This article was downloaded by: [Moskow State Univ Bibliote]On: 06 November 2013, At: 10:43Publisher: Taylor & FrancisInforma Ltd Registered in England and Wales Registered Number: 1072954 Registeredoffice: Mortimer House, 37-41 Mortimer Street, London W1T 3JH, UK

Journal of Immunoassay andImmunochemistryPublication details, including instructions for authors andsubscription information:http://www.tandfonline.com/loi/ljii20

IMMUNOASSAY OF MODIFIED FORMS OFHUMAN LOW DENSITY LIPOPROTEININ ISOLATED CIRCULATING IMMUNECOMPLEXESGabriel Virella a , Joan Colglazier b , Charlyne Chassereau b , Kelly J.Hunt c , Nathaniel L. Baker c & Maria F. Lopes-Virella d ea Department of Microbiology and Immunology , Medical Universityof South Carolina , Charleston , South Carolina , USAb Department of Medicine , Medical University of South Carolina andRalph H. Johnson VA Medical Center , Charleston , South Carolina ,USAc Medical University of South Carolina, Medicine , Charleston , SouthCarolina , USAd Department of Medicine , Medical University of South Carolina ,South Carolina , USAe Department of Laboratory Services , Ralph H. Johnson VA MedicalCenter , Charleston , South Carolina , USAAccepted author version posted online: 13 Apr 2012.Publishedonline: 16 Jan 2013.

To cite this article: Gabriel Virella , Joan Colglazier , Charlyne Chassereau , Kelly J. Hunt ,Nathaniel L. Baker & Maria F. Lopes-Virella (2013) IMMUNOASSAY OF MODIFIED FORMS OF HUMAN LOWDENSITY LIPOPROTEIN IN ISOLATED CIRCULATING IMMUNE COMPLEXES, Journal of Immunoassay andImmunochemistry, 34:1, 61-74, DOI: 10.1080/15321819.2012.683500

To link to this article: http://dx.doi.org/10.1080/15321819.2012.683500

PLEASE SCROLL DOWN FOR ARTICLE

Taylor & Francis makes every effort to ensure the accuracy of all the information (the“Content”) contained in the publications on our platform. However, Taylor & Francis,our agents, and our licensors make no representations or warranties whatsoever as tothe accuracy, completeness, or suitability for any purpose of the Content. Any opinionsand views expressed in this publication are the opinions and views of the authors,and are not the views of or endorsed by Taylor & Francis. The accuracy of the Contentshould not be relied upon and should be independently verified with primary sources

of information. Taylor and Francis shall not be liable for any losses, actions, claims,proceedings, demands, costs, expenses, damages, and other liabilities whatsoever orhowsoever caused arising directly or indirectly in connection with, in relation to or arisingout of the use of the Content.

This article may be used for research, teaching, and private study purposes. Anysubstantial or systematic reproduction, redistribution, reselling, loan, sub-licensing,systematic supply, or distribution in any form to anyone is expressly forbidden. Terms &Conditions of access and use can be found at http://www.tandfonline.com/page/terms-and-conditions

Dow

nloa

ded

by [

Mos

kow

Sta

te U

niv

Bib

liote

] at

10:

43 0

6 N

ovem

ber

2013

IMMUNOASSAY OF MODIFIED FORMS OF HUMANLOW DENSITY LIPOPROTEIN IN ISOLATED CIRCULATINGIMMUNE COMPLEXES

Gabriel Virella,1 Joan Colglazier,2 Charlyne Chassereau,2 Kelly J. Hunt,3

Nathaniel L. Baker,3 and Maria F. Lopes-Virella4,5

1Department of Microbiology and Immunology, Medical University of South Carolina,Charleston, South Carolina, USA2Department of Medicine, Medical University of South Carolina and Ralph H. JohnsonVA Medical Center, Charleston, South Carolina, USA3Medical University of South Carolina, Medicine, Charleston, South Carolina, USA4Department of Medicine, Medical University of South Carolina, South Carolina, USA5Department of Laboratory Services, Ralph H. Johnson VA Medical Center, Charleston,South Carolina, USA

& Modified lipoproteins are able to induce inflammatory reactions through innate immunitypathways and are immunogenic, leading to an autoimmune response that results in the formationof proinflammatory immune complexes. The measurement of circulating oxidized lipoproteins andcorresponding antibodies has, therefore, been proposed as an approach to assess the risk for compli-cations in patients with diabetes and for the risk of cardiovascular disease in the general popu-lation. However, the majority of modified low density lipoprotein (LDL) in the peripheralcirculation exists in the form of immune complexes, and this is a significant obstacle for themeasurement of modified LDL and the corresponding antibodies. In this manuscript, we describein detail the methodology developed by our group for isolation and fractionation of circulatingimmune complexes (IC), allowing the accurate assay of different LDL modifications. Thisapproach has resulted in several studies showing that the levels of modified LDL are risk factorswith a stronger association to diabetic retinopathy, nephropathy, and macrovascular disease.Ongoing research is focused on evaluating the predictive power of modified LDL levels for the devel-opment or progression of atherosclerotic cardiovascular disease in other patient populations and onthe simplification of the assay to make it more applicable to diagnostic laboratories.

