REPORTS

Immunophenotyping of the Cutaneous Infiltrate and of the Mononuclear Cells in the Peripheral Blood in Patients With Atopic DerUlatitis

Rosemary Lever, M.D., M.R.C.P., Marlyn Turbitt, M.Sc., Ph.D., Andrew Sanderson, M.l. BioI, and Rona MacKie, M.D., F.R.C.P., FRC Path Department of Dermatology , University of Glasgow (RL, MT, RM) , Glasgow, U.K ., and Department of Zoology, University of Edinburgh (AS), Edinburgh, U.K.

Fourteen adult patients with chronic atopic dermatitis and active skin lesions had a skin biopsy and venous blood sample taken on the same day. Absolute numbers of circulating lymphocytes were normal in all patients. Fluo­rescence-activated cell sorter (FACS) analysis revealed nor­mal numbers of total T lymphocytes and T -helper and T­suppressor subsets (helper: suppressor ratio, 2: 1) in the atopic patients' peripheral blood, but an increase in circulat­ing B lymphocytes and in HLA-D-related antigen-bearing

Multiple immunologic abnormalities have been re­ported in patients with atopic dermatitis . These have been well summarized by Hanifin [1 ,2] . Some of these abnormalities are seen only tran­siently when the skin is acutely inflamed. One

of the more commonly reported abnormalities is an alteration in the numbers of circulating T lymphocytes. A deficiency in both the total T cells, as measured by E-rosette formation [3-5], and numbers of T -suppressor cells, as measured by either the Fc re­ceptor [5-8] or by using a monoclonal marker such as OKT8, has been reported (7,9-13]. However, not all studies have re­corded similar changes, and some were unable to demonstrate any abnormali ty in either T -suppressor cell numbers or function [14-16].

Histologic examination of the skin of patients with atopic der­matitis (AD) has revealed a dermal lymphocytic infiltrate [17,18] that was shown to be composed predominantly ofT lymphocytes [19]. This infiltrate was later phenotypically identified as T -helper lymphocytes [20,21] . It is difficlllt to interpret the significance of this cutaneous infiltrate without knowledge of the absolute I1lIlTI­

bers and relative proportions of T-cell subsets in a concnrrent blood sample, and we therefore decided to examine the numbers of T lymphocytes and subsets in both the skin and peripheral blood in the same patients on the same day.

Manuscript received May 5, 1986; accepted for publication October 22, 1986.

This study received financial support from the National Eczema Society and from the Greater Glasgow Health Board Research Support Group .

Reprints not ava ilable. Correspondence: Dr. Rosemary Lever, Department of Dermatology,

Western Infirmary, Glasgow, U.K. Gl1 6N T Abbreviations:

AD: atopic dermatitis FACS: fluorescence-activated cell sorter MNC: mononuclear cells NK: natural ki ller

cells. The skin biopsy showed a dermal infil tra te of predomi­nantlyT -helperlymphocytes (helper: suppressor ratio, 7: 1). These cells showed strong HLA-DR plasma membrane staining. There was no HLA-DR staining in the membranes of epidermal keratinocytes. Using a monoclonal antihuman IgE, positive staining was observed in the dermis, though none was identified in the epidermis. The dermal anti-lgE staining was concentrated around clusters of T lympho­cytes. ] Invest Dermatol 89:4-7, 1987

An elevated total serum IgE is seen in 80% of patients with AD, and positive prick tests , presumably due . to tissue-bound IgE, are also found in a similar number of patients. Because of this, we looked for the presence of IgE staining in the skin and subsequently used double-staining immunoperoxidase techniques to determine its association with cells in the dermal 'infiltrate .

MATERIALS AND METHODS

Fourteen adults (6 men, 8 women), aged 18-47 years (mean 27.2 years) gave their informed consent. All patients had classic chron­ically relapsing AD w ith recurrent severe flares. To enable the necessary coordination of the project, patients were generally studied in a chronicall y active state requiring moderate-strength topical steroids such as c1obetasone butyrate (0.05%) rather than in an acute flare, with 1 exception. Other diagnostic criteria in­cluded: onset of dermatitis when the patient was under 5 years old (13 patients, 93%); personal history of other atopic disease (13 patients, 93%); a positive fam ily history of atopy (10 patients, 71 %). In addition, 12 patients had either an elevated IgE, positive RAST tests, or positive prick tests . The study was approved by the hospital ethi cs committee.

