Britishjournal of Dermatology (1979) iOO, 499.
Thymosin-inducible 'null' cells in atopic eczema N.A.BYROM*, R.C.D.STAUGHTONt, MARY-ANN CAMPBELL*, D.M. T I M L I N t , M.CHOOI*, A.M.LANE§, P.W.M.COPEMANf AND J.R.HOBBS* Departments of *Chemical Pathology, tDermatology, §Haematology and {Computer Medicine, Westminster Hospital and Medical School, London SWi Accepted for publication 27 October 1978
SUMMARY
Thirty children with atopic eczema were compared with an equal number of age-matched healthy children. The mean peripheral blood T-lymphocyte level was lower in the eczema group (mean 1,197/mm^ as against 1,702/mm^; P = o 003). This difference was abolished in vitro by thymosin, a thymic hormone extract. Positive correlations were found between eczema severity and: eosinophilia; hyperimmunoglobulinaemia E; but not T lymphopaenia. Thymosin-inducible T-cell (TJ counts correlated with plasma IgE levels, suggesting that these Tj cells may be immature suppressor T cells. If this T-cell deficiency represents inadequate suppression of IgE responses, then a trial of treatment with thymosin appears to be warranted.
T lymphopaenia occurs frequently in atopic disorders, e.g. in atopic asthma (Byrom et al, i91^a), although such findings are controversial. Defects in in vitro T-lymphocyte function (Lobitz, Honeyman & Winkler, 1972; Andersen & Hjorth, 1975; McGeady & Buckley, 1975; Hovmark, 1975, 1977; Thestrup-Pedersen et al, 1977), depressed delayed cutaneous hypersensitivity reactions (Rajka, 1968, 1970; Grove, Reid & Forbes, 1975; Palacios, Fuller & Blaylock, 1976) and increased severity of cutaneous infections (Hanifin & Lobitz, 1977) hate been reported in patients with atopic eczema; similar defects in cell-mediated immunity (CMI) have been noted in patients with atopic asthma (Grove et al, 1975b). Defects in CMI have not, however, been found by all authors. Elevated serum/plasma IgE levels are associated with certain disorders with T-lymphocyte abnormalities, and these include primary T-cell immunodeficiencies such as the Wiskott-Aldrich syndrome (Berglund et al, 1968; Waldmann et al, 1972), the Di George syndrome (Polmar, Waldmann & Terry, 1972), and thymic alymphoplasia (Kikkawa et al, 1973). Hyperimmunoglobulinaemia E is also a frequent finding in atopic disease (Johansson, 1969), e.g. atopic asthma (Byrom et al, 1978a). Correspondence Dr N.A.Byrom, Department of Chemical Pathology, Westminster Medical School, 17 Page Street, London SWiP 2AR. 0007-0963/79/0500-0499S02.00 © 1979 British Association of Dermatologists
499
500
N.A.Byrom et al
Eosinophilia is also a frequent finding in patients with atopic disorders (Stickney & Heck, 1944). Immunological intervention, with the in vitro and in vivo use of transfer factor, has been found to restore depressed T-lymphocyte levels in children with asthma (Strannegard et al, 1975; Khan et al, 1976a; Khan, Sellars & Thometz, 1976) and with the Wiskott-Aldrich syndrome (Khan et al, 1975; Sellars & South, 1975) to normal levels, with concomitant marked clinical improvement. The present study was undertaken to assess any in vitro capacity of thymosin to restore T-lymphocyte numbers to normal levels in children with atopic eczema, as has already been reported for the Wiskott-Aldrich syndrome (Wara et al, 1975) and atopic asthma (Byrom et al, 1978a). The ability of a lymphocyte to form a spontaneous (E) rosette with sheep red blood cells (SRBC) identifies that cell as a thymus-dependent (T) lymphocyte (Jondal, Holm & Wigzell, 1972). Since fetal calf serum (FCS) has thymosin-like activity and may result in the overestimation of T-lymphocyte levels in atopic subjects (Byrom et al, 1977, 1978a; Byrom, 1978), all E-rosette tests for this study were set up in the absence of FCS, or any other serum. The effect of thymosin is dosedependent (Byrom et al, I978a,b), so in vitro thymosin responses were assessed over a wide logarithmic dilution range of the drug. PATIENTS AND METHODS
