relative roles of icam-1 and vcam-1 in the pathogenesis of experimental radiation-induced intestinal...

doi:10.1016/S0360-3016(03)00523-6

BIOLOGY CONTRIBUTION

RELATIVE ROLES OF ICAM-1 AND VCAM-1 IN THE PATHOGENESIS OFEXPERIMENTAL RADIATION-INDUCED INTESTINAL INFLAMMATION

MERITXELL MOLLA, M.D.,*† MERITXELL GIRONELLA, PH.D.,* ROSA MIQUEL, M.D.,‡

VICTORIA TOVAR, PH.D.,§ PABLO ENGEL, M.D.,� ALBERT BIETE, M.D.,† JOSEPM. PIQUE, M.D.,* AND

JULIAN PANES, M.D.*

Departments of *Gastroenterology,†Radiation Oncology, and‡Pathology and§Liver Unit and �Immunology Unit, Hospital Clıńic,Institut de Investigacions Biome`diques August Pi i Sunyer (IDIBAPS), University of Barcelona, Barcelona, Spain

Purpose: Cell adhesion molecules mediate leukocyte recruitment into the irradiated organs; modulation of thisprocess may protect from radiation damage. Our objective was to characterize the requirement for intercellularadhesion molecule 1 (ICAM-1) and vascular cell adhesion molecule 1 (VCAM-1) in intestinal inflammatoryresponse after abdominal irradiation.Methods and Materials: Endothelial ICAM-1 and VCAM-1 expression was determined using radiolabeledantibodies in mice 24 h and 14 days after irradiation with 10 Gy, or sham radiation. Leukocyte–endothelial cellinteractions in intestinal venules were assessed using intravital microscopy, and the function of ICAM-1 andVCAM-1 in this process by using blocking antibodies and ICAM-1�/� mice.Results: The number of adherent leukocytes significantly increased 24 h after irradiation and remained elevatedat 14 days. Treatment with anti–ICAM-1 antibodies and ICAM-1 genetic deficiency significantly reducedleukocyte adhesion 24 h after irradiation. At 14 days after irradiation, both wild-type and ICAM-1�/� mice hadan upregulation of VCAM-1, expression, and VCAM-1 immunoneutralization, but not ICAM-1 immunoneu-tralization, significantly reduced leukocyte adhesion. In ICAM-1�/� mice, regeneration of the intestinal epithe-lium was enhanced relative to wild-type mice.Conclusions: ICAM-1 plays a key role in leukocyte recruitment at early time points after abdominal irradiation,whereas VCAM-1 is the main molecular determinant of leukocyte recruitment at late time points. © 2003Elsevier Inc.

Radiation, Leukocyte, Endothelium, Adhesion molecules.

INTRODUCTION

The acute and late effects of irradiation on the intestine arecritical dose-limiting factors during treatment for pelvic andabdominal tumors. Vascular changes occurring early afterirradiation, such as endothelial cell apoptosis (1) and in-creased leukocyte recruitment in tissue exposed to ionizingradiation (2), are viewed as important events in the patho-genesis of radiation damage. The development of an inflam-matory response is a finely regulated process that involvessequential leukocyte–endothelial cell interactions, namelyleukocyte rolling, activation, firm adhesion, and emigrationinto the surrounding tissue. Different families of cell adhe-sion molecules (CAMs) have been shown to participate inthe process of leukocyte recruitment (3). Intercellular adhe-sion molecule 1 (ICAM-1) and vascular cell adhesion mol-

ecule 1 (VCAM-1) are endothelial CAMs of the immuno-globulin (Ig) superfamily with a critical role in mediatingthe firm adhesion of leukocytes to endothelial cells in var-ious acute and chronic inflammatory diseases (4–6). Amajor determinant of the relative contribution of VCAM-1and ICAM-1 to the recruitment of leukocytes in a givenpathologic condition is the density of expression of theseCAMs on the surface of endothelial cells. Studies performedon monolayers of cultured endothelial cells have shown thatICAM-1 and, to a much lesser extent, VCAM-1 are consti-tutively expressed on the surface of these cells (7, 8).Activation of cultured endothelial cells with various cyto-kines increases ICAM-1 and VCAM-1 expression, but themagnitude and kinetics of expression of these CAMs seemto vary in an organ- and stimuli-specific manner (4, 9).

Reprint requests to: Juliań Pane´s, M.D., Gastroenterology De-partment, Hospital Clıńic, Villarroel 170, 08036, Barcelona, Spain.Tel: (34) 932275418; Fax: (34) 932279387; E-mail: [email protected]

This study was supported by grants from Instituto de SaludCarlos III FIS 00/00574, FIS 01/0099–01, and C03/02. Dr. M.Molla is the recipient of a grant from Ministerio de Sanidad y

Consumo, and Meritxell Gironella is the recipient of a grant fromMinisterio de Educacioń y Cultura, Spain.

None of the authors of this paper have any potential conflict ofinterest with the information presented in this manuscript.

Received Sep 18, 2002, and in revised form Apr 9, 2003.Accepted for publication Apr 14, 2003.

Int. J. Radiation Oncology Biol. Phys., Vol. 57, No. 1, pp. 264–273, 2003Copyright © 2003 Elsevier Inc.