An elliptical skin biopsy was taken from the most active area, excluding the face, and a venous blood sample withdrawn from all patients between 9 and 11 A.M. in order to minimize any variation in the lymphocyte subsets due to the circadian rhythm.

Four milliliters of the venous sample was taken into a seques­trene bottle for a standard full blood count by Coulter analysis , and the remainder placed in heparinized containers (Evans bottles, Speke, Liverpool, U .K,) . A similar blood sample was taken from age- and sex-matched controls with no personal or family history of atopy. The mononuclear cells from both patients and controls were sepa rated using a Ficoll gradient (Lymphoprep, Nyegaard, Oslo, N orway) and were centrifuged at room temperature for 30 min at 400 g. The mononuclear cells (MN C) were then harvested, washed 3 times in RPMI (Gibco, Paisley, Scotland), resuspended in a known volume, and counted. The yield ofMNC was checked

0022-202X/87/S03.50 Copyright © 1987 by The Society for Investigative Dermatology, Inc.

4

VOL. 89. NO. I JULY 1987

Table I. U se of Monoclonal Antibodies

Antibody

Antigen

Pan T cc ll

Helper T ce ll

Skin

Used ill the S/l/dy Leu-4 UC HTI OKT43

Suppressor-cytotoxic T cell OKTS

Regulatory T ce ll B ce ll

HLA-DR antigcn

Natural killer cell

Leu-8 Lcu-1 2

13231

Leu-lib

Used EXc/IISiIJe/y ill the Skill Nal 34"

Pcriphcral Blood

Lcu-4 UC HTI O KT4 Leu-3a OKT8 UCHT4 Leu-8 Lcu-12 B2-Coulter O Kb b 231 Lcu-ll a Leu-lib Leu-7

Langerh ans ce lls 19E lnterleukin 2 (IL-2)

receptor Transfcrrin receptor

Anti-lgE (New England Nucl~a r)

Anti human IL-2 (Bccton-Dickinson)

O KTY

Most of th e anti bod ies were obtained commerciall y: O KT series-O rrho. High Wycombe. Bucks. U.K .. Leu series-Becton-Dickinson. Lab lm pex. Middlesex. U.K.; UC HT series-Unipath. Bedford. U.K.; T he Ia 23 1 was kind ly provided by Dr. V. vO n Heynin gen. Edinblll'gh. U . K.

"Nal 34-T his antibod y stains identica l numbers of Lange rh ans cells as OKT6 and Leu-6 1221.

on 22 specimens and was found to vary from 46-65. 9% (mean 53.5% ) for th e atopic patients and 33.3- 64% (mea n 51.3%) fo r the controls. Mononuclea r cells (0.5 x 10(,) were then in cubated w ith 3 JLl of each of th e monoclonal antibodies (Tab le I) [22] at 4°C for 30 min: T he cells were then washed twice and incubated w ith F(abh for a furth er 30 min at 4°C. The MN C were then rewashed twice and res uspended in 1 % para formaldehyde. The cells were counted on a fluorescence-activated cell sorter (FACS) 4 autoanalyser (Becton-Dickinson) . All tes ts were done in du­plicate .

The skin biopsy was divided into two. Half was placed into 5% formol sa line, processed in the usual way , and stained wi th H & E for ro utin e histo logic exa mination . T he other half was snap-frozen for immunocytochemistry and sections cut at 5 JLm . These were stained by the indirect immuno pero xidase technique using the monoclonal antibodies shown in T able I. Leu-8 was used to stain the reg ul atory T cell [23] and Leu-11 and Leu-7 for the natural killer (NK) cells [24,25] . Additional antibodies were used exclusively in the skin (see Tab le J) . A doubl e-staining im­munoperoxidase techniqu e was also used in which the first an­tibody was labeled in the usual way using the indirect peroxidase method , th e section was then relabeled w ith a second an tibody fo llowed by an alkaline phos phatase co mplex (APAAP-Dako) revealed by incubation w ith naptho l AS-MX (Sig ma) and Fas t Blue BB salts [26].

