clin icam-1 expression in renal disease · icam-1 positivity on damaged tubules...

I8 Clin Pathol 1992;45:880-884

ICAM-1 expression in renal disease

J Chow, R B Hartley, C Jagger, S A Dilly

AbstractAim: To determine the role ofintercellularadhesion molecule-I (ICAM-1) in im-mune mediated damage in glomerulo-nephritis, and whether its expressioncorrelates with disease activity.Methods: Fifty three renal biopsy speci-mens from a range of non-immune renaldisease and low grade and high gradeglomerulonephritides were stained withICAM-1. Positivity was assessed in thetubules. Tubular damage and accompany-ing interstitial inflammation were alsonoted. The ICAM-1 positivity in damagedand undamaged tubules was correlatedwith the three groups of renal disease.Results: ICAM-1 positivity in undamagedtubules was observed in glomerulo-nephritis, and this showed a strong corre-lation with disease activity. In contrast,ICAM-1 positivity on damaged tubulescorrelated with evidence of chronic tubu-lar damage, and was seen in a largeproportion of cases, regardless of theunderlying disease. There was no correla-tion between ICAM-1 positivity and alocal lymphocytic infiltrate.Conclusion: ICAM-1 probably has animportant role in the pathogenesis ofglomerulonephritis. Its expression may besecondary to cytokines released by cellsparticipating in the glomerular damage.As these cytokines also influence tubulefunction, tubular ICAM-1 expression maybe a marker of the extent of tubulardisturbance in glomerulonephritis.

(7 Clin Pathol 1992;45:880-884)

by the ability of cytokines such as tumournecrosis factor (TNF), interleukin-1 (IL-1),and y-interferon (y IF) to upregulate ICAM-1expression.6

Previous studies of ICAM-1 expression inthe kidney have demonstrated its role in T cellmediated rejection of allografts" and in mur-ine lupus nephritis.

MethodsFifty three renal biopsy specimens were stud-ied. The histological diagnoses were made onparaffin wax sections and by using immuno-fluorescence, together with relevant clinicalinformation.Frozen sections were cut and stained with

mouse antihuman ICAM-1 (Serotec, Kidling-ton, Oxford) at a dilution of 1 in 200, using anindirect immunoalkaline phosphatase tech-nique. Negative controls were provided byexclusion of the primary antibody in the renalbiopsy specimens and ICAM-1 staining offrozen sections of normal skin, in whichkeratinocytes do not stain with ICAM- 1. " Thestained sections were examined and scoredwithout knowing the clinical diagnosis. Posi-tivity was assessed in the following structures:glomeruli, tubules, and blood vessels. The stateof the tubules-that is, whether they werenormal or exhibiting atrophy and peritubularfibrosis-was recorded. The presence of alymphocytic infiltrate in the interstitium wasalso noted. The results of ICAM-I positivitywere then correlated with the clinicopatho-logical diagnosis.

Tests for significance on proportions weredone using the x2 test or Fisher's exact testwhere appropriate.

Department ofHistopathology, StGeorge's HospitalMedical School,LondonJ ChowC JaggerS A DillyDepartment ofClinical Microscopy,Guy's Hospital,LondonR B HartleyCorrespondence to:Dr J Chow, HistopathologyDepartment, St George'sHospital Medical School,Cranmer Terrace, LondonSW17 ORE

Accepted for publication20 March 1992

Cellular adhesion molecules have an importantrole in the immune response by modulatingcell-cell interactions after antigen presentation.Lymphocyte function associated antigen-I(LFA-1) is one of many cellular adhesionmolecules on lymphocytes. Studies on mono-clonal antibody inhibition have shown thatLFA-1 has a role in cytotoxic T lymphocytemediated lysis,' 2 the proliferation of helper Tcell populations,3 antibody responses to Blymphocytes4 and natural killer cell mediatedcytotoxicity.l

Intercellular adhesion molecule-i (ICAM-1), a cell surface glycoprotein, has a role in theregulation of interactions among immunecells.5'6 This interaction with target cells ismediated through its ligand LFA-1 7-10 Thekey role of ICAM-1-LFA-1 interaction in theregulation of immune reactions is also shown