Printed in the USA. All rights reserved0360-3016/03/$–see front matter

264

ICAM-1 expression has been shown to increase in vari-ous organs in response to irradiation, and blockade of thisadhesion molecule diminishes leukocyte infiltration in thehours or days immediately after irradiation, but the role ofthis adhesion molecule at late time points has not beenassessed so far (10, 11). On the other hand, the role ofVCAM-1 in radiation-induced leukocyte recruitment hasyet to be clarified. VCAM-1 is an important modulator oflymphocyte and monocyte trafficking. Heckmann et al. (12)reported that VCAM-1 expression in human dermal micro-vascular endothelial cells is upregulated after irradiation, ina time- and dose-dependent fashion. In contrast, Hallahan etal. (10) found that the expression of VCAM-1 was notincreased by radiation in human umbilical vein endothelialcells during the initial 24 h after irradiation, and Quarmby etal. (13) extended this observation in human dermal micro-vascular endothelial cells up to 72 h with similar results.Changes in expression of VCAM-1 at late time points havebeen evaluated in an experimental study, showing that afterthoracic irradiation, upregulation of VCAM-1 occurs laterthan that of ICAM-1, but VCAM-1 expression was stillincreased 1 week after irradiation (14). However, the func-tional significance of the increase in VCAM-1 expressionhas not been determined so far.

In the present study, we characterized the changes inexpression of VCAM-1 and ICAM-1 in the intestine afterabdominal irradiation at early (24 h) and late (14 days) timepoints. To assess the functional significance of changes inexpression of these CAMs in the process of leukocyterecruitment, we used the intravital microscopy technique tostudy leukocyte–endothelial cell interactions in intestinalvenules in irradiated mice, along with functional blockadeof ICAM-1 or VCAM-1 by means of monoclonal antibodies(mAbs). Using ICAM-1–deficient mice, we evaluatedwhether prolonged abrogation of ICAM-1 function amelio-rates late radiation-induced damage in the intestine.

METHODS AND MATERIALS

Animal model and radiation procedureMale C57BL/6 mice weighing 24–28 g were obtained

from Iffa Credo (Lyon, France). ICAM-1–deficient mice ona C57BL/6 background (15) were obtained from JacksonLaboratory (Bar Harbor, ME). Principles of laboratory an-imal care (EC guideline 86/609/CEE), as well as the guide-lines and procedures for animal experiments from the Gen-eralitat de Catalunya, were followed.

Before irradiation, the mice were anesthetized with s.c.ketamine (Ketolar, Parke-Davis, Morris Plains, NJ), 50 mg/kg. Abdominal irradiation was performed with a linearaccelerator producing 6-MV photons at a focus-to-surfacedistance of 100 cm. Radiation was delivered at 2.0 Gy/min.Tissue-equivalent bolus was placed on both sides of theirradiated tissue to ensure electronic buildup. A total dose of10 Gy was delivered. The mice were studied 24 h or 14 daysafter irradiation. Control mice were treated in an identicalfashion, but were not irradiated (sham irradiation).

Quantification of ICAM-1 and VCAM-1 expressionAt the time of the study, mice were anesthetized by s.c.

ketamine (100 mg/kg) and xylazine (7.5 mg/kg). The rightcarotid artery and right jugular vein were cannulated. En-dothelial ICAM-1 and VCAM-1 expression were deter-mined in wild-type mice receiving sham radiation or at 24 hand 14 days after abdominal irradiation with 10 Gy. Endo-thelial VCAM-1 expression was also determined in ICAM-1–deficient mice in basal conditions and at 24 h and 14 daysafter irradiation with 10 Gy. Each group included 5 or 6mice.

The mAbs used for in vivo assessment of ICAM-1 andVCAM-1 expression were as follows: YN1/1.7.4 (scaled upand purified at Pharmacia/Upjohn Laboratories, Kalamazoo,MI), a rat IgG2b against mouse ICAM-1 (16), and M/K-2(R&D Systems Inc., Minneapolis, MN), a rat IgG1 againstmouse VCAM-1. The control immunoglobulin used wasUPC-10, a nonbinding murine IgG2 (Sigma, Quımica, Ma-drid, Spain) (17). Binding mAbs directed against ICAM-1or VCAM-1 were labeled with 125I, whereas the nonbindingmAb (UPC-10) was labeled with 131I (17) (AmershamIberica, Madrid, Spain). Radioiodination of mAbs was per-formed by the iodogen method (18). Labeled mAbs werestored at 4°C and used within 3 weeks after the labelingprocedure. The specific activity of labeled mAbs was 0.5mCi/mg.

To measure endothelial ICAM-1 expression, a mixture of10 �g of 125I-YN1/1.7.4 and 40 �g unlabeled YN1/1.7.4,and 10 �g of 131I-UPC was given through the jugular veincatheter. To measure endothelial VCAM-1 expression, amixture of 10 �g of 125I-M/K-2 and 20 �g of unlabeledM/K-2 and 10 m �g of 131I-UPC was used. The dose ofanti–ICAM-1 and anti–VCAM-1 has been shown to besaturating under stimulated conditions in previous assays(19). Blood samples were obtained through the carotidartery catheter 5 min after injection of the mAb mixture.Thereafter, the mice were heparinized (1 mg/kg sodiumheparin i.v.) and rapidly exsanguinated. Entire organs werethen harvested and weighed.

125I (binding mAb) and 131I (nonbinding mAb) activitiesin each organ and in 100-�L aliquots of cell-free plasmawere counted in a Cobra II �-counter (Packard; Meridien,Canberra, Australia), with automatic correction for back-ground activity and spillover. The injected activity in eachexperiment was calculated by counting a 5-�L sample ofthe mixture containing the radiolabeled mAbs. The accu-mulated activity of each mAb in an organ was expressed asng of binding antibody per g of tissue. The formula used tocalculate ICAM-1 and VCAM-1 expression has been pre-viously described (9).

