Accelerated Intimal Thickening in CarotidArteries of Balloon-Injured Rats AfterImmunization Against Heat Shock Protein 70Jacob George, MD,* Shai Greenberg, MSC,* Iris Barshack, MD,† Madhavir Singh, PHD,‡Sara Pri-Chen, PHD,* Shlomo Laniado, MD,* Gad Keren, MD*Tel Aviv and Tel Hashomer, Israel and Braunschweig, Germany

OBJECTIVES The goal of this study was to test the hypothesis that induction of an immune response to heatshock protein (Hsp) 70 would increase intimal thickening in a rat carotid-injury model.

BACKGROUND Restenosis resulting from intimal thickening poses a major limitation to the long-term successof coronary angioplasty. Several studies have proposed that infectious agents increaserestenosis. Heat shock proteins are highly conserved structures, produced by all cells inresponse to nonspecific forms of stress. Infectious agents are known to contain Hsp70, whichis markedly immunogenic and can elicit a strong immune response.

METHODS To investigate whether Hsp70 immunity can affect neointimal thickening, we immunized ratswith either Hsp70 (n � 11), bovine serum albumin ([BSA] n � 9) or with a control adjuvant(n � 10). Three weeks later, rats were boosted using the same regimen to achieve a sustainedimmune response to Hsp70 after which carotid injury was applied to all animals.

RESULTS Arterial injury was associated with upregulation of Hsp70, 3, 7 and 14 days after inductionof the injury as evidenced by Western blotting and immunohistochemistry. Intimal area andintimal/medial ratio was significantly increased in Hsp70-immunized rats in comparison withBSA or control-injected rats.

CONCLUSIONS Our results imply that upregulation of Hsp70 in balloon-injured arteries can serve as a targetfor anti-Hsp70 immune response, thereby facilitating enhanced intimal thickening. Theseobservations may provide a possible mechanism that explains the accelerated intimalthickening that has been associated with the occurrence of infectious pathogens. (J Am CollCardiol 2001;38:1564–9) © 2001 by the American College of Cardiology

The long-term effectiveness of balloon angioplasty is stilllargely hampered by the occurrence of late lumen loss afterintimal thickening (1,2). Although several experimentalstrategies have provided some success in reducing intimalthickening in animals (3,4), clinical trials in humans per-formed thus far failed to accomplish significant improve-ment. The reason for the limited effectiveness of appropriateprevention lies in the yet unresolved processes that mediateformation of the neoimtima and remodelling by the cellularparticipants (i.e., monocyte/macrophages, vascular smoothmuscle cells [SMC], T lymphocytes) and their products.

In recent years, considerable data have accumulated tosupport a role for infectious agents in the progression ofatherosclerosis and restenosis (5–7). The data derive fromseveral lines of study: 1) evidence for the presence of theinfectious agents within lesioned arteries (5); 2) the presenceof antibodies to the pathogen in the sera of the patients(5,8); 3) experimental induction of lesions in infectedanimals (5,9); and 4) in vitro assays demonstrating theability of the organism to influence the properties of thecellular components of the arterial wall (5,10,11).

As both atherosclerosis and restenosis basically display aform of response of the arterial wall to injury (12), an

inflammatory phenotype set by an infectious agent is likelyto influence the milieu within which atherosclerosis andrestenosis occur. However, why certain infections appear tohave a more detrimental effect on the response to injury thanothers is still controversial and requires further study.

Heat shock proteins (Hsp) are produced by all cells inresponse to several forms of stress (thermal, mechanical,irradiation, etc.) (13,14). Heat shock protein 70 membershave been known to provide an essential function inpreventing aggregation and assisting refolding of misfoldedproteins. However, they also have a principal role undernormal conditions including assisted folding of newly trans-lated proteins, guiding translocating proteins across or-ganellar membranes, disassembling oligomeric structuresand facilitating proteolytic degradation of unstable proteins(13,14). It has recently been noticed that members of theHsp family contained within infectious organisms are anti-genic and immunodominant, thus eliciting strong cellularand humoral immune responses (15,16). This is specificallytrue for the Hsp70, which appears to trigger a particularlybrisk immune response (17).

