Characterization of skin-infiltrating cells during acute graft-versus-host disease following bone marrow transplantation using unrelated marrow donors

Characterization of Skin-Infiltrating Cells During Acute Graft-versus-Host Disease Following Bone Marrow Transplantation Using Unrelated Marrow Donors Afzal Nikaein, Twyala Poole, Randi Fishbeck, Guido Ordonez, Lori Dombrausky, Marvin J. Stone, Robert H. Collins, Jr., and Joseph W. Fay

A B S T R A C T To characterize skin-infiltrating T lymphocytes during acute GVHD, skin biopsies were obtained from two patients who rece,ved unrelated marrow matched for HLA-A, -B, -DR, and -DQ but mismatched for -DP A total of 120 T-cell clones were generated Phenotype analysis of the clones showed that the malorlty of cells were CD4 + and expressed or/J3 TCR HLA-DP ohgonucleotide genotypmg of the clones revealed the presence of lymphoid chimerism PLT assay showed the lack of HLA specificity, including mismatched HLA-DP However, mAb to HLA antigens blocked proliferation of the malorlty of the clones, indicating that the clones recogmzed HLA-assoclated molecules Interestingly, prol,feratlon of two CD4 + T-cell clones was inhibited by class

ABBREVIATIONS APC antigen-presenting cell CMV cytomegalovlrus EBV Epstem-Barr virus GVHD graft-versus-host disease IL-2 interleukin 2 LCL lymphoblastmd cell line

I mAb A few of the clones revealed augmented prohferation in the presence of CMV antigens and a few revealed cytolytic activity The above study suggests that (a) CD4 ÷ helper T cells may be primarily responsible for lmmunopathogenesls of skin manifestations during acute GVHD, (b) there is a mixed lymphoid chimerism in skin during acute GVHD, (c) HLA-DP may not be a factor contributing to the development of acute GVHD, (d) the peptide of the HLA groove or superantigen associated with HLA molecules may be the snmulatory antigen, and (e) CMV antigens appear to stimulate some of the skininfiltrating T lymphocytes Human Immunology 40, 68-76 (I994)

mAb MHC PBL PBS PLT TCR

monoclonal antibody malor hlstocompatlblhty complex peripheral blood lymphocyte phosphate-buffered saline primed lymphocyte testing T-cell receptor

INTRODUCTION From the Transplant Immunology Department (A N , T P , R F L D , M J S ), Immunology Laboratory (G 0 , M J S ) andBone Marrow Transplantatmn Clmlcal Research Department (M J S , R H C J W F ), Charles Sammons Cancer Center, Baylor Unwerstty Medtcal Center, Dallas, Texas U S A Address reprmt requests to Dr A Nlkaem, Transplant Immunology, Baylor Unwerslty Medical Center 3500 Gaston Avenue, Dallas, T X 75246, USA Recetved ( U ) September 1, 1993, acceptedNovember 9, 1993

68 0198-8859/94/$7 00

Acute graft-versus-host &sease ( G V H D ) remains a slgmficant problem m chnlcal bone marrow transplantation [ 1 - 3 ] The g e n e n c factors that provoke the development of acute G V H D are presumably major and m i n o r hlstoc o m p a t l b l h t y anngens [ 4 - 9 ] However, the m e c h a m s m whereby T lymphocytes become actlvated and infiltrate various organs, predominantly the skin, is not clear Several stu&es suggest the involvement of m i n o r h~stoHuman Immunology40, 68-76 (1994) © AmerJcanSocietyfor Hlstocompat~blhtyand Immunogenetlcs, 1994

Immunologic Monitoring of Skin Durmg Acute GVHD

compatlblhty antigens m the activation of T lymphocytes [10-12] However, the nature of these antigens remains unclear To study the pathogenic mechamsm of acute G V H D , we have focused on charactenzmg the skm-infiltratmg T lymphocytes Unrelated bone marrow transplants were selected to fac,htate the detection of malor or minor h i s t o c o m p a t l b l h t y antigens S k i infiltrating T lymphocytes were cloned by the hmlting dilution method Analysis of these clones suggests that T lymphocytes may recognize HLA-assoclated molecules, perhaps the peptlde of the HLA groove or superantlgen Prehmmary study revealed that cytomegalovirus (CMV) antigens contributed to ant~gemc stlmulatlon of some of the T-cell clones MATERIALS AND METHODS

