W W S V Z Z Z V S L H G L J L W D O O L E U D U \ R U J W H U P V R I X V Hvariance of measured spectra in each histologic category, two parameters were determined for eachspectrum: (1) the position of the fluorescence maximum and (2) the fluorescence band-width at halfmaximal amplitude (FWHM). Standard deviations (SD) of the parameters (1) and (2) from the meanvalues for Stary categories I (no or minimal changes) and V (lipid free fibrous caps) were very small (nm). Slightly larger values ()()&#x 10 ;&#xnm w;re ;&#xobta;&#xined;&#x for;&#x the;&#x Sta;&#xry c; teg;&#xory ;&#xII f; tty;&#x str;êks;&#x. In;&#x con;&#xtras;&#xt,00;for Stary category IV (mixed fibro-lipid plaques), standard deviations were large (100 nm), reflecting thecompositional variance of the lesions. Therefore, measured fluorescence spectra belonging to Starycategories 1,11, and V were averaged, but for flbro-fatty lesions we report here only selected examples offluorescence spectra.The statistical significance of differences (p-values) between mean values of fluorescence maximumposition for Stary groups I, II, and V was evaluated by the analysis-of-variance procedure using theSAS/STAT statistics program. The generalized linear model (GLM) employing a multiple comparison testyielded p-values for each pair of Stary categories I, II, and V. P -values of 0.01were considered torepresent significant intergroup differences. Fluorescence spectra for rabbit aorta samples and differentlipoprotein preparations were analyzed by the same statistical procedure.3.1. Human samplesCl)C0C00C00Cl)00U-0�Co3. RESULTSThe increased fluorescence of oxidized LDL at visible wavelengths is associated with a change in theabsorption spectrum. Differences between the average absorption spectra from 6 native and 6 oxidizedLDL samples in the wavelength-range 240 nm -800nm are illustrated in Figure 3. The absorption bandspresent in native samples at 450, 480, and 520 nm are absent after oxidative modification. Qualitative andquantitative differences between spectra from native and oxidized LDL very similar to those depicted inFigures 2 and 3 were observed in a total of six preparations from six different plasma pools.34ISPIE Vol. 1878 (1993)Averaged fluorescence spectra from 6 human LDL pools before and after their oxidativemodifications are shown in Figure 2. The native LDL profile exhibits a single fluorescence peak atapproximately 340 nm (Fig. 2, continuous line). Oxidative modification of LDL produces a prominentspectral shift toward the red w
ith a maximum at about 430 nm. Standard
ith a maximum at about 430 nm. Standard deviations of the position offluorescence maxima from the mean value do not exceed 4 nm for native and 6 nm for oxidized LDL. Inaddition, there is a significant broadening of the spectrum consistent with an overlap of severalfluorescence bands (Fig. 2, dashed line). The fluorescence band-width for native and oxidized LDL were55±4 nm and 145±6 nm, respectively.16FIGURE2.12Fluorescenceprofile of native(nonoxidized) human LDL (continuousline). There is a fluorescence maximumat approximately 340 nm with no8appreciableadditional features.Fluorescence profile of LDL oxidized byincubation with endothelial cells (dashed4line).Compared