Statistical analysis was by un pa ired Student's I-tes t.

Table II. Full B lood Count-Coulter Counter Res ults

Hemoglobin (g/d l) White cell count (x 109)/L % Lymphocytes Total lymphocytes (x 10")/L Platelets ( x IO")/ L

Normal (n = 16)

13.S :t 0.3 6.3 :t 0.5

32. 1 ± 2.2 2.0 :t 0. 1 272 :t 16.3

Atopic Dermatitis (n = 14)

13. S ± 0.4 6.5 :t 0.5

29. 1 ± 1.7 1. 9 :t 0.2

310 ± 23.2

IMMUN O PH ENOTYPING IN DERMATO LOGY 5

Table III. Fluo rescence-A ctivated Cell Sorter (FACS) Analys is Results

Lymphocytes

Pan T ce ll T helper T supprcssor B cell HLA-DR antigcn Natural ki ller-Lcu-7

Normal

66.4 ± 2.4( 16) 43 .4 :t 2.5 (17) 19.7 ± 0.9 (17) 3.9 :t 0.4 (12) 6.3 :t 0.7 (17) 9.S :t 2.0 (7)

AtOpi c Dermatitis

66 .5 ± 3.3 (13) 43.6 ± 2.9 (14) 20.6:t 1.2 (14)

-Leu-II Regulatory T cell (Lcu-S)

I J.l ± 1. 0 (16) 54.5 :t 3.7 (12)

6.3 ± 0.7 (12)" 9.9 ± 0.9 (14)1.

10.0 ± 1. 5 (S) 10.S ± 2.0 (13) 54.4 ± 2.6 (9)

Ites ul,s give th e percentage of cells labeled with each anti body. Numbers in parentheses rcpresellt the number of p:lticllts/ coll tro ls in c:lch g ro up .

"1' < (l.O I. /'1' < O.DO ! .

RESULTS

T o tal serum IgE levels were elevated in 12 of 14 patients with AD (> 1000 lU / ml in 7 and 550- 950 lU / ml in 5). All contro l subj ects had total serum IgE levels in the no rm al range, and 14 of the 15 had values under 30 1U / ml. The results of the Coulter ana lys is (Table II) showed no abnormality in the peripheral blood of A D patients by comparison w ith age- and sex-matched control subj ects. In particular, there was no evidence of a to tal lymp ho­penia in the patients with AD.

T he results o f the FACS analys is are shown in Table 1I1. There was no difference in to tal T-lymphocyte numbers, in absolute numbers or ratio of helper and suppressor T cells (helper : suppressor ratio, 2: I ) or in reg ul ato ry T cells (Leu-8) between the patients w ith A D and th e no rmal controls. In the patients w ith AD there w as, however, a statistica ll y signifi ca nt increase in the percentage numbers of circul ating B cells (AD patients 6.3%, controls 3 .9%; p = 0.01) and in the numbers of HLA-DR antigen-bearin g cells (AD patients 9.9% , controls 6.3%; p < 0.001) . T he numbers of HLA-DR antigen-bea rin g cells in bo th patients and contro ls were hi gher than numbers of cells stained by B l or Leu-1 2, ind ica ting th at not all were B lymphocytes. When the antibod y Leu-7 was used as a m arker of NK cells, the num bers of th ese cells appeared to be red uced in patients compared with the controls but this difference did not reach statisti cal signifi ca nce. In contras t, when Leu-II was used , no difference was seen between numbers of circulating NK cells in patien ts with AD and the controls.