ResultsThe cases were subdivided on clinicopatho-logical grounds into three categories: non-immune renal parenchymal disease (group 1, n= 14); low grade glomerulonephritis togetherwith interstitial nephritis (group 2, n = 20);and high grade glomerulonephritis (group 3, n= 19). Details of these cases are listed in table1.Non-immune renal parenchymal disease

included conditions such as acute tubularnecrosis and ischaemic damage. The case ofnon-Hodgkin's lymphoma was not.affected bytumour in the frozen tissue assessed.Low grade glomerulonephritis and inter-

stitial nephritis comprised the immune com-plex glomerulonephritides with a slow andvariable course and two cases of interstitial

880

on August 19, 2021 by guest. P

rotected by copyright.http://jcp.bm

j.com/

J Clin P

athol: first published as 10.1136/jcp.45.10.880 on 1 October 1992. D

ownloaded from

http://jcp.bmj.com/


Table 1 Histological diagnosis and tubule staining for ICAM-1

ICAM-1 positivity ICAM-I positivityon damaged on undamaged

Group n tubules tubules

I Non-immune renal parenchymal disease (n = 14):Hypertension 1 0 0Thin membrane disease 1 0* 0Acute tubular necrosis 3 1 0Non-Hodgkin's lymphoma 1 1 1Renal dysplasia 1 1 0Reflux nephropathy 1 0 0Diabetes mellitus 1 1 0Amyloid 1 1 0Ischaemia 3 2 0Remnant nephropathy 1 1 0

2 Low grade glomerulonephrins and interstitial nephritts (n = 20):Minimal change disease 1 0* 1Focal and segmental glomerulosclerosis 4 4 1IgA nephropathy 4 2 0Membranous glomerulonephritis 2 1* 1Type 1 Mesangiocapillary glomerulonephritis 4 3 4Mesangioproliferative GN (lupus II b) 1 0* 0Mesangioproliferative GN 1 0* 0Henoch-Schonlein purpura (no crescents) 1 0 0Interstitial nephritis 2 0 0

3 High grade glomerulonephritis (n = 19):Henoch-Schonlein purpura (30-80% crescents) 3 1* 3Diffuse proliferative crescentic GN 5 4* 3Lupus nephritis (class IV) (4 crescentic) 7 4* 5Type 1 Mesangiocapiliary glomerulonephritis (crescentic) 1 1 tMicroscopic vasculitis with crescents 2 2 2Haemolytic uraemic syndrome 1 1 0* includes cases with no damaged tubules for assessment.t undamaged tubules for assessment. GN = glomerulonephritis.

nephritis. The case of mesangioproliferativeglomerulonephritis represented a recoverystage from a postinfectious proliferative glo-merulonephritis a year earlier. Minimal changenephropathy was included in this groupbecause, although the aetiology is uncertain,some evidence implicates aT cell disorder.High grade glomerulonephritis comprised

cases with a rapidly aggressive clinical course,

tk it. X tVg a

Figure 1 Endothelial cells in the peritubular vessels and medium size arteries areconsistently positive with ICAM-1.

and included all cases of crescentic nephritis,WHO class IV lupus nephritis, and a case ofclinically and histologically extremely activehaemolytic uraemic syndrome (HUS).ICAM-1 was consistently positive in the

endothelium of small and medium sized ves-sels; this was used as the positive internalcontrol (fig 1). The pattern was that of diffusecytoplasmic staining. Cells within the glomer-uli also stained, but it was difficult to differ-entiate the cell types on frozen sections (fig1).