with native LDL, thespectrum exhibits a marked red shift.Additionally, there is a broadening of thespectrum, consistent with multiple0fluorophores.300350400450500550Wavelength (nm)600 ' R Z Q O R D G H G ) U R P K W W S V Z Z Z V S L H G L J L W D O O L E U D U \ R U J F R Q I H U H Q F H S U R F H H G L Q J V R I V S L H R Q 7 H U P V R I 8 V H K W W S V Z Z Z V S L H G L J L W D O O L E U D U \ R U J W H U P V R I X V HFigure 4a (continuous line) shows an average fluorescence profile elicited from 3 samples of a 12year old child aorta. These samples had no structural signs of atherosclerosis, including an absence offoam cells and stainable extracellular lipid. The nonatherosclerotic intima exhibited an abundance offmely reticulated material known to represent proteoglycan ground substance (Stary type I lesion15). Thespectrum from this sample reveals a broad emission band with a fluorescence maximum at about 370 nm,smaller peaks at 382 nm, 390 nm, and 430 nm, and a broad shoulder extending into the yellow range.The FWHM value is about 105 nm. Virtually identical spectra were obtained from the aorta of a 10 yearold child and two young adults. The average spectrum from the juvenile aortas with minimal changes(Stary type I, 10 samples from 4 aortas) was very similar to that depicted by the continuous line in Figure4a except that its major maximum was shifted 4 nm toward longer wavelengths..FIGURE3.2-.Absorptionspectra of native (continuousline)and oxidized (dashed line) LDL insolution. Oxidative modification resultsin a loss of absorption bands in the 430 -540nm wavelength region.800Wavelength (nm)Figure4a (dashed line) depicts an average fluorescence spectrum taken from an adult aortacontaining abundant foam cells and numerous extracellular Oil Red 0 -positiveparticles and pronouncedthickening of the aortic wall, but no collagenous deposits (Stary lesion type II, 9 lesions from 7 aortas).Compared with the non-atherosclerotic aorta (continuous line), there is a broadening of the spectrum witha marked shift towards the red and a maximum occurring at about 430±9 nm. Lipid-rich lesions (Starytype II) collected from a total of five aortas yielded similar spectra with a FWHM value of 170±10 nm.The p-value for the significance of the difference in position of the fluorescence maximum between lipid-rich plaques versus non-atherosclerotic (normal) human aorta walls was 0.0007.An average spectrum of 3 crude preparations of human aortic elastin is depicted in Figure 4b(continuous line). The broad spectral band extends from the ultraviolet to the yellow range (FWHMequals 105nm) with a maximum at 407±3 nm. Similar
spectra have been obtained from elastin
spectra have been obtained from elastin withthe use of a He-Cd laser9. Compared with elastin, the fluorescence spectrum of the non-atheroscleroticaorta appears to be shifted toward shorter wavelengths. This may reflect non-elastin matrix proteins suchas glycosaminoglycans abundant in L-tryptophan (Figure 4b, dashed line), an amino acid not contained inelastin28. The XeCl laser induced fluorescence spectrum of tryptophan was characterized by a position ofthe maximum at 362±3 nm and a FWHM value of 65±3 nm.Figure 5 shows an average fluorescence