Keywords diabetic complications, immune complexes, modified LDL capture assays,modified lipoproteins, oxidized LDL, risk factors

Address correspondence to Gabriel Virella, Department of Microbiology and Immunology, MedicalUniversity of South Carolina, 173 Ashley Avenue, Charleston, SC 29425, USA. E-mail: virellag@musc.edu

Journal of Immunoassay and Immunochemistry, 34:61–74, 2013Copyright # Taylor & Francis Group, LLCISSN: 1532-1819 print/1532-4230 onlineDOI: 10.1080/15321819.2012.683500

Dow

nloa

ded

by [

Mos

kow

Sta

te U

niv

Bib

liote

] at

10:

43 0

6 N

ovem

ber

2013

INTRODUCTION

Our understanding of the pathogenesis of human atherosclerosisunderwent a significant evolution during the past three decades. In 1976,the main emphasis was on the response to (endothelial) injury as the mainunderlying pathogenic mechanism, and on the possible role of hyperlipide-mia as a primary insult leading to endothelial damage.[1] By 1993, the roleof inflammatory processes triggered by endothelial and smooth muscle celldamage was postulated to play a most significant role,[2] and a few yearslater atherosclerosis was defined as an inflammatory disease[3]—a conceptthat enjoys general acceptance at the present time.[4,5]

Once it was accepted that a chronic inflammatory process was key to theprogression of atherosclerosis, the search for insults that triggered theinflammatory reaction became a focal point. Copper-oxidized and mini-mally modified low density lipoprotein (LDL), two models of sponta-neously oxidized LDL (oxLDL), were given the leading role in the late1980s and early 1990s.[6–10] It was soon recognized that different formsof modified LDL (mLDL) were immunogenic and would induce thesynthesis of specific antibodies.[11–14] The implication of mLDL antibodiesin vascular inflammation was strongly supported by findings supportingthe formation of antigen–antibody complexes in the arterial wall byYla-Herttuala et al.[15,16] The detection of oxLDL and oxLDL-specific IgGantibodies in atherosclerotic lesions by this group retains its critical signifi-cance as the best circumstantial evidence pointing to a significant pathogenicrole of antigen–antibody complexes in atherosclerosis 20 years after theywere initially reported. Strong support for the pro-inflammatory propertiesof mLDL-containing immune complexes (IC) has also been obtained inex vivo experiments using antigen–antibody complexes prepared withcopper-oxidized human LDL and purified human oxLDL antibodies.[17–19]

A surge of interest in the role of mLDL antibodies led to numerous andoften contradictory reports of human and animal studies, as summarized byus in previous publications.[20,21] The contradictory nature of the dataobtained in these studies results from two main factors: first, mLDL assaysare adversely affected by IC formation, and there has been no significanteffort to account for this interference and standardize the assays; second,there are significant differences in lipoprotein metabolism and in thehumoral immune response between humans and rodents.[20,22]

In the last 15 years, we have developed methodology for the isolationand fractionation of circulating immune complexes[13] and capture assaysfor different LDL modifications.[23] The combination of these two techni-ques has allowed us to measure modified LDL present in isolated IC, andthis parameter has shown to have significant correlations with macro- andmicrovascular disease in diabetes,[24–28] and particularly with different

62 G. Virella et al.

Dow

nloa

ded

by [

Mos

kow

Sta

te U

niv

Bib

liote

] at

10:

43 0

6 N

ovem

ber

2013

indices of cardiovascular disease.[24–26] In this manuscript, we describe indetail the methodology that has evolved over the years in our laboratoryfor the isolation and analysis of human circulating IC containing differentmodifications of LDL.

METHODS

Patients and Blood Samples

The methodology described in this manuscript was developed usingserum samples obtained from healthy human volunteers and from subjectsfrom the Diabetes Control andComplications Trial=Epidemiology of DiabetesInterventions and Complications (DCCT=EDIC) cohort, as described in themanuscripts reporting the data generated with our methodology.[24–26]

Recently we have also studied a cohort from the Veterans Affairs Diabetes Trial(VADT) study.[29,30] The DCCT, EDIC, and VADT studies were approved bythe Institutional Review Board of all participating centers, and all participants,as well as all healthy volunteers that were not part of the DCCT=EDIC studies,provided written informed consent.

Isolation of Circulating Immune Complexes

Polyethylene glycol (PEG) precipitation is a well established techniquefor the isolation of soluble circulating immune complexes.[31–33] We havechosen a concentration of 3.5% (w=v) of PEG 8000 for our precipitationstep, because under these conditions we did not observe significant precipi-tation of radiolabeled, aggregate-free, human LDL.[34] It needs to be notedthat there are lot-to-lot variations in PEG 8000, so it is advisable to confirmthe optimal conditions for each different batch.

