Eur J Immunol 1991 21 2839-2844 Hctcrogeneic expression of non MHC antigens on keratinocytcs 2839 Marken De Bueger,
Astrid Bakker, Jon J. Van Rood and Eis Goulmy
Department of
Immunohaematology, University Hospitaä Leiden, Leiden
Minor histocompatibility antigens, defined by
graft-vs.-host disease-derived cytotoxic
Τ iymphocytes, show variable expression on
human skin cells*
Little IS known on the etiectoi raechamsms inducing the cutaneous lesions observed dunng acute graft vs -host diseasc (aGvHD) alter allogeneic bone marrow tiansplantation (BMT) Histological findings havc indicated that infiltrating CD8+ Iymphocytes probably play a role We addressed the question of whethei host minor histocompaübihty (mH) antigen-ieactive cytotoxic Τ Iym-phocytes (CTL) could account for this phenomenon via direct lysis ot the epidermal cell layer Six CTL clones, obtamed hom penphcral blood Iympho-cytes of patients suflenng from aGvHD, each recogm/mg a well chaiactcuzed MHC class I-restncted mH antigen epitope, weic tested on cultured keratmo-cytes of nmc MHC and mH antigen typed donors Four ol six mH antigen-speciüc CTL clones lysed unstimulated MHC class I expressmg, as well as recombinant mterferon-γ (rIFN y)-activated, ICAM-1, MHC class I- and II-expiessmg keia-tinocytes Two strongly cytolytic CTL clones showed no recogmtion of keiatino cytes of donors whose phytohemagglutimn-activated Tccll lines were leadily lysed With respect to aGvHD, the results imply that some class I-restncted CTL obtamed from penpheial blood Iymphocytes of aGvHD patients have the in vitro potcntial to destroy resting as well as IFN-y-activatcd epidermal cells, wheieas others do not In othei words, CTL-defmed human mH antigens vary with lespect to then expression in the skin it is intnguing that those minor Η antigens which cannot be detccted on human keratmocytes in vitro are those known to be associated with the occurrence ot GvHD
1 Introduction
The nature of the effector cells causmg damage to lymphoid ind epithehal tissues in GvH disease (GvHD) alter allo-geneic BM transplantation (BMT) is unknown [1,2] Munne GvHD modeis have indicated the mvolvement of vanous cell populations [3-7] and have shown that besides donor recipient dispanty for MHC class I, II or non-MHC antigens, other lactors such as genetic background influence which subpopulations and eftector mechanisms predominate [8] In human GvHD, host antigen-speciiic cytotoxic Τ cells (CTL) are consideied to be the local eftectors because of an inverse CD4/CD8 Τ cell latio in the epithehal target organs [9-11] and the presence oJ charac tenstic CTL target cell conjugates wnhin GvHD-atfected tissues such as the epidermal tissue ot the skin [12, 13] It is well estabhshed that host-reartive CTL are piesent in the penpheral blood ol patients sulfermg Irom acute or chrome GvHD upon grafting with MHC genotypically identical BM [14-17] These host-speciftc CTL were genei-ally found to exert class I MHC restneted lecogmtion ol [I 9622] ' This work was supported by the Maciopa Foundation and the
Τ Α Cohen Institute for Radiopathology and Radiation Protcc-tion (IRS)