profile from fibrous caps and representative fluorescencespectra from mixed fibro-fatty plaques. The average fluorescence spectrum for well developed fibrouscaps without stainable superficial lipid (Stary type V, 18 lesions from 12 aortas) is drawn with acontinuous line. The spectrum is narrower than that of the non-atherosclerotic human aorta (Fig. 4a,continuous line), lacking the broad shoulder in the yellow range (FWHM is 85±5nm).The position of themaximum for the fibrous plaque fluorescence is at 382±4 nm. The spectrum of fibrous plaques resemblesSPIEVol. 1878 (1993) / 35Oxi-LDL200300400500600700 ' R Z Q O R D G H G ) U R P K W W S V Z Z Z V S L H G L J L W D O O L E U D U \ R U J F R Q I H U H Q F H S U R F H H G L Q J V R I V S L H R Q 7 H U P V R I 8 V H K W W S V Z Z Z V S L H G L J L W D O O L E U D U \ R U J W H U P V R I X V HU)CCU)0CU)0(I)U)0U)�U)�U)CU)CU)0CU)U)0U-U)�CUU)Figure5 (dashed lines) depicts also spectrafromirradiation sites containing both collagenous andabundant lipidic deposits (fibro-fany plaques, Stary type IV lesions). Fibro-fatty plaques were the mostfrequent lesions in our autopsy material (79 Stary type IV lesions in 15 aortas). The spectra from suchlesions appear to combine features of pure fatty lesions and lipid-free collagenous caps.Figure 5 (long dashed line) depicts a fluorescence spectrum with two major bands obtained from alesion with a thick collagenous cap and abundant subjacent lipid. The intense band with a maximum at384 nm and a FWHM value of about 83 nm resembles the fluorescence spectrum of pure fibrous caps,whereas the broad band with a maximum at about 427 nm and a FWHM value of about 175 nm issuggestive of a lipid-rich lesion. Figure 5 (short dashed line) shows another example of a mixed fibro-lipid lesion with a thin collagenous cap incompletely masking subjacent lipid. The structural and chemicalvariability of fibro-fatty lesions is reflected in marked spectral variability that precludes useful statisticalanalysis.3.2. Rabbit sanplesCompared with human LDL, oxidative modification of rabbit 3-VLDL produced a noticeable but lessprominent spectral change, a phenomenon that may reflect VLDLs decreased susceptibility to oxidation19.36ISPIE Vol. 1878 (1993)1612that of crude collagen8'9. The p-values of the significance of the difference in the position of fluorescencemaxima between fibrous plaques versus normal human aorta, and between fibrous versus lipid-richplaques were 0.0012 and 0.0016, respectively.FIGURE 4.(a) Fluorescence spectrum from humanaortas: Continuous line -spectrumfromthe aortic surface of a 12 year old child(Stary lesion type I -slightintimalthickening with ground substanceaccumulation, but without microscopicsigns of atherosclerosis). Dash
ed line -spectrumfrom the aorta of a you
