Xenotransplant Cardiac Chimera: ImmuneTolerance of Adult Stem CellsTakayuki Saito, MD, PhD, Jin-Qiang Kuang, MD, Bindu Bittira, MD,Abdulaziz Al-Khaldi, MD, and Ray C.-J. Chiu, MD, PhDDivision of Cardiac Surgery, McGill University Health Center, Montreal, Quebec, Canada, and Division of CardiovascularSurgery, Nagoya City University Medical School, Nagoya, Japan

Background. Bone marrow stromal cells have beenshown to engraft into xenogeneic fetal recipients. In viewof the potential clinical utility as an alternative source forcellular and gene therapies, we studied the fate of xeno-geneic marrow stromal cells after their systemic trans-plantation into fully immunocompetent adult recipientswithout immunosuppression.

Methods. Bone marrow stromal cells were isolatedfrom C57B1/6 mice and retrovirally transduced with LacZreporter gene for cell labeling. We then injected 6 � 106

labeled cells into immunocompetent adult Lewis rats.One week later, the recipient animals underwent coro-nary artery ligation and were sacrificed at various timepoints ranging from 1 day to 12 weeks after ligation.Hearts, blood, and bone marrow samples were collectedfor histologic and immunohistochemical studies.

Results. Labeled mice cells engrafted into the bonemarrow cavities of the recipient rats for at least 13 weeks

after transplantation without any immunosuppression.On the other hand, circulating mice cells were positiveonly for the animals with 1-day-old myocardial infarc-tion. At various time points, numerous mice cells couldbe found in the infarcted myocardium that were not seenbefore coronary ligation. Some of these cells subse-quently showed positive staining for cardiomyocyte spe-cific proteins, while other labeled cells participated inangiogenesis in the infarcted area.

Conclusions. The marrow stromal cells are adult stemcells with unique immunologic tolerance allowing theirengraftment into a xenogeneic environment, while pre-serving their ability to be recruited to an injured myo-cardium by way of the bloodstream and to undergodifferentiation to form a stable cardiac chimera.

(Ann Thorac Surg 2002;74:19–24)© 2002 by The Society of Thoracic Surgeons

Bone marrow stromal cells (MSCs) are adult stem cellscapable of differentiating into cells of both mesen-

chymal and nonmesenchymal lineages [1– 6]. Whentransfused intravenously, MSCs have been reported toengraft into the bone marrow compartment of allogeneicrecipients without significant toxicity [7]. Liechty andcolleagues [8] recently reported that human MSCs couldengraft into the fetus of sheep after transplantationintraperitoneally, and persisted in the host organs for aslong as 13 months without immunosuppression. Thusalthough precise mechanisms are still unknown, MSCsseem to have unique immunologic characteristics thatallow their persistence in a xenogeneic fetal environ-ment. However, whether or not these xenogeneic MSCscan engraft into fully immunocompetent adult animalsafter systemic administration and whether or not thesecells, if they survive, still possess their abilities to migrateand to differentiate have not been demonstrated. Inthis study, we intravenously injected mice MSCs into ratsand found their engraftment in the bone marrow of therecipients. Furthermore, we confirmed that these cells

were capable of being recruited to an injured myocar-dium and undergo differentiation into several pheno-types, thus participating in the repair process.

Material and Methods

AnimalsFemale C57Bl/6 mice and male Lewis rats were used inthis study as the donors and recipients, respectively. Allanimals received humane care in compliance with the“Guide for the Care and Use of Laboratory Animals”prepared by the Institute of Laboratory Animal Re-sources, National Research Council, and published bythe National Academy Press, revised 1996, and the“Guide to the Care and Use of Experimental Animals” ofthe Canadian Council on Animal Care.