T he results of the immunoperoxidase sta ining of th e skjn are summarized in T able IV. T he dermal in ftl trate was composed predominan tl y of T -helper cells (Fig 1) and these w ere strongly HLA-DR anti gen-positi ve. Relatively few T suppressor-cyto­toxic cell s were present in the in fi ltrate and the helper : suppresso r ratio was 7 : 1. In contrast to a number of o th er chroni c inflam­m atory skin diseases w ith a lymphoid infiltrate such as allerg ic

Table IV. II11Il1l1n o pero xidase Staining in the Sk in

Antibody

Pan T Helper Suppressor Regulatory-Leu-8 T ransfcrr in-O KT9 HLA-DR-dermis

B ce lls NK ce lls IgE

-epidermis

Interleukin 2 recepto r Langcrhans cells

77.7 6S.5 10.4 21. 0 3S.0 S7.0

Mean

No positive keratinocytcs Occasional positi ve onl y Occasional positivc onl y 13/ 14 patients positi ve dermal stainin g S/9 paticnts few positive cells 36.5

Range

50-90 50-80 5-20 5-40 2-75

50-100

25-58

Ly mph ocytes arc expressed :lS :l pero:llt:lge of rhe dcrll1:l 1 infilt r:lrc. L:lIl gc rh:lll s cells arc counted over 200 basa l cel ls.

6 LEVER ET AL

Figure 1. Double-sta ining irnmulloperox idasc techniquc, illustrating th c presence of T-helper ce lls (brown stainin g) and blue sta inin g in dica ting the presence of IgE, which is seen mainl y on the surface of a cl uster of cells to the right of the illustration.

contact dermatitis, li chen planus, and m ycosis fungoides, no HLA­D R antigen-s tained epiderm al keratin ocy tes were seen , alth ough Langerh ans cells were cl ea rl y stained . Numbers of epiderm al La ngerhans cell s were in creased, w ith a range of25-58 cell s (mea n 36.5) overl ying 200 basal cells (no rm al range in o ur laborato ry is 20- 30) 127]. Very few cells in either the epidermis o r dermis showed positive staining w ith either Leu-1 2 (B cell ) o r Leu-11 b (NK cell) . Cell s bearin g the interleukin 2(IL-2) recepto r we re fo und in all sa m ples exa min ed alth o ug h onl y in sm all numbers.

Posit ive IgE stainin g was seen in th e dermi s in 13 of th e 14 pat ients . T he 1 pati ent w ho did no t show pos iti ve IgE staining had bo th a no rmal to tal serum IgE and a negati ve RA ST screen . U sing a dou ble- stainin g technique, th e IgE was m ainl y seen in associatio n w ith clusters o f T lymphocytes in th e dermis. N o IgE stainin g was seen on the epiderma l Langerh ans cells no r on th e dendri t ic cells in the demus.

D ISCU SSIO N

It is poss ible th at th e observed in crease in B- cell nu m bers in th e peri pheral blood may be responsibl e fo r th e in creased level of serum IgE th at is seen in patients w ith AD. O ur findin gs w ith rega rd to T ce lls are at variance w ith a number of repo rts of a reduction in num bers o r fun cti on of T supp resso r cell s. Som e of these reports used the Fc recepto r as a m arker o f suppressor ce lls and , as it is recogni zed th at Fc cell s and O KT8 cell s do no t m ark identica l populati o ns, o ur p resent repo rt is no t st ri ctl y co m pa rable w ith such studies .

T here arc two possible ca uses of th e o bserved differen ces be­tween o ur stu dy an d those th at also used O KT8 as a suppresso r­cell m arker. Firs t, in 2 of th e studi es th e low suppressor cell numbers repo rted were in children . Whereas redu ced suppresso r. ce ll s m ay be a fea ture of AD in children , in neither o f these 2 stu dies we re the controls age-matched. In the first , patients had a m ean age of 9 and cont rol subj ects a m ean age o f 17 [9]. In th e second, the m ea n age o f th e patients w as 5. 5 yea rs and th e co nt ro ls 32 yea rs [1 0]. A second possible explanati on m ay Jie in diffe rences in th e cl inica l ac tiv ity o f the disease w ith redu ced numbers of circul atin g suppressor cell s seen onl y in m ore acute phases of AD

