Cross-Reactivity of Anti-CagA Antibodies WithVascular Wall Antigens

Possible Pathogenic Link Between Helicobacter pyloriInfection and Atherosclerosis

Francesco Franceschi, MD; Antonia R. Sepulveda, MD; Antonio Gasbarrini, MD; Paolo Pola, MD;Nicolò Gentiloni Silveri, MD; Giovanni Gasbarrini, MD; David Y. Graham, MD; Robert M. Genta, MD

Background—Helicobacter pylori-CagA positive strains have been shown to be associated with atherosclerosis. However,the pathogenesis is still undetermined. The aim of this study was to determine whether anti-CagA antibodies cross-reactwith antigens of normal and atherosclerotic arteries.

Methods and Results—Eight umbilical cord sections, 14 atherosclerotic artery sections, and 10 gastrointestinal tractsections were examined by immunohistochemistry using polyclonal anti-CagA antibodies. Five atherosclerotic and 3normal artery samples were also lysed in ice-cold lysis buffer containing protease inhibitors and were immunoprecipi-tated using the same antibodies. Anti-CagA antibodies reacted with cytoplasm and nuclei of smooth muscle cells inumbilical cord and atherosclerotic vessel sections, cytoplasm of fibroblasts-like cells in intimal atherosclerotic plaques,and the cell membranes of endothelial cells. Anti-CagA antibodies also specifically immunoprecipitated 2 highmolecular weight antigens of 160 and 180 kDa from both normal and atherosclerotic artery lysates.

Conclusions—Anti-CagA antibodies cross-react with antigens of both normal and atherosclerotic blood vessels. Wespeculate that the binding of anti-CagA antibodies to those antigens in injured arteries could influence the progressionof atherosclerosis in CagA-positive H pylori-infected patients. (Circulation. 2002;106:430-434.)

Key Words: atherosclerosis � infection � antibodies

Atherosclerotic vascular disease, the major cause of mor-tality and morbidity in industrialized countries, is a

chronic multifactorial disorder.1 Ischemic heart disease(IHD), cerebrovascular disorders, and gangrene of the ex-tremities are among the most common and devastatingconsequences of this condition.2 Epidemiological studieshave linked atherosclerosis to hypertension, cigarette smok-ing, diabetes, hyperlipidemia, hypercoagulability, and hyper-homocystinemia.3 Nevertheless, no condition or exposure,either individually or in combination, completely explains theoccurrence and progression of the disease. Other factors arelikely to be involved.4–8

Both humoral and cellular immune reactions participate inatherogenesis.9–12 Fibroblasts and macrophages are an activecomponent of atherosclerotic plaques, where they play acrucial role in accumulating lipids. Immunoglobulin comple-ment deposits and activated CD4� T cells have also beenassociated with atherosclerotic plaques.9–10 These latter find-ings have marked the beginning of a new era in the field ofatherosclerosis research. The emphasis has shifted from the

traditional view of atherogenesis as an essentially mechanicalprocess (the “response to injury hypothesis”)9 to the “inflam-matory hypothesis.” According to this view, both the initia-tion and progression of atherosclerosis, including its throm-botic and thromboembolic complications, may be linked to aninflammatory process taking place within the arterial wall andthe atherosclerotic lesion. The activation of inflammatorycells could promote plaque irregularity and rupture, elicitingischemic events.9–11 Although the mechanisms of activationof inflammatory cells within atherosclerotic lesions have notbeen elucidated, there is evidence for a role of autoimmunityas well as infections,13–15 particularly those caused by Cyto-megalovirus, herpes simplex, Chlamydia pneumoniae, and,most recently, Helicobacter pylori.

Since 1994, more than 30 studies have been publishedreporting an association between H pylori infection andischaemic heart disease (IHD), the most common clinicalmanifestation of atherosclerosis.16,17 The majority of thesestudies were seroepidemiological, on the basis of the detec-tion of antibodies against H pylori in patients with IHD. H

Received March 26, 2002; revision received May 8, 2002; accepted May 9, 2002.From the Departments of Pathology (F.F., R.M.G.) and Medicine (D.Y.G., R.M.G.), VA Medical Center and Baylor College of Medicine, Houston,

Tex; Department of Pathology (A.R.S.), University of Pittsburgh, Pittsburgh, Pa; and Department of Internal Medicine (F.F., A.G., P.P., N.G.S., G.G.),Catholic University of Rome, Italy.