Pattents Two bone marrow transplant recipients who received marrow from unrelated donors matched for HLA-A, -B, -C, -DR, and -DQ by serology and for HLA-DR and -DQ by ohgonucleotide genotyping were studied Both patients' marrow donors were mismatched for HLA-DP Both patients developed grade III acute G V H D , and skin biopsy specimens were obtamed usmg a 2-mm punch biopsy Patient 1 developed a skin rash 19 days after transplantation and a skin biopsy was performed on day 20 She is ahve 3 years after transplantation Patient 2 developed a skin rash 10 days after transplantation and a s k i biopsy was performed on day 12 Patient 2 died of hepatic failure secondary to G V H D 2 weeks after the biopsy was performed Both patients' CMV serology status prior to the transplant mdicated CMV infection The results of donor serology tests were negative

Cultures Skm-mfiltratmg cells from the biopsy specimens were expanded m culture accordmg to the method described prewously [13] The biopsy specimens were cut into small pieces and placed m RPMI medium supplemented with 10% human AB serum and 25% Interleukin 2 (IL-2) Half of the biopsy specimens were cultured in the presence of irradiated donor perlpheral blood lymphocytes (PBLs), and the other half were cultured m the presence of i r r a d i a t e d recipient PBLs Skinmfiltratmg T lymphocytes from the patients grew In culture, whereas those from controls (healthy mdlviduals) did not Two weeks after mltlal culture, the cells were cloned usmg the h m l t m g dilution assay [ 14], placmg 1 cell per well in 96-well mlcrotiter plates Clones were cultured m the presence of media supplemented with IL-2 and the same irradiated stimulator cells (1 × 104/well) When a large number of cells were obtamed, the contents of each well were transferred into two wells Cells were fed every 2 days with fresh media and every 7 days with feeder cells

69

Phenotype analysts of T-cell clones T cells (2 × 105) from those clones that grew to a large number (67 clones) were mcubated with 10 ~l of two-color conlugated CD4/CD8 monoclonal antibody (mAb) (Coulter clones, Coulter, FL, USA) or, 20 ~1 anti-T-cell receptor (TCR) mAbs (TCR-1, oLWT31 to detect TCR-o~ or antl-TCR-~ mAb to detect TCR-~/, Becton Dickmson, CA, USA) Cells were incubated for 20 minutes at room temperature, and those that were incubated with CD4/CD8 Coulter clones were washed with 1 ml phosphate-buffered saline (PBS) Cells mcubated with antl-TCR were washed twice wlth PBS and then incubated with 10 ~l of FITC-conlugated goat-anti-mouse lmmunoglobuhn (TAGO) for 30 mmutes at 4°C Controls were purified FITC-conlugated mouse IgG (Becton Dickinson) Ceils were fixed with 0 3 ml of 1% paraformaldehyde m PBS The percentage of FITC- and phycoerythrm-stamed cells was determmed usmg an Ortho cytofluorograf 2s

Okgonucleottde Genotypmgfor HLA-DP D N A from each clone was extracted and then amplified m the polymerase chain reaction with HLA-DP-specific primers (Table 1) by using methodology previously described [15] The resultant amplified D N A adhered to nylon membrane The membrane then was incubated with 32P-labeled single-strand ohgonucleotlde probes speclfic to HLA-DP polymorphic regions (Table 1) These probes detect 33 HLA-DP speclficltles The membrane filters were exposed to Kodak x-ray film (Rochester, NY, USA) and developed after 8- or 18-hour Intervals F U N C T I O N A L ASSAYS

Przmed lymphocytetestmg The primed lymphocyte testing (PLT) assay was performed on 109 clones as previously described [16] Clones were IL-2 starved for 3 days and feeder starved for 7 days prior to the PLT assay Clones were plated as responders at 1 × 104 cells/well In triphcate wells of 96-well round-bottom plates (Cornmg, NY, USA) Stimulators mcluded autologous, donor, and a panel of HLA-typed PBLs Irradiated stlmulator cells were plated at 1 X 105/well The assay was then incubated at 37°C, 5% CO 2 On day 2, 1 IxC1/well of [3H]thymldme was added The assay was harvested on day 3 by using a Tomtec Harvester 96 Mach II and counted in a Wallac 1205 Betaplate liquid scintillation counter The results were analyzed using the Rider calculations at the 75th percentile for positive and negative results and were represented as stimulator normalized value (SNV) The PLT assay was performed twice for all clones and three times for some of the clones

mAb mhtbttton The specific reactivity of skin-infiltrating T-lymphocyte clones m PLT was assayed with class-I and class-II-specific mAbs as described previously [16]