Intravital microscopyIntravital microscopy was used to quantify rolling and

adhesion of leukocytes in lamina propria and submucosalintestinal postcapillary venules from control wild-type andICAM-1–deficient mice at 24 h and 14 days after irradiationwith 10 Gy, or 24 h after sham irradiation, as previously

265ICAM-1 and VCAM-1 in radiation-induced inflammation ● M. MOLLA et al.

described (20). A midline abdominal incision was made toallow exteriorization of a section of the small bowel, andwas maintained moisturized by covering the section with abuffer of bicarbonate saline–soaked gauze. An invertedmicroscope (Nikon Diaphot 300, Tokyo, Japan) with a CFFluor 40� objective lens (Nikon) was used. A 3CCD cam-era (DXC-930P, Sony, Tokyo, Japan) mounted on the mi-croscope projected the image onto a color monitor (Trini-tron KX-14CP1, Sony, Tokyo, Japan), and the images werecaptured on videotape for off-line analysis. Leukocyteswere in vivo labeled by i.v. injection of acridine orange (200�g/ml; Sigma) dissolved in sterile saline and filteredthrough a 0.22-�m filter (Millipore, Molsheim, France)(21). Single unbranched venules with diameters rangingbetween 25 and 35 �m were studied. The number of rollingand adherent leukocytes, as well as the rolling velocity, wasmeasured off-line during playback of videotaped images. Aleukocyte was considered to be adherent to the vessel wallif it was stationary for more than 30 s. A rolling leukocytewas defined as one that marginates along the vessel wall andis dissociated from the bulk blood flow. The flux of rollingleukocytes was measured as the number of rolling leuko-cytes passing a certain point of the vessel during 1 min.Leukocyte rolling velocity was calculated as the mean of 10rolling leukocyte velocity measurements and expressed in�m/s. Venular blood flow was calculated from the mean ofthe velocity of 3 free-flowing leukocytes, using the follow-ing empiric relationship: Venular Blood Flow � Free-Flowing Leukocytes/1.6 (9). Venular wall shear rate wascalculated assuming cylindrical geometry and using theNewtonian definition � � 8 (Venular Blood Flow/Diame-ter) (22). In each mouse, 3 to 5 random venules wereexamined, and results were calculated as the mean of eachparameter in all venules examined.

To assess the functional role of ICAM-1 and VCAM-1 inleukocyte–endothelial interactions, groups of wild-typemice (n � 6 per group) were treated with blocking antibod-ies against ICAM-1 (YN1/1.7.4, 2 mg/kg) or VCAM-1(M/K-2, 1.2 mg/kg). Additional groups of ICAM-1–defi-cient mice were treated with anti–VCAM-1 (MK-2, 1.2mg/kg). mAbs were administered 12 h before the intravitalmicroscopy studies, which were carried out at 24 h or 14days after irradiation. The dose of anti–ICAM-1 and anti–VCAM-1 mAbs used in this study was based on previousexperiments (23).

Histologic damageTo evaluate histologic damage, small-bowel tissue sam-

ples from wild-type and ICAM-1–deficient mice were har-vested 24 h and 14 days after irradiation, fixed in 4%formaldehyde, and embedded in paraffin. Sections 5 �mthick were stained with hematoxylin and eosin followingstandard procedures. The number of apoptotic bodies withinthe intestinal crypt, the number of mitotic figures, number ofPaneth cells, and number of proliferating cells (other thanPaneth cells) were counted in 10 random crypts per sample.The arithmetic means per mouse were considered single

values for statistical calculations. Stains were assessed in ablinded fashion by a pathologist (R.M.). Because intravitalmicroscopy experiments do not allow one to define the celltype adhering to endothelial cells, tissue sections were alsoexamined to determine this aspect.

Peripheral blood leukocyte countsBlood samples were obtained from all groups of mice at

the time of the study, and the total leukocyte count wasdetermined in a Sysmex SE-9000 hemocytometer (ToaMedical Electronics Co., Kobe, Japan).

Statistical analysisAll data were analyzed using analysis of variance with

Bonferroni as a post hoc multiple comparison test, andStudent’s paired or unpaired test where appropriate. Allvalues are reported as means � SEM. Statistical signifi-cance was set at p value less than 0.05.

RESULTS

Study 1: Acute inflammatory responseICAM-1 and VCAM-1 expression. In wild-type mice,

ICAM-1 was constitutively expressed in the intestinal vas-cular endothelium, and the expression of ICAM-1 increased3-fold (p � 0.05) after irradiation with 10 Gy at 24 h (Fig.1A). VCAM-1 expression in the intestinal vascular endo-thelium was very low under baseline conditions and was notupregulated in wild-type mice at 24 h after irradiation. Bycontrast, a significant increase in VCAM-1 expression wasnoted in ICAM-1–deficient mice 24 h after irradiation rel-ative to both basal levels and to wild-type irradiated mice(Fig. 1B).

Leukocyte rolling in small-bowel venules. The absoluteflux of rolling leukocytes did not increase 24 h after ab-dominal irradiation in wild-type or in ICAM-1–deficientmice (Fig. 2A). At this time, a marked decrease in thenumber of circulating leukocytes in peripheral blood wasdetected, this reduction being higher in wild-type comparedto ICAM-1–deficient mice (Table 1). Because the lack ofincrement in rolling interactions might be related to thedecreased peripheral leukocyte counts, we calculated theflux of rolling leukocytes relative to peripheral leukocytecounts, observing an increase in this parameter in bothwild-type and ICAM-1–deficient mice relative to sham-irradiated mice (Fig. 2A). Leukocyte rolling velocity wasalso reduced in both groups of mice, but shear rate de-creased only in wild-type irradiated mice (Table 1).