Guest editor for this article was Joseph B. Muhlestein, MD, LDS Hospital, Salt Lake City, Utah.Correspondence to Robert M. Genta, MD, Pathology –113, VAMC, 2002 Holcombe Blvd, Houston, TX 77030. E-mail rmgenta@bcm.tmc.edu© 2002 American Heart Association, Inc.

Circulation is available at http://www.circulationaha.org DOI: 10.1161/01.CIR.0000024100.90140.19

430

by guest on June 5, 2018http://circ.ahajournals.org/

Dow

nloaded from

pylori has been proposed to increase the levels of lipids andfibrinogen in response to a low-grade persistent inflammatorystimulation and to raise anti-heat shock protein 65 (hsp65)antibodies, a known marker of atherosclerosis.18–21 Somestrains of H pylori express the cytotoxin-associated gene-A(cag A), which encodes for a hydrophilic, surface-exposedprotein named CagA.22 The size of the protein may varybetween 116 and 140 kDa in different strains. H pylori strainsexpressing CagA and carrying the cag pathogenicity islandinduce an inflammatory response in the gastric mucosagreater than that induced by strains lacking the pathogenicityisland.22 CagA-positive strains have been found to be signif-icantly more prevalent among patients with IHD than incontrols,23 suggesting that the enhanced immunologic re-sponse evoked by such strains may influence plaque forma-tion and/or activation. Despite the plausibility of this hypoth-esis, however, the exact mechanisms by which CagA-positivestrains would contribute to the progression of atherosclerosishave not been elucidated.

Autoimmune responses have been shown to participate inboth the initiation and progression of atherosclerosis.12–15 Wehypothesized that antigenic mimicry between H pylori anti-gens and structural elements of blood vessels could partici-pate in this process. Therefore, we designed a study aimed atdetermining whether antibodies against CagA cross-reactedwith antigens of normal and atherosclerotic arteries, provid-ing a possible pathogenic link with atherosclerosis.

MethodsVascular, Gastrointestinal, and Other TissuesFresh segments of atherosclerotic arteries (5 samples) and normalarteries (3 samples) were collected from 8 patients, either duringautopsies performed shortly after death or from surgically amputatedextremities. Formalin-fixed paraffin-embedded sections of umbilicalcord (8 samples) and atherosclerotic tibial arteries (14 samples) wereretrieved from the files of the Baylor College of Medicine Depart-ment of Pathology. Paraffin-embedded sections of stomach infectedwith CagA-positive H pylori strains (6 samples), colon (4 samples),liver (10 samples), and skin (4 samples) were also used as positiveand negative controls for the immunohistochemical studies. All theprocedures followed were in accordance with institutionalguidelines.

ImmunohistochemistryImmunostaining was performed on 5-�m formalin-fixed paraffin-embedded sections by a standard avidin-biotin detection system(Dako Autostainer, Dako Corporation). Briefly, tissue sections weremounted onto slides, deparaffinized in xylene, and rehydrated ingraded ethanol solutions. Antigen retrieval was done by steaming thesamples for 20 minutes. After washing with PBS, sections wereincubated in 10% goat serum for 20 minutes to reduce non-specificantibody binding. Sections were then incubated with the primarymouse polyclonal anti-CagA antibodies (Acambis Inc, Cambridge,Mass) at a 1:300 dilution. Sections were then incubated withbiotinylated rabbit anti-mouse IgG for 10 minutes, washed 3 timeswith PBS, treated with streptavidin-peroxidase reagent for 10 min-utes, and washed again with PBS. Gastric sections from patients withH pylori infection that were previously shown to express CagAantigens by Western blot analysis were used as the positive control.Negative controls were stained without primary antibody. To con-firm the specificity of the immunoreactivity, all immunostaining wasrepeated after incubation of the primary antibody with purified CagAantigen (Acambis Inc, Cambridge, Mass).