70

A Nlkaem et al

TABLE 1

Primers and probes wlth their sequences and amino acid correspondence Primers

Probe ID 1 2 3 4 5

6 7 8 9 10 11 12 13 14 15 16

PP4 PP5 PP6 PP7 PP8 PP9 PP10 PP12 PP13 PP 14 PP15 PP16 PP18 PP 19 PP20 PP21

DPBAMP-B 5' GCCGGCCCAAAGCCCTCACTC-3' DPBAMP-G5 5 'CCCCTCCCCGCAGAGAATTAC-3' Amino acid Sequences sequence CGGCCTGCTGCGGAGTAC CGGCTGATGAGGAGTAC CGGCCTGAGGCGGAGTAC ACATCCTGGAGGAGAAG ACATCCTGGAGGAGGAGC CGGCCTGATGAGGACTAC ACCTCCTGGAGGAGAGGC CTTTTCCAGGGACGGCAG GTGTACCAGTTACGGCAG GTGCACCAGTTACC~CAG GTGTACCAGGGAGGCAG ACCTCCTGGAGGAGAAGC CAGGATGTGCAGACACAA CAGGGTATGCAGACACAA CGGGCCCATGACCCTGCA CGAGGCCGTGACCCTGCA

Briefly, ~rradmted pretransplant recipient stimulator cells (1 × 103 cells/well) were incubated with either 1 100 dilution of w6/32 m A b (specific for the monomorphlc port,on of class I) or 1 10 dilution of L227 (specific for the monomorphlc portmn of class II) These dilutions of m A b were chosen based on a previous screenm g assay for the selection of o p t i m u m dilution to ensure a blocking effect Both mAbs were supernatants of tissue culture hybndomas (American Tissue Culture Collectlon, Rockville, MD, USA) Cultures were incubated at 37°C, 5% CO2 for 30 minutes T-cell clones were IL-2 starved for 3 days, feeder starved for 7 days prior to the assay, and were added to the stimulator cells at 2 × 104 cells/well The assay was Incubated at 37°C, 5% CO 2 On day 2, 1 >Cdwell of [3H]thymldlne was added, on day 3, the assay was harvested and counted as described above Percent mhlb*t,on of proliferation in the presence of either class-I- or Class-II-speclfic mAbs was determined by the following formula % mhlbltzon = 100

__ ( cp__m_wlth mA.__bb 100 ~ \ c p m wlt-~out m A b ×

/

RPAAEY RPDEEY RPEAEY ILEEK ILEEE RPDEDY LLEER LFQGRQ VYQLRQ VHQLRQ VYQGRQ RMCRHN RMCRHN RVCRHN GPMTLQ EAVTLQ

Pos~t~on 53-59 53-59 53-59 64-70 64-70 53-59 64-70 8-14 8-14 8-14 8-14 74-80 74-80 74-80 84-90 84-90

(supernatant of tissue culture not infected with CMV) were plated at both 1 10 and 1 20 dilutions Cultures were incubated for 30 minutes at 37°C, 5% CO 2 The clones, which were IL-2 starved for 3 days and feeder starved for 7 days, were then added to the cultures The assay was incubated at 37°C, 5% CO 0, for 3 days On day 2, 1 ~C1/well of [3H]thymldme was added On day 3, the assay was harvested and counted as previously described in the PLT section above

Cell-medtated lympholysls The clones' cytotoxlclty was determined against the E p s t e l n - B a r r Virus (EBV)transformed lymphoblastold cell llne (LCL) of the donor, recipient, third-party HLA-mlsmatched, and K562 cell hne LCLs were treated wlth 100 [xCI of C h r o m i u m 5 1 (~lCr) per 1 × 106 cells for 90 minutes at 37°C, 5% CO 2 The cells were washed twice with PBS and plated m 96-well V - b o t t o m plates (Nunc, Denmark) The 51Cr_labeled LCLs were Incubated with T-cell clones at a ratio of 25 1 The assay was incubated in 5% colostrumfree bovine s e r u m - R P M I (Irvme, CA, USA) at 37°C, 5% CO 2, for 4 hours The supernatant of the wells was then harvested Radioactivity was measured in a Wallac 1205 Betaplate hquld scintillation counter Percent of cytotoxlclty was measured by the following formula

E-S CMV assay CMV antigens were used to stimulate T-cell clones Donor, recipient, and third-party cells completely HLA mismatched with donor or recipient were used as antigen-presenting cells (APCs) Irradiated APCs were plated at 2 × 104 cells/well CMV (Whltaker Blo Products, Walkersvllle, MD, USA) or control reagents