Thirty children (sixteen male, fourteen female), their ages ranging from 1-14 (mean 8-8 + 4-3) years, were studied (Table i). All patients undoubtedly had atopic eczema, with the age of onset ofthe symptoms ranging from birth to 11 years. At least 22/29 (76%) ofthe patients had first-degree relatives with atopy. All patients had been referred to our Dermatology Out-Patient Clinic with active disease. Six patients were judged on clinical grounds to have secondary cutaneous bacterial infection at the time of testing. No patient had received systemic or strong topical corticosteroids during the previous year. One of us assessed the grade of severity of each patient's eczema on an arbitrary scale. Mild (five patients). Atopic eczema confined to two or less ofthe 'classical' sites (antecubital fossae, popliteal fossae, and extremities). Moderate (eight patients). Atopic eczema confined to the 'classical' sites. Severe (seventeen patients). Confluent atopic eczema, extending beyond the 'classical' sites. Thirty normal children (nineteen male, eleven female) with no known history or family history of atopic disease and with no evidence of skin disease were also selected, their ages ranging from 5-16 (mean 9-9 + 30) years. None had received treatment during the 2 weeks preceding the study. Methods Haematology. Total white blood cell counts were performed on peripheral blood samples using a Coulter model'S' counter. Blood films, fixed with methanol, then stained using a modified JennerGiemsa technique, were all interpreted by the same individual (A.M.L.) to determine the absolute levels of lymphocytes and eosinophils per mm^ of blood. Isolation of lymphocytes. As described by Byrom et al (1978b), using Lymphoprep (Nyegaard, Oslo). T-cell (E) rosettes. T lymphocytes were enumerated by spontaneous sheep erythrocyte rosette formation, in the absence of serum, as described previously (Byrom et al, 1978b). Briefiy, lymphocytes were pre-incubated in round-bottom plastic tubes with either tissue culture medium (control) or with thymosin fraction V (Hoffmann-LaRoche, New Jersey, batch no. C 98035). o-i ml of a lymphocyte
'Null' cells in atopic eczema
501
suspension of approximately 5 x 10* cells/ml initial concentration was mixed with an equal volume of medium or diluted thymosin (log dilution steps from o-oo8 to 8,000 fig/ml, initial concentration, in medium) and incubated at 37°C for 10 min. 0-2 ml of a sheep red blood cell (SRBC) suspension (10^ cells/ml) was then added to each tube, so that the final concentrations of thymosin ranged from 0002 to 2,000 fig/ml. After thorough mixing, the tubes were gently centrifuged for 2-5 min at 150 g to bring the cells into close contact, then incubated at 37°C for 15 min and then at 4°C overnight. Next day, the rosettes were gently resuspended and read by phase-contrast microscopy. B lymphocytes. The numbers of surface immunoglobulin (Slg)-bearing lymphocytes were measured as an in vitro correlate of B-lymphocyte levels. A polyvalent antiserum raised against human IgG plus IgA plus IgM (H- and L-chains) (Behringwerke AG, Marburg-Lahn; catalogue no. OTKG; batch no. 686A), or a monovalent rabbit antiserum against human IgE (e-chain) (Behringwerke; catalogue no. OTKI; batch no. 687D), both conjugated with fluorescein, was used to stain the lymphocytes by the direct technique, as previously described (Byrom et al, 1978b). The specificity of both conjugates was checked against myeloma cells. Plasma immunoglobulins. Heparinized peripheral blood plasma was assayed for IgA and IgM by a modification of the 'Rocket' technique (Slater, 1975) and IgG by the automated immunoprecipitin (AIP) method (Technicon, 1975).
3-00
FIGURE I. Plot of logio plasma IgE level and level of thymosin-inducible T cells (T,) in a population of thirty children with atopic eczema.
502
N.A.Byrom et al.
Total plasma IgE levels were determined by the radioimmunosorbent technique (RIST) (Wide, 1969). The anti-IgE for the RIST was raised against purified IgE from myeloma patient N.D. (a kind gift from Drs S. G. O. Johansson and H. Bennich) and IgE isolated from the serum of patient P.S. (the kind gift of Dr O. Ross McIntyre). The IgE standard was WHO 68/341 (Rowe et al., 1970). Statistical analyses Data from the patient group were compared with control values by the Mann-Whitney U test. Standard deviation values are given throughout this paper, although the data are not necessarily normally distributed. Many authors have found correlations between such parameters as eczema severity, serum/plasma IgE level, and levels of eosinophils and of T lymphocytes. In this present study, non-parametric regression analyses (Spearman) were used to test for correlations between these parameters, and the results are summarized in Table i and Figs 1-3.