Staining of the tubular epithelium was vari-able, depending on whether it was damaged,and the type of disease. Staining on undam-aged tubules did indicate differences amongthe groups. The pattern of staining in all threegroups was similar, showing a cytoplasmicpositivity, mainly of the luminal portion of thetubular epithelial cells, and membrane stainingof the luminal aspect. However, there werestrong differences in the proportion of casespositive in the three groups (table 2). Only onecase in group 1 showed positivity in theundamaged tubules; this was the case of non-Hodgkin's lymphoma (7%). In group 2ICAM-1 positivity affected undamaged tubu-lar epithelium in 35% of cases. This included acase each of minimal change disease, focal andsegmental glomerulosclerosis, membranousglomerulonephritis, and four cases of type 1MCGN (fig 2). The percentage of ICAM-1positivity in low grade glomerulonephritis wassignificantly different from that in non-immune renal disorders (p < 0004). In con-trast, 72% of cases in group 3 showed ICAM-1positivity in undamaged tubules (fig 3). Thiswas significantly different from that in boththe cases of non-immune renal diseases(p < 0002) and low grade glomerulonephritis(p < 0-02).The percentage of cases exhibiting positivity

881



j.com/

J Clin P


ownloaded from

http://jcp.bmj.com/

Chow, Hartley, _'agger, Dilly

Figure 2 Undamaged proximal tubules in a case of crescentic diffiuse proliferativeglomerulonephritis shows cytoplasmic positivity with ICAM-1, mainly on the luminalside.

Figure 3 A case of type I mesangioproliferaive glomerulonephrits exhibiting stainingwith ICAM-I in the undamaged proximal tubules.

Table 2 ICAM-1 positivity in tubular epithelium in non-immune and immune-mediated renal disease

Damaged Undamagedtubulest tubules

1 Non immune renalparenchymal disease (n = 14) 8 (67%) Id (7%)

2 Low gradeglomerulonephritis andinterstitial nephritis (n = 20) 10 (71%) 7' (35%)

3 High gradeglomerulonephritis (n = 19) 13 (100%) 13*bC (72%)

* one case did not have undamaged tubules for assessment.t some cases did not have damaged tubules.a 2 v 1 - p < 0 0004 (Fisher's exact test)b 3 v 1 - p < 0-0002 (Fisher's exact test)c 3 v 2 - p <0-02 X2 test)d (3 + 2) v 1 - p < 0-0002 (Fisher's exact test).

in damaged tubules in non-immune renalparenchymal disease (group 1), low gradeglomerulonephritis (group 2), and high gradeglomerulonephritis (group 2) was 67%, 71%,and 100%, respectively (table 2). Hence dam-aged tubules, which were mainly atrophiedwith peritubular fibrosis, showed no significantdifference in positivity among these threegroups (fig 4). The positivity was that of astrong diffuse granular cytoplasmic pattern.The correlation of tubule ICAM-1 positivity

with lymphocytic infiltration was assessed byfirst identifying the number of cases exhibitingboth ICAM-1 tubule positivity and an asso-ciated increase in local lymphocytes and relat-ing this to the number of cases showing eitherICAM-1 positivity or a lymphocytic infiltrate.No lymphocytic infiltrate was seen in thevicinity of positively stained undamaged tubulesin any of the groups.

In contrast, the percentage of cases exhibit-ing both positivity in damaged tubules and alymphocytic infiltrate was 25%, 14%, and23%, respectively, in groups 1, 2 and 3. Thepercentage of cases with positivity in damagedtubules (irrespective of lymphocytic infiltrate)was 67%, 71%, and 100%, in the three groups,respectively, while the percentages of caseswith a lymphocytic infiltrate (irrespective ofpositivity in damaged tubules) were 36%, 20%and 21% in groups 1, 2, and 3, respectively.Thus the number of cases exhibiting bothpositivity in damaged tubules and a lympho-cytic infiltrate was not greater than would beexpected by chance. In particular, both casesof interstitial nephritis with a dense lympho-cytic infiltrate were completely negative withICAM- 1. It therefore seems that ICAM- 1positivity in damaged tubules correlated withevidence of chronic tubular damage ratherthan an inflammatory infiltrate, and thatICAM-1 positivity in undamaged tubules wascorrelated with disease activity and not withthe presence of local lymphocytes.