% cytotoxlclty - T - S )< 100 E is the experimental 51Cr release, S ~s the spontaneous 1Cr release from target cells alone, and T is the total release by incubating the target cells wlth 0 1 ml of 10% sodium dodecyl sulfate

Immunologic Momtormg of Skin During Acute GVHD

RESULTS

Clones A total of 120 T-cell clones were generated from skin b,opsles of two patients who rece,ved marrow from an unrelated donor Patient 1 (HLA type A3,-, B8,35, DRB1-0101,DRB1-0301, DRB3-0101, DQ2,5, DPB1-0401) received bone marrow from a donor matched for A, B, and DR, but mismatched for HLADP (DPB1-0202,DPB1-0402) Patient 2 (HLA type A3,23, B18,65, DRB1-0102,DRBl-1501, DRB50101, DQ1,2, DPB1-0201,DPBl-17) received marrow transplant from a donor matched for A, B, and DR and m,smatched for HLA-DP (DPB1-0402, DPB1-0201) Thirty clones were generated from patient 1, and 90 clones were generated from pat*ent 2 Clones 1-99 were generated m the presence of rec,p~ent's stimulator cells and ~ 100 m the presence of donor's stimulator cells

Phenotypes of the clones Phenotypes of the clones were determined by staining them with mAbs to CD4, CD8, and the {3 and ~/chain of TCRs A total of 65 (97%) of 67 tested clones were CD4 +, and two (3%) were CD8 ÷ Of 17 clones tested, 15 expressed o~/~3 TCR and two expressed neither ~x/~3 nor ~//~ TCR (Table 2) These results indicate that the malonty of the infiltrating T lymphocytes m the skin of GVHD pat*ents are CD4 +, o~/~3 TCR positive

HLA-DP genotype of the clones To determine the source of T-cell clones, the,r HLA-DP genotypes were determined by ohgonucleot~de genotypmg A total of 38 (64%) of 59 clones expressed the donor's HLA-DP, whereas 21 (36%) expressed the rectp,ent's HLA-DP, indicating the presence of m~xed lympho,d chlmer,sm m skin during the cutaneous GVHD Functional Assays

PLT assay PLT assays using 25 stimulators, including donor and recipient cells, were performed twlce and, for some clones, three ttmes E~ghteen patterns were obtained by 19 clones from patient 1 (data not shown), and 53 patterns were obtained by 88 clones from patient 2 (F,g 1) The closest pattern of react,vlty :s shown m one TABLE 2

Phenotypes of clones

mAb

Total

Patient 1

Patient 2

No positive (%)

CD4 CD8 TCR~x/[3 TCR"//8

67 67 17 17

14 0 ND ~ ND

51 2 15 0

65 (97) 2 (3) 15 (88) 0

ND, not done

71

column No spec,fic pattern of react*wty was obtained by clones that were expanded with recipient or donor stimulator cells For instance, clone 6, which was expanded m the presence of rec~p,ent's cells, had a slm~lar pattern of reactlwty to clone 120, which was grown m the presence of donor cells (column 4) Similarly, clones 8 and 124 had the same pattern of reactlwty (column 5) The patterns of reaet~vlty showed no restriction to donor or recipient HLA, including HLA-DP Stimulator cells BC 136 and BC 185 stimulated the ma 1orlty of the clones, although they did not share HLA-DP with the recipient or donor Stimulator cells BC194, BC146, or BC148, which shared HLA-DP with the reclplent, and stimulator cells BC109 and BC135, which shared HLA-DP with the donor, stimulated a few of the clones These results present the lack of malor hlstocompatlblhty complex (MHC) restriction, including the mismatched HLA-DP

mAb mhlbttlon A total of 39 randomly selected clones from both patients were tested m the mAb mhlbltlon assay, using w6/32 and L227 Results showed that prohferatlon of clones 160, 6, 153, 101, 145, and 52 was strongly mhlblted by class I w6/32 mAb (Fig 2), whereas prohferatlon of clones 133, 161, 131, 12, 174, 167, 5, 103, and 16 was strongly inhibited by class II L227 mAb (Fig 3) Clones 104, 15, 107, 115, 128, 141, 46, 127, and 130 were weakly mhlblted by L227 mAb All of the above clones, which were mhlblted by class II mAb, were CD4 ÷ Of those clones that were inhibited by class I mAb, only clones 160 and 52 were tested for phenotype and, interestingly, both were CD4 ÷ The two CD8 ÷ clones were not tested m the mAb inhibitor assay Prohferatlon of 13 clones was not affected by either of the above mAbs (data not shown) These results indicate the lack of MHC restriction by PLT assay, but suggest that the antigen is assocmted with the HLA molecules as mAbs inhibited T-cellantigen interactions Such antigens may be the peptldes of the HLA groove or superantlgen bound to the HLA molecules Prohferatmn of clones to CMV antzgen Figure 4 shows the results of T-cell clone prohferatlon to CMV antigen Of 17 clones tested, six (7, 15, 104, 113, 115, and 175) responded m an augmented fashion to CMV antigen when they were cultured m the presence of donor PBL The response of three of the above clones (104, 113, and 115) to CMV was augmented m the presence of the reclplent's stimulator PBLs Clone 113 was CD8 ÷ and clones 7, 15, 104, and 175 were CD4 ÷ PLT response of clones 15, 104, and 115 was inhibited by class II mAb Clone 7 was not inhibited by either of the mAbs Clones 113 and 175 were not tested m the mAb inhibitor assay, because we lost the clones Clones 113 and