Equal volumes (usually 1mL) of serum and 7% of PEG 8000 in 0.1Mborate-buffered saline, pH 8.4 (w=v), are mixed by adding PEG to theserum dropwise under gentle stirring in a vortex. Glass tubes are preferablefor the precipitation step, but polystyrene tubes can also be used. The sam-ples are then incubated for 18 h at 4�C to allow maximal precipitation, andat the end of this incubation period, the samples are centrifuged at2060� g for 20min at 4�C. After carefully removing the supernatant, theprecipitate is washed with a volume equal to sixfold the initial sample vol-ume of 3.5% PEG in 0.1M borate-buffered saline, pH 8.4 (w=v), and recen-trifuged as described above. A volume equal to that of the initial serumsample of 0.05M tris-borate buffer containing 21mM EDTA, pH 9.3, pre-warmed at 37�C, is added to the washed precipitate, and the centrifugetubes are gently tapped to resuspend the precipitate, incubated at 37�C

Immunoassay of Human LDL 63

Dow

nloa

ded

by [

Mos

kow

Sta

te U

niv

Bib

liote

] at

10:

43 0

6 N

ovem

ber

2013

for 20min, and gently tapped once more. A small aliquot of the resus-pended precipitate is saved for measurement of total cholesterol. Theremaining resuspended precipitate is used for isolation of modified LDLpresent in precipitated IC and corresponding antibodies, if necessary.The resuspended precipitates can be stored at 4�C for one week. Iflong-term storage is necessary, freezing at �20�C is preferred.

Fractionation of Isolated IC

Variations of the methodology for the fractionation of isolated IC byaffinity chromatography using either staphylococcal protein A or strepto-coccal protein G have been previously described.[14,33,35,36] The procedurethat we are describing in this manuscript is the one used for isolation ofmodified forms of LDL from isolated IC. The rationale for our procedureis the difference in the binding affinity of human autoantibodies to oxLDL(Kd of 1.02�8mol=L, range 0.26 to 4.79, n¼ 30[18]) and the binding affinityof IgG to protein G (Kd ranging from 2.0 to 6.1�9ml=L for human IgG ofsubclasses 1 to 4[37]). This difference allows the elution of modified LDLfrom isolated immune complexes using mild conditions that dissociatethe IgG-modified LDL IC without affecting the binding of IgG to protein G.

To achieve this dissociation of modified LDL-containing IC, we start by dia-lyzing the aliquot of resuspended precipitate set aside for chromatography(usually 0.8mL) overnight against 0.02M sodium phosphate buffer contain-ing 0.5M NaCl, pH 7.3. The same buffer is used to equilibrate 5mL proteinG-sepharose columns (protein G sepharose 4 fast flow, GE Healthcare Bio-Sciences, Piscataway, NJ, USA) used in the next step. The resuspended and dia-lyzed precipitate is added to the column, and after the sample has diffused intothe gel, 0.02M sodium bicarbonate buffer containing 0.5M NaCl, pH 7.3 buf-fer is used to obtain a first eluate containing modified forms of LDL and otherproteins not trapped by protein G. These first eluates are next dialyzed againstphosphate buffered saline, pH 7.4, containing 279mM=mL of EDTA andstored in screw-capped cryovials for up to a week at 4�C and at �20�C whenstored for longer periods of time. The concentrations of cholesterol and ApoBin those eluates are determined prior to the capture assay. The concentrationof ApoB is used to adjust the concentration added in capture assays, to avoidadding insufficient or excessive concentrations of modified LDL.

The protein G-bound IgG is eluted with 0.1M glycine-HCL, pH 2.7.This second eluate contains IgG antibodies co-precipitated with the IC,and should be immediately neutralized with 1M tris-HCL buffer pH 9.0,dialyzed, and stored as described for the first eluate. The assay of IgGand IgM oxLDL antibodies isolated from circulating IC has been reportedin a separate publication.[36]

64 G. Virella et al.

Dow

nloa

ded

by [

Mos

kow

Sta

te U

niv

Bib

liote

] at

10:

43 0

6 N

ovem

ber

2013

Quantitation of Modified LDL Isolated from IC

Under the conditions of our isolation method, the LDL eluted from theprotein G was dissociated from circulating immune complexes formed withautoantibodies specific for the different types of modifications that occurspontaneously. By using antibodies to the different modifications, we havedeveloped capture assays for oxLDL, advanced glycation end-point-modi-fied (AGE-LDL), and malondialdehyde-modified LDL (MDA-LDL).[23]