Correspondence: Marleen De Bueger, Dcpaitment of Immuno-haematology and Bloodbank Room L3 37, Umveisity Hospital Leiden Rijsburgerweg 10 2333 AA Leiden The Ncthciiands
Abbreviations:
plantation
tK: Cullurcd keratinocyte BMT: BM tians
polymorphic non MHC antigens, or mmor histocompati-bihty (mH) antigens [14-17] Unlike the CTL response to murme mH antigens, it is not ceitain to which extent and via which mechanisms the CTL responses to Single human mH antigens influence the outcome ot BMT and in particular the onset of GvHD [17] One of the commonly proposed mechanisms is that host mH antigen specilic CTL would be directly lesponsible foi the local tissue destiuction in the GvHD affected organs
The aim of this study was to substantiate this hypothesis, which has been solely based on the histopathological obseivations descubed above We used a number of estabhshed, strongly cytolytic mH antigen-specitic CTL clones obtamed from PBL of patients suffeung fiom GvHD [18] Their cytolytic capacity toward human skin was tested in vitro in a modilied ^Ci lelease assay using cultmed skin-denved keiatmocytes as taiget cells [19] Wc found that keiatmocytes weie susceptible to lysis by some but not all mH antigen-specific CTL These Undings stiess that in addition to the male speciüc Η Yantigen [20] othei, but apparently not all, mH antigens can seive as taiget deteiminants tor CTL in the human skin Fuitheimoi e, the observed dilfeience m tissue expression of distinct mH antigens may be inteipreted as an indication ot stiuctuial diversity between human CTL-defined mFI antigens
2 Materials and methods
2.1 mH antigen and allo-dass I-specific CTL clones
2842 Μ. De Bueger, Α. Bakker, J. J.Van Rood and E. Goulmy
Table 3. Differential lysis of cK by class I-reslricted mH antigen-specific CTL
Eur. J. Immunol. 1991. 21: 2839-2844 CTL clone/ speeificity 3E7cVallo-A2 3HA15/HA-1 5H17/HA-2 5G30/HA-4 d.2/aIlo-Al 5HO11/HA-3 Korl8/allo-B7 C1.21/HA-6 cl.o/HA-7 E : T ratio 20 2 0.2 20 2 0.2 20 2 0.2 20 2 0.2 20 2 0.2 20 2 0.2 20 2 0.2 20 2 0.2 20 2 0.2 Ind. la> T") cK cK+ 76d» 33 55 69 9 9 33 0 0 72 72 13 89 83 30 97 72 31 -8 -8 -6 —3 —3 -4 18 5 0
—1
_4
-6
-2 -4 -4 56 30 9 -5 -5 -5 -7 -7 -7 -7 -8 _2 -6 -6 -3 -8 -8 -4 -6 -8 -2 -7 -8 —1 — — 0 — -_3 _ _ -1 -1 8 0 -1 3 -3 -3 0 -6 -6 -6 % Speciflc lysis Ind. 2 Τ cK cK+ 83 33 86 62 17 52 32 0 17 -3 -4 0 Λ - 1 - 2 —3 - 3 - 2 - 2 77 77 26 -4 - 3 - 4 | Äj - 4 —5 - 4 17 15 0 0 -3 - 3 - 5 0 0 - 5 —3 - 8 - 6 0 —3 0 55 30 91 55 14 80 68 11 6 _3 -4 14 3 3 49 12 -5 74 28 2 Ind. 3 Τ cK cK+ —5 - 4 _4 -5 -4 - 4 —3 - 2 -3 -2 -5 -3 -2 0 - 1 - 4 -2 _4 -6 -6 -6 -2 -6 -6 -4 1 —2 - 4 -4 -4 -6 7 7 Ο Ο Οο
- 1 2 Ο 89 51 71 61 23 55 10 7 26 86 67 16 37 16 2 63 5827 a) HLA types of individuais 1, 2 and 3 arc: A2, B27, 60, Cw2,3; DR11J2; A2,29; B7, w57, Cw6,7 DR2; AI, 24, B8, w57, DR3,7.
b) T, cK and cK+ indicate the different cell types of cach individual used as targets: T, PHA-induccd Τ cell lines; cK, culturcd keratinocytes and cK^, eultured keratinocytes preineubated for 24 h with 100 U/ml rIFN-γ. c) See Table 1.
d) Mean values of % speeifie lysis in a 4-h 51Cr-release assay.