T H E jOUHNA L OF INVESTIGATIVE J)E I~MATOLOGY

ac ti vity. In general , the pati ents we exa mined had disease in a chroni ca ll y acti ve stage rather th an in acu te Aare, but the 1 pati ent in acute Aare had th e lowest number of circul atin g suppresso r cell s (15.5% vs our m ea n of 20.6% in thi s st ud y) . O ur results fo r the qu antitati ve analys is o f NK cells va ry w ith th e antibod y used (Leu-7 o r Leu- 11 ). Patients w ith ato pi c derm atitis are rec­ogni zed to have redu ced N K cell fun ction 128,291. Leu-ll is sa id to " bette r" defin e NK cell s than Leu-7 124]. If thi s is accurate, it wo uld im pl y that the redu cti on in N K act ivity in patients w ith A D is d ue to a fun cti o nal defect ex pressed by a no rm al number of NK ce lls.

O u.r stu dy confirms previo us repo rts th at the derm al in fi ltrate in atopic derm atitis is predo minantl y com posed o fT-h elper ce ll s 130 1 and that m any o f th ese cell s ex press HLA-D R antigen [21] . The rati o of T - helper to T-suppresso r-cyto tox ic cell s o bse rved in the de rm al inftltrate was 7: 1, co mpared w ith th e 2: 1 ratio seen in the peripheral blood , impl yin g a se lecti ve sequ es trati o n of T ­helper lymphocy tes in the derm al infiltrate. T his is in contras t to th e fi nd in gs in acute co ntact derm atitis w herein the derm al in­filtrate also shows the sa m e 2 : 1 helper: suppressor ra tio as th at seen in the pel'ipheral blood [31]. In contrast to studi es o n allerg ic conta ct derm ati tis, we saw no la stJ inin g of epiderm J I keratin o­cy tes in biops ies of ato pi c skin . An in crease in numbers of epi­derm al Lan ge rhan s ce ll s in th e skin of pat ients w ith AD has been repo rted prev io usly 120 1. A recent stud y has repo rted IgE staining on epide rm al dend ri t ic ce ll s shown to be Langerh ans cells, bo th by do ub le stainin g w ith anti-l gE and O KT6 and by immun ogold ultras tru ctural techni ques 132 1. We did no t o bser ve any epiderm al IgE stainin g in our patien ts . .

O ur o bser va ti on o f a spati al rel ati onshi p between derm al IgE stainin g and T-helper lymphocy tes m ay be due to the fac t they these are T lymphocytes th at bea r the epsilo n receptor 133]. w hich co uld be in vo lved in m o dul atin g the response to IgE. T ly m­phocy tes J re kn own to pro du ce so luble IgE bindin g fac to rs [341 . N o direct rcl ati on o f derm al dendriti c ce lls was seen in relatio n to th e IgE, alth oug h an indirect associati on m ay ex ist w ith the T ly mph ocy tes rela tin g to bo th the IgE and the dendriti c cells in th e dermis. The fin d in gs in this stu dy suggest a possible link between th e well- recogni zed but as ye t un ex pl ained coexistence of type I IgE-medi ated and type IV delayed hy persensiti v ity ab­no rm aliti es in pati ents w ith AD .

ADDENDUM

T his stud y showed that the derm al infiltrate in ato pi c derm atitis has the predo mi na nt pheno ty pe o f Leu 3 + 8 - (inducers o f h elp). Leu 8 stained o nl y 21 % of th e cell s in stead of the ex pected 60-70% . T his is in con tras t w ith recent results in all ergic contac t derm atitis w here the in filt ra tin g cell was predo min antl y Leu 3 + 8 + (inducers of su ppress ion). W ood GS, Vo lterra A S, Abel EA , Ni ckolo ff BJ , Ad am s RM: All erg ic contact dermatitis: novel im­muno histo logic fea tures. J In ves t D erm ato l 87 :688- 693, 1986.

We .~ rat ~fi '{l )' nclmo lVlc~Qc Dnko, Ltrl. , ElIglnlld jar nVsorbillg the cost of thc color prillt .