The antibodies used in our assays were raised in our laboratory. NewZealand white rabbits were immunized with oxLDL, AGE-LDL, and MDA-LDL. After exsanguination, the sera were aliquoted and stored at �20�C.To obtain specific antibodies, we first fractionated serum aliquots by affinitychromatography in protein G-sepharose columns (Protein G sepharose 4fast flow, Amersham, Piscataway, NJ, USA), equilibrated, and washed with0.1M sodium bicarbonate buffer containing 0.5M sodium chloride, pH8.3. The protein G-bound IgG is eluted with 0.1M glycine-HCL, pH 2,and immediately neutralized with 1M tris-HCl buffer, pH 9.0. The neutra-lized eluate is next dialyzed against 0.01M sodium bicarbonate buffer, pH8.3, and then run though a native LDL–sepharose column equilibratedwith the same buffer to absorb antibodies to unmodified ApoB.[23] Theantibody recovered in the column washout is tested for specificity by EIA,checking that it reacts specifically with the modification used to immunizethe rabbit. The reactivity of our rabbit antibodies to oxLDL, MDA-LDL, andAGE-LDL with unmodified LDL was insignificant. The capture assayinvolves coating EIA plates with antibodies specific for each modification(usually 3.5 mg=well, established by running a panel of combinations of anti-body concentrations coating the plates and different concentrations ofmodified LDL) and adding 100 mL=well of the first Protein G eluates,adjusted to contain 2mg=mL of ApoB, based on a previous ApoB assay.The captured mLDL is detected using horseradish peroxidase-labeled goatanti-human ApoB (Academy Biomedical Co., Houston, TX, USA), dilutedto 1:1600 in 1% BSA-PBS. 2,20-Azinobis [3-ethylbenzothiazoline-6-sulfonicacid]-diammonium salt (ABTS) is used as substrate, and the plates are readat 414 nm. The assays are calibrated with known concentrations of the dif-ferent modifications of LDL prepared in our laboratory chosen for optimallinearity under the general conditions of the assays (32 mg=mL for oxLDL,44 mg=mL for MDA-LDL, and 2.16 mg=mL for AGE-LDL).

Because the recovery of mLDL from IC is not uniform for every serumsample, total cholesterol is measured in the washed and resuspended pre-cipitates, and total cholesterol and ApoB are measured in the first proteinG eluate that contains modified forms of LDL. The concentrations of thedifferent LDL modifications measured by the capture assays are then cor-rected to reflect the concentration of a given LDL modification=mL of

Immunoassay of Human LDL 65

Dow

nloa

ded

by [

Mos

kow

Sta

te U

niv

Bib

liote

] at

10:

43 0

6 N

ovem

ber

2013

serum, according to the following formulas (all concentrations expressedin mg=mL):

�mLDL in 1st eluate=ApoB concentration in sample vol:

used for the capture assay

�

total cholesterol in isolated IC

or:

�mLDL in 1st eluate=cholesterol concentration in sample vol:

used for the capture assay

�

total cholesterol in isolated IC

In vivo modified LDL is likely to contain a combination of epitopes,such as MDA-lysine (present in low concentrations in oxLDL and in veryhigh concentrations in MDA-LDL prepared in the laboratory), CML-lysine(present in high concentrations in AGE-LDL and low concentrations inoxLDL prepared in the laboratory), and others that are not characterized,particularly in oxLDL and AGE-LDL, as suggested by absorption studiescarried out with these modified forms of LDL and the corresponding anti-bodies.[14] Therefore, the results of the assay of mLDL are to be consideredas a reflection of the proportions of different LDL modifications containedin the IC isolated from a given individual.

RESULTS AND DISCUSSION

Production of Antibodies to Different LDL Modifications

Rabbits are an excellent source of polyclonal antibodies to differenthuman LDL modifications, and the resulting antibodies have high avid-ity[21] and specificity.[14] The specificity of the antibodies is evaluated in avariety of ways, as previously described.[14,23] One of the tests we routinelyrun consists of coating an EIA plate with an antibody specific for a givenLDL modification, then adding serial dilutions of several different modi-fied forms of human LDL, and determining their binding to the immobi-lized antibody with a peroxidase-labeled goat anti-human ApoB antiserum.Figure 1 reproduces the result of a capture assay to test the specificity of arabbit anti-MDA-LDL obtained in our laboratory. There was no detectablebinding of native LDL (nLDL) or AGE LDL and only minimal binding ofoxLDL, which contains MDA-lysine and other epitopes.[14] We also per-form inhibition assays, in which we determine the difference in bindingof the antibody to a given modified LDL (MDA-LDL antibody in the

66 G. Virella et al.

Dow

nloa

ded

by [

Mos

kow

Sta

te U

niv

Bib

liote

] at

10:

43 0

6 N

ovem

ber

2013

example given in Figure 2) after absorption with the immunogen and otherforms of modified LDL. The results of this competition study show that thereactivity of MDA-LDL antibody was almost 100% inhibited by absorptionwith MDA-LDL, while oxLDL, known to contain a much lower density ofMDA-lysine epitopes,[23] only reduced the reactivity with MDA-LDL byabout 35%. The fact that MDA-BSA significantly reduced the reactivityagainst MDA-LDL by 85% suggests that the MDA-lysine epitope is immuno-dominant and is not very affected by the configuration of the protein whereis located.

Because approximately 50mL of rabbit antiserum can be collectedfrom a rabbit by exsanguination, we had sufficient antibody from a givenrabbit to perform all necessary assays in any large cohort of patients. Whenthe supply of a given antibody was exhausted, we used serum from a secondrabbit with a similar pattern of reactivity and similar antibody titer. If neces-sary, the antibody concentration was adjusted to obtain a calibration curvethat was super-imposable to the one obtained with the previous antibodybatch. We decided to use rabbit polyclonal antibodies because their higheravidity is advantageous in enzymoimmunoassays and because we had noproblems obtaining antisera of almost identical characteristics in differentrabbits immunized with the same preparation of modified LDL.