ImmunoprecipitationFive fresh atherosclerotic samples and 3 normal artery samples werelysed in an equal volume of ice-cold PBS buffer containing 1%nonylphenol polyoxyethylene 40 (NP-40), 100 mmol/L Na3VO4,5 mmol/L EGTA, 10 mmol/L Ma4P2O7, 100 mmol/L phenyl-methylsulfonyl fluoride (PMSF), and 2 �g/mL each of leupeptin,pepstatin, and aprotinin. Samples were then incubated on ice for 30minutes, cleared of insoluble debris by centrifugation at 13 000g for20 minutes at 4°C, and diluted with PBS to obtain a final concen-tration of NP-40 of 0.5%. Immunoprecipitation was performed byincubating the lysates overnight at 4°C with the same mouse-raisedpolyclonal anti-CagA antibodies used for immunohistochemistry(Acambis Inc, Cambridge, Mass) or with normal mouse IgG ascontrol (Santa Cruz Biotechnology, Santa Cruz, Calif), followed byincubation with 100 �L protein A/G Agarose (Santa Cruz Biotech-nology) for 1 hour at the same conditions. After 3 washes with PBSbuffer containing protease inhibitors, precipitated proteins wereseparated by SDS-PAGE and revealed by silver staining.

To determine the molecular weights of the CagA present inbacterial strains isolated from patients with gastric cancer, we alsoimmunoprecipitated bacterial lysates with anti-CagA antibodies(Acambis Inc, Cambridge, Mass).

ResultsImmunohistochemistryAnti-CagA antibodies showed immunoreactivity with bothumbilical cord and atherosclerotic artery antigens. Specifi-cally, they recognized nuclei (and to a lesser extent, thecytoplasm) of smooth muscle cells, as well as nuclei andcytoplasm of endothelial cells in all umbilical cord sections(Figure 1). Furthermore, in all atherosclerotic vessel sections,anti-CagA antibodies strongly reacted with smooth-musclecells, the cytoplasm of fibroblasts-like cells in intimal athero-sclerotic plaques, and endothelial cells, including the endo-thelium in re-canalized segments of vessel (Figure 2). Noimmunoreaction was observed after incubation of the arterialsections with the primary antibody previously treated with thepurified CagA antigen (Figure 3A and 3B).

Anti-CagA antibodies intensely stained CagA-positive Hpylori organisms in the stomach (Figure 4A); however the

Figure 1. Anti-CagA antibodies immunoreacted with smoothmuscle cells (SMC) and endothelial cells (EC) in an umbilicalcord artery. Immunoreactivity is seen in both the cytoplasm andnuclei of endothelial cells, and mostly in the nuclei of smoothmuscle cells.

Franceschi et al Helicobacter pylori and Atherosclerosis 431

by guest on June 5, 2018http://circ.ahajournals.org/

Dow

nloaded from

same antibodies did not exhibit immunoreactivity with anygastrointestinal mucosal or submucosal structures, includingsmall vessels and smooth-muscle cells (Figure 4B). Noimmunoreactivity was observed between anti-CagA antibod-ies and small vessels of either liver or skin. This suggests thatCagA cross-reacts only with vascular peptides in large- andmedium-size arteries.

ImmunoprecipitationAnti-CagA antibodies immunoreacted with all artery sam-ples. Specifically, anti-CagA antibodies recognized 2 vascu-lar antigens of 160 K and 180 K from both normal andatherosclerotic artery lysates, as revealed by silver staining(Figure 5). The same antigens were not precipitated bynormal mouse IgG, confirming the specificity of the reaction.Bacterial lysates immunoprecipitated with anti-CagA anti-bodies consistently showed a band at approximately 130 K.

DiscussionAtherosclerosis is the principal cause of death in Westernsocieties, but its pathogenesis remains to be elucidated.24

Many patients develop atherosclerosis in the absence ofconventionally recognized risk factors, a finding which sug-gests that unrecognized mechanisms may also be determinantin the pathogenesis of the disease.4–7 To explain such dis-crepancies, several authors have proposed additional riskfactors for atherosclerosis, including H pylori infection.13–21

H pylori infection has been associated with anti-hsp65 pro-duction21 and with increased fibrinogen and serum lipids,18–20

but inconsistent results have been reported by differentstudies.25,26 Recently, cross-reactivity between rabbit-raisedhyperimmune anti-H pylori serum and antigens from athero-sclerotic carotid arteries has been demonstrated.27 In thisstudy, however, the interpretation of the results was limitedby the absence of inclusion of normal arteries and the lack ofidentification of the bacterial antigens mimicking vascularpeptides.