Experimental DesignFigure 1 shows an experimental design of this study.Isolated mice MSCs (3 � 106) were injected into thepenile vein of the recipients for 2 consecutive days. Oneweek after the second injection, either left coronaryartery ligation (MSC group) or sham operation (controlgroup) was performed on the recipients. The animalswere then sacrificed at various intervals after the proce-dure ranging from 1 day to 12 weeks.

Accepted for publication March 5, 2002.

Address reprint requests to Dr Chiu, Division of Cardiac Surgery, TheMontreal General Hospital, MUHC, 1650 Cedar Ave, Suite C9-169,Montreal, Quebec H3G 1A4, Canada; e-mail: rchiu@po-box.mcgill.ca.

© 2002 by The Society of Thoracic Surgeons 0003-4975/02/$22.00Published by Elsevier Science Inc PII S0003-4975(02)03591-934

Marrow Stromal Cell Isolation and Labeling WithLacZBone marrow cells were harvested from female C57Bl/6mice by flushing the femurs and tibias with Dulbecco’sModified Eagle’s Medium supplemented with 10% fetalbovine serum and 50 U/mL penicillin-streptomycin ac-cording to the method of Wakitani and associates [4].Whole marrow was plated in tissue culture dishes, and 5to 7 days later, the nonadherent hematopoietic cells werediscarded and the adherent bone MSCs were cultured at37°C with 5% CO2. The LacZ reporter gene was trans-fected into MSCs with pMFG-LacZ retrovirus-mediatedgene transfection as described previously [9]. The result-ing LacZ� MSCs were expanded for 4 weeks beforetransplantation.

Marrow Stromal Cells Staining With 5-bromo-4-chloro-3-indoyl-�-d-galactoside for Detection of �-galactosidase ActivityCells were plated in 35-mm dishes and were fixed in 1%glutaraldehyde for 5 minutes at room temperature, thenthey were washed with phosphate-buffered saline. Stain-ing solution at pH of 7.8 to 8.0 [10], which contains1 mg/mL 5-bromo-4-chloro-3-indoyl-�-d-galactoside (X-gal), 1 mmol/L ethyleneglycol-bis(�-aminoethyl-ether)-N,N�-tetraacetic acid, 5 mmol/L K3Fe(CN)6, 5 mmol/LK4Fe(CN)6O � 3 H2O, 2 mmol/L magnesium chloride, and0.01% sodium deoxycholate [9], was added. Then cellswere incubated at 37°C and protected from light for 16hours.

Transplantation of Marrow Stromal Cells and Ligationof the Left Coronary ArteryMale Lewis rats (225 to 275 g) were used in this study asrecipient animals. The animals were anesthetized withisoflurane (MTC Pharmaceuticals, Cambridge, Ontario,Canada), intubated, and mechanically ventilated at 80breaths/min. Mice MSCs (3 � 106 suspended in 150 �L ofDulbecco’s Modified Eagle’s Medium) were injected intothe penile vein and were reinjected 24 hours later in thesame manner. Total number of MSCs injected was 6 �106 per animal. One week after the second injection, theanimals were again anesthetized as mentioned and un-derwent either coronary artery ligation or sham opera-tion (left thoracotomy only) as previously described [9].

Animals were divided into two groups: the MSCs groupreceived MSCs intravenously followed by coronary ar-tery ligation (n � 22); the control group received MSCsintravenously followed by sham operation (n � 5). Mor-tality of the coronary artery ligation and sham procedurewas 31.8% and 0%, respectively. Surviving animals(MSCs group, n � 15; control group, n � 5) weresacrificed at 1 day and 2, 4, 8, and 12 weeks afteroperation; thus 3 MSC and 1 control hearts were exam-ined at each time point.