ed line -spectrumfrom the aorta of a young adultat a site containing a non-collagenouslipid-rich lesion (Stary type II), (b)Spectrum from a crude preparation ofhuman aortic elastin (continuous line) andfrom L-tryptophan in aqueous solution(dashed line).300350400450500550600Wavelength (nm) ' R Z Q O R D G H G ) U R P K W W S V Z Z Z V S L H G L J L W D O O L E U D U \ R U J F R Q I H U H Q F H S U R F H H G L Q J V R I V S L H R Q 7 H U P V R I 8 V H K W W S V Z Z Z V S L H G L J L W D O O L E U D U \ R U J W H U P V R I X V H�U)CU)CU)0CU)U)0U-U)�U)U)a:�1U)CU)CU)0CU)U)SU-U)�U)U)a:SPIE Vol. 1878 (1993) / 37300350400450500550Figure 6 (continuous line) shows the average spectrum of the 3-VLDL fraction (d)from plasma of 6 normocholesterolemic rabbits. Figure 6 (long and short dashed lines) depicts averagefluorescence profiles from 6 unoxidized and 6 oxidized J-VLDL samples prepared from pooled plasma of6 hypercholesterolemic rabbits. The rabbit lipoprotein spectra are asymmetrical, suggesting that more thanone fluorophore contributed to the total fluorescence pattern. The spectrum of nonnocholesterolemicrabbit plasma exhibits a major peak at 410±5 nm and minor peaks at about 450 nm and 468 nm (Figure 6,continuous line). These minor peaks become the major emission maxima in unoxidized frVLDL preparedfrom hypercholesterolemic rabbit plasma (Figure 6, long-dashed line).FIGURE 5._Continuousline -Averagefluorescencespectrum of of 18 thick fibrous capscontaining no Oil-Red-O-positive materialDashed lines -Fluorescencespectra ofadvanced fibro-fatty plaques (Stary typeIV). One lesion was abundant in extra-and intracellular lipid and had a thincollagenous cap (short-dashed line);another lesion had a thicker collagenouscap with deposits of extracellular Oil-Red-O-positive material (long-dashedline). Note that the spectra of mixedfibro-fatty plaques have features in600commonwith spectra from histologicallypure fatty and pure fibrous lesions.Wavelength(nm)Oxidationof -VLDL produces a broadening of the spectrum, which is less marked than thatobserved with oxidation of human LDL. Oxidation changes the FWHM value from 86±4 nm to 100±8nm that broadens the fluorescence band-width, but only 14 nm. In contrast to the oxidation of humanLDL, oxidative modification of rabbit f-VLDL shifts the fluorescence spectrum toward the blue with theappearance of new maxima at approximately 382±3 and 394±3 nm (Figure 6, short-dashed line).16FIGURE6.12Fluorescenceprofile of native(nonoxidized) VLDL isolated from a poolof normocholesterolemic rabbit plasma8(continuousline). Fluorescence spectrumof f3-VLDL isolated from a pool ofhypercholesterolemic rabbit plasma (long-4dashedline). Oxidative modification of J3-VLDLproduces a spectral shift towardsthe shorter wavelength range (short-dashed line).300350400450500550Wavelength (nm)600 ' R Z Q O R D G H G ) U R P K W W S V Z Z Z V S L H G L J L W D O O L E U D U \ R U J F R Q I H U H Q F H S U R F H H G L Q J V R I V S L H R Q 7 H U P V R I 8 V H K W W S V Z Z Z V S L H G L J L W D O O L E U D U \ R U J W H U P V R I X V H�(1)CC00C000LI0�00Themajor finding of this study is that arteries undergoing early atherosclerotic changes exhibit analtere
d autofluorescence that may reflect the