72

A NlkaeIn et al

11 1 1 1 HLA

Stimulator A DONOR 3 23 PATIENT 2 3 23 BCI~ 224 BC144 6C154 BC1~6 BC126 RC194 BC138 BC201 BC148 BC106 BC185 8C109

3 11 13 3 28 1 28

12 2 25 328 22 //

B

DR

DQ

65 18 6235 3.

12 13 13 13

7 42 8

23

1 2 402,17 12 402401 12 401 112 ND 401 1 2 401 402

3. 7:

78 44 60 738 1444 27:76

22

3

2 30

6 12o

5744,o 78

33 310

62:/

4 447

1444 18 44 45 60 62 81 1381

13 1 113

201 201 17 201 40132

7

4o2

22 1

.o

11111111111111111111111111111111111111

11111111111111111111111111111

m •

i•

•

• II~

ll

•

•

I

•l

•lib

ii•

Ill

•

•

m

•

m

[,

•m

ml

Iron

m

m

I

I

•

-

•

I001

• ~l

I

322011001 401402 23 1101 7 12 2 7 NO 71113 172 301 201 117 23 NO

i•

I

|

28 1~ 401. 23 12 ND I l l l •

BC202

BC167 BC146 231 2 31 BC135 1131 31 BCl17 232 ~5;9 BC197 25;! 80199 BC200 2 32 3l BC193 230 2

DP

281,1ool,2Ol

24 26 26 211

11

~77~2~2~1111~11~1112244~113334~55555~8348~e87777~8~11341112222223333444445~e 234513~4341423~3~1259823237~572372~125361236846789~45~7678~1~4~915e7~9148~1345e4~75

m

•

I i

m

m m

m

FIGURE 1 Patterns of PLT response of patient 2 clones clones tested m prohferatlon assays against 25 PBL stimulator cells Clones are indicated across the top of the figure Clones with numbers < 100 were expanded m the presence of reczp~ent's stimulator PBL and clones with numbers />100 were expanded in the presence of donor stimulator cells Bla~k squares represent strong poslt~ve st,mulatlon (stamulator normalized value, SNV t> 100%) Gray shaded squares indicate weak response (SNV 759~-100%)

165 were donor origin and the rest of the clones, with the exception of clones 2, 7, and 138, are recipient origin The origin of the latter clones is unknown Augmented growth of the recipient origin clones w~th donor cells and CMV antigens suggest the CMV-related response by these clones None of the clones proliferated in the presence or absence of CMV when third-party HLAmismatched stimulator cells were used

FIGURE 2 Monoclonal antibody Inhibition of clones from patients 1 and 2 performed using w6/32 class I HLA mAb Inhlbltlon >50% was considered as a strong blocking effect

Cytotoxtcassay O f 40 clones tested, only clone 138 from patient 1 demonstrated cytotoxlclty toward the recipient, and one (clone 81) demonstrated toxicity toward the donor EBV-transformed cell line Clones 82 and 119 from patient 1 weakly lysed the K562 cell line (Fig 5) No cytolytlc activity was observed by patient 2 clones (all cytolytic activity < 2 0 % was considered negative) The lack of cytotoxic function by the malority of the clones suggests that the cytotoxIc T cells may not be primarily involved in the development of immunopathologic changes in skin G V H D

DISCUSSION The findings in this study indicate that (a) T cells milltratmg the skin during acute G V H D were of both donor and recipient origin, (b) the malorlty of the infiltrating

FIGURE 3 Monoclonal antibody inhibition of clones from patients 1 and 2 performed using L227 class II HLA mAb Inhibition >50% was considered as a strong blocking effect % INHIBITION

%INHIBITION

120

120 100f80

100

100

100___100_____

80 66 60

_5Z

__

60

45 40

3023

20 0

40

f

li)