Tissue Processing and Staining for �-galactosidaseActivityOn sacrifice of the recipient animals, blood, bone mar-row, and hearts were collected. The blood samples werediluted sevenfold with Dulbecco’s Modified Eagle’s Me-dium containing 10% fetal bovine serum, plated in35-mm dishes, and cultured for 2 weeks. Culture mediumwas changed twice a week, and most hematopoietic cellswere discarded during this procedure. Bone MSCs wereisolated from recipients’ femurs and tibias and thencultured for 1 week. Adherent cells derived from both theblood samples and bone marrow specimens were stainedwith X-gal staining solution as mentioned above. Thehearts were rinsed with phosphate-buffered saline andperfusion fixed in 2% paraformaldehyde in phosphate-buffered saline. The staining for �-galactosidase activitywas performed as described above, but with the additionof 0.02% Nonidet P-40 and 0.01% deoxycholate to thestaining solution [9]. After X-gal staining, the hearts werecut longitudinally and embedded in paraffin.

Histology and ImmunohistochemistryHeart sections 5 �m in thickness were processed foreither hematoxylin and eosin staining or immunohisto-chemical staining. Immunohistochemical staining wasperformed for anti-� smooth muscle actin (Sigma Labo-ratories, St. Louis, MO), troponin I-C (Santa-Cruz Bio-technology Inc, Santa Cruz, CA), and sarcomeric myosinheavy chain molecules with MF20 as described previ-ously [6, 9].

Results

Transfection Efficiency and Intravenous Injection ofMice Marrow Stromal CellsTo trace the fate of mice MSCs after transplantation, welabeled them in vitro with a retrovirus carrying the LacZgene. 5-Bromo-4-chloro-3-indoyl-�-d-galactoside stain-ing of these cells revealed nearly 100% of MSCs ex-pressed �-galactosidase activity (Fig 2A). We trans-planted 6 � 106 labeled mice MSCs intravenously intoLewis rats without immunosuppression. There was nei-ther transplant-related mortalities nor morbidities asso-ciated with immunorejection. Bone marrow specimens ofthe recipient rats were collected at various intervalsranging from 1 to 13 weeks after transplantation. Labeledmice MSCs could be identified in the rat bone marrowspecimens of both MSCs and control groups at all timesstudied (Fig 2B).

Fig 1. Experimental design. (LCA � left coronary artery; MSCs �marrow stromal cells.)

20 SAITO ET AL Ann Thorac SurgXENOGENEIC MSCs REGENERATE HOST MYOCARDIUM 2002;74:19–24

Migration to an Injured MyocardiumGross examination of the hearts revealed selectivelylocalized blue discoloration of X-gal stain in the area ofinfarcted myocardium, whereas adjacent noninfarctedmyocardium remained unstained (Fig 3). Histologic ex-amination of serial cross sections confirmed the bluecolor seen on the gross heart specimens was indeedowing to the presence of labeled mice MSCs (Fig 4). Suchcells were identified in the myocardium of all recipientrats. In the hearts harvested 1 day after coronary arteryligation, most labeled MSCs were found in the perivas-cular zone of noninfarcted myocardium (Figs 4C and 4D).However, in the hearts obtained after 2 weeks, mostlabeled MSCs were seen in the infarcted myocardium(Fig 4B). To confirm that these MSCs migrated from bonemarrow to the heart by way of the bloodstream, wecollected blood samples just before coronary artery liga-tion and 1 day and 2, 4, 8, and 12 weeks after coronaryartery ligation. Mice MSCs could be detected only in theblood collected 1 day after ligation (Fig 2C).

In the sham-operated rats that received mouse MSCsintravenously but without coronary artery ligations, X-gal stains of the gross and microscopic specimens wereall negative, with no evidence of labeled cells in theirmyocardia.