RESULTS
T-lymphocyte numbers in medium without thymosin (T^;^) (Table 2) In the eczema group, the range for T^;^ was 2OO-3,469/mm^ Six patients, all with severe eczema, were T lymphopaenic: S.P. (455/mm3); D.C. (227/mm3); B.H. (2oo/mm3); R.B. (24o/mm3); J.S. (5ii/mm^)j A.B. (325/mm^). Only five patients had a T-lymphocyte count higher than the normal mean value of I,74o/mm^ The overall mean T-cell level in the eczema group was much lower than that in the control group (P = 0-0009) (Table 2). The mean 'null' (non-T, non-B) lymphocyte level for the eczema group was raised in comparison with the control group mean, although a significant difference could not be shown. None of the normal controls was T lymphopaenic (normal range for T^i^ = 470-1,8oo/mm-'). T^•,„ levels failed to correlate with other parameters tested (see Table i). TABLE I. Grid showing correlations between some parameters of a population of thirty children with atopic eczema. Results obtained by Spearman rank correlation coefficient method (r = correlation coefficient) Log plasma IgE Log plasma IgE
Eosinophils/mm ^ Eczema severity No correlation
P< 00002
r = 064 (Fig. 3) Eosinophils/mm'
No
P = 0 01 r = 0 44 (Fig. 2)
correlation T cells in medium (T^in/mm^) T cells with thymosin (T ma,/mm^) Thymosin-inducible T cells (Ti/mm^)
No
No
correlation
correlation
correlation
No
No
No
correlation P = 004 r = 0 38 (Fig. I)
correlation
correlation
No
No
correlation
correlation
No
'Null' cells in atopic eczema
503
3000
m
2000
3
1000
Eczema severity grade
FIGURE 2. Plot of absolute eosinophil level and disease severity grade in a population of thirty children with atopic eczema.
T-lymphocyte numbers in medium with thymosin (T^^,) {Table 2) Thymosin, in vitro, abolished the apparent T-lymphocyte deficit in the eczema patient population, i.e. it increased the levels of E-rosette-forming cells to normal. In the eczema group, the range for T ^ ^ was 1,048-4,11 i/mm^. All six of the T-lymphopaenic patients had normal levels of E rosettes after in vitro lymphocyte induction with thymosin, and sixteen patients had a T-cell level higher than the normal mean for T^,,,. Correspondingly, the difierence in null cell levels between patient and control groups was abolished with in vitro thymosin. The entire deficit of E-rosette-forming cells could be corrected by thymosin. Although the controls had a small number of thymosin-inducible T lymphocytes (T,), the mean level for the patients was more than ten times higher. In vitro incubation of lymphocytes with thymosin at low concentration (mean 0-02 /ig/ml) was sufficient to increase the mean T-lymphocyte level of the Tlymphopaenic patients into the normal range. No single thymosin concentration produced maximal rosette formation for all of the patients or control subjects. At above-optimum concentrations, thymosin actually inhibited E rosetting. Tn,3, levels failed to correlate with other parameters tested (see Table i). Thymosin-inducible T lymphocytes {T-^ = T^.^^ — T^-^^ (Table 2) All children with eczema had Tj cells (Tindu 480/mm^), and there was a correlation with eczema severity (P = o-oi) (Fig. 2). No other correlations with eosinophil levels were found (Table i). B lymphocytes (Table 2) Ofthe B lymphocytes, only IgE-bearing lymphocytes were increased in eczema (P = 0002; range 1-21%; mean 7-5%). Plasma IgE (Table 3) The mean plasma (total) IgE level was significantly elevated in the patient group (P = o-oi). Plasma IgE levels (range 20-64,000 iu/ml) exceeded 1,000 iu/ml in eighteen (60%) and 10,000 iu/ml in six (20%) ofthe patients. Eleven (37%) had normal IgE levels (
Thymosin-inducible 'null' cells in atopic eczema N.A.BYROM*, R.C.D.STAUGHTONt, MARY-ANN CAMPBELL*, D.M. T I M L I N t , M.CHOOI*, A.M.LANE§, P.W.M.COPEMANf AND J.R.HOBBS* Departments of *Chemical Pathology, tDermatology, §Haematology and {Computer Medicine, Westminster Hospital and Medical School, London SWi Accepted for publication 27 October 1978
SUMMARY
Thirty children with atopic eczema were compared with an equal number of age-matched healthy children. The mean peripheral blood T-lymphocyte level was lower in the eczema group (mean 1,197/mm^ as against 1,702/mm^; P = o 003). This difference was abolished in vitro by thymosin, a thymic hormone extract. Positive correlations were found between eczema severity and: eosinophilia; hyperimmunoglobulinaemia E; but not T lymphopaenia. Thymosin-inducible T-cell (TJ counts correlated with plasma IgE levels, suggesting that these Tj cells may be immature suppressor T cells. If this T-cell deficiency represents inadequate suppression of IgE responses, then a trial of treatment with thymosin appears to be warranted.