DiscussionIn this study we found that 7% of biopsyspecimens from cases of non-immune renaldisease were positive with ICAM-1. Geneticmake-up is a possible explanation for ICAM-1expression in a few normal proximal tubules.'"This differs from early studies of ICAM-1expression in human tissues where proximal

882



j.com/

J Clin P


ownloaded from

http://jcp.bmj.com/


A-<t<X;'2s2

c....}.; ,y 4,; :..s. -.-

Figure 4 A focus of atrophic tubules with peritubular fibrosis shows strong cytoplasmicpositivity with ICAM-1. Note absence of a lymphocytic infiltrate in the interstitium.

tubules of kidneys were uniformly negative6 l1but is supported by the findings of Smithand Thomas,'3 who demonstrated ICAM-1expression in proximal tubules in the singlepiece of renal tissue included in their study.Damaged tubules were frequently

(67%-100%) positive and showed no correla-tion with disease activity. In contrast, ICAM-1expression on undamaged tubules did relate tothe underlying disease, with 35% of low gradeglomerulonephritides and 72% of high gradeglomerulonephritides being positive. The cases

of glomerulonephritis exhibiting ICAM-1 pos-itivity on undamaged tubules showed diffuse or

focal segmental proliferation in the glomeruli,but no lymphocytic infiltrate was seen in thevicinity of these undamaged tubules. The

correlation of ICAM-1 expression and diseaseseverity is highlighted by the three cases ofHenoch-Schonlein purpura with crescentswhich exhibited ICAM-1 positivity; a lowgrade Henoch-Schonlein purpura (histolog-ically and clinically) was negative. In addition,five out of seven cases of class IV lupusnephritis exhibited ICAM-1 positivity inundamaged tubules whereas a case of class IIlupus nephritis was negative.

It is uncertain if the cases of ICAM-1expression in low grade glomerulonephritisrepresent an initial event in T cell activation,where the appearance of such lymphocytes isstill not yet evident or in which cytokines maybe released by circulatingT lymphocytes. Thiswould agree with Ooi et al's findings oflymphocytoxins in patients with minimalchange glomerulonephritis. 4 The absence of a

surrounding lymphocytic infiltrate in the vicin-

ity of these ICAM-1 positive tubules contra-dicts the findings of Weetman et al inautoimmune thyroiditis."5 The likely explana-tion for this is that ICAM-1 expression bythyrocytes renders them more susceptible tolymphocyte mediated cytotoxicity; but renaltubules are not the primary target of cellmediated lysis in glomerulonephritis. Bothcases of interstitial nephritis examined wereICAM-1 negative.Although there is no evidence of immuno-

logical damage to the tubules in cases ofglomerulonephritis, the ICAM-1 expressionnone the less would be consistent with therelease of cytokines in the microenvironmentof the nephron as a result of primary glo-merular immune complex disease. There is abody of evidence to suggest that the changes intubular function, which accompany glomer-ulonephritis and other non-immune renal dis-orders, may be mediated by cytokines such asIl-1 and TNF, and also arachidonic acidmetabolites."6 These factors alter the sodium-potassium-adenosinetriphosphatase pump andhaemodynamics, and their presence wouldupregulate tubular ICAM-1 expression in glo-merulonephritis. Because they are secreted bymacrophages, it is not suprising that ICAM-1expression correlates with proliferative activityin glomeruli and the presence of crescents. Onthis note, it is also interesting that althoughtype 1 MCGN (without crescents) is a lowgrade disease, all the cases studied showedICAM-1 positivity; this is presumably relatedto the mesangial proliferation and infiltrationby macrophages.The consistent presence of ICAM-1 in a

significant proportion of damaged tubuleswhich are atrophic with thickening of thetubular basement membrane and peritubularfibrosis is interesting. Inflammatory mediatorsreleased as a result of ischaemia may be thecause of this upregulation of ICAM.6"1 Alter-natively, any damage promotes attempts atregeneration which may non-specifically upre-gulate a variety of surface molecules. Obvi-ously, diagnostically, it is crucial to distinguishbetween this non-specific staining and thestaining on undamaged tubules.

In conclusion, we have found that ICAM-1is expressed in renal tubules in cases ofglomerulonephritis, and that the degree ofpositivity correlates with disease activity.ICAM-1 expression on tubules does not corre-late with the presence of an adjacent inflamma-tory infiltrate, raising the possibility thatICAM- 1 expression may be secondary tocytokines released by cells participating in theglomerular damage. Because these cytokinesalso influence tubule function, tubule ICAM-1expression may be a marker of the extent oftubular disturbance in glomerulonephritis.