Differentiation of Mice Marrow Stromal Cells in anInfarcted MyocardiumIn the infarct scar area, some MSCs were seen in thefibrous layer. Morphologically, these MSCs had a myo-fibroblast-like appearance and may have contributed toscar formation (Fig 4E). Near the infarcted myocardium,there were areas of neoangiogenesis as could be expectedafter myocardial infarction. Some of this neovasculaturecontained labeled MSCs that expressed �-smooth muscleactin in their cytoplasm, and some were integrated into

Fig 2. Histochemical staining for �-galactosidase activity of micemarrow stromal cells in cultures. The transfected marrow stromalcells showed positive staining for �-galactosidase activity (blue col-or). (A) Culture-expanded mice marrow stromal cells before implan-tation. (B) Bone marrow specimen of the recipient rat to which micemarrow stromal cells were transplanted intravenously 13 weeks be-fore. LacZ� mice marrow stromal cells (arrow) were identified,whereas host marrow stromal cells were unstained. (C) Blood sam-ple obtained at 1 day after ligating the left coronary artery in rats towhich mice marrow stromal cells were transplanted intravenously 1week before the ligation. LacZ� mice marrow stromal cells (arrows)were also identified, whereas host cells were unstained. (Originalmagnification [A through C], �100.)

Fig 3. Histochemical staining for �-galactosidase activity of micemarrow stromal cells in the gross heart. Gross heart specimen of therecipient rat was harvested at 12 weeks after coronary artery liga-tion. Mice marrow stromal cells were transplanted intravenously 1week before the ligation. After fixation, the heart was stained for�-galactosidase activity of mice marrow stromal cells. Bluish dis-coloration can be seen on the infarcted myocardium, whereas non-infarcted myocardium remains unstained. (A) Frontal view. (B) Lat-eral view.

21Ann Thorac Surg SAITO ET AL2002;74:19–24 XENOGENEIC MSCs REGENERATE HOST MYOCARDIUM

vessel walls (Fig 4F). Moreover, some MSCs in theinfarcted area stained positively for the sarcomeric myo-sin heavy chain (Fig 4G) and for the cardiomyocyte-specific protein, troponin I-C (Fig 4H). However, up until2 weeks, the MSC-derived cells showed an immatureappearance with a large nucleus to cytoplasm ratio andwere negatively stained for these proteins.

Comment

In this study, we demonstrated that culture expandedxenogeneic MSCs were capable of homing into the bonemarrow of immunocompetent adult recipients after sys-temic transplantation. The apparent lack of immuneresponse indicated by the persistence of mice MSCs inthe rats is intriguing. Although the murine MSCs havebeen reported to express major histocompatibility com-

plex class I on their cell surfaces and lack class II [11], thetolerance of xenogeneic cells in a fully immunocompe-tent recipient is perplexing according to the classic self–non-self immune response paradigm of Burnet [12] andBillingham and colleagues [13]. Nevertheless, our find-ings could be consistent with the prediction of the morerecent “danger” model of immune tolerance proposed byMatzinger [14] and Anderson and Matzinger [15]. Thismodel posits that the antigen-presenting cells require thepresence of “danger signals” as costimulant with theantigen to induce immune activation of T cells. Antigenspresented without concomitant danger signal shouldresult in tolerance. Thus minimal tissue damage associ-ated with injection of mice MSCs in our experiments mayhave allowed for their survival. Recent in vitro and invivo studies also reveal MSCs can exert multiple effectsdirectly on the immune system [11], although precise

Fig 4. Histology and immunohistochemistryof the hearts with myocardial infarction. Micemarrow stromal cells were transplanted intra-venously 1 week before coronary artery liga-tion. (A and B) Heart sample harvested at 8weeks after ligation, and stained for �-galac-tosidase activity and hematoxylin and eosin.Numerous mice marrow stromal cells wereidentified mostly in the infarcted myocardium.(C) Heart sample harvested at 1 day after li-gation. LacZ� mice marrow stromal cells (ar-rows) were seen near vessel (asterisk) in anintact area. (D) Heart sample harvested at 1day after ligation. LacZ� cell (arrow) wasseen in the interstitium of an intact myocar-dium. (E) Heart sample harvested at 12 weeksafter ligation. LacZ� cell (arrow) was mor-phologically integrated into the scar forma-tion. (F) Heart sample harvested at 12 weeksafter ligation, and immunostained for�-smooth muscle actin. LacZ� cell (arrow)was seen in the vascular wall in the infarctedmyocardium. (G) Heart sample harvested at12 weeks after ligation, and immunostainedfor MF20. LacZ� cells (arrows) had immuno-reactivity for sarcomeric myosin heavy-chainmolecules in their cytoplasms. (H) Heart sam-ple harvested at 12 weeks after ligation, andimmunostained for troponin I-C. LacZ� cells(arrows) had immunoreactivity for troponinI-C in their cytoplasms. Original magnifica-tions: (A) �100; (B) �200; (C through F)�400; (G and H) �1000.