FIGURE 1 Study of the specificity of a rabbit MDA-LDL antibody by capture assay. The concentration ofanti-MDA-LDL used to coat the plate was 35 mg=mL=well and the different LDL modifications weretested at concentrations ranging from 1 mg=mL to 0.06mg=mL. The immobilized MDA-LDL wasdetected using horseradish peroxidase-labeled goat anti-human ApoB, diluted to 1:1600 in 1% BSA-PBS.ABTS was used as substrate and the plates were read at 414 nm.

Immunoassay of Human LDL 67

Dow

nloa

ded

by [

Mos

kow

Sta

te U

niv

Bib

liote

] at

10:

43 0

6 N

ovem

ber

2013

Development of Capture Assays for Modified LDL

The availability of antibodies to ox, MDA, and AGE-LDL enabled us todevelop capture assays for these LDL modifications. As illustrated byFigure 3, using MDA-LDL antibody as an example, immobilized MDA-LDLantibody bound MDA-LDL preferentially over oxLDL, and basically did notcapture other tested LDL modifications or non-modified (n-LDL) over awide concentration range. Also using the MDA capture assay as an example,Figure 4 shows that although the absolute values of MDA determined byGC=MS and capture assay were different (a direct consequence of the dif-ferent calibrations of the two different assays), there was a satisfactory cor-relation between the obtained values.

Isolation and Fractionation of Immune Complexes

Our isolation and fractionation technique for circulating IC has evolvedand improved over the last decade. With the development of specific cap-ture assays for different forms of mLDL, we proved that the majority of thecirculating molecules of modified LDL are bound to the correspondingantibodies and precipitate as immune complexes (Figure 5). This is a very

FIGURE 2 Reactivity of rabbit MDA-LDL antibodies. After IgG isolation from the serum of a rabbitimmunized with human MDA-LDL and absorption of the isolated IgG with native LDL to remove anti-bodies to unmodified ApoB, the anti-MDA-LDL IgG was tested for specificity. The enzymoimmunoassayplates were coated with MDA-LDL at 0.75mg=well and 100mL aliquots of the MDA-LDL antibody weretested without absorption or after absorption with different modified human LDL preparations,MDA-BSA, and native human LDL (n-LDL), all at a concentration of 200mg=mL. The results depictedin the figure are expressed as the mean reduction of reactivity after incubation with the indicated pro-teins relative to the unabsorbed sample (100% reactivity).[14]

68 G. Virella et al.

Dow

nloa

ded

by [

Mos

kow

Sta

te U

niv

Bib

liote

] at

10:

43 0

6 N

ovem

ber

2013

significant observation, because it provides very strong corroboration to thereservations that we have expressed for the accuracy of the assays of bothmodified forms of LDL and their corresponding antibodies in serum orplasma. The formation of immune complexes is a very significant causeof error in both assays, and to our knowledge none of the proposed assaysfor circulating modified LDL of mLDL antibodies has considered this prob-lem and included a strategy that would eliminate the interference of IC.

FIGURE 4 Comparison of the values calculated for MDA-LDL isolated from IC obtained from nine dif-ferent individuals by gas chromatography=mass spectroscopy and by our capture immunoassay by linearregression analysis.

FIGURE 3 Capture assay of MDA-LDL. The enzymoimmunoassay plate was coated with MDA-LDL anti-body at 35 mg=well. The plot represents the O.D. (414nm) obtained after incubating different modifica-tions of LDL and a native LDL preparation at variable concentrations with the immobilized antibody.Peroxidase-labeled goat anti-human ApoB was used to detect the captured LDL.[23]

Immunoassay of Human LDL 69

Dow

nloa

ded

by [

Mos

kow

Sta

te U

niv

Bib

liote

] at

10:

43 0

6 N

ovem

ber

2013

Taking these results in consideration, we have proposed that the assayof modified LDL isolated from PEG-precipitated IC is the only accurateassay that has been developed so far. This approach has been used in anumber of clinical studies that have consistently shown that high levels ofmodified forms of LDL in IC indicate increased risk for the developmentof diabetic complications, including cardiovascular, renal, or ophthalmicdisease.[24–28] In a more recent study, we studied a cohort of patients fromthe VADT study, on whom we measured the concentrations of both differ-ent LDL modifications (oxLDL, MDA-LDL, and AGE-LDL), as well as ofthe IgG and IgM antibodies reacting with oxLDL (Figure 6).