RE FERE N CES

I. HanifinJM : Atopic dermati tis. JAm Acad Derm ato I 6: 1- 13, 1982 2. Hanifin JM: Atopic de rm atiti s. J Allergy Clin ImmLlnol 73:211 -222,

1984 3. l3 yrom NA, Timlin OM: Immune status in atopic eczema: a survey.

Br J Derm atol 100:49 1-498, 1979 4. Byrom NA , Staughton RCD, Ca mpbell MA, Timlin OM. Chooi

M, Lanes AM, Copeman PWM, Hobbs JR: T hymosin-indu cible 'null ' cells in atopic exce ma. Br J Dermatol 100:499-510, 1979

5. Stingl G, Gazze LA , Cza rnecki N, Wolff K: T cell abnorm alities in atopic derm atitis patients: imbaian ces in T ce ll subpopuiations and impaired generation of Con A-induced suppressor cells. J In ves t DernJatol 76:468-473, 1981

VOL. 89, N O, I JULY 1987

6. Schuster DL, Bongiovanni BA, Pierson DL, Barbaro JF, Wong DTO , Levinson AI : Selective defi ciency of a T cell subpopulation in active atopic dermatitis. J ImmunoI1 24: 1662-1667, 1980

7. Strannegard 0, Strannegard I-L, Kang K, Cooper KD, Hanifin JM : Fel gG receptor-bearing lymphocytes and monoclonal ant ibody­defined T cell subse ts in atopic dermatitis: effect of treatment w ith thymopoietin penta peptide (TP-5). Int Arch Allergy Appl Im­munol 69:238-244, 1982

8. Jensen JR, T hest rup-Pedersen K: Modulating effec ts of enzy mes on T> and T". cel ls in patients with atopic derm atitis. J Inves t Dermatol 80:53- 55, 1983

9. Leung DYM, Rhodes All. , Geha RS: E numeration of T cell subse ts in atopic dermatitis using monoclonal antibodies. J Allergy C lin Immunol 67:450-455, 198 1

10. Butler M , Atherton D, Levinsk y RJ: Q uantitative and fun ctional defi cit of suppressor T cells in children with ato pic dermatitis. C lin Exp Immunol 50:92-98, 1982

11 . Faure MIl. , Gaucherand , MA , T hivo let J , Czernielewski JM, Ni­cho las J F: Decreased levels ofT- celis and ce lls w ith suppressor T­cel l pheno type as defined by specific monoclonal antibodies in patients w ith atop ic dermatitis. CJin Exp Dennato l 7:5 13-518, 1982

12. Kang K, Cooper KD, Hanifin JM : Th ympoietin penrapeptide (TP-5) improves clinica l para meters and lymphocyte subpo pulations in atopic dermatitis. J Am Acad Dermatol 8:372-377, 1983

13. Zachary CB, MacDonald DM: Quantitative analys is ofT-lympho­cyte subsets in atop ic eczema, usin g monoclonal antibodies and flow cytofluorimetry. Br J Dermatol 108:411 -422, 1983

14. Birkeland SA, Larsen P0, Larsen FS: Subpopulations of lympho­cytes and lymphocyte transformation tests in ato pic dermatitis: evaluation of a sys temic treatment with a new chromone com­pound and co mparison with a no rmal g roup . J Inves t Dcrm atol 76:367-370, 1981

15. Beard LJ, Thong YH , Turner TW: The immunologica l status of children w ith atopic dermatitis. Acta Paediatr Scand 70:55 1-555, 1981

16. ArmitsteadJG , Ewan PW: Suppressor T cell numbers and fun ction in atopic patients with high serum IgE levels. J C lin Lab In vest 12: 179-185, 1983

17. Mihm MC, Soter NA, Dvorak HF, Austen KF: The st ructure o f normal skin and the morpho logy of acute eczema. J Invest Der­matol 67:305-312, 1976

18. Russell Jones R: The histogenesis of eczema. C lin E xp Dermatol 8: 213-225,1983

19. Braathen LR, F0rre 0, Natvig JB , Eeg-Larsen T: Predominance of T lymphocytes in the dermal infiltrate o f atopic dermatitis. Br J Dermatoll00:511-519,1979