Several studies have pointed to an important role for Hsp60 kd and 70 kd in the initiation of experimental autoim-mune diseases (18–21). In a set of studies performed in theprevious decade, it has been shown that Hsp60/65 couldserve as an antigenic target of an immune-mediated re-sponse that serves to enhance atherosclerosis in experimen-tal animals and in humans (22–25). The hypothesis claims

From the *Department of Cardiology and the Cardiovascular Research Laboratory,Tel Aviv Medical Center, Tel Aviv, Israel; †Institute of Pathology, Sheba MedicalCenter, Tel Hashomer, Israel; and ‡GBF-Braunschweig, Braunschweig, Germany.

Manuscript received December 19, 2000; revised manuscript received June 14,2001, accepted July 12, 2001.

Journal of the American College of Cardiology Vol. 38, No. 5, 2001© 2001 by the American College of Cardiology ISSN 0735-1097/01/$20.00Published by Elsevier Science Inc. PII S0735-1097(01)01579-0

that immunity against infectious agents cross-reacts withself-expressed Hsp60, thereby promoting lesion formation.Studies have also revealed expression of Hsp70 in athero-sclerotic plaque (26,27), and similar mechanisms may beoperative for Hsp70.

As Hsp70 is the predominant Hsp that is upregulated inresponse to stress on one hand (13,14) and that it is highlyimmunogenic (17) on the other hand, we tested the hypoth-esis that induction of an immune response to Hsp70 in aninjured rat carotid artery would influence neointima forma-tion. Such an effect could account for an important mech-anism mediating infection-related acceleration of restenosis.

METHODS

Animals. Eight-week-old male Wistar rats (aged two tothree months) were purchased from Tel Aviv University andraised at the local animal house. All animals were fed anormal diet and allowed free access to water and food.Experimental design. Rats were immunized with either30 �g Hsp70 mycobacterial Hsp70 (GBF, Braunschweig,Germany; n � 11), bovine serum albumin ([BSA] n � 9)emulsified in incomplete Freund’s adjuvant (IFA) or withIFA alone (n � 10; “control” group). Rats were boostedafter three weeks with the same antigen, and balloon injuryof the left carotid artery was performed two weeks later.Rat carotid injury model. Animals were anesthetized byintraperitoneal injection of Ketamin (80 mg/kg) and Xyla-zine (5 mg/kg). Endothelial denudation and vascular injurywere performed in the left common carotid artery, asdescribed (28).Histologic assessment of intimal lesions. Serial crosssections (5 �m thick) were used throughout the entirelength of the carotid artery for histologic analysis (average offive per animal). All samples were routinely stained withhematoxylin & eosin or Masson-Trichrome stains.Quantification of intimal lesions in sections of carotidarteries. Five equally spaced cross sections were used in allrats to quantify intimal lesions. Using image analysis soft-ware (Prof. I. Hammel, Department of Pathology, Tel AvivUniversity), total cross-sectional medial area was measuredbetween the external and internal elastic laminae. Totalcross-sectional intimal area was measured between theendothelial cell monolayer and the internal elastic lamina.