Early radiation-induced leukocyte adhesion: Role ofVCAM-1 and ICAM-1. In wild-type mice, the absolute num-ber of adherent leukocytes significantly increased 24 h afterirradiation (6.5-fold) relative to sham-irradiated mice (Fig.3A). Examination of fixed tissue sections revealed thatadherent cells were predominantly polymorphonuclear leu-kocytes (63%), and the rest (37%) corresponded to lympho-cytes (Fig. 4A); no other cell types were identified. Todetermine whether ICAM-1 or VCAM-1 is required for the

266 I. J. Radiation Oncology ● Biology ● Physics Volume 57, Number 1, 2003

infiltration of leukocytes into the irradiated small bowel,blocking anti–ICAM-1 and anti–VCAM-1 antibodies wereadministered. Treatment with anti–ICAM-1 mAb com-pletely abrogated leukocyte adhesion 24 h after irradiationin wild-type mice. This observation contrasts with the par-tial reduction in leukocyte adhesion observed in ICAM-1–deficient mice. To discern whether this discrepancy mightbe related to residual ICAM-1 expression in knock-out miceor the possible existence of compensatory mechanisms ingenetically deficient mice, we tested the effects of ICAM-1–blocking antibodies in ICAM-1–deficient mice. Leuko-cyte adhesion was not significantly modified by treatmentwith the blocking anti–ICAM-1 antibody, and remainedsignificantly higher than in wild-type mice treated with the

same antibody, suggesting that compensatory mechanismsmay be operating in ICAM-1–deficient mice (Fig. 3A).

In wild-type mice, VCAM-1 immunoneutralization atten-uated, but did not completely prevent, leukocyte adhesion inintestinal venules 24 h after irradiation, whereas in ICAM-1–deficient mice, blockade of VCAM-1 abrogated irradia-tion-induced leukocyte adhesion (Fig. 3A). Immunoneutral-ization of ICAM-1 or VCAM-1 did not influence shear rateor leukocyte rolling at this time point (data not shown).

Fig. 1. Expression of ICAM-1 and VCAM-1 in the intestine underbaseline conditions and 24 h and 14 days after abdominal irradi-ation. (A) ICAM-1 expression was determined in wild-type mice,and (B) VCAM-1 was determined in (solid bars) wild-type and(hatched bars) ICAM-1–deficient mice. Expression is measured asspecific accumulation of a monoclonal antibody per gram of tissue.*p � 0.05 vs. sham-irradiated mice.

Fig. 2. Leukocyte rolling in intestinal venules after irradiation in(solid bars) wild-type and (hatched bars) ICAM-1–deficient mice.(A) The flux of rolling leukocytes was increased only 2 weeks afterradiation. (B) However, when this value was corrected by thenumber of circulating leukocytes, an increase was noted both at 24h and 2 weeks after irradiation. *p � 0.05 vs. sham-irradiatedmice, #p � 0.05 vs. 24 h after irradiation.


Study 2: Chronic inflammatory responseICAM-1 and VCAM-1 expression. ICAM-1 expression in

intestinal vascular endothelium was not upregulated in wild-type mice 14 days after irradiation (Fig. 1A). In contrast,there was a 2-fold increase in VCAM-1 expression at thislate time point. ICAM-1–deficient mice had also an upregu-lation of VCAM-1 2 weeks after irradiation of similarmagnitude to that of wild-type irradiated mice (Fig. 1B).

Leukocyte rolling in small-bowel venules. Compared withsham-irradiated mice, there was a significant increase in theabsolute flux of rolling leukocytes in wild-type and ICAM-1–deficient mice 14 days after irradiation (Fig. 2A). Theincrease relative to the flux or rolling leukocytes detectedearly (24 h) after irradiation is probably because of therecovery of total peripheral leukocyte counts at this timepoint (Table 1). In fact, the flux of rolling leukocytes rela-tive to the peripheral leukocyte counts 14 days after irradi-ation in wild-type mice was reduced relative to valuesobserved 24 h after irradiation, but was still significantlyhigher than that observed in sham-irradiated mice, both inwild-type and ICAM-1–deficient mice (Fig. 2B). Leukocyterolling velocity returned to baseline values by 2 weeks afterirradiation (Table 1), although shear rate was still signifi-cantly reduced in wild-type, but not in ICAM-1–deficientmice, as observed at 24 h after irradiation.

Late leukocyte adhesion: Role of VCAM-1 and ICAM-1.In wild-type mice, the number of adherent leukocytes 14days after irradiation was significantly lower than that ob-served 24 h after irradiation, but still significantly elevatedrelative to sham-irradiated mice (3.3-fold). In ICAM-1–deficient mice, the number of adherent leukocytes 14 daysafter irradiation was similar to that observed 24 h afterirradiation, and significantly higher (2.5-fold) than the num-ber of adherent leukocytes observed in intestinal venules ofsham-irradiated mice (Fig. 3B). Examination of fixed tissuesections revealed that at this time point, most of the adherentcells were lymphocytes (88%), with only a few (12%)

adherent polymorphonuclear lymphocytes detected. Noother adherent cell types were identified (Fig. 4B).