d autofluorescence that may reflect the fluorescent properties of the lipoproteins accumulating in thetissue.In the case of human atherosclerosis, lipoprotein retained in the arterial intima belongs predominantlyto the low density fraction (LDL)21'22, possibly because of its high binding affinity to intimalglycosaminoglycans. Owing to its abundance in polyunsaturated cholesterol esters, LDL is particularlysusceptible to oxidation19. It is of interest that several of the features observed in the spectra ofoxidatively modified LDL, such as position of maximum, fluorescence band-width, and nonsymmetricalcomplex profile are also seen in the spectra from pure fatty lesions in human aorta.The opposite oxidation-shifts of the fluorescence spectra of human LDL and rabbit -VLDL is aninteresting finding that may be explained on the basis of the different compositions of these twolipoproteins. Compared with LDL, -VLDL is much less abundant in polyunsaturated cholesterol estersand it contains apoproteins other than apo-B (apo-E and apo-C's). These differences may account fordisparate oxidative modifications and may explain the opposite shifts.38ISPIE Vol. 1878 (1993)Figure 7 (continuous line) depicts the average spectrum from the aortas of 6 rabbits withouthypercholesterolemia. As opposed to the non-atherosclerotic human aorta with a well developed intimallayer abundant in glycosaminoglycans, the native rabbit intima is extremely thin with the internal elasticmembrane in close apposition to the endothelial cell layer. The fluorescence spectrum has a band-width of103±4 nm and shows a major fluorescence peak at 407±3 nm, a fluorescence signature characteristic ofelastin (Fig. 4a)9.Figure 7 (dashed line) depicts the average spectrum from the thoracic aortas of 6 cholesterol-fedrabbits at the site of an advanced foam cell lesion. The spectrum exhibits a major peak at 382±2 nm andlesser peaks at 394±2 nm, 450±4 and 468±4 nm, features found in the spectrum of oxidized -VLDL(Figure 6, short-dashed line). The general profile of fluorescence spectra and FWHM values for lipid richplaques in rabbits strongly resemble that of the oxidized f3-VLDL. Spectra from rabbit aortas occupied byfoam cell lesions were very reproducible and were seen in all rabbits both at the level of the aortic arch anddescending thoracic aorta. The p-value of the significance of the difference in position of fluorescencemaxima between lipid plaques and normal rabbit aorta walls was 0.0011.1612FIGURE7.Fluorescence spectra from the aorticsurface of a normocholesterolemic rabbit8(continuousline) and from that of ahypercholesterolemic rabbit at a sitecontaining a multilayered foam cell lesion4(short-dashedline). The spectrum fromthe diseased rabbit aorta has similaritieswith that from oxidized -VLDL (Figure6, short-dashed line).300350400450500550600Wavelength (nm)4. DISCUSSION ' R Z Q O R D G H G ) U R P K W W S V Z Z Z V S L H G L J L W D O O L E U D U \ R U J F R Q I H U H Q F H S U R F H H G L Q J V R I V S L H R Q 7 H U P V R I 8 V H K W W S V Z Z Z V S L H G L J L W D O O L E U D U \ R U J W H U P V R I X V HThe similarities of the fluorescence spectra for oxidized lipoproteins in solution and lipid-rich plaquesin both rabbits and humans support the view that oxidized lipoproteins are important determinants of
thefluorescence of lipid-rich atheroscl