T lymphopaenia occurs frequently in atopic disorders, e.g. in atopic asthma (Byrom et al, i91^a), although such findings are controversial. Defects in in vitro T-lymphocyte function (Lobitz, Honeyman & Winkler, 1972; Andersen & Hjorth, 1975; McGeady & Buckley, 1975; Hovmark, 1975, 1977; Thestrup-Pedersen et al, 1977), depressed delayed cutaneous hypersensitivity reactions (Rajka, 1968, 1970; Grove, Reid & Forbes, 1975; Palacios, Fuller & Blaylock, 1976) and increased severity of cutaneous infections (Hanifin & Lobitz, 1977) hate been reported in patients with atopic eczema; similar defects in cell-mediated immunity (CMI) have been noted in patients with atopic asthma (Grove et al, 1975b). Defects in CMI have not, however, been found by all authors. Elevated serum/plasma IgE levels are associated with certain disorders with T-lymphocyte abnormalities, and these include primary T-cell immunodeficiencies such as the Wiskott-Aldrich syndrome (Berglund et al, 1968; Waldmann et al, 1972), the Di George syndrome (Polmar, Waldmann & Terry, 1972), and thymic alymphoplasia (Kikkawa et al, 1973). Hyperimmunoglobulinaemia E is also a frequent finding in atopic disease (Johansson, 1969), e.g. atopic asthma (Byrom et al, 1978a). Correspondence Dr N.A.Byrom, Department of Chemical Pathology, Westminster Medical School, 17 Page Street, London SWiP 2AR. 0007-0963/79/0500-0499S02.00 © 1979 British Association of Dermatologists
499
500
N.A.Byrom et al
Eosinophilia is also a frequent finding in patients with atopic disorders (Stickney & Heck, 1944). Immunological intervention, with the in vitro and in vivo use of transfer factor, has been found to restore depressed T-lymphocyte levels in children with asthma (Strannegard et al, 1975; Khan et al, 1976a; Khan, Sellars & Thometz, 1976) and with the Wiskott-Aldrich syndrome (Khan et al, 1975; Sellars & South, 1975) to normal levels, with concomitant marked clinical improvement. The present study was undertaken to assess any in vitro capacity of thymosin to restore T-lymphocyte numbers to normal levels in children with atopic eczema, as has already been reported for the Wiskott-Aldrich syndrome (Wara et al, 1975) and atopic asthma (Byrom et al, 1978a). The ability of a lymphocyte to form a spontaneous (E) rosette with sheep red blood cells (SRBC) identifies that cell as a thymus-dependent (T) lymphocyte (Jondal, Holm & Wigzell, 1972). Since fetal calf serum (FCS) has thymosin-like activity and may result in the overestimation of T-lymphocyte levels in atopic subjects (Byrom et al, 1977, 1978a; Byrom, 1978), all E-rosette tests for this study were set up in the absence of FCS, or any other serum. The effect of thymosin is dosedependent (Byrom et al, I978a,b), so in vitro thymosin responses were assessed over a wide logarithmic dilution range of the drug. PATIENTS AND METHODS
Thirty children (sixteen male, fourteen female), their ages ranging from 1-14 (mean 8-8 + 4-3) years, were studied (Table i). All patients undoubtedly had atopic eczema, with the age of onset ofthe symptoms ranging from birth to 11 years. At least 22/29 (76%) ofthe patients had first-degree relatives with atopy. All patients had been referred to our Dermatology Out-Patient Clinic with active disease. Six patients were judged on clinical grounds to have secondary cutaneous bacterial infection at the time of testing. No patient had received systemic or strong topical corticosteroids during the previous year. One of us assessed the grade of severity of each patient's eczema on an arbitrary scale. Mild (five patients). Atopic eczema confined to two or less ofthe 'classical' sites (antecubital fossae, popliteal fossae, and extremities). Moderate (eight patients). Atopic eczema confined to the 'classical' sites. Severe (seventeen patients). Confluent atopic eczema, extending beyond the 