Treatment with an anti–ICAM-1 blocking mAb did notprevent radiation-induced leukocyte adhesion 14 days afterirradiation in wild-type mice, indicating that molecular de-terminants of leukocyte adhesion at late time points differfrom those involved in early leukocyte recruitment (Fig.3B). As expected, the antibody had also no effect in ICAM-1–deficient mice.

The leukocyte adhesion response in intestinal venules 14days after abdominal irradiation was completely abrogatedby treatment with an anti–VCAM-1 antibody both in wild-type and in ICAM-1–deficient mice (Fig. 3B), indicatingthat VCAM-1 plays a central role in leukocyte adhesion ina later phase after irradiation. Immunoneutralization ofICAM-1 or VCAM-1 did not influence shear rate or leuko-cyte rolling at 2 weeks after irradiation (data not shown).

Histologic changesThe acute effects of radiation on the intestinal mucosa

are generally ascribed to inhibition of epithelial mitosisin the crypts that results in a reduction of cell prolifera-tion. Another abnormality that occurs hours to days afterirradiation is the presence of apoptotic bodies. We ob-served that 24 h after irradiation, there was a markedincrease in the absolute number of apoptotic bodies, areduction in the number of proliferating cells, and in thenumber of mitotic cells within the intestinal crypt, both inwild-type and in ICAM-1–deficient mice as comparedwith control mice.

At 14 days after irradiation, the number of apoptoticbodies decreased significantly as compared with changesobserved at 24 h. The number of proliferating cells in-creased, in association with a higher number of mitoticcells. These findings indicate that stem cells in the cryptwere not totally damaged by irradiation, and the intestinalepithelium was recovering. In comparison with wild-type

Table 1. Effects of irradiation on shear rate (s�1), rolling leukocyte velocity (�m/s), andperipheral blood leukocyte counts (peripheral cells · 109/L)

Control

Irradiation

24 h 14 days

Shear rate (s�1)ICAM-1 deficient 414 � 16 355 � 39 354 � 21Wild-type 500 � 63 317 � 26* 307 � 28*

Leukocyte rolling velocity (�m/s)ICAM-1 deficient 64 � 5 44 � 5* 68 � 5Wild-type 62 � 6.8 37 � 3.4* 58 � 3

Peripheral blood leukocyte counts(peripheral cells · 109/L)ICAM-1 deficient 3.8 � 1 1.5 � 0.3* 1.6 � 0.2*Wild-type 2.9 � 0.4 0.5 � 0.02*‡ 1.8 � 0.1*†

* p � 0.05 vs. control.† p � 0.05 vs. 10 Gy/24 h.‡ p � 0.05 vs. ICAM-1–deficient mice.


mice, ICAM-1–deficient mice had a similar number ofapoptotic bodies and Paneth cells, but a significantly highernumber of proliferating cells in the crypts and of mitoticcells 14 days after irradiation, suggesting that a sustainedblockade of leukocyte infiltration into the intestine as aresult of ICAM-1 deficiency enhances healing of the intes-tinal epithelium (Fig. 5).

Study 3: Adhesion molecule expression in extraintestinalorgans.

To assess a possible tissue type dependence of ICAM-1and VCAM-1 expression in response to abdominal irradia-

tion, as has been previously described in response to otherproinflammatory stimuli (24), we determined expression ofthese adhesion molecules in organs included in the radiationfield, such as the kidneys, and in relevant organs outside theradiation field, such as heart and lung. The results aresummarized in Fig. 6. Changes in adhesion molecule ex-pression in the kidney closely paralleled those observed inthe intestine. In wild-type mice, ICAM-1 expression mark-edly increased at 24 h after radiation and had returned tobasal levels 14 days after irradiation. As for kidneyVCAM-1 expression, a significant increase was observed24 h after radiation in only ICAM-1–deficient mice, but notin wild-type mice, whereas VCAM-1 expression 14 daysafter irradiation was significantly increased in both wild-type and ICAM-1–deficient mice. In organs not included inthe radiation field, such as heart and lung, ICAM-1 expres-sion was not significantly modified at any time point, but amarked upregulation of VCAM-1 was observed 14 daysafter irradiation in both wild-type and ICAM-1–deficientmice (Fig. 6).

DISCUSSION

The present study demonstrates that the molecular deter-minants of leukocyte recruitment differ in early (24 h)compared to late (2 weeks) radiation-induced inflammatoryresponse. Whereas ICAM-1 plays a pivotal role in leukocyteadhesion to venular endothelium at early time points,VCAM-1 acts as the key determinant of leukocyte adhesionin more chronic phases of radiation-induced inflammation.

The concept that ICAM-1 is a determinant of leukocyterecruitment at early time points after irradiation was con-firmed by the two approaches used to test ICAM-1 function,namely blockade of ICAM-1 by means of monoclonal an-tibodies and ICAM-1–deficient mice. This concept is inkeeping with previous observations in irradiated lung basedon administration of blocking monoclonal antibodies (10),or use of genetically deficient mice (25). However, upregu-lation of adhesion molecule expression varies widely amongdifferent organs exposed to the same inflammatory stimuli(4, 9), and the role of a particular determinant of leukocyterecruitment in response to an inflammatory stimulus in anorgan cannot be generalized to other organs (26).