14

104 18133181131 12 174107167116 8 10316 128141160 6 1631011484812713082129 i Jl I m

PT 2 CLONES

PT 1 CLONES

20 0

104 18138161131 12 174107167116 6 108 16 128141160 6 1831011484812718062129 [

II

m

PT 2 CLONE8

P T 1. C L O N E 8

I

Immunologic Monitoring of Skin During Acute GVHD

12o

STIMULATOR NORMALIZE VALUES ReCll:)lent s t i m u l a t o r

cells

lOO

80

60

40

20

0 2

7

15

16 104 118 115 123 131 138 141 158 159 161 165 167 175

STIMULATOR NORMALIZE VALUES

I

Donor stimulator cells

100 •

80

60

40

20

0 2

120

7

15

16 104 113 115 123 131 138 141 158 159 161 166 167 175

STIMULATOR NORMALIZE VALUES Unrelated stimulator cells

100 80 60

40

0 2

7

15 16 104 113 115 123 131 138 141 158 169 161 165 167 175 ~No

CMV

mBIcMv

FIGURE 4 PLT response of clones from patient 2 to CMV Proliferation of clones measured in the presence or absence of CMV antigens and recipient stimulator cells (top), donor stimulator cells (middle), or third-party HLA-mIsmatched stimulator cells (bottom) Results were calculated as stimulator normalized values Soltd bars represent proliferation of clones in the absence of CMV antigens (control was the supernate of cell culture without CMV) Hatched bars represent prohferatlon of clones in the presence of CMV T cells were CD4 + and expressed ot/[3 TCR, and (c) the H L A - D P did not appear to play a role in the development of antihost or ant~donor alloreactive T-cell response Instead, It appeared that activated T cells m acute G V H D les~ons may have recogmzed the peptldes of

73

the HLA groove or the superantlgen bound to the HLA molecules In bone marrow transplantation, the severity of acute G V H D correlates with the degree of HLA disparity [6, 17-20] In this study, we attempted to determine the antigen(s) that may have contributed to the development of acute G V H D in two recipients of unrelated bone marrow transplants Using m vitro funct,onal assays to measure prohferatlon against a panel of stimulators that shared various HLA antigens with the donor or the recipient, no specific HLA-restrlcted T lymphocytes, even restricted to the mismatched HLA-DP, were found This suggests that H L A - D P may not be involved in the development of acute cutaneous G V H D , which is in contrast to the observations by others [18, 19, 21] who have shown that H L A - D P mismatches are associated with a higher incidence of acute G V H D Two of the three stud*es were performed with a small group of patients, however, and the third was performed with only one patlent [21] The clonahty of our clones have not been estabhshed by any other methods Blocking of T-cell prohferatlon by either antl-HLA class I or class II mAbs suggests that T cells may recograze HLA-assoc,ated molecules, perhaps the peptldes of the groove [22, 23] The second possibility IS the recogmtlon of superantlgens by these T cells Superantlgens are suggested as binding to M H C products [ 2 4 - 2 6 ] The third possibility is that T cells recogmze non-HLA (minor HLA antigens), as reported by others [ 1 0 - 1 2 , 27], however, the minor HLA antigens reported by these Investigators might be the same as the peptides of the HLA groove The fourth posslbllaty is the recogmtlon of tissue antigens that may cross-react with HLA molecules Binding of HLA mAbs with those antigens therefore blocks the PLT reactivity The existence of such antigen(s), however, has not yet been reported The lack of blocking of 13 clones by either of the mAbs may be caused by nonspeclfic attraction of these clones into the skin Whether the CD4 + T cells are the primary source of cytokines that attract nonspec~fic T cells to the skin is not clear An incidental finding was that two of the clones that were inhibited by class I m A b were CD4 + Thls supports our hypothesis of peptlde or superant~gen recogmt,on, as these clones may recognize the peptldes of the class I HLA groove or the superantlgens, independent of the HLA molecules Thls needs to be clarified by further study, as it may describe a new process of antigen recogmtlon by T lymphocytes Peptldes of the HLA groove could be exogenous or endogenous In the former case, it may be possible to define the peptldes by using a variety of antigens that often are components of infectious agents m these pat,ents Super antigens can also be the components of