thefluorescence of lipid-rich atherosclerotic plaques 20The fluorescence properties of LDL have been ascribed to the tryptophan residues of apoprotemB19,3 1-33Oxidative modification of LDL is known to alter amino acid residues in apo B.Oneimportant change involves lysine residues which undergo covalent linkage with lipid peroxidationproducts21'32'3'. It has been suggested that peroxidation-related changes in LDL autofluorescence reflect4-hydroxynonenal-amino acid condensation products32. It is important to note, however, that theabundance and state of oxidation of polyunsaturated fats might by themselves contribute to changes inlipoprotein fluorescence32,34,35Whereasthe autofluorescence of early lipid-rich lesions in humans and rabbits may reflect thefluorescence properties of the dominant lipoprotein accumulated in tissue, the autofluorescence of arteriescontaining no superficial lipid deposits may express mainly the fluorescence of matrix proteins directlyunderlying endothelial surfaces. In non-atherosclerotic arteries with a thin intima, the fluorescence mayreflect the abundance of elastin, whereas in arteries with advanced lesions, collagen may become thedominant determinantThe fluorescence spectrum of normal human aorta is not identical to that of elastin, although itdisplays some important features of elastin fluorescence such as the long broad shoulder in the range from450 nm to 550nm.Althoughelastin contributes to the fluorescence of normal aorta, other fluorophores inthe well developed human aortic intima may be important. In contrast, in the rabbit aorta where the intimais very thin with the internal elastic membrane located directly under the endothelial layer, the fluorescencespectrum is similar to that of elastin (compare continuous line profiles at Fig. 4b and Fig.7).In the normal human aorta, the position of the fluorescence maximum differs from that of pureelastin (370 nm vs 407 nm), which probably reflects other tryptophan-containing matrix proteins in thesubendothelial space. The similarity between fluorescence spectra of normal aorta and elastin reported byDeckelbaum et ai9 may have resulted from the use of a longer excitation wavelength (325nmvs 308 nmin this study). The longer effective penetration depth of He-Cd laser radiation may lead to the induction offluorescence from deeper structures including the inner elastic membrane.Overall, our findings confirm earlier reports that fluorescence spectroscopy may be a useful tool tocharacterize the composition of atherosclerotic lesions143fluorescence profiles of normal aorta walland various plaques excited by the laser pulse at 308 nm differ significantly.Because of its abilityto effectively excite numerous tissue fluorophores, the XeC1 laser radiation at 308 nm may be particularlyuseful for diagnostic applications. Angioscopic detection of arterial tissue containing peroxidizedlipoproteins might prove important, since they may signal arteries at risk for the development of advancedlesions. Lesions with dense collagen may have a limited potential for true regression (absolute rarefactionof collagen) 36 Therefore, identification of atherosclerosis before dense collagenization has occurred maybe important for the prevention of advanced disease and its inevitable complications. According to currentconcepts, atherosclerotic plaques at risk for thrombotic complications m
ay contain abundant foam cells 37.Theref