Detection of anti-Hsp70 antibodies. Recombinant myco-bacterial Hsp70 (1 �g/ml) in phosphate-buffered saline([PBS] pH 7.2) was coated onto flat bottom 96-wellenzyme-linked immunosorbent assay (ELISA) plates(Nunc, Maxisorp, Denmark) by overnight incubation at 4°Cas previously described, and sera (dilution of 1:100) wereprobed using ELISA as previously described (25).Western blot for detection of Hsp70 in rat carotidarteries. Fresh balloon injured and normal carotid arterieswere homogenized by homogenizer (polytron) in PBScontaining 2 mmol of a protease inhibitor (i.e., phenyl-methylsulfonylfluoride; Sigma, Rehovot, Israel). Superna-tants of the respective injured and of nonmanipulatedarteries were applied on 10% acrylamide SDS-gel underreduced conditions, transferred to nitrocellulose paper andprobed with an anti-Hsp70 antibody. As a control marker,we used recombinant Hsp70.Detection of Hsp70 by immunhistochemistry. Immuno-histochemical staining for Hsp70 was performed on 5�m-thick frozen sections of the rat carotid arteries employ-ing polyclonal mouse anti-Hsp70 antibody.Elution and characterization of bound immunoglobulinG from rat carotid arteries. In a separate experiment, rats(n � 6) were immunized with Hsp70, and arterial injurywas performed as described in the preceding text. Tissuesamples were obtained and homogenized by a Polytronhomogenizer in Tris-buffered saline (TBS) supplementedwith protease inhibitors ([PI] 0.02 �M, pH, 7.4). Thehomogenate was washed three times in TBS � PI bycentrifugation. A minimal volume (0.3 to 0.5 ml) of 0.1 MGlycine-HCl buffer was (pH: 2.8 to 3.5) added to the pelletby vortexing. Subsequently, the tube was centrifuged, andthe supernatant was collected and neutralized with 1 M ofTris (pH 7.4). The supernatant was later dialyzed againstTBS overnight at 4°C. A bicinchoninic acid kit (Pierce,Rockford, Illinois) was used for protein determination.Cell culture and measurement of SMC proliferation.Smooth muscle cells were obtained from the carotid arteriesof Wistar rats by use of the collagenase and elastasedigestion method. Proliferation in the presence of immu-noglobulin G (IgG) anti-Hsp70 or control rat antibodies(10 �g/ml and 100 �g/ml) was assessed by [3H]-thymidineincorporation into DNA was measured as described previ-ously (29).Statistical analysis. Differences between groups were com-pared using a one-way analysis of variance test followed byFisher protected least significant difference. P �0.05 wasaccepted as statistically significant. Values represent mean �SD.

RESULTS

Kinetics of Hsp70 antibody production. Our previousexperience obtained from preliminary studies indicated,similar to others (17), that Hsp70 was highly immunogenicand elicited a pronounced immune response as manifested

Abbreviations and AcronymsBSA � bovine serum albuminELISA � enzyme-linked immunosorbent assayHsp � heat shock proteinIFA � incomplete Freund’s adjuvantIgG � immunoglobulin GPBS � phosphate-buffered salinePI � protease inhibitorSMC � smooth muscle cellTBS � Tris-buffered saline

1565JACC Vol. 38, No. 5, 2001 George et al.November 1, 2001:1564–9 Hsp70 and Intimal Thickening

by a strong IgG anti-Hsp70 antibody response within 9 to10 days of the initial immunization (Fig. 1). Hyperimmunesera from mycoabacterial Hsp70-immunized rats displayedcross-reactivity with human Hsp70 by ELISA (67% of thebinding to mycobacterial Hsp70, data not shown).Detection of expressed Hsp70 in rat arteries by Westernblot. By employing a polyclonal mouse anti-Hsp70 anti-body, we observed that constitutive Hsp70 expression wasevident in relatively low levels in normal arteries, as has beenshown by others (29). Three days after injury, Hsp70 wasmore abundantly expressed in the injured arteries as com-pared with the noninjured arteries (Fig. 2). Interestingly,

injured arteries obtained seven days and 14 days afterballoon injury still maintained higher expression levels ofHsp70 in comparison with the nonlesioned arteries fromthe same animals.Detection of Hsp70 by immunohistochemistry. By usinga high affinity antibody against Hsp70, we observed thatnoninjured arteries displayed mild Hsp70 expression thatwas principally distributed within the SMC in the medialregions (Fig. 3). However, not all normal carotid arteriesappeared to express this protein. In the balloon-injuredarteries, Hsp70 was abundantly present at the subendothe-lial regions and within SMC in the media.The extent of intimal thickening in rat arteries. Theimmunization regimen did not influence animal weight,which was similar in all groups (280 � 25 g for the Hsp70group, 286 � 24 g for the BSA group and 277 � 28 g forthe control group). Intimal area in the rats immunizedagainst Hsp70 was considerably larger (mean � SD: 0.18 �0.03 mm2) in comparison with the BSA (0.081 � 0.025mm2) control-immunized rats (0.075 � 0.032 mm2;p � 0.001) (Figs. 4A and 5). Medial area did not differsignificantly between the Hsp70-immunized, BSA andcontrol-immunized rats (Fig. 4B). The intimal/medial ratiowas also considerably higher in the animals immunized withHsp70 (2.2 � 0.33) as compared with their BSA (1.0 �0.12; p � 0.001) control-immunized littermates (1.05 �0.14; p � 0.001).IgG deposition in rat carotid arteries. Elution of IgGfrom rat arteries revealed that anti-Hsp70 IgG was presentin the arteries from both injured and noninjured rats. Thisfinding is not surprising, as constitutive Hsp70 is known toexist and has also been recently described by other authors(30). However, binding of eluted IgG anti-Hsp70 antibod-ies from injured carotid arteries (mean optical density �SEM: 0.98 � 0.066) was approximately twice as high asthose eluted from normal arteries of Hsp70-immunized rats(0.445 � 0.05; p � 0.0001) (Fig. 6). Immunoglobulin G