20. Rocha C, de Maubeuge J , Sarfa ti M , Song M , Dclespesse G: C har­acte riza tion of cellular infiltrates in skin les ions of atop ic eczema by means of monoclonal antibodies. Dermatolog ica 169:330-338, 1984

IMMUNO PH ENOTYPIN G IN DERM ATOLOGY 7

2 1. Zachary CB , All en MH, MacDonald DM: In situ quanti fication of T -l ymphocyte subsets and Langerhans cells in the inflamm atory infiltrate of atopic eczema. Br J Dermatol 11 2:1 49-156, 1985

22. MacKie RM : A monoclonal antibody technique to demonstrate an increa e in Langerhans cells in cutaneous lesions of mycos is fun­goides . C lin Exp Dermato l 7:43-47, 1982

23. Lanier LL, Engelman EG, Gatenby P, Babcock GF, Warner NL, Herzenberg LA : Correlation of fun ctional properties of hum an lymphoid cell subse ts and surface marker pheno types using mul­tiparameter ana lysis and flow cytometry. Immunol Rev 74:143- 160. 1983

24. Lanier LL, Lc AM. Phillips JH , Warner NL, Babcock GF: Subpop­ulations of human natural killer cells defi ned by express ion o f the LCll-7 (HNK-1 ) and Leu- II (N K-1 5) antigens. J Immunol 13 1: 1789-1 796, 1983

25. Lanier LL. Loken MR: Hum an lymphocyte subpopulations identi­fied by using three-color immunofluo rescence and flow cytom etry analys is: correlation of Leu-2, Leu 3, Lcu 7, Leu 8, and Leu-II ce ll surface express ion. J ImmunoI1 32: 151-156, 1984

26. Cordell JL, Falini B. Erber WN, Ghosh AK, Abdu laz iz Z, Mac­donald S. Pulfo rd KAF, Stein H . Mason DY: Immunoenzy matic labellin g of monoclonal antibodies using immune complexes of alka line phosphatase and monoclona l anti-alka line phosphatase (APAAP co mplexes). J Histiochem Cytochem 32:219- 224, 1984

27. MacKie RM , Turbitt ML: Quantitation of dendritic ce ll s iIi no rm al and abnormal hum an epidermis usin g m onoclonal antibodies di­rected aga inst la and HTA antigens. J Inves t Dermato l 81 :216- 220, 1983

28. Kusiami NT, Trentin JJ: Natural cell-mediated cytotoxic activity in the peripheral blood of patients w ith atopic dermatiti s. Arch Der­mato l '118:568- 57 1, 1982

29. Lever RS, Lesko MJ. MacKie RM, Parrott DMV : N atural killer cell activ ity in ato pic dermatitis. Clin Allergy 14:483-490, 1984

30. Sillevis Smitt JH , Bos JD, Hulscbosch H-J , Krieg SR: In situ im­munopheno typing of antigen presentin g ce lls and T cel l subsets in atopic dermatiti s. C lin Exp Dermatol 11 :1 59-168, 1986

3 1. Ferguson J , Gibbs JH , Swanson Beck J: Lymphocyte subsets and Lan gerhans cells in allergic and irritant patch test reactions: his­tiometric studies . Contact Dermatitis 13: 166-174. 1985

32. Bru ynzeel- Kooin en C , van Wichen DF, Toonstra J, Ben'ens L, Bru ynzeel PLB: The presence of IgE molecules on epidermal Langerhans cells in patients with atopic dermatitis. Arch Dermatol Res 278: 199- 205, 1986

33 . T hompson LF, Mellon MH. Zeiger RS, Spiegelberg HL: C harac­teri za tion with monoclonal antibodies of T Iym phocyres bearing Fc rec<! pto rs fo r IgE (T E cells) and IgE (Ty ce ll s) in ato pic patients. J Immunol 131:2772-2776, 1983

34. Ishizaka K, Sandberg K: Formation of lgE binding factors by human T lymphocytes. J Immunol 126:1692-1696, 1981