1 Krensky AM, Sanchez-Madrid F, Robbins E, Nagy JA,Springer TA, Burakoff SJ. Functional significance, dis-tribution and structure of LFA-I, LFA-2 and LFA-3: cellsurface antigens associated with CTL-target interactions.J Immunol 1983;131:611-16.

2 Sanchez-Madrid F, Krensky AM, Ware CF, et al. Threedistinct antigens associated with human T-lymphocyte-mediated cytolysis: LFA-1, LFA-2, LFA-3. Proc NatlAcad Sci USA 1982;79:7489-93.

883



j.com/

J Clin P


ownloaded from

http://jcp.bmj.com/

Chow, Hartley, Jagger, Dilly

3 Shaw S, Ginther Luce GE. The lymphocyte function-associated antigen (LFA-1) and CD2/LFA-3 pathways ofantigen-independant human T-cell adhesion. Immunol1987;139: 1037-45.

4 Tedder TF, Schmidt RE, Rudd CE, Kornachi MM, Ritz J,Schlossman SF. Function of the LFA-1 and T4 moleculesin the direct activation of resting human B-lymphocytesby T-lymphocytes. Eur Immunol 1986;16:1539-43.

5 Rothlein R, Dustin ML, Marlin SD, SpringerTA. A humanintercellular adhesion molecule (ICAM-1) distinct fromLFA-1. Immunol 1986;137:1270-4.

6 Dustin ML, Rothlein R, Bhan AK, Dinerello CA, SpringerTA. Induction of IL-1 and interferon-y: tissue distribu-tion, biochemical and function of a natural adhesionmolecule (ICAM-1). Immunol 1986;137:245-52.

7 Dougherty CJ, Murdoch S, Hogg N. The function ofhumanintercellular adhesion molecule-I (ICAM-1) in the gen-eration of an immune response. Eur Immunol 1988;18:35-40.

8 Marlin S, Springer TA. Purified intercellular adhesionmolecule-I (ICAM-1) is the ligand for lymphocytefunction-associated antigen-l (LFA-1). CeUl 1987;51:813-19.

9 Dustin ML, Singer KH,Tuck DT, SpringerTA. Adhesion ofT-lymphoblasts to epidermal keratinocytes is regulated byinterferon-y and is mediated by intercellular adhesion

molecule (ICAM-1). Exp Med 1988;167:1323-40.10 Makgoba MW, Sanders ME, Ginther-Luce GE, et al.

Functional evidence that intercellular adhesion molecule-1 (ICAM-1) is a ligand for LFA-1-dependent adhesion inT-cell mediated cytotoxicity. Eur Jf Immunol 1988;18:637-40.

11 Faull RJ, Russ GR. Tubular expression of intercellularadhesion molecule-I during renal allograft rejection.Transplantation 1989;48:226-30.

12 Wuthrich RP, Jevnikar AM, Takei F, Glimcher LH, KelleyVE. Intercellular adhesion molecule-I (ICAM-1) expres-sion is upregulated in autoimmune murine lupus nephri-tis. Am Pathol 1990;136:441-50.

13 Smith MEF, Thomas JA. Cellular expression of lymphocytefunction associated antigens and the intercellular adhe-sion molecule-I in normal tissue. Clin Pathol 1990;43:893-900.

14 Ooi BS, Orline AR, Masaitis L. Lymphocytotoxins inprimary renal disease. Lancet 1974;ii:1348-50.

15 Weetman AP, Cohen S, Makgoba MW, Borysiwicz LK.Expression of an intercellular adhesion molecule, ICAM-1, by human thyroid cells. Endocrinol 1989;122:185-91.

16 Schreiner GF, Kohan DE. Regulation of renal transportprocesses and hemodynamics by macrophages and lym-phocytes. Am _7 Physiol 1990;258(4 Pt 2):F761 -7.

884



j.com/

J Clin P


ownloaded from

http://jcp.bmj.com/

clin icam-1 expression in renal disease · icam-1 positivity on damaged tubules...

Documents