skin-derived keratinocyte eultures. Of each individual, cK eilher untreated or after a 24-h ineubation with 100 U/ml rIFN-γ, as well as PH Α blasts were ineubated as target cells with HLA class I-specific and mH antigen-specific CTL clones and assayed for 5 1Cr release. Results of three informative donors are given in Table 3. The HLA-A2, -AI and -B7 speeifie CTL clones were found to lyse both untreated and, somewhat more efficiently, IFN-γ treated cK, consistent with lysis of the PHA blasts and the serological typing of the donors (see legend of Table 3). Of the three HLA-A2-restricted mH antigen-specific CTL clones tested, clonc 5G30 was found to lyse either both the Tcell line and cK of a given donor (e.g. Ind. 1), or none (e.g. Ind. 2,3). In contrast, clones 3HA15 and 5H17 were negative ( < 3% speeifie lysis) with cK of donors whose Tcell lines were readily lysed by the HA-1- or HA-2 speeifie CTL (e.g. Ind. 1, 2).The CTL clone speeifie for the Al-restricted mH antigen HA-3 was strongly positive (>30% at E : T = 20) with cK of HA-3+ donors (.e.g Ind. 3). The HLA-B7-restricted HA-6 and HA-7-reactive CTL clones were also indicated to recognize cK, although percentages of speeifie lysis of uninduced cK were low in all experiments (cl.21: ränge 6%-18%; cl.6 ränge 14%-19% at Ε : Τ = 20). In contrast to the variable suseeptibility of basal cK to lysis by distinet anti-minor CTL clones, IFN-γ pretreated keratinocytes (cK+, Table 3) were efficiently
lysed by HA-3 as well as HA-4-, HA-6- and HA-7-reactive CTL. However, ineubation with IFN-γ or other keratino-cyte-activating lymphokines (not shown) could never induce lysis by the HA-L- or HA-2-specific, HLA-A2-restricted CTL clones (Table 3, Ind. 1,2). Analogous test-ing of cK and PHA blasts of six other donors was fully consistent with the data shown in Table 3, indicating that expression of the mH antigens, HA-3, HA-4, HA-6 and HA-7 but not HA-1 and HA-2 was dctectable on cK in vitro.
3.4 Cell surface antigens involved in mH antigen-specific iysis of keratinocytes
Eur J Immunol 1991 21 2839-2844 Hctcrogeneic cxprcssion ot non MHC antigcns on keratinocyte 2843
clone 5HO11
1000 00 10 0 1000
rec procal ant body concentration
Figurc 2 Inhibitory effect of mAb agamst vanous cell surfacc antigcns on IIA 3 and HA 7 mll antigen specific lysis of rIFN γ treatcd keratinocytcs 5lCi release valucs measurcd in the picscncc of antibody arc represented as percentagc of the unblocked icsponsc at Ε Τ = 20
anti-LFA-1 (36% -54%) antibodies Although complete Inhibition could never be obtamed at satmating concentra-tions, the observed Inhibition by anü-CD8 anti-MHC class I anti-LFA 1 and anti ICAM 1 mAb mdicated the bmding ot TcR/CD8 to MHC class I, and of LFA 1 to ICAM-1 is involvcd in MHC class I-specific as well as class I restncted lysis ot IFN γ activated keratmocytes Α lepresentativeblockingexpenmenlusinglFN γ pretreated cK and the mH antigen specific CfL clones 5HO11 (anü-HA 3/A1) and cl 6 (anti-HA-7/B7 is shown m Fig 2)
4 Discussion