74

A Nlkaexn et al

PATIENT 1 CLONES 40 46 52 82 112 119 130 137 138 41

L I

i_____

'

l

m.., r~iA

5O 51 68

75

81 105 106 127 128 132

RECIPIENT

t

DONOR [~B UNRELATED MISMATCH

t

mlB K562

0

I

I

I

I

I

10

20

3O

40

50

60

% CYTOTOXICITY PATIENT 2 CLONES

2 7 11 15 16 101 104 105 107 m 111 125 136 138 148 k,m 148 149 159 160 165 166

7__ RECIPIENT

Bm DONOR

I

~

I

I

I

I

0

10

20

30

40

50

% CYTOTOXIClTY infectious agents W e examined CMV, as It xs a common cause of mfect~on in bone marrow transplant recipients The augmented proliferation response of some of the clones in the presence of CMV antigens suggests that the clones are either primed to CMV or to ant,gens with homologous sequences Prohferatlon of three of these clones was inhibited by class II mAb, suggesting the recognition of CMV ant,gen by these clones was restricted to class II HLA antigen The above observation is supported by other investigators who showed that wral antigens are recognized in association with class II HLA molecules [23] Two of these five clones (15 and 104)

60

FIGURE 5 Cytotoxiclty of clones from patient 1 assayed using recipient, donor unrelated HLA mismatches, LCLs, or K562 cell hne as targets Cytotoxacity of clones from pat,ent 2 was measured using only rec,pient or donor LCL as targets Results are expressed as percent of cytotox,clty compared with maximum 51Cr release

were recipient-originating T cells and two (115 and 113) were donor-originating T cells, indicating that perhaps both T-cell populations responded to CMV antigens Only one reclplent-specific cytotoxic T-cell clone was identified m our study It is posslble that most cytotoxic T cells did not grow m m vitro culture The finding of a few clones with natural killer activity may support the findings of others [ 2 8 - 3 0 ] or may be due to the expansion of these cells in IL-2, as we have seen it in liver transplant recipients [31] The other unique finding of this study was the presence of lymphoid chimerlsm in skin during acute cuta-

Immunologic Monitoring of Skin During Acute GVHD

neous G V H D It is not clear, however, whether the rec,plent cells were the skin resldenual T cells or were attracted to the skin from the periphery W e could not find any umque phenotypes or functional patterns with e~ther the donor- or rec,p~ent-denved clones S~mllar to findings in sohd organ transplant recipients [32, 33], however, interact,on of donor and recip,ent T cells in skin may contribute to the induct,on of tolerance m bone marrow transplant recipients Because th,s is a hm,ted investigation m two patients, further study of the function of the infiltrating cells m regard to their interaction with each other and with host keratmocytes may prowde a better understanding of the pathogenes,s of acute G V H D or the development of tolerance

ACKNOWLEDGMENTS

This work was supported by funding from Baylor University Medical Center Foundation The secretarial assistance of JoAnn Noe' and Rebecca Green IS appreciated

REFERENCES 1 Grunnet N, Krastensen T Cell mediated lympholysls (CML) in man HLA (LA and FOUR) identical siblings versus Identical unrelated individuals In HIstocompat,bihty Testing, 1975, p 862 2 Thomas ED, Storb R, Chft RA, et al Bone marrow transplantation N Engl J Med 292 895, 1975 3 Storb R, Thomas ED Allogenelc bone~marrow transplantation Immunol Rev 71 77, 1983 4 Mathe G, Prltchard LL, Halle-Pannenko O Mismatching for minor hlstocompatibihty antigens m bone marrow transplantation consequences for the development and control of severe graft-versus-host disease Transplant Proc l l 235, 1979 5 Korngold R, Sprent J Lethal GVHD across minor hlstocompatibihty barriers nature of the effector cells and role of the H-2 complex Immunol Rev 71 5, 1983 6 Galewskl JL, Ho WG, Feig SA, Hunt L, Kaufman N, Champhn RE Bone marrow transplantation using unrelated donors for patients with advanced leukemia or bone marrow failure Transplantation 50 244, 1990 7 Anasettl C, Amos D, Beatty PG, et al Effect of HLA compatiblhty on engraftment of bone marrow transplants in patients with leukemia or lymphoma N Engl J Med 320 197, 1989 8 Cecka M, Hopfield J, Hunt L, Champlln R, Terasaki PJ, Gajewskl JL HLA-DP lncompatlblhtles and severe graftversus-host disease in unrelated bone marrow transplants Transplantation 52 374, 1991