We and others (10–12) have previously reported thatICAM-1 expression is upregulated by irradiation, increasesin ICAM-1 expression being evident from 6 h up to 1 weekafter irradiation. The changes in ICAM-1 expression afterirradiation in the intestine documented in the current studyare consistent with a role for ICAM-1 in the early inflam-matory response, because expression of this adhesion mol-ecule was increased at early time points only and returned tobaseline levels by 2 weeks after irradiation. At this late timepoint, blockade of ICAM-1 by monoclonal antibodies hadno effect on leukocyte adhesion. This observation is inkeeping with previous evidence from in vitro and in vivostudies. Gaugler et al. (27) characterized the expression ofICAM-1 in irradiated human umbilical vein endothelial

Fig. 3. Leukocyte adhesion in intestinal venules in (solid bars)wild-type and (hatched bars) ICAM-1–deficient mice. (A) Theincrease in leukocyte adhesion 24 h after irradiation is dependenton ICAM-1 function, as determined by use of blocking monoclo-nal antibodies and ICAM-1–deficient mice, whereas (B) adhesionat 2 weeks after irradiation is dependent on VCAM-1 function. *p� 0.05 vs. sham-irradiated mice. #p � 0.05 vs. ICAM-1–deficientmice.


cells for 10 days, and documented an upregulation ofICAM-1 from Days 1 to 6. In an in vivo study of lungirradiation in mice (14), an increase in ICAM-1 expressionwas observed only in the initial 48 h, whereas expression atlater time points, up to 8 weeks, did not differ from baseline.Although ICAM-1 is an absolute requirement for leukocyteadhesion to intestinal venular endothelium in the initialperiod after irradiation, VCAM-1 has also a contributoryrole, because blockade of this adhesion molecule signifi-cantly reduced leukocyte adhesion, but to a much lowerextent than ICAM-1 immunoblockade.

One intriguing observation in our study is that the reduc-tion in leukocyte adhesion 24 h after irradiation varieddepending on the method used to abrogate ICAM-1 func-tion. Firm leukocyte adhesion was reduced by 97% whenblocking antibodies were used, but it was reduced by only56% when ICAM-1–deficient mice were used. Other stud-ies assessing adhesion molecule function also suggested thatresults may depend on the methods used to block function(19, 28). There is evidence indicating that chronic defi-ciency of some CAMs in knock-out mice results in pertur-bation of basal and cytokine-induced endothelial cell sur-face expression of other CAMs. In particular, it has beenshown that ICAM-1 knock-out mice have an enhanced

VCAM-1 (29) and E-selectin (30) upregulation in responseto endotoxin challenge compared to wild-type mice. Webelieve that in the current study, compensatory mechanismsin ICAM-1–deficient mice may have underestimated thefunction of ICAM-1 during the acute inflammatory process.We observed that under baseline conditions, expression ofVCAM-1 was similar in ICAM-1–deficient and wild-typemice, but 24 h after irradiation, only ICAM-1–deficientmice had an upregulation of VCAM-1, which contributed tosustained adhesion, because treatment with anti–VCAM-1antibodies completely abrogated leukocyte adhesion 24 hafter irradiation in only ICAM-1–deficient mice.

For the first time, we characterized the adhesion molec-ular determinants of late leukocyte recruitment in the irra-diated intestine. We observed that this is a VCAM-1–de-pendent process, in which ICAM-1 seems to play no role,because the increase in leukocyte adhesion is abrogated byVCAM-1 immunoblockade and unaffected by ICAM-1 im-munoblockade. This differential function of VCAM-1 andICAM-1 at late time points is in keeping with the observedkinetics of VCAM-1 expression; in wild-type mice, upregu-lation of this adhesion molecule was present only at latetime points (2 weeks) after irradiation, whereas by this timeICAM-1 expression had returned to baseline values. This

Fig. 4. Representative small-bowel section from an irradiated wild-type mouse (A) 24 h and (B) 14 days after abdominalirradiation. (A) At early time points, adherence of both polymorphonuclear leukocytes (arrowhead) and lymphocytes(arrow) was observed in small venules (v). As expected, no adhesion was observed in arterioles (a). (B) Two weeks afterirradiation, most of the adherent cells were lymphocytes (arrows) with only scattered polymorphonuclear leukocytesadhering to the venular endothelium (arrowhead). Hematoxylin and eosin (200�, A; and 400�, B).


observation is in agreement with previous in vitro studies,showing no increase in VCAM-1 expression after irradia-tion for periods of 1 to 10 days (10, 13, 27). In the currentstudy, we provide evidence also that the pattern of adhesionmolecule expression after irradiation is quite similar indifferent organs included in the radiation field. Thus, asignificant but transient increase in ICAM-1 and a markedbut late upregulation of VCAM-1 were observed in thekidney. By contrast, in organs not included in the radiationfield, such as lung and heart, no significant changes inICAM-1 expression were detected, but a significant increasein VCAM-1 expression was present also 2 weeks afterirradiation. Increases in VCAM-1 expression in distant or-gans associated with the presence of intestinal inflammationhave been previously reported in experimental models ofinflammatory bowel disease (31). In keeping with the results ofthe current study, ICAM-1 expression in extraintestinal organsin these models was not modified. These changes have beenattributed to increased circulating cytokine levels originating inthe inflamed intestine (31). The kinetics of lung VCAM-1expression after abdominal irradiation differ from those ob-served in a model of lung irradiation, in which a significantincrease in expression of VCAM-1 can be detectable as earlyas 4 h after irradiation, reaching a peak at 12 h, with asubsequent progressive reduction in VCAM-1 expression, re-turning to baseline values by Week 8 (14).