ay contain abundant foam cells 37.Therefore, the occurrence in close proximity of fluorescence spectra indicative of fibrous and lipid-rich locimight help in the detection of advanced plaques with inflammatory cellular components, lesioncharacteristics thought to be risk-factors for rupture and trombosis.S. ACKNOWLEDGEMENTSThis work was supported by NIH grants ROl-HL40884 and RO1-HL36894. We should like tothank Drs. D. Weilbaecher and S.L. Thomsen for providing the human artery samples.SPIEVol. 1878(1993)139 ' R Z Q O R D G H G ) U R P K W W S V Z Z Z V S L H G L J L W D O O L E U D U \ R U J F R Q I H U H Q F H S U R F H H G L Q J V R I V S L H R Q 7 H U P V R I 8 V H K W W S V Z Z Z V S L H G L J L W D O O L E U D U \ R U J W H U P V R I X V H(66L) 9L91 1on 3IdSIOt LLT-LT : 8L6T VSfl !dSPt'dVlWN 3OJ 'uiiOidOdii iCisup-Mo-odE Jo iipu2oiiq ppui-u2iCxj jpj iAuiqosjj ' qnoprnj ' nos 6T 91 :o zc61 joiqoiaipj ywj V13 LI1 U! SP!(1!I P!PPO'' J ous.id .11 2oIopd uiwnq U! SpIXO1dOdIj Jo jO1 qi uo spru Uc[ 'ax ussç 'f usuiuij ' UUuUH 'ç PUIAJO 8 ZSOC-LPO : O66T VPVVf s!su2o.Iq11 suiioidodr :mç wmzuj, 'j 2iqu LT LLT-f'9T :0! O66 s!soJaldsou?UV 'UOI31UU.OJ )j.tjS I(UIJ 'j uBwnqUou q:i U! jAj oj ump ssu2oiqiy �:j �ssoj ' pnsj 9j CE-I - 611 6 686T s.lsoJ?l3soy?JJv Sflpi 2unoA pu uarpjrqo jo siui £rtuwoo u suo!sI jo uo!ssi2o.id pu UOt1flIOA :3}J "ns ç j 9cLT-LfLT :g 686T UO!W!fl3JD q21p c.nrnoioo UjjXtS U! P UOUOJtJU! Jp.roo1cuI nog UT SUUt? £flflIOlOO qinotth iovui inoj qi m snbiqd onwjoso.iq Jo UOtflSOdWOO qi jo s!sAI1u opLuoqtho}4 DM suqo'j 'ii su!M 'OS Appi 'HV p21 'N 6EST -9cT :9- '066T UO43 I nbiqd jo soisoup ouosionjj PonPU! 1S 'N 21qUA 'S 21qLflA '11 UI1?1U1S '1 UOSUOf 'S SP2Ua-UOSS1PUV 'J ci-coz : 'Ø33T PS aJ:17 's!Iu uuinq OO1jOSO1tfl1? WO1J OUOSJ.OfljJO2tW ponpu! 1W!OX Ud A1UH 'cW1 UO.fl 'a Aiqini 'Nd PU!J "! !"d 'HO 2!11d n: TLc-cc : 6861 pajr ns SJ9SI7TJ 'sUo!l!pUoo Arnjq -uoU pu Aiiiq uotssiuioioqd nss pii pAuo isiq itmox mu gØ JO S32S.I3JtII4D d fflZ '1 I0tI 'd i.unqxn 'H !U1 ' 1?pJOfl){ '44 ipqon ' iin'a 'o pujjo ' . T i ZEZI-ZZZT :9 6861 8uuaaui8u pawoig s'uv.ij q'q'qj ssoijosoiqiy o uotnoijddy :nssu urnnnq u ssp jo ssou2p ioj uosionjj pnpu-isj jo ppow iij -uo s pi "ai ' jjip 'j' uoj, ' ouniwzitj 'pj gg "j �wuo-spitqoj j TO6168T :g 686T UO!W!fl3lD .ouos1onU p!.IJB o!:o1I3soJq1 ptm jIUJiOU UM1( 3U.IJJIp qi .IoJ ssq :y wnqppj 'fj pU.Lfl9 'io vuio "ij is 'i uic 'ri wi 6 66-c6 :sr 686T V3V vaiuiipo4aadS '° ""i v ouos.ionjj isj joi.tq s PPd 'DI !.TSU 'D P!N 'cR1 'TI 2UOj 'V p1S44 '{ UlçT 'Q I'd 'd !'0L 'If 1U.IL ' S osc-zLc :j 6861 p'jv 8ing sjaszYi uopui is1? mu 9Li psno jo sopsiio.nqo jiuiods jo uoiijy S} PPd "ai: 'd !Acxd 'D IPU1N '1 AiC1qfljO 'pj �wnuo,j-spiiqo'j 'jj sqpiwq L 6coT-EcoT :9gg6T S?!3J0 s!1ug umunq Jo OUOSJOflJJO2fl : IUH "J iCaiqin 'Nd PURL '1W 'JL 'D1 Uiq 'S 'Po1'N "TO .uPPA ' '44 UOUE 9 SSz-cLz :z 8861 PS aJ:r7 .sllgM PSSA p°°iq 3:Io1Joso14:1 uiumq Jo sudoid i'ods s'i u.mqo 'MI UTUOIj)J 'VV 'DA '[S AOUJI2 'SA AO)JOW 'y I)JSAJØ ç EL-OL dd 'L861 I'°A MN - 2'LPP!H UTj1fl :2iji 2uudS 'SU3 SI 'UOSJ1d 'M pu 1qU1?As iqp 'jjj - (doaso4aadg .13SV7 :u 'SpssA oioipsoiqn uw.unq Jo ssAjiui iuoods JS :s U!IfltIO)IV 'yr AAI[ 'S AOUU2)J 'DA 0)IUUkPU1O 'SA AO1J)jO 'yr c)jSAEJ() 'f L6L1-I6LT :cz..a L861 UO43 UVnj f pipui ui
nb!u1p u i :siu uwnq ojosoiqu jo sdiu 'U