Figure 3. Umbilical cord arteries. A, Negative control (absenceof primary antibody). Similarly, no reactivity was found after pri-mary antibodies were incubated with purified CagA antigen (B).EC indicates endothelial cells; SMC, smooth muscle cells.

Figure 2. Anti-CagA antibodies reacted with endothelium (EC),smooth muscle cells (SMC), and plaque fibroblasts (FB) in anatherosclerotic artery.

Figure 4. A, Anti-CagA antibodies stained CagA-positive H pylori organisms, visible within the gas-tric pit and in the mucous layer adjacent to thegastric surface. B, Low-power photomicrograph ofthe same gastric biopsy specimen depicted in A.No structures (including small vessels and smooth-muscle cells) showed immunoreactivity whenstained with anti-CagA antibodies. The reactivityseen in inflammatory cells is non-specific; it iscommon to virtually all immunoperoxidase-basedimmunostains and is due to reactivity with second-ary reagents.34

432 Circulation July 23, 2002

by guest on June 5, 2018http://circ.ahajournals.org/

Dow

nloaded from

Our demonstration that anti-CagA antibodies are capableof reacting with both bacterial CagA and proteins present inthe wall of medium- and large-size arteries provides evidenceof molecular mimicry between CagA and vascular antigens.Specifically, we found that anti-CagA antibodies, which havebeen shown to be increased in patients with IHD,23 exhibitedimmunohistochemical cross-reaction with antigens expressedby cells involved in the atherogenic process, such as vascularsmooth muscle cells, fibroblasts-like cells in intimal athero-sclerotic plaques, and cell membrane of endothelial cells. Toevaluate the specificity of the immunohistochemical reactiv-ity, we also performed immunoprecipitation of artery lysatesusing the same anti-CagA antibodies used for immunohisto-chemistry. Our results showed that these antibodies immuno-precipitated 2 vascular antigens with molecular weights ofapproximately 160 K and 180 K, from both normal andatherosclerotic artery lysates, respectively, whereas the sameantigens were not precipitated by normal mouse IgG.

Anti-CagA antibodies precipitated CagA and vascularantigens at different molecular weights; 120 K for CagA and160-180 K for arterial proteins. This finding indicates that theimmunoprecipitated proteins extracted from vessels are struc-tural vascular components different from the CagA protein.Therefore, it seems unlikely that the reactivity detected invessels and their extracts would represent bacterial CagAdeposited within vascular walls.

Recently, some studies reported the visualization of Hpylori genomic material28 and H pylori organisms withinatherosclerotic plaques.29 In the latter study, the detection oforganisms was performed by immunohistochemical methodsusing specific anti-H pylori antibodies.29 The authors specu-lated that H pylori infection might affect atherosclerosisthrough a direct colonization of arterial walls. In our view,several points militate against the plausibility of thishypothesis.

First, the published figures29 do not show specific morpho-logical characteristics that would allow an observer to inde-pendently identify H pylori. Second, there was no evidence,until now, of H pylori’s ability to spread from its gastric nichevia the blood circulation. Third, the presence of genomicmaterial of H pylori within the arterial wall of atheroscleroticarteries could be related to a non-specific deposition ofcirculating DNA fragments, rather than to the presence ofintact H pylori organisms inside plaques. Finally, otherstudies that specifically addressed this issue failed to confirmthese findings.30

Our study does not establish whether anti-CagA antibodieswould recognize the mimicking vascular antigens in normalarteries in vivo. It is tempting, however, to speculate that suchphenomena may happen in injured arteries. Antigens nor-mally made inaccessible by the endothelial integrity couldbecome exposed to circulating antibodies after the initiationof the atherogenic process damages the arterial wall. Anti-CagA antibodies could then bind the exposed vascular anti-gens and further contribute to the activation of inflammatorycells within lesions. Production of cytokines and other in-flammatory mediators by activated macrophages and fibro-blasts31–33 could then lead to the destabilization of atheroscle-rotic plaques, possibly triggering ischemic events.