Figure 1. Kinetics of anti-heat shock protein (Hsp) 70 antibody produc-tion. Rats immunized with Hsp70, bovine serum albumin (BSA) or controlwere bled before (t � 0), 10 days after primary immunization (t � 1), at thetime of injury (t � 2) or at sacrifice (t � 3). Anti-Hsp70 antibody levelswere determined by enzyme-linked immunosorbent assay. Values representmean � SEM of OD from all animals in each group. Upward-pointingsolid triangle � BSA; downward-pointing solid triangle � control; solidcircle � Hsp70.

Figure 2. Time dependence of heat shock protein (Hsp) 70 expression ininjured rat carotid arteries. Injured or noninjured carotid arteries wereharvested from animals at different time points before and after ballooninjury. Heat shock protein 70 expression was determined by Western blotcombined with enhanced chemiluminescence, employing polyclonal mouseanti-Hsp70 (A). Representative samples taken 3 and 14 days after arterialinjury are shown. Densitometric analysis of reactivity with Hsp70 ofinjured arteries (average of three vessel per bar) from panel A (B).

Figure 3. Immunohistochemical detection of heat shock protein (Hsp) 70in rat carotid arteries. Frozen sections of injured (A) and noninjured (B)carotid arteries were frozen and embedded in OCT. Immunohistochemicaldetection of expressed Hsp70 was carried out employing mouse anti-Hsp70 polyclonal antibody as a primary antibody. The arrow indicatesneointima-media border.

1566 George et al. JACC Vol. 38, No. 5, 2001Hsp70 and Intimal Thickening November 1, 2001:1564–9

antibodies recovered from Hsp70-immunized arteries didnot appear to react with other irrelevant proteins such asBSA (OD: 0.016 for the injured rats, Hsp70 immunizedrats and an OD of 0.01 for the noninjured immunized rats).No Hsp70 reactivity was evident in the small IgG amountsrecovered from IFA-injected rats after carotid injury (datanot shown).SMC proliferation assays. No differences were evidentwith carotid artery SMC from all three groups with regardto their basal proliferation evident by thymidine uptake.

Incubation of SMC with IgG Hsp70 (100 �g/ml) resultedin an increase of 58 � 12% in proliferation in comparisonwith incubation with control IgG. No effect of SMCthymidine incorporation was evident by incubation with 10�g/ml of anti-Hsp70 antibodies.

DISCUSSION

The following study tested the hypothesis that the elicita-tion of an immune response to Hsp70 would influenceneointimal formation in the rat carotid artery. We havefound that rats initially immunized with recombinantHsp70 developed a brisk and sustained immune responsetowards the antigen. The strong reaction was evident by theproduction of IgG anti-Hsp70 antibodies starting from day8 after the primary immunization. When compared withcontrols, injured carotid arteries from Hsp70-immunizedrats exhibited a marked increase of neointimal area and amore pronounced increase in the intimal/medial ratio.Infections, Hsp and restenosis. Evidence has recentlybeen put forward to incriminate infectious agents in theinitiation and progression of atherosclerosis (5–7). As such,seroepidemiologic studies have pointed toward Herpes vi-ruses and Chlamydia pneumoniae as candidate pathogens.

Figure 4. Intimal thickening in rats immunized with heat shock protein(Hsp) 70. Rats were immunized and boosted with Hsp70, bovine serumalbumin (BSA) or phosphate-buffered saline (PBS) (control). Two weeksafter boost, balloon injury was applied to carotid arteries of both groups.Morphometric evaluation of intimal (A) and medial (B) area wereperformed using PhotoShop software. The ratio of intima to medial areawas also calculated (C). Values represent mean � SD of all animals in eachgroup. *p � 0.001 as compared with control.