nb!u1p u i :siu uwnq ojosoiqu jo sdiu 'Ud 1(1UH �'j suqo' 'i pun. 's ipoqn "j Caiqin 'js uoii c-oE: :,,, L86T PJ4J 8ng SJaSZrJ '3UOS.IOflJj ponpuisq Xq m.io ojosoiqi pU puuou jo :j4 uo' 'SN qqnj 'SH 'NI tu" 'j' mnqppj 'z jg-Øg : çgj idO iddV 'ousionjj 2usn ssoijoswpi jiu snoiq Jo ssou2ij 'SI'1 PPd 'O1 1T!V 'III UJN 'HN JJ!PI '3 SOj P "m 2UIM '3 Jfl1)j S33NII '9 ' R Z Q O R D G H G ) U R P K W W S V Z Z Z V S L H G L J L W D O O L E U D U \ R U J F R Q I H U H Q F H S U R F H H G L Q J V R I V S L H R Q 7 H U P V R I 8 V H K W W S V Z Z Z V S L H G L J L W D O O L E U D U \ R U J W H U P V R I X V Hlt'/(U661) RL9I 1on 31d 9ir8 :(lllddns)Z8 O66T UOfl1jfl31!D SgS!p L''I ouJqos U jflSU 11JflOSA LflUOJOO Jo MIA OJOUI pui OJO1U y :fl SIA L 89T-LN: 8L6T ttat S7SOJdPSOJdZPV UOiSSaIaI U! flSSt3 At1DUUOD :'ij UOUSUU\T 9 9c-6c :s 8L6T d05' W??/3 l!O "Vf MflA1 :uouipixo pdj jo iuui.insi :ff A1U9 ç 8O9O9 :9 L861 UY3 l!g I sonpoid uosodwoop pxo.id p!di Aq up1oidodJod jo snpsai u!SAI io UOIWZUAUp U S1jflSJ u!O1dOd!j Cisup oj urnnnq jo uornpix :fl iqoJqui1 oc-c6P :sz L86J S?! P!d!7 I spi(qpi jo uonuu pu uunn!A P'-' sP! 'n piiurrnsunAjod jo sso :upoidod JcJ!sup MOJ uuinq Jo uOp!xO1fly :a iio '0 i2iqupn '0 SUJflf 'H o:i:-izi: :9 D!861 'P!'!7 SYd uiayj fl-up1oJdodJod-uou u!1wdod!j isup MOj J 1OJJ[ sioMqns ii dAi pu oiiupid!Iouuou moij suoJdod!I Asup MOj JO SUOT P!P°! cq uqounb ouosaionjj :a AID "J 1UI13 'd ZIfl-1SflOU '& 89c-T9c :jj T66T qWOYL S'lSOJd/dSOUdJJl/ UO1D1U1U -uiioidodrj Asup oj oqi jo ouods 'is 'jj insiq "ç ij 'j unpuqurr . j 6o:-co: :ss 986J Sd! .01 iooz; Nd i)pojq wn!oIgo qii iuui.ii Aq iqqiu pj-jo.iisjoqo u uogixpi mflosA iuputhp-umpqiopu Jo uoinusoij c[d £IUH 'us iisp.ioj ' swj 's sjpj ' iissoi 'ar q'qH o 086T 'LZ-ZTZ dd 'j iidiqo 'ssaij iC1!SJA!Ufl 11S MO] 'SpOq4 O1SPfl DAt '-"4°°D 'M'D JOOpU 6Z ç9-Øçdd ' A 'UOj 'UO1j 1OO[ :ssij J3 M PU S1jSpWJ J Aq p 'ssiis pui u!1sgIL :uj uusiiijo siudoid powqzoq pu ' uojos "ti uqo'j 'j qoo 8Z cLc-ccc :rjj Z6T 1 w'V Cdosoiorui uoiiop Aq pur sinsp JjA1 SI? S.IqJ O1S?p Jo uouizuioimqo ouIqo pui inion.us qj Cosdu 'y xop 'jj pqiuso'j 'jv uswr LZ cLz-c9z :j Tç6T WdllJ /oig I iu2ai jouqd ujoj qip iuuioinsiw uioi1.j �:çj "y J.r 'IN qnoiqso 'HO '"°'ii 9Z z91:09T :,tE O66J dJflJVN suiwdodq Aisup-io ppou u uupoosXj Aq uogxpi iupubp -wn!Ioq1opu Jo IUUUU?dUJI Ud (IUH ' suqo 'cj ns.uoj 'y sui 'j unAin ç Z6Z98Zfr :sz 686T idO idd 'Injog uuinq jo uosionjjoini :pu pqJn:I Jo Cdoosoiiods ouosionjj si r'ii "is snboç "jj wnioj-spiqoj 'j iz1ij p 06t'8 :(T)ZL 66T Y"'! UO?Wfl3Jl3 suiToidod!J pzipixo.id pu suosj qo.z-pdj uiq sppjus o!dooso.r1ods :si.n oioipsoiqu jo ouosionjj ponpu isij ix :Ani}j ffd H1 'Aaiqin v'i 'iuni 'snboç TS '2!2d H0 'VV 'AjSA1U 66-88 :j O66J f UVH inq rnio pu puuou UT suiioJdod!'T 'Tf mfl1ZU 'Q 1qu91 'JA4 JUSO' 'A PfSUd 'S fl1!-'H-IA ç-ç :01 S so souaJ.w uiio.idodq isup MOj jo uonoJTpow AI11PIXO 2uunp ppuua sdotd ioj oU!ods spoqiut prnojoououi pu iusiuy m UJflIZUM 'fl SStIJflD ' Cq1U1?S1fl41J1 'VS qoo 'MS ipttq 'EIA PPJ S0?1 'S fl1!.1H!A 'M J 66-06 :g 166T isatuJui!jJ' sionpp uto1d-p!d!I oiods-uouppco u!inuoo pui 'suploidod!! A1tsup-Moj Jo uorn?plxo iowoid 'supoidod!! pJpouJ pmi2p suo!sI o1oJoso1q1 1qqmi i.uoij pujos Ajqsij sipo unoJ p3Ap-aqdo1o1y4 ff wn1z1I 'M i)JSuflJ 'S '[ II!}1 'DI OO) '3J,44 pIJUSO