In conclusion, this study provides experimental evidence ofmolecular mimicry between the CagA antigen and vascularwall peptides. This finding yields biological plausibility to thetheory that H pylori infection may play a role in thepathogenesis of atherosclerosis. The verification of thesehypotheses and the elucidation of the mechanisms involvedwill require further studies, ideally using combined animalmodels of atherosclerosis and H pylori infection.

AcknowledgmentThis work was supported by a Merit Review Grant from theDepartment of Veterans Affairs (Dr Genta).

References1. Murray CJ, Lopez AD. Global mortality, disability, and the contribution

of risk factors: Global Burden of Disease Study. Lancet. 1997;349:1436–1442.

2. Libby P. Changing concepts of atherogenesis. J Int Med. 2000;247:349–358.

3. Ross R. The pathogenesis of atherosclerosis: a perspective for the 1990s.Nature. 1993;362:801–809.

4. Stehbens WE. Epidemiological risk factors of coronary heart disease arenot causal in atherosclerosis. Clin Exp Hypertens. 2000;22:445–453.

5. Heller RF, Chinn S, Pedone HD, et al. How well can we predict coronaryheart disease? Findings in the United Kingdom Heart Disease PreventionProject. BMJ. 1984;288:1409–1411.

6. Wald NJ, Law M, Watt HC, et al. Apolipoproteins and ischaemic heartdisease: implications for screening. Lancet. 1994;343:75–79.

7. Chaturvedi N, McKeigue PM, Marmot MG. Relationship of glucoseintolerance to coronary risk in Afro-Carribeans compared withEuropeans. Diabetologia. 1994;37:765–778.

8. Yao C, Wu Z, Wu Y. The changing pattern of cardiovascular diseases inChina. World Health Stat Q. 1993;46:113–118.

9. Wick G, Perschinka H, Xu Q. Autoimmunity and atherosclerosis. AmHeart J. 1999;138:444–449.

10. Hansson GK. Immune and inflammatory mechanisms in the developmentof atherosclerosis. Br Heart J. 1993;69(suppl 1):S38–S41.

11. De Boer OJ, van der Wal AC, Becker AE. Atherosclerosis, inflammationand infection. J Pathol. 2000;190:237–243.

12. Biasucci LM, Liuzzo G, Buffon A, et al. The variable role of inflam-mation in acute coronary syndromes and in restenosis. Semin IntervCardiol. 1999;4:105–110.

13. Rothwell PM, Villagra R, Gibson R, et al. Evidence of a chronic systemiccause of instability of atherosclerotic plaques. Lancet. 2000;355:19–24.

14. George J, Harats D, Shoenfeld Y. Autoimmunity in atherosclerosis: therole of autoantigens. Clin Rev Allergy Immunol. 2000;18:73–86.

15. George J, Afek A, Gilburd B, et al. Autoimmunity in atherosclerosis:lessons from experimental models. Lupus. 2000;9:223–227.

16. Danesh J, Peto R. Risk factors for coronary heart disease and infectionwith Helicobacter pylori: meta-analysis of 18 studies. BMJ. 1998;316:1130–1132.

17. Kusters JG, Kuipers EJ. Helicobacter and atherosclerosis. Am Heart J.1999;138(5 pt 2):S523–S527.

18. Laurila A, Bloigu A, Nayha S, et al. Association of Helicobacter pyloriinfection with elevated serum lipids. Atherosclerosis. 1999;142:207–210.

Figure 5. Mouse-raised polyclonal anti-CagA antibodies (AntiCagA) specifically recognized 2 high molecular weight antigensof approximately 160 kDa and 180 kDa from both atheroscle-rotic and normal artery lysates. The same antigens were notrecognized by normal mouse IgG (Contr IgG).

Franceschi et al Helicobacter pylori and Atherosclerosis 433

by guest on June 5, 2018http://circ.ahajournals.org/

Dow

nloaded from

19. Patel P, Carrington D, Strachan DP, et al. Fibrinogen: a link betweenchronic infection and ischaemic heart disease. Lancet. 1994;343:1634–1635.