Figure 5. Representative injured carotid arteries immunized with heatshock protein (Hsp) 70 or control. Representative hematoxylin & eosinstaining of carotid arteries from Hsp70-immunized rats (A) as comparedwith control-injected rats (B) two weeks after balloon injury. Originalmagnification �25.

1567JACC Vol. 38, No. 5, 2001 George et al.November 1, 2001:1564–9 Hsp70 and Intimal Thickening

Parallel animal and in vitro studies strengthened the notionthat infectious agents can alter the properties of the arterialwall favoring increased lesion formation (5). In this context,it is worthwhile to investigate the role of Hsp in influencingthe response to injury in the arterial wall. In a series ofstudies, it has been shown that experimental induction ofarteriosclerosis could follow immunization with Hsp65 inrabbits (22–24). Lesions from rabbits have been shown toexpress mammalian Hsp60 and to contain rich infiltrates ofT lymphocytes (23). These data were confirmed in mice feda high fat diet and immunized against Hsp65 (25).

In humans, an association was noticed between anti-Hsp65 antibodies and sonographically confirmed athero-sclerotic plaques (24). These results were substantiated byBirnie et al. (30) showing that anti-Hsp 65 titers correlatedwith the severity and extent of coronary atherosclerosis.Very recently, Zhu et al. (31) demonstrated that a signifi-cant association existed between levels of antihuman Hsp60antibody levels and the presence and severity of coronaryartery disease. These studies reinforce the association be-tween humoral immunity to infectious agent-related Hsp65(cross-reactive with human Hsp60) and carotid atheroscle-rosis.Hsp expression in restenotic carotid arteries. The ratio-nale for favoring an immunization with Hsp70 over Hsp60is that it is the preferential Hsp induced during mechanicalstress (32). Indeed, we have found that Hsp70 was signifi-cantly expressed in the injured tissue as compared with thenonmanipulated arteries as evidenced by the Western blotstudies and by immunohistochemistry. Interestingly, Hsp70was expressed in the injured arteries 3, 7 and 14 days afterballoon inflation in comparison with the normal vessels. Weassume that the mechanical injury precipitated by theballoon injury elicited an ongoing low-grade inflammatory

response with consequent local production of cytokine/chemokines and other mediators that may serve to con-stantly stress the SMCs and the leukocytes, thereby main-taining Hsp70 expression.Immunity to Hsp70 and the association with intimalthickening. We have also observed in this study thatinjured arteries of rats immunized with Hsp70 containedsignificantly larger Hsp70-reactive IgG in comparison withnonmanipulated arteries (from similarly immunized rats).Thus, a conceivable explanation for the larger deposition ofHsp70 IgG could be that injured tissues produced andexpressed higher levels of endogenous Hsp70 that attractedcirculating anti-Hsp70 antibodies. These antibodies mayhave functional effects on the endothelial cells as has beenshown for anti-Hsp65 antibodies (33).An additional possible mechanism for the acceleration ofneointimal formation in the cellular immune response.As Hsp70 is a powerful immunogen, it is possible thatanti-Hsp70 T-cells were generated that localized to theareas of preferential Hsp70 expression (i.e., the arterialinjury domains) where ligation of their receptor could havetriggered a local production of cytokines that could promotesmooth muscle migration and leukocyte chemoattraction.Conclusions and possible implications of the study. Inconclusion, we have observed that immunization of ratswith a recombinant Hsp70 led to a pronounced increase inneointimal formation in a balloon-injury model. By analogy,we assume that pathogens (harboring Hsp70) can elicit,upon infecting the host, an anti-Hsp70 immune response.When arterial injury is induced, self-Hsp70 is upregulatedthat eventually serves to direct trafficking of either anti-Hsp70 antibodies or T-cells to the lesioned area. Theinteraction of self-overexpressed Hsp70 with the cross-reactive anti-Hsp70 IgG or lymphocytes may act to facili-tate events that result in SMC migration and enhancedintimal thickening. Additionally, IgG anti-Hsp70 antibod-ies may also influence SMC proliferation as shown in thecell culture assay, and this property may also explain amechanism by which these antibodies could increase neo-intimal hyperplasia.