Matuie donor Τ cells reactive with host antigens arc known to be lelevant in murine acute GvHD atter allogeneic BMT [3] Depending on the antigcnic barner and the genetic background ot the strams analyzed, either the Ly-2 [3-5,8], or the Ly 2 as well as the L3T4 Tcell subpopulations [6, 8] arc known to be involved in the mduction of lethal GvHD I h e obseived lesions in the GvHD atlectcd organs are beheved to result tiom dcstruction οι basal epidermal cells [24] For this to occur, the local piesence of inriammatory lymphokmes such asTNF-α has been implied to be essential [25, 26], whereas lt is still unclcar whethei NK or large granulär lymphocytes [7], CD4 [26] or CD8 Τ cells [27] ultimately mediate the local pathogenesis
In human BMT, selective depletion of matureTcell subsets mdicated that both the CD4 and the CD8 Τ cell populations aie generally involvcd in the induction oi GvHD [28], although in sotne reports preüeatmcnt ot BM with anti-CD8 mAb alonc could sigmficantly leduce the seventy of the disease [29] The local pathogenesis of GvHD affected tissues in human is, as in expenmental modeis charactei ized by apoptosis of mostly ICAM 1 and MHC class II-exprcssing keratmocytes [13] in the basal epithelial cell layei [30], accompamed by a moderate deimal and epidei-mal Infiltrate of mainly CDSf Tctlls [9-11J On the basis ot
these histopathological data and the common detection oi host-ieacüve CTL in PBL of GvHD patients eaily atter BMT [14-17], the hypothesis was commonly held that cell-medidted lysis of keratmocytes by anti-host CTL would be impoitant in the mduction of tissue lesions dunng GvHD The data presented in this lepoit now showed that MHC class I lestucted CTL clones speeihe tor the human mH antigens HA 3, 4, 6 and 7, but not toi the mH antigens
HA 1 and HA-2 were capable of lysmg keratmocytes in vitro From this lt follows that at least some mH antigen-specific CTL populations obtained from PBL of GvHD patients have the in vitro potential to destroy epithelial cells in an MHC class I restncted, mH antigen-specific fashion For this Τ cell recogmtion to occur, IFN-y-mduced ICAM 1 cxpression was not per se required However, the data presented here cannot be dnectly extrapolated to an in vivo role of CTL agamst these mH Ag as effectors in GvHD-atfected skin As mdicated in some lecent repoits, CD4 Τ cells displaying in vitro host antigen-specific prohferation can also be obtained from GvHD-affected skin [31, 32] Furthermore, the recently observed TNF α and IFN-γ dependent destruction of human skin epithelmm in vitro indicates that a direct lole of inflammatory lymphokmes in the local GvHD pathogenesis cannot be exeluded [33] Clearly, only a more direct assessment like in murine modeis [26] could help to answer the qucstion which host antigens are most impoitant in evoking the inflammatory reaction ot prohferative and cytolyticT cells into the tissues and which antigen-specific oi nonspecitic mechamsms ultimately mflict the local damage
lt must be süessed that though dispanty foi non-HLA antigens forms an estabhshed risk factoi for GvHD after MHC identical BMT, lt is virtually impossible to assess and prove the in vivo importance of smgle human mH antigens in transplantation This is in contrast with the Situation in murine GvHD modeis wheie by using congemc strams the relative contribution ot each Single mH antigenic dispanty to the development of GvHD could be evaluated [26, 34-36] An indication of the in vivo relevance of some ot the human mH antigens discussed here was obtained by retrospectively companng antigemc dispanty and gratt outcome in a panel of MHC genotypically identical do-noi/recipient pairs