75

MHC-restrlcted cytotoxic T lymphocytes during graftversus-host disease Nature 302 159, 1983 11 Goulmy E, Termijtelen A, Bradley BA, Van Rood JJ Y-antigen killing by T cells of women is restricted by HLA Nature 266 544, 1977 12 Remsmoen NL, Kersey JH, Bach FH Detection of HLA restricted and minor hlstocompatIblhty antigen(s) reactive cells from skin GVHD lesions Hum Immunol 11 249, 1984 13 Nlkaeln A, Phillips C, Gilbert SC, et al Characterization of skin infiltrating lymphocytes in patients with psoriasis J Invest Dermatol 96 3, 1991 14 Irle C, Beatly PG, MIchelson E, Thomas ED, HansenJA Alloreactlve T cell responses between HLA-ldentmcal slbhngs detection of anti-minor histocompatibihty T cell clones induced m vlvo Transplantation 40 329, 1985 15 Salkl RK, Gelfand DH, Stoffel S, et al Primed-directed enzymatic amplification of DNA with a thermostable DNA polymerase Science 239 487, 1988 16 Nikaem A, Boyer B, McQueen K, Ryan D Development of a class I specific helper-inducer T cell clone Hum Immunol 19 205, 1987 17 Anasett, C, Amos D, Beatty P, et al Effect of HLA compatlblhty on engraftment of bone marrow transplants in patients with leukemia or lymphoma N Engl J Med 320 197, 1989 18 AI-Daccak R, Lolseau P, Rablan C, et al HLA-DR, DQ, and/or DP genotypic mismatches between recipientdonor pairs in unrelated bone marrow transplantation and transplant chnical outcome Transplantation 50 960, 1990 19 Odum N, Platz P, Jakobsen BK, et al HLA-DP and bone marrow transplantation DP Incompatibility and severe acute graft versus host disease Tissue Ant*gens 30 213, 1987 20 Fleischhauer K, Kernan NA, O'ReIlly RJ, DuPont B, Yang SY Bone marrow-allograft relection by T lymphocytes recognizing a single amino acid difference in HLAB44 N Engl J Med 323 1818, 1990 21 Duquesnoy RJ, Zeevi A, Marrari M, Hackbarth S, Camita B Bone marrow transplantation for severe aplastlc anemia using a phenotyplcally HLA-identical SB-compatIble unrelated donor Transplantation 35 566, 1987 22

Blorkman PJ, Saper MA, Samraoui B, Bennett WS, Stromingen JL, Wiley DC Structure of the human class I histocompatlbihty antigen, HLA-A2 Nature 329 506, 1987

9 FerraraJLM, Deeg HH Graft versus host disease N Engl J Med 324 667, 1991

23 Eckels DD, Sell TW, Long EO, Sekaly RP Presentation of influenza hemagglutimn peptide in the presence of limited allo stimulation by HLA-DR1 transfected human fibroblasts Hum Immunol 21 173, 1988

10 Goulmy E, Gratama JW, Blokland E, Swain FE, Van Rood JJ A minor transplantation antigen detected by

24 Marrack P, Kappler J The staphylococcal enterotoxins and their relatives Science 248 705, 1990

76

A Nlkaem et al

25 Johnson HM, RussellJK, Pontzer CH Super-antigens in human diseases Immunol Today 13 92, 1992

function in the spontaneously diabetic BB/W rat J Immunol 137 1862, 1986

26 Pontzer CH, Russell JR, Johnson HM Structural basis for differential binding of staphylococcal enterotoxm A and toxic shock syndrome toxin to class II malor histocompatlblhty molecules Proc Natl Acad Scl USA 88 125, 1991

30 Rhoades JL, Clbull ML, Thompson JS, et al Role of natural killer cells in the pathogenesls of human acute graft-versus-host-disease Transplantation 56 113, 1993

27 Mackelson EM, Beatly PG, Storb R, Hansen JA Immune responses m an untransfused patient with aplastic anemia analysis of cytolytic and prohferatlve T cell clones Hum Immunol 10 189, 1984 28 Woda BA, Like AA, Padden C, McFadden ML Deftciency of phenotyplc cytotoxlc-suppressor T lymphocytes m the BB/W rat J Immunol 136 856, 1986

29 Woda BA, Blron CA Natural killer cell number and

31 Nlkaem A, Colhns RH, Khntmalm G, Poole T, FayJW, Stone MJ Immunologic momtormg of graft-versus-host disease following orthotoplc liver transplantation Transplantation 57 637, 1994 32 Starzl TE, Demetrl AJ, Murase N, Ilstad S, Rlcord, C, Trucco M Cell m,gratlon, chlmerlsm and graft acceptance Lancet 339 1579, 1992 33 Starzl TE, Demetrls AJ, Trucco M, et al Chimerism after liver transplantation for type IV glycogen storage disease and type I Gaucher's disease N Engl J Med 328 745, 1993