22 SAITO ET AL Ann Thorac SurgXENOGENEIC MSCs REGENERATE HOST MYOCARDIUM 2002;74:19–24

mechanisms are still poorly understood. Most intriguingis the finding that MSCs can suppress the function ofmature T cells, either directly or by stimulating suppres-sor T cells, and thus are tolerogenic. They also speculatedthat MSCs recruited to an area of tissue injury maydownregulate inflammatory response to initiate tissuerepair. But even if the MSCs are immune privileged, whydo they not get rejected when they have differentiated?Perhaps by the time they differentiate, which may takeseveral weeks, the danger signals from myocardial in-farction have subsided. Clearly such speculations need tobe confirmed by future studies. Nevertheless, the thera-peutic importance of immunologic tolerance of adultstem cells cannot be overemphasized. This study dem-onstrated the tolerogenic properties of xenogeneic cells,and there is some evidence that allogeneic cells behavesimilarly [16]. The rationale for therapeutic cloning toobtain embryonic stem cells, a highly controversial issue,is to avoid immune rejection of these cells by the recip-ients. The reason for using autologous myoblasts andMSCs, which is currently undergoing early clinical trials,is that autologous cells can avoid immune rejection.However, use of autologous cells faces the disadvantagesof a time lag between cell harvesting and implantation aswell as the logistic complexity to transport them to andfrom a central laboratory for quality control and forisolation and expansion of donor cells. In contrast, use oftolerogenic allogeneic or xenogeneic adult stem cells willavoid the need for cloning, and frozen donor cells can besupplied to be readily available for clinical use withoutdelay.

Another unique characteristic of MSCs is their capacityto migrate to an injured site in the body [17, 18]. To assessthis capability in xenogeneic MSCs, we created myocar-dial infarction in rats by ligating the left coronary artery1 week after mice MSCs had been intravenously trans-planted. Although sham-operated hearts did not containmice MSCs, the hearts with myocardial infarction hadMSCs in the heart tissues. Interestingly, in the heartsharvested 1 day after coronary artery ligation, most of thelabeled MSCs were found in the perivascular zone ofnoninfarcted myocardium, possibly because the coronaryartery that supplied the infarcted area had been perma-nently occluded. However, in the hearts obtained after 2weeks, most of the labeled MSCs were seen in theinfarcted myocardium, suggesting they had migrated tothe injured site. To further confirm our suggestion thatthese MSCs migrated from bone marrow to the heartthrough the bloodstream, we collected blood samplesjust before coronary artery ligation and 1 day and 2, 4, 8,and 12 weeks after coronary artery ligation. Mice MSCscould be detected only in the blood collected 1 day afterligation. Such findings appear to be consistent with thescenario that a signal or signals were released from thedamaged tissue shortly after injury, initiating MSCsrecruitment from the bone marrow and their migration tothe injured area by way of the bloodstream.

To examine the normal physiologic response of MSCsto tissue injury in vivo, we did not artificially purify thesecells to clonal subgroups before their transplantation in

this study. Thus the multiple phenotypes expressed bythese cells at the infarct sites may represent the existenceof progenitor cells for different lineages in this cellpopulation or the multipotential adult stem cells re-sponding to different in situ signals in the microenviron-ment [17, 19].