Our findings suggest that anti-Hsp70 antibody levels mayprove as a marker for increased risk of restenosis in patientsundergoing percutaneous transluminal coronary angio-plasty. Studies to address this assumption are ongoing.Furthermore, generation of an anti-Hsp70 immune re-sponse may serve to explain a mechanism by which infec-tious pathogens alter the response to injury in the arterialwall.

Reprint requests and correspondence: Dr. Gad Keren, Depart-ment of Cardiology and the Cardiovascular Research Laboratory,Ichilov Hospital, Elias Sourasky, Tel Aviv Medical Center, 6Weizman Street, Tel Aviv, Israel. E-mail: kereng@tasmc.health.gov.il.

Figure 6. Heat shock protein (Hsp) 70 reactivity of immunoglobulin Gfractions eluted from injured and noninjured carotid arteries of Hsp70immunized rats. Immunoglobulin G eluted from carotid arteries of injured(striped bar) and noninjured (black bar) Hsp70-immunized animals werecharacterized by assessing their binding to solid phase coated Hsp70 or anirrelevant antigen (bovine serum albumin). Values represent mean � SEMof three tissue samples in each group.

1568 George et al. JACC Vol. 38, No. 5, 2001Hsp70 and Intimal Thickening November 1, 2001:1564–9

REFERENCES

1. Holmes DR, Vilestra RE, Smith HC, et al. Restenosis after percuta-neous transluminal coronary angioplasty (PTCA): a report from thePTCA registry from the National, Heart, Lung, and Blood Institute.Am J Cardiol 1984;53:77C–81C.

2. Serruys PW, Luijten HE, Beatt KJ, et al. Incidence of restenosis aftersuccessful coronary angioplasty: a time-related phenomenon: a quan-titative angiographic study in 342 consecutive patients at 1, 2, 3 and 4months. Circulation 1988;77:361–71.

3. Ferns GA, Raines EW, Sprugel KH, et al. Inhibition of intimalsmooth muscle cell accumulation after angioplasty by an antibody toPDGF. Science 1991;253:1129–32.

4. Powell JS, Clozel JP, Muller RK, et al. Inhibitors of angiotensin-converting enzymes prevent myointimal proliferation after vascularinjury. Science 1989;245:186–8.

5. Libby P, Egan D, Scarlatos S. Roles of infectious agents in athero-sclerosis and restenosis: an assessment of the evidence and need forfuture research. Circulation 1997;96:4095–103.

6. Cook PJ, Lip GYH. Infectious agents and atherosclerotic vasculardisease. Q J Med 1996;89:727–35.

7. Capron L. Chlamydia in coronary plaques-hidden culprits or harmlesshobo? Nat Med 1996;8:856–7.

8. Zhou YF, Leon MB, Wclawiw MA, et al. Association between priorcytomegalovirus infection and the risk of restenosis after coronaryatherectomy. N Engl J Med 1996;335:624–30.

9. Hajjar DP, Fabricant CG, Minick CR, et al. Virus induced athero-sclerosis: herpes virus infection alters aortic cholesterol metabolism andaccumulation. Am J Pathol 1986;122:62–70.

10. Visser MR, Tracy PB, Vercellotti G, et al. Enhanced thrombingeneration and platelet binding on herpes simplex virus-infectedendothelium. Proc Natl Acad USA 1988;85:8227–30.

11. Gaydos CA, Summrsgill JT, Shney NN, et al. Replication of Chla-mydia pneumonia in vitro in human macrophages, endothelial cells, andaortic artery smooth muscle cells. Infect Immun 1996;64:1614–20.

12. Ross R. The pathogenesis of atherosclerosis: a perspective for the1990s. Nature 1993;362:801–9.

13. Lindquist S, Craig EA. The heat shock proteins. Annu Rev Genet1988;22:631.

14. Bukau B, Horwich AL. The Hsp70 and Hsp60 chaperone machines.Cell 1998;92:351–66.

15. Young DB, Mehlert A, Smith DF. Stress proteins and infectiousdisease. In: Morimoto RI, Tissiers A, Georgoupolus C, editors. StressProteins in Biology and Medicine. Cold Spring Harbor, NY: ColdSpring Harbor Press, 1990:131–65.