Contraiy to what one would expect on the basis of the data piesented here, a donor/iecipient dispanty lor HA-1 and/or HA 2 was mdicated to be associated with an increased risk ot GvHD ([14],Tablc 1) In contrast, üansplantation across a HA-3 bamer was not a statistically sigmficant risk tactor for GvHD [ 14], whereas the impact ot BMT across a dispanty toi HA 4 6 oi -7 on the development of GvHD has not yet been analy/ed The in vitro observations leported here, reveahng recogmtion ot mH antigcns on keratmocytes by anti-HA-3, but not anti HA-1 or HA 2 CTL, thus stress that, in analogy to reports in the mouse [37], the potential of mH antigen specific CTL to lyse keiatinocytes in vitro does not simply correlate with the pioposed immunogenicity ot the mH antigen in vivo
In addition to the circumstantial Information regardmg the role of anti-mH antigen CTL in cutaneous GvHD, the data piesented heie piovide insight into the tissue expression ot human mH antigens Fiom the observed absence of lysis by the stiongly cytolytic anti-HA-1 and HA 2 CTL clones m a sensitive 5lCi-ielease assay, it may well be concluded that
2844 Μ De Bueger, Α Bakker, J J Van Rood and Ε Goulmy Eur J Immunol 1991 21 2839-2844 mH antigens HA-3, 4, 6 and 7 have an MHC class I-like
tissue distnbution, whereas HA-1 and HA-2 are only expressed on cells of hematopoietic ongm (De Bueger et al , in preparation)
This observed differential expression does provide an indication for heterogeneity withm the group of CTL-defmed human mH antigens, which so far had been ldenti-fied and studied predommantly on PHA blasts and EBV-BLCL [14-18, 38] Since to date no data are available on the immunochemical nature of and genes encodmg for the mH antigen epitopes, lt IS difficult to speculate by which mechamsm the restncted tissue expression of HA-1 and HA-2 complexes is achieved [39] Interestingly enough, an analogous heterogeneity seems to exist for munne mH antigens Wettstein and Colombo [34] and Korngold and Wertstem [35] reported that in contrast to the mH antigens
deüned in congemc strams which have a class I-like tissue distnbution, mH antigens discovered in the H-2b
-compati-ble, B6 to BALB Β GvHD model appeared to be expressed
only on lymphoid tissues Attempts to isolate and identify the human cell surface products as well as to map the encodmg genes are currently being performed for a number of mH antigens and will hopefully provide msight into the molecular basis of heterogeneity of human mH antigens Combmed with Information on the relative lmmunodomi-nance of non-MHC antigens in vivo and in vitro [18] this will hopefully provide sufficient knowledge to match for rele-vant mH antigemc Systems and thereby avoid at least one nsk factor m human BMT
The authon thank Dr R Teepe for obtaimng skin specimens, Μ Ponec for advice and cntical reading of the manustnpt and J Kempenaar for expert technical advice
Received June 3 1991, in final revised form August 20, 1991
5 References
1 Renkonen, R and Hayry, Ρ, Bone Marrow Transplant 1987 2 333
2 Ringden Ο , Deeg, Η J , Bcschorner, W and Slavin, S ,
Transplanl Proc 1987 1 2758
3 Korngold, R and Sprent, J , Immunol Rev 1983 71 5 4 Okunawick, J P, in Gale, R Ρ and Champlin, R (Eds ),
Progress in Bone Marrow Transplantation, Alan Liss, New York
1987, ρ 199
5 Hamilton,B C,Bevdn,M J andParkman,R P,7 Immunol 1981 126 621