Although the sample size of hearts examined at eachpoint in this study is modest, it should be noted thatexperimental animals sacrificed at various times aftermyocardial infarction (n � 15) were all positive forimplanted labeled cells without signs of rejection. WhenX-gal stain for �-galactosidase is performed at a pH of 7.8to 8.0 as described above [10], no false-positive stain ofmyocardial scar has been seen, not only in the 5 controlanimals in this study, but also in all sham-operatedcontrol animals in other series of experiments that wehad previously reported [20].

In summary, our findings indicate that (1) xenogeneicMSCs can home in and survive in the bone marrowcavities of hosts without immunosuppression, (2) theyare capable of being recruited to the injured myocardiumthrough the bloodstream, and (3) they can differentiateinto various phenotypes such as vascular smooth musclecells and cardiomyocytes in the infarcted myocardium,resulting in a stable cardiac chimera without the need forimmunosuppressive therapy. How such a chimera mayaffect cardiac function is an interesting and clinicallyrelevant question that deserves further investigation.

We deeply appreciate the technical assistance of Minh Duong,BS.

References

1. Johnstone B, Hering TM, Caplan A, Goldberg VM, Yoo JU. Invitro chondrogenesis of bone marrow-derived mesenchymalprogenitor cells. Exp Cell Res 1998;238:265–72.

2. Ferrari G, Cusella-De Angelis G, Coletta M, et al. Muscleregeneration by bone marrow-derived myogenic progeni-tors. Science 1998;279:1528–30.

3. Peterson B, Bowen W, Patrene K, et al. Bone marrow as apotential source of hepatic oval cells. Science 1999;284:1168–70.

4. Wakitani S, Saito T, Caplan AI. Myogenic cells derived fromrat bone marrow mesenchymal stem cells exposed to 5-aza-cytidine. Muscle Nerve 1995;18:1417–26.

5. Makino S, Fukuda K, Miyoshi S, et al. Cardiomyocytes canbe generated from marrow stromal cells in vitro. J ClinInvest 1999;103:697–705.

6. Wang JS, Shum-Tim D, Galipeau J, Chedrawy E, EliopoulosN, Chiu RCJ. Marrow stromal cells for cellular cardiomyo-plasty: feasibility and potential clinical advantages. J ThoracCardiovasc Surg 2000;120:999–1006.

7. Devine SM, Bartholomew AM, Mahmud N, et al. Mesenchy-mal stem cells are capable of homing to the bone marrow ofnon-human primates following systemic infusion. Exp He-matol 2001;29:244–55.

8. Liechty KW, MacKenzie TC, Shaaban AF, et al. Humanmesenchymal stem cells engraft and demonstrate site-specific differentiation after in utero transplantation insheep. Nat Med 2000;11:1282–6.

9. Wang JS, Shum-Tim D, Chedrawy E, Chiu RCJ. The coro-nary delivery of marrow stromal cells for myocardial regen-eration: pathophysiologic and therapeutic implications.J Thorac Cardiovasc Surg 2001;122:699–705.

23Ann Thorac Surg SAITO ET AL2002;74:19–24 XENOGENEIC MSCs REGENERATE HOST MYOCARDIUM

10. Al-Khaldi A, Lachapelle K, Galipeau J. Endogenous �-galac-tosidase enzyme activity in normal tissues and ischemicmyocardium: a comparison study with prokaryotic �-galac-tosidase reporter enzyme detection. Card Vasc Regeneration2000;1:283–90.

11. McIntosh K, Bartholomew A. Stromal cell modulation of theimmune system. A potential role for mesenchymal stemcells. Graft 2000;3:324–8.