The demonstration that VCAM-1 is the key determinantof leukocyte infiltration in the irradiated intestine at latetime points is in keeping with the concept that this adhesionmolecule is of pivotal importance in chronic inflammatoryconditions of the intestine (32) and other organs (33, 34).The reduction in leukocyte adhesion and intestinal lesions at2 weeks in ICAM-1–deficient mice may be related to sus-tained blockade of ICAM-1–dependent leukocyte recruit-ment in the early phase of radiation-induced inflammatorydamage. The increased number of proliferating and mitoticcells observed in ICAM-1–deficient mice 2 weeks afterradiation with a dose of 10 Gy suggests that reduction ofleukocyte recruitment into the intestine as a result of abro-gation of ICAM-1 function ameliorates radiation enteropa-thy by accelerating epithelial barrier restitution (increasednumber of proliferating and mitotic cells). This notion issupported by the findings of a recent study assessing theeffects of ICAM-1 genetic deficiency on radiation-inducedlung injury. The authors observed that in ICAM-1–deficientmice, leukocyte infiltration in lung tissue at 5 weeks afterirradiation was markedly reduced, although not completelyabrogated, and this was followed by a reduction in the inci-dence of respiratory distress and pulmonary fibrosis (35).

Current clinical radiotherapy schedules employ fraction-

Fig. 5. Histologic changes in wild-type (solid bars) and ICAM-1–deficient (hatched bars) mice 24 h and 14 days after abdominalirradiation. The results express the arithmetic means of 10 intes-tinal crypts per mouse, (n � 6 per group). *p � 0.05 vs. sham-irradiated mice, #p � 0.05 vs. 24 h after irradiation.


ated doses to minimize normal tissue damage, instead of asingle dose as used in the present study. Constraints of theexperimental model, such as the need of repeated anesthesiaand manipulation-induced stress, preclude the use of dosefractionation. However, it can be estimated according to thelinear-quadratic approach (31), assuming an a/� ratio for theintestine of 10 � 4 (32), that a single dose of 10 Gy isequivalent to a biologic dose of 19 Gy with a standardfractionation of 2 Gy per day. Therefore, we believe that thefindings of the current study are clinically relevant. Never-

theless, additional studies may be required to confirm thekey role of VCAM-1 in leukocyte adhesion in a fractionatedschedule of irradiation.

From the data presented in this study, we may concludethat strategies aimed at preventing intestinal radiation-in-duced inflammatory response based on adhesion moleculeblockade should target ICAM-1 in the early phase afterirradiation, whereas modulation of the chronic phase ofradiation-induced inflammation should be based on abroga-tion of VCAM-1 function.

REFERENCES

1. Paris F, Fuks Z, Kang A, et al. Endothelial apoptosis as theprimary lesion initiating intestinal radiation damage in mice.Science 2001;293:293–297.

2. Panes J, Anderson DC, Miyasaka M, et al. Role of leukocyte-endothelial cell adhesion in radiation-induced microvascu-lar dysfunction in rats. Gastroenterology 1995;108:1761–1769.

3. Panes J, Granger DN. Leukocyte-endothelial cell interactions:Molecular mechanisms and implications in gastrointestinaldisease. Gastroenterology 1998;114:1066–1090.

4. Henninger DD, Panes J, Eppihimer M, et al. Cytokine-inducedVCAM-1 and ICAM-1 expression in different organs of themouse. J Immunol 1997;158:1825–1832.

5. Norris PG, Barker JN, Allen MH, et al. Adhesion molecule

Fig. 6. Expression of ICAM-1 and VCAM-1 in kidney, lung, and heart under baseline conditions and 24 h and 14 daysafter abdominal irradiation. ICAM-1 expression (left panels) was determined in wild-type mice, and VCAM-1 (rightpanels) was determined in wild-type (solid bars) and ICAM-1–deficient (hatched bars) mice. Expression is measuredas specific accumulation of a monoclonal antibody per gram of tissue. *p � 0.05 vs. sham-irradiated mice.


expression in polymorphic light eruption. J Invest Dermatol1992;99:504–508.

6. Poston RN, Haskard DO, Coucher JR, et al. Expression ofintercellular adhesion molecule-1 in atherosclerotic plaques.Am J Pathol 1992;140:665–673.

7. Henseleit U, Steinbrink K, Sunderkotter C, et al. Expressionof murine VCAM-1 in vitro and in different models of inflam-mation in vivo: Correlation with immigration of monocytes.Exp Dermatol 1994;3:249–256.

8. Myers CL, Wertheimer SJ, Schembri KJ, et al. Induction ofICAM-1 by TNF-alpha, IL-1 beta, and LPS in human endo-thelial cells after downregulation of PKC. Am J Physiol 1992;263:C767–C772.

9. Panes J, Perry MA, Anderson DC, et al. Regional differencesin constitutive and induced ICAM-1 expression in vivo. Am JPhysiol 1995;269:H1955–H1964.

10. Hallahan D, Kuchibhotla J, Wyble C. Cell adhesion moleculesmediate radiation-induced leukocyte adhesion to the vascularendothelium. Cancer Res 1996;56:5150–5155.

11. Molla M, Panes J, Casadevall M, et al. Influence of dose-rateon inflammatory damage and adhesion molecule expressionafter abdominal radiation in the rat. Int J Radiat Oncol BiolPhys 1999;45:1011–1018.

12. Heckmann M, Douwes K, Peter R, et al. Vascular activationof adhesion molecule mRNA and cell surface expression byionizing radiation. Exp Cell Res 1998;238:148–154.