16. Kaufmann SHE, Schoel B. Heat shock protein antigens in immunityagainst infection and self. In: Morimoto RI, Tissiers A, GeorgoupolusC, editors. Stress Proteins in Biology and Medicine. Cold SpringHarbor, NY: Cold Spring Harbor Press, 1994:495–531.

17. Bonorin C, Nardi N, Zhang X, et al. Characteristics of the strongantibody response to mycobacterial Hsp70: a primary, T-cell depen-dent IgG response with no evidence of natural priming or �� T cellinvolvement. J Immunol 1998;161:5220–16.

18. Pearson CM, Wood FD. Studies of polyarthritis and other lesions

induced in rats by injection of mycobacterial adjuvant. I. Generalclinical and pathologic characteristics and some modifying factors.Arthritis Rheum 1959;2:440–59.

19. Elias D, Markovits D, Reshef T, et al. Induction and therapy in anonobese diabetic (NOD/Lt) mouse by a 65-kDa heat shock protein.Proc Natl Acad Sci U S A 1990;87:1576–80.

20. Salvetti M, Butinelli C, Ristori G, et al. T lymphocyte reactivity to themycobacterial 65 and 70 kDa heat shock proteins in multiple sclerosis.J Autoimmun 1992;5:691.

21. Minota S, Cameron B, Welch WJ, et al. Autoantibodies to theconstitutive 73kd member of the Hsp70 family of heat shock proteinsin systemic lupus erythematosus. J Exp Med 1988;162:1475.

22. Xu Q, Dietrich H, Steiner HJ, et al. Induction of arteriosclerosis innormocholesterolemic rabbits by immunization with heat shock pro-tein 65. Arterioscl Thromb 1992;12:789–99.

23. Xu Q, Kleindienst R, Waitz H, et al. Increased expression of heatshock protein 65 coincides with a population of infiltrating T lym-phocytes in atherosclerotic lesions of rabbits specifically responding toheat shock protein 65. J Clin Invest 1993;91:2693–702.

24. Xu Q, Willeit J, Marosi M, et al. Association of serum antibodies toheat shock protein 65 with carotid atherosclerosis. Lancet 1993;341:255–9.

25. George J, Shoenfeld Y, Afek A, et al. Enhanced fatty streak formationin C57BL/6J mice by immunization with heat shock protein 65.Arterioscl Thromb Vasc Biol 1999;19:505–10.

26. Johnson AD, Berberian PA, Tytell M, et al. Differential distribution of70-kd heat shock protein in atherosclerosis: its potential role in arterialSMC survival. Arterioscl Thromb Vasc Biol 1995;15:27–36.

27. Berberian PA, Meyers W, Tytell M, et al. Immunohistochemicallocalization of heat shock protein 70 in normal appearing andatherosclerotic specimens of human arteries. Am J Pathol 1990;136:71–80.

28. Clowes AW, Reidy MA, Clowes MM. Kinetic of cellular proliferationafter arterial injury. Lab Invest 1983;49:327–34.

29. Sudhir K, Wilson E, Chatterjee K, Ives HE. Mechanical strain andcollagen potentiate mitogenic activity of angiotensin II in rat vascularsmooth muscle cells. J Clin Invest 1993;92:3003–7.

30. Birnie DH, Holme ER, McKay IC, Hood S, McColl KE, Hillis WS.Association between antibodies to heat shock protein 65 and coronaryatherosclerosis: possible mechanism of action of Helicobacter pylori andother bacterial infections in increasing cardiovascular risk. Eur Heart J1998;19:387–94.

31. Zhu J, Quyyumi AA, Rott D, et al. Antibodies to human heat-shockprotein 60 are associated with the presence and severity of coronaryartery disease: evidence for an autoimmune component of atherogen-esis. Circulation 2001;103:1071–5.

32. Neschis DG, Safford SD, Raghunath PN, et al. Thermal precondi-tioning before arterial balloon injury: limitation of injury and sustainedreduction of intimal thickening. Arterioscl Thromb Vasc Biol 1998;18:120–6.

33. Schett G, Xu Q, Amberger A, et al. Antibodies against heat shockprotein 60 mediate endothelial cytotoxicity. J Clin Invest 1995;96:2569–77.

1569JACC Vol. 38, No. 5, 2001 George et al.November 1, 2001:1564–9 Hsp70 and Intimal Thickening