6 Hamilton, Β C , / Immunol 1987 139 2511
7 Guillen, F J , Ferrara, J , Hancock, W W, Mossadi, D , Fonferko, Ε , Burakoff, S J and Murphy, G F , Lab luvest 1986 55 35
8 Hamilton, Β , in Gale R Ρ and Champlin, R (Eds ), Bone
Marrow Transplantation, current controversies, Alan R Liss,
New York 1989, ρ 405
9 Guyotat, D , Maudmt, G Chouvet, Β , Kamtakis, J ,Van Vu, Η , Fiere, D and Thivolet, J , Transplantation 1986 41 340
10 Kaye, V Ν , Neumann, Ρ Μ , Kersey, J , Goltz, R W, Baldndge,B D , Michael, Α F and Platt, J L ,Am J Pathol 1984 116 436
11 Atkmson, K , Munro, V, Vasak, Ε and Biggs, J , Β J
Dermatol 1986 114 145
12 Säle, G Ε , Gallucci, Β Β , Schubert, Μ Μ , Suhvan, Κ Μ and Thomas, Ε D , Arch Pathol Lab Med 1987 111 333
13 Lampert, Ι Α , Janossy, G , Suiters, Α J , Bofill, Μ , Palmer, S , Gordon-Smith, Ε ,Prentice,H G and Thomas, J Α , Chn
Exp Immunol 1982 50 123
14 Goulmy, Ε , in Morris, Ρ J andTilney, Ν L (Eds), Transplant
Rev 1989 2 29
15 Tsoi, Μ S , Storb, R , Santos, Ε and Thomas, Ε D ,
Transplant Proc 1983 15 1484
16 Irle C , Beatty, Ρ G , Mickelson, Ε , Thomas, Ε D and Hansen, J Α , Transplantation 1985 40 329
17 Van Eis, C A C M , Bakker, Α , Zwinderman, Α Η ,Zwaan, F Ε ,VanRood,J J and Goulmy, Ε Transplantation 1990 50 62
18 Van Eis, C Α C Μ , D'Amaro, J , Pool, J , Blokland, Ε , Bakker, Α ,Van den Elsen, Ρ ,Van Rood, J J and Goulmy, Ε ,
J Immunol , in press
19 De Bueger, Μ Μ ,VanEls,C A C M , Kempenaar, J , Ponec, Μ and Goulmy, Ε , / Immunol Methods 1990 127 117 20 Van Eis, Α C Μ , De Bueger, Μ Μ , Kempenaar, J , Ponec,
Μ and Goulmy, Ε , / Exp Med 1990 170 1469 21 Rheinwald, J and Green, Η , Cell 1975 6 331
22 Dustin, Μ L , Springer, Κ Η ,Tuck, D Τ and Springer,Τ Α ,
J Exp Med 1988 167 1323
23 Schmidt, Η , Gekeler,V, Haas, Η , Engler Blum, G , Steiert, F , Probst, Η and Müller, C Α , lmmunogenetics 1990 31 245
24 Ferrara, J , Guillen, F J , Sleckman, Β , Burakoff, S J and Murphy, G F , / luvest Dermatol 1986 86 371
25 Piguet, Ρ F, Grau, G Ε , Allet, Β and Vassalh, P, J Exp
Med 1987 766 1280
26 Micconet, 1 , Huchet, R , Bonardelle, D , Motta, R , Canon, C , Garey-Rojas, Ε , Kress, Μ , Reynes, Μ , Halle-Pannenko, Ο and Bruley-Rosset, Μ , / Immunol 1990 145 2123 27 Piguet, Ρ F , Jamn-Mercicr, Α , Assalh, Ρ and Saurat, J Η , /
Immunol 1987 139 406
28 Maramnchi, D , Mawas, C , Guyotat, D , Reiffers, J ,Vernant, J P, Gratecos, Ν , Hirn, J and Novakovitch, G , Transplant
Int 1988 1 91
29 Champlin, R , Gajewski, J , Feig, S , Giorgi, J , Lyddane, Ν , Lee, Κ , Schmidt, I ,Winston, D , Ho, W and Reichert, Τ ,
Transplant Proc 1989 21 2947
30 Säle, G Ε , Shulman, Η Μ , Gallucci, Β Β and Thomas, E
Ό , Am J Pathol 1979 118 278
31 Reinsmoen, Ν L , Kersey, J Η and Bach, F-H , Human
Immunol 1984 11 249
32 Kasten-Sportes, C , Masset, Μ , Varnn, F , Devergie, Α and Gluckman, Ε , Transplantation 1989 47 621
33 Dickinson, Α Μ , Sviland, L , Dünn, J , Carey, Ρ and Proctor, S J , Bone Marrow Transplantation 1991 7 209
34 Wettstein, Ρ J and Colombo, Μ Ρ, .1 Immunol 1987 139 2166
35 Korngold, R and Wettsem, Ρ J , J Immunol 1990 145 4079
36 Fast, L D , / Immunol 1990 144 4177
37 Stemmuler, D , Smder, Μ Ε , Noble, R L and Waldschmidt, Τ J , Transplantation 1990 50 663
38 Voogt, P, Goulmy, Ε ,Veenhof,W F J , Hamilton, Μ , Fibbe, W Ε ,Van Rood, J J and Falkenburg, J Η F , / Exp Med 1988 168 2337