13. Quarmby S, Hunter RD, Kumar S. Irradiation induced expres-sion of CD31, ICAM-1 and VCAM-1 in human microvascularendothelial cells. Anticancer Res 2000;20:3375–3381.

14. Tsujino K, Kodama A, Kanaoka N, et al. Expression ofpulmonary mRNA encoding ICAM-1, VCAM-1, and P-selec-tin following thoracic irradiation in mice. Radiat Med 1999;17:283–287.

15. Sligh JE, Jr., Ballantyne CM, Rich SS, et al. Inflammatory andimmune responses are impaired in mice deficient in intercel-lular adhesion molecule 1. Proc Natl Acad Sci USA 1993;90:8529–8533.

16. Kumasaka T, Quinlan WM, Doyle NA, et al. Role of theintercellular adhesion molecule-1 (ICAM-1) in endotoxin-induced pneumonia evaluated using ICAM-1 antisense oligo-nucleotides, anti-ICAM-1 monoclonal antibodies, andICAM-1 mutant mice. J Clin Invest 1996;97:2362–2369.

17. Bickel U, Kang YS, Yoshikawa T, et al. In vivo demonstrationof subcellular localization of anti-transferrin receptor mono-clonal antibody-colloidal gold conjugate in brain capillaryendothelium. J Histochem Cytochem 1994;42:1493–1497.

18. Fraker PJ, Speck JC. Protein and cell membrane iodinationwith a sparingly soluble chloramine. Biochem Biophys ResCommun 1978;80:849–856.

19. Gironella M, Molla M, Salas A, et al. The role of P-selectin inexperimental colitis as determined by antibody immunoblock-ade and genetically deficient mice. J Leukoc Biol 2002;72:56–64.

20. Molla M, Gironella M, Salas A, et al. Role of P-selectin inradiation-induced intestinal inflammatory damage. Int J Can-cer 2001;96:99–109.

21. Ma L, Raycroft L, Asa D, et al. A sialoglycoprotein fromhuman leukocytes functions as a ligand for P-selectin. J BiolChem 1994;269:27739–27746.

22. Janssen GH, Tangelder GJ, Oude Egbrink MG, et al. Sponta-neous leukocyte rolling in venules in untraumatized skin ofconscious and anesthetized animals. Am J Physiol 1994;267:H1199–H1204.

23. Soriano A, Salas A, Salas A, et al. VCAM-1, but not ICAM-1or MAdCAM-1, immunoblockade ameliorates DSS-inducedcolitis in mice. Lab Invest 2000;80:1541–1551.

24. Panes J, Perry MA, Anderson DC, et al. Regional differencesin constitutive and induced ICAM-1 expression in vivo. Am JPhysiol 1995;269:H1955–H1964.

25. Hallahan DE, Virudachalam S. Intercellular adhesion mole-cule 1 knockout abrogates radiation induced pulmonary in-flammation. Proc Natl Acad Sci USA 1997;94:6432–6437.

26. Lloyd CM, Gonzalo JA, Salant DJ, et al. Intercellular adhe-sion molecule-1 deficiency prolongs survival and protectsagainst the development of pulmonary inflammation duringmurine lupus. J Clin Invest 1997;100:963–971.

27. Gaugler MH, Squiban C, van der Meeren A, et al. Late andpersistent up-regulation of intercellular adhesion molecule-1(ICAM-1) expression by ionizing radiation in human endo-thelial cells in vitro. Int J Radiat Biol 1997;72:201–209.

28. Doerschuk CM, Quinlan WM, Doyle NA, et al. The role ofP-selectin and ICAM-1 in acute lung injury as determinedusing blocking antibodies and mutant mice. J Immunol 1996;157:4609–4614.

29. Schneeberger EE, Vu Q, LeBlanc BW, et al. The accumula-tion of dendritic cells in the lung is impaired in CD18-/- butnot in ICAM-1-/- mutant mice. J Immunol 2000;164:2472–2478.

30. Eppihimer MJ, Russell J, Anderson DC, et al. Endothelial celladhesion molecule expression in gene-targeted mice. Am JPhysiol 1997;273:H1903–H1908.

31. Sans M, Kawachi S, Soriano A, et al. Brain endothelialadhesion molecule expression in experimental colitis. Micro-circulation 2001;8:105–114.

32. Sans M, Panes J, Ardite E, et al. VCAM-1 and ICAM-1mediate leukocyte-endothelial cell adhesion in rat experimen-tal colitis. Gastroenterology 1999;116:874–883.

33. Atsuta J, Sterbinsky SA, Plitt J, et al. Phenotyping and cyto-kine regulation of the BEAS-2B human bronchial epithelialcell: Demonstration of inducible expression of the adhesionmolecules VCAM-1 and ICAM-1. Am J Respir Cell Mol Biol1997;17:571–582.

34. Propst SM, Denson R, Rothstein E, et al. Proinflammatory andTh2-derived cytokines modulate CD40-mediated expressionof inflammatory mediators in airway epithelia: Implicationsfor the role of epithelial CD40 in airway inflammation. J Im-munol 2000;165:2214–2221.

35. Hallahan DE, Geng L, Shyr Y. Effects of intercellular adhe-sion molecule 1 (ICAM-1) null mutation on radiation-inducedpulmonary fibrosis and respiratory insufficiency in mice.J Natl Cancer Inst 2002;94:733–741.


relative roles of icam-1 and vcam-1 in the pathogenesis of experimental radiation-induced intestinal...

Documents