20. Torgano G, Cosentini R, Mandelli C, et al. Treatment of Helicobacterpylori and Chlamydia pneumoniae infections decreases fibrinogen plasmalevel in patients with ischemic heart disease. Circulation. 1999;99:1555–1559.

21. Birnie DH, Holme ER, McKay IC, et al. Association between antibodiesto heat shock protein 65 and coronary atherosclerosis: possiblemechanism of action of Helicobacter pylori and other bacterial infectionsin increasing cardiovascular risk. Eur Heart J. 1998;19:387–394.

22. Figura N, Vindigni C, Presenti L, et al. New acquisitions in Helicobacterpylori characteristics. Ital J Gastroenterol Hepatol. 1998;30(suppl3):S254–S258.

23. Pasceri V, Cammarota G, Patti G, et al. Association of virulent Helico-bacter pylori strains with ischemic heart disease. Circulation. 1998;97:1675–1679.

24. Nicoletti A, Caligiuri G, Hansson GK. Immunomodulation of atheroscle-rosis: myth and reality. J Int Med. 2000;247:397–405.

25. Wald NJ, Law MR, Morris JK, et al. Helicobacter pylori infection andmortality from ischaemic heart disease: negative result from a large,prospective study. BMJ. 1997;315:1199–1201.

26. Rosenstock SJ, Jorgense T, Andersen LP, et al. Association of Heli-cobacter pylori infection with lifestyle, chronic disease, body-indices,

and age at menarche in Danish adults. Scand J Public Health. 2000;28:32– 40.

27. Cammarota G, Figura N, Cianci R, et al. Is there mimicry betweenarteriosclerotic lesions and H pylori antigens? Clin Biochem. 2000;33:419–421.

28. Farsak B, Yildirir A, Akyon Y, et al. Detection of Chlamydia pneumoniaeand Helicobacter pylori DNA in human atherosclerotic plaques by PCR.J Clin Microbiol. 2000;38:4408–4411.

29. Ameriso SF, Fridman EA, Leiguarda LC, et al. Detection of Helicobacterpylori in human carotid atherosclerotic plaques. Stroke. 2001;32:385–391.

30. Radke PW, Merkelbach-Bruse S, Messmer BJ, et al. Infectious agents incoronary lesions obtained by endatherectomy: pattern of distribution,coinfection, and clinical findings. Coron Artery Dis. 2001;12:1–6.

31. Cassatella MA, McDonald PP. Interleukin-15 and its impact on neu-trophil function. Curr Opin Hematol. 2000;7:174–177.

32. Heim A, Zeuke S, Weiss S, et al. Transient induction of cytokine pro-duction in human myocardial fibroblasts by coxackievirus B3. Circ Res.2000;86:753–759.

33. Meyer JE, Berner I, Teran LM, et al. RANTES production by cytokine-stimulated nasal fibroblasts: its inhibition by glucocorticoids. Int ArchAllergy Immunol. 1998;117:60–67.

34. Van der Loos CM, Houtkamp MA, de Boer OJ, et al. Immunohisto-chemical detection of interferon-�: fake or fact? J Histochem Cytochem.2001;49:699–710.

434 Circulation July 23, 2002

by guest on June 5, 2018http://circ.ahajournals.org/

Dow

nloaded from

Silveri, Giovanni Gasbarrini, David Y. Graham and Robert M. GentaFrancesco Franceschi, Antonia R. Sepulveda, Antonio Gasbarrini, Paolo Pola, Nicolò Gentiloni

Pathogenic Link Between Helicobacter pylori Infection and AtherosclerosisCross-Reactivity of Anti-CagA Antibodies With Vascular Wall Antigens: Possible

Print ISSN: 0009-7322. Online ISSN: 1524-4539 Copyright © 2002 American Heart Association, Inc. All rights reserved.

is published by the American Heart Association, 7272 Greenville Avenue, Dallas, TX 75231Circulation doi: 10.1161/01.CIR.0000024100.90140.19

2002;106:430-434; originally published online July 8, 2002;Circulation. 

http://circ.ahajournals.org/content/106/4/430World Wide Web at:

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

  http://circ.ahajournals.org//subscriptions/

is online at: Circulation 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 this process is available in the

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

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

by guest on June 5, 2018http://circ.ahajournals.org/

Dow

nloaded from