In the VADT study, the concentrations of the different modifications ofLDL in isolated IC were divided into quartiles, and patients with concentra-tions of MDA in isolated IC in the highest quartile at baseline had a signifi-cantly higher hazard ratio for acute myocardial infarction [HR¼ 2.44 (95%CI: 1.03, 5.77)] than those in the lowest quartile. No significant correlationsbetween endpoints and the concentrations of oxLDL or AGE-LDL in ICwere observed in this cohort, although the levels of oxLDL and AGE-LDLin IC were markedly higher than those measured in patients with type 1 dia-betes. These results have raised the interesting question as to what extentdoes the predominant modification of the LDL molecules involved in ICformation contribute to phagocytic cell activation, and to what extent this

FIGURE 5 Comparison of the concentrations of AGE-LDL and oxLDL in the precipitate and super-natant resulting from incubation of the sera obtained from 15 individuals with 3.5% PEG 8000. Thesupernatants were assayed directly for IgG and IgM antibodies, while the precipitates were fractionatedin a sepharose protein G column to dissociate modified LDL from the corresponding IgG antibodies.The first eluate, containing modified LDL but devoid of IgG antibodies, was obtained with 0.02Msodium phosphate buffer containing 0.5M NaCl, pH 7.3. A second eluate, obtained with 0.1Mglycine-HCL, pH 2.7, was used to measure IgG antibodies.

70 G. Virella et al.

Dow

nloa

ded

by [

Mos

kow

Sta

te U

niv

Bib

liote

] at

10:

43 0

6 N

ovem

ber

2013

contribution can lead to different outcomes, such as cell proliferation insome cases and apoptosis in others.

The distribution of IgG and IgM antibodies to oxLDL in isolated circu-lating IC shows a marked predominance of antibodies of thepro-inflammatory IgG isotype. Only in very exceptional cases (less than1.5%) do the concentrations of IgM antibodies equal or exceed those ofantibodies of the IgG isotype. Thus, the postulated protective role of IgMantibodies to modified LDL,[38–40] which is largely based on animal studies,is likely to be a very exceptional event in humans.

In conclusion, our methodology has allowed us to generate data thatstrongly support the pathogenic role of modified forms of LDL and ofthe immune complexes formed with autoantibodies specific for the corre-sponding LDL modifications. Measuring modified LDL in isolated IC hasemerged as a risk factor for cardiovascular disease with high predictivevalue. However, the methodology involved is complex and needs to be sim-plified for adoption in diagnostic laboratories. The lower affinity of thebinding between modified LDL and the corresponding autoantibodiesrelative to the binding affinity between IgG and protein G could be thebasis for such simplification. After precipitation, modified LDL-IC couldbe incubated with beads of any other surface coated with immobilized pro-tein G, and by careful choice of eluting conditions, it should be possible toseparate a fraction containing unblocked modified LDL by simple centrifu-gation, for example, and use it for the assay of different LDL modifications.Alternatively, we are also investigating the possibility of taking advantage ofthe much higher affinity of rabbit polyclonal antibodies to native and

FIGURE 6 Comparative distribution of IgG (panel A) and IgM (panel B) oxLDL antibodies in immunecomplexes isolated from 937 patients with type 2 diabetes by precipitation with 3.5% PEG 8000. Theresuspended precipitates were fractionated by affinity chromatography in sepharose-protein G. IgMwas measured in the fraction eluted with 0.02M sodium phosphate buffer containing 0.5M NaCl, pH7.3; IgG was measured in the second eluate, obtained with 0.1M glycine-HCL, pH 2.7.

Immunoassay of Human LDL 71

Dow

nloa

ded

by [

Mos

kow

Sta

te U

niv

Bib

liote

] at

10:

43 0

6 N

ovem

ber

2013

modified LDL compared to that of the human modified LDLauto-antibodies (1.02� 10�8 vs. 9.34� 10�11mol=L, respectively[21]) todevise a protocol that would allow the capture of modified LDL, the elutionof human antibodies, and the assay of modified LDL that should remainassociated with the rabbit-specific antibody without need for an affinitychromatography step.

ACKNOWLEDGMENTS

The development of the methodology reported in this manuscriptwould not have been possible without the collaboration of dedicated tech-nical personnel—Francesco Wagner, Brooks Derrick, Matt Rhettt, andVirginia Pate—to whom we are greatly indebted.

REFERENCES

1. Ross, R.; Harker, L. Hyperlipidemia and atherosclerosis. Science 1976, 193, 1094–1100.2. Ross, R. The pathogenesis of atherosclerosis: A perspective for the 1990s. Nature 1993, 362, 801–809.3. Ross, R. Atherosclerosis—an inflammatory disease. N. Engl. J. Med. 1999, 340, 115–126.4. Libby, P.; Theroux, P. Pathophysiology of coronary artery disease. Circulation 2005, 111, 3481–3488.5. Hansson, G. K. Inflammation, atherosclerosis, and coronary artery disease. N. Engl. J. Med. 2005,

352, 1685–1695.6. Drake, T. A.; Hannani, K.; Fei, H.; Lavi, S.; Berliner, J. A. Minimally oxidized low-density lipo-protein

induces tissue factor expression in cultured human endothelial cells. Am. J. Pathol. 1991, 138,601–607.

7. Kusuhara, M.; Chait, A.; Cader, A.; Berk, B. C. Oxidized LDL stimulates mitogen-activated proteinkinases in smooth muscle cells and macrophages. Arterioscler. Thromb. Vasc. Biol. 1997, 17, 141–148.

8. Liao, F.; Berliner, J. A.; Mehrabian, M.; Navab, M.; Demer, L. L.; Lusis, A. J.; Fogelman, A. M. Mini-mally modified low density lipoprotein is biologically active in vivo in mice. J. Clin. Invest. 1991, 87,2253–2257.