12. Burnet FM. The clonal selection theory of acquired immu-nity. Nashville, TN: Vanderbilt University Press, 1959.

13. Billingham RE, Brent L, Medawar PB. Actively acquiredtolerance of foreign cells. Nature 1953;172:603–6.

14. Matzinger P. An innate sense of danger. Semin Immunol1998;10:399–415.

15. Anderson CC, Matzinger P. Danger: the view from thebottom of the cliff. Semin Immunol 2000;12:231–8.

16. Caparrelli DJ, Cattaneo S, Shake JG, et al. Cellular cardio-myoplasty with allogeneic mesenchymal STEM cells resultsin improved cardiac performance in a swine model ofmyocardial infarction. Circulation 2001;104(Suppl):II-599.

17. Chen J, Li Y, Wang L, Zhang Z, et al. Therapeutic benefit ofintravenous administration of bone marrow stromal cellsafter cerebral ischemia in rats. Stroke 2001;32:1005–11.

18. Bittner RE, Schofer C, Weipoltshammer K, et al. Recruitmentof bone-marrow-derived cells by skeletal and cardiac musclein adult dystrophic mdx mice. Anat Embryol 1999;199:391–6.

19. Orlic D, Kajstura J, Chimentl S, et al. Bone marrow cellsregenerate infarcted myocardium. Nature 2001;410:701–5.

20. Bittira B, Kuang J-Q, Piquer S, Shum-Tim D, Al-Khaldi A,Chiu RCJ. The pathophysiological roles of bone marrowstromal cells (MSCs) in myocardial infarction. Circulation2001;104(Suppl)II-523.

INVITED COMMENTARY

In this very interesting paper from Dr Ray Chiu’s group,Saito and colleagues report that intravenously-injectedmouse bone marrow stromal cells successfully engraftedinto the bone marrow of rats in the absence of anyimmunosuppressive therapy. Furthermore, following li-gation of one of the rat’s coronary arteries, mouse cellscould be detected in the infarcted myocardium. In simi-larly treated rats that had not undergone coronary liga-tion, mouse cells could not be identified in the heart. Theconclusion drawn is that the marrow stromal cells wereadult stem cells with a unique tolerance to immunologicinjury. This allowed their engraftment in a xenogeneicenvironment while preserving their ability to be re-cruited to an injured myocardium (via the bloodstream)and to undergo differentiation to form a stable cardiacchimera. These results are certainly remarkable but, aswith all experimental studies, leave many questionsunanswered.

How did the mouse cells escape immunologic injury?The authors invoke the “danger model” hypothesis thatis currently topical, and suggest that the cells might nothave presented “danger signals” to the rat recipient, andthus escaped immune destruction. However, one wouldanticipate that the presence of any foreign cells would beinterpreted by the host as being potentially dangerous.Why were xenogeneic cells chosen for infusion, and notallogeneic or autologous cells? The authors discuss the

logistic disadvantages of the use of autologous cells, butdo not adequately explain why they chose xenogeneicover allogeneic cells. One could hypothesize that xeno-geneic cells are so different that they escape an immuneresponse, but this has not been the case in other mouse-to-rat models. Their failure to be injured by the host,therefore, must presumably be related to the nature ofthe “stem cells” used. Was the number of mouse cellsdetected in the infarcted rat myocardium sufficient tohave a clinically-useful effect on its recovery or on thestrength of its coordinated myocardial contractions? Notruly quantitative data are provided but, from the paper’sexcellent histochemical figures, the answer is “probablynot.” Nevertheless, despite the intriguing and, as yet,unanswered questions, the study provides some encour-agement that we shall one day not only be able to treatmyocardial infarction by a cell therapy approach, but thatwe may also be able to overcome the barriers toxenotransplantation.

David K. C. Cooper, MD

Transplantation Biology Research CenterMassachusetts General HospitalHarvard Medical SchoolBoston, MA 02129

24 SAITO ET AL Ann Thorac SurgXENOGENEIC MSCs REGENERATE HOST MYOCARDIUM 2002;74:19–24

© 2002 by The Society of Thoracic Surgeons 0003-4975/02/$22.00Published by Elsevier Science Inc PII S0003-4975(02)03691-3