'classical' sites. Thirty normal children (nineteen male, eleven female) with no known history or family history of atopic disease and with no evidence of skin disease were also selected, their ages ranging from 5-16 (mean 9-9 + 30) years. None had received treatment during the 2 weeks preceding the study. Methods Haematology. Total white blood cell counts were performed on peripheral blood samples using a Coulter model'S' counter. Blood films, fixed with methanol, then stained using a modified JennerGiemsa technique, were all interpreted by the same individual (A.M.L.) to determine the absolute levels of lymphocytes and eosinophils per mm^ of blood. Isolation of lymphocytes. As described by Byrom et al (1978b), using Lymphoprep (Nyegaard, Oslo). T-cell (E) rosettes. T lymphocytes were enumerated by spontaneous sheep erythrocyte rosette formation, in the absence of serum, as described previously (Byrom et al, 1978b). Briefiy, lymphocytes were pre-incubated in round-bottom plastic tubes with either tissue culture medium (control) or with thymosin fraction V (Hoffmann-LaRoche, New Jersey, batch no. C 98035). o-i ml of a lymphocyte
'Null' cells in atopic eczema
501
suspension of approximately 5 x 10* cells/ml initial concentration was mixed with an equal volume of medium or diluted thymosin (log dilution steps from o-oo8 to 8,000 fig/ml, initial concentration, in medium) and incubated at 37°C for 10 min. 0-2 ml of a sheep red blood cell (SRBC) suspension (10^ cells/ml) was then added to each tube, so that the final concentrations of thymosin ranged from 0002 to 2,000 fig/ml. After thorough mixing, the tubes were gently centrifuged for 2-5 min at 150 g to bring the cells into close contact, then incubated at 37°C for 15 min and then at 4°C overnight. Next day, the rosettes were gently resuspended and read by phase-contrast microscopy. B lymphocytes. The numbers of surface immunoglobulin (Slg)-bearing lymphocytes were measured as an in vitro correlate of B-lymphocyte levels. A polyvalent antiserum raised against human IgG plus IgA plus IgM (H- and L-chains) (Behringwerke AG, Marburg-Lahn; catalogue no. OTKG; batch no. 686A), or a monovalent rabbit antiserum against human IgE (e-chain) (Behringwerke; catalogue no. OTKI; batch no. 687D), both conjugated with fluorescein, was used to stain the lymphocytes by the direct technique, as previously described (Byrom et al, 1978b). The specificity of both conjugates was checked against myeloma cells. Plasma immunoglobulins. Heparinized peripheral blood plasma was assayed for IgA and IgM by a modification of the 'Rocket' technique (Slater, 1975) and IgG by the automated immunoprecipitin (AIP) method (Technicon, 1975).
3-00
FIGURE I. Plot of logio plasma IgE level and level of thymosin-inducible T cells (T,) in a population of thirty children with atopic eczema.
502
N.A.Byrom et al.
Total plasma IgE levels were determined by the radioimmunosorbent technique (RIST) (Wide, 1969). The anti-IgE for the RIST was raised against purified IgE from myeloma patient N.D. (a kind gift from Drs S. G. O. Johansson and H. Bennich) and IgE isolated from the serum of patient P.S. (the kind gift of Dr O. Ross McIntyre). The IgE standard was WHO 68/341 (Rowe et al., 1970). Statistical analyses Data from the patient group were compared with control values by the Mann-Whitney U test. Standard deviation values are given throughout this paper, although the data are not necessarily normally distributed. Many authors have found correlations between such parameters as eczema severity, serum/plasma IgE level, and levels of eosinophils and of T lymphocytes. In this present study, non-parametric regression analyses (Spearman) were used to test for correlations between these parameters, and the results are summarized in Table i and Figs 1-3.
RESULTS
T-lymphocyte numbers in medium without thymosin