9. Quinn, M. T.; Parthasarathy, S.; Fong, L. G.; Steinberg, D. Oxidatively modified low density lipopro-teins: A potential role in recruitment and retention of monocyte=macrophages during atherogen-esis. Proc. Nat. Acad. Sci. USA 1987, 84, 2995–2998.

10. Yla-Herttuala, S. Biochemistry of the arterial wall in developing atherosclerosis. Ann. N. Y. Acad. Sci.1991, 623, 40–59.

11. Palinski, W.; Yla-Herttuala, S.; Rosenfeld, M. E.; Butler, S. W.; Socher, S. A.; Parthasarathy, S.; Curtiss,L. K.; Witztum, J. L. Antisera and monoclonal antibodies specific for epitopes generated duringoxidative modification of low density lipoprotein. Arteriosclerosis 1990, 10, 325–335.

12. Witztum, J. L.; Steinbrecher, U. P.; Kesaniemi, Y. A.; Fisher, M. Autoantibodies to glucosylated pro-teins in the plasma of patients with diabetes mellitus. Proc. Natl. Acad. Sci. USA 1984, 81, 3204–3208.

13. Mironova, M.; Virella, G.; Lopes-Virella, M. F. Isolation and characterization of human antioxidizedLDL autoantibodies. Arterioscler. Thromb. Vasc. Biol. 1996, 16, 222–229.

14. Virella, G.; Thorpe, S.; Alderson, N. L.; Derrick, M. B.; Chassereau, C.; Rhett, J. M.; Lopes-Virella, M.F. Definition of the immunogenic forms of modified human LDL recognized by human autoanti-bodies and by rabbit hyperimmune antibodies. J. Lipid Res. 2004, 45, 1859–1867.

15. Yla-Herttuala, S.; Palinski, W.; Rosenfeld, M. E.; Parthasarathy, S.; Carew, T. E.; Butler, S.; Witztum, J.L.; Steinberg, D. Evidence for the presence of oxidatively modified low density lipoprotein in ather-osclerotic lesions of rabbit and man. J. Clin. Invest. 1989, 84, 1086–1095.

72 G. Virella et al.

Dow

nloa

ded

by [

Mos

kow

Sta

te U

niv

Bib

liote

] at

10:

43 0

6 N

ovem

ber

2013

16. Yla-Herttuala, S.; Palinski, W.; Rosenfeld, M. E.; Steinberg, D.; Witztum, J. L. Lipoproteins in normaland atherosclerotic aorta. Eur. Heart J. 1990, 11, 88–89.

17. Saad, A. F.; Virella, G.; Chassereau, C.; Boackle, R. J.; Lopes-Virella, M. F. Oxidized LDL immunecomplexes activate complement and induce cytokine production by MonoMac 6 cells and humanmacrophages. J. Lipid Res. 2006, 47, 1975–1983.

18. Virella, G.; Atchley, D. H.; Koskinen, S.; Zheng, D.; Lopes-Virella, M. Pro-atherogenic andpro-inflammatory properties of immune complexes prepared with purified human oxLDL anti-bodies and human oxLDL. Clin. Immunol. 2002, 105, 81–92.

19. Abdelsamie, S. A.; Li, Y.; Huang, Y.; Lee, M.-H.; Klein, R. L.; Virella, G.; Lopes-Virella, M. F. OxidizedLDL immune complexes stimulate collagen IV production in mesangial cells via Fc gamma receptorI. Clin. Immunol. 2011, 139, 258–266.

20. Lopes-Virella, M. F.; Virella, G. Clinical significance of the humoral immune response to modifiedLDL. Clin. Immunol. 2010, 134, 55–65.

21. Virella, G.; Lopes-Virella, M. F. Lipoprotein autoantibodies: Measurement and significance. Clin.Diagn. Lab. Immunol. 2003, 10, 499–505.

22. Virella, G.; Lopes-Virella, M. F. Humoral immunity and atherosclerosis. Nature Med. 2003, 9,243–244.

23. Virella, G.; Derrick, M. B.; Pate, V.; Chassereau, C.; Thorpe, S. R.; Lopes-Virella, M. F. Developmentof capture assays for different modifications of human low-density lipoprotein. Clin. Diagn. Lab.Immunol. 2005, 12, 68–75.

24. Lopes-Virella, M. F.; McHenry, M. B.; Lipsitz, S.; Yim, E.; Wilson, P. F.; Lackland, D. T.; Lyons, T.;Jenkins, A. J.; Virella, G. Immune complexes containing modified lipoproteins are related to theprogression of internal carotid intima-media thickness in patients with type 1 diabetes. Atherosclerosis2007, 190, 359–369.

25. Lopes-Virella, M. F.; Baker, N. L.; Hunt, K. J.; Lachin, J.; Nathan, D.; Virella, G. Oxidized LDLimmune complexes and coronary artery calcification in type 1 diabetes. Atherosclerosis 2011, 214,462–467.

26. Lopes-Virella, M. F.; Hunt, K. J.; Baker, N. L.; Lachin, J.; Nathan, D. M.; Virella, G. The levels of oxi-dized LDL and AGE-LDL in circulating immune complexes are strongly associated with increasedlevels of carotid intima-media thickness and its progression in type 1 diabetes. Diabetes 2011, 60,582–589.

27. Lopes-Virella, M. F.; Carter, R. E.; Baker, N. L.; Lachin, J.; Virella, G.; DCCT=EDIC study group.Highlevels of oxidized LDL in circulating immune complexes are associated with increased odds ofdeveloping abnormal albuminuria in Type 1 diabetes. Nephrol. Dial. Transplant. 2012, 27,1416–1423.

28. Lopes-Virella, M. F.; Baker, N.; Hunt, K.; Lyons, T.; A. J.; Virella, G.; DCCT=EDIC study group. Theconcentration of AGE-modified LDL and Oxidized LDL in circulating immune complexes is asso-ciated with progression of retinopathy in type 1 diabetes. Diabetes Care 2012, 35, 1333–1340.

29. Abraira, C.; Duckworth, W.; McCarren, M.; Emanuele, N.; Arca, D.; Reda, D.; Henderson, W. Designof the cooperative study on glycemic control and complications in diabetes mellitus type 2: VeteransAffairs Diabetes Trial. J. Diabetes Complications 2003, 17, 314–322.

30. Duckworth, W.; Abraira, C.; Moritz, T.; Reda, D.; Emanuele, N.; Reaven, P. D.; Zieve, F. J.; Marks, J.;Davis, S. N.; Hayward, R.; Warren, S. R.; Goldman, S.; McCarren, M.; Vitek, M. E.; Henderson, W. G.;Huang, G. D. Glucose control and vascular complications in veterans with type 2 diabetes. N. Engl.J. Med. 2009, 360, 129–139.

31. Digeon, M.; Laver, M.; Riza, J.; Bach, J. F. Detection of circulating immune complexes in human seraby simplified assays with polyethylene glycol. J. Immunol. Meth. 1977, 16, 165–183.

32. Zubler, R. H.; Perin, L. H.; Creighton, W. D.; Lambert, P. H. Use of polyethylene glycol (PEG) to con-centrate immune complexes from serum or plasma samples. Ann. Rheum. Dis. 1977; 36(Suppl.), 23–26.

33. Virella, G.; Kilpatrick, J. M.; Chenais, F.; Fudenberg, H. H. Isolation of soluble immune complexesfrom human serum: Combined use of polyethylene glycol precipitation, gel filtration, and affinitychromatography on protein A-sepharose. Methods Enzymol. 1981, 74, 644–663.

34. Mironova, M.; Virella, G.; Virella-Lowell, I.; Lopes-Virella, M. F. Anti-modified LDL antibodies andLDL-containing immune complexes in IDDM patients and healthy controls. Clin. Immunol. Immuno-pathol. 1997, 85, 73–82.

Immunoassay of Human LDL 73

Dow

nloa

ded

by [

Mos

kow

Sta

te U

niv

Bib

liote

] at

10:

43 0

6 N

ovem

ber

2013

35. Chenais, F.; Virella, G.; Patrick, C. C.; Fudenberg, H. H. Isolation of soluble immune complexes byaffinity chromatography using staphylococcal protein A-sepharose as substrate. J. Immunol. Methods1977, 18, 188–192.

36. Virella, G.; Carter, R. E.; Saad, A.; Crosswell, E. G.; Game, B. A.; Lopes-Virella, M. F. Distribution ofIgM and IgG antibodies to oxidized LDL in immune complexes isolated from patients with type 1diabetes and its relationship with nephropathy. Clin. Immunol. 2008, 127, 394–400.

37. Akerstrom, B.; Bjorck, L. A physicochemical study of protein G, a molecule with unique immuno-globulin G-binding properties. J. Biol. Chem. 1986, 261, 10240–10247.

38. Yamashita, T.; Freigang, S.; Eberle, C.; Pattison, J.; Gupta, S.; Napoli, C.; Palinski, W. Maternalimmunization programs postnatal immune responses and reduces atherosclerosis in offspring. Circ.Res. 2006, 99, e51–64.

39. Su, J.; Georgiades, A.; Wu, R.; Thulin, T.; de Faire, U.; Frostegard, J. Antibodies of IgM subclass tophosphorylcholine and oxidized LDL are protective factors for atherosclerosis in patients withhypertension. Atherosclerosis 2006, 188, 160–166.

40. Faria-Neto, J. R.; Chyu, K. Y.; Li, X.; Dimayuga, P. C.; Ferreira, C.; Yano, J.; Cercek, B.; Shah, P. K.Passive immunization with monoclonal IgM antibodies against phosphorylcholine reduces acceler-ated vein graft atherosclerosis in apolipoprotein E-null mice. Atherosclerosis 2006, 189, 83–90.

74 G. Virella et al.

Dow

nloa

ded

by [

Mos

kow

Sta

te U

niv

Bib

liote

] at

10:

43 0

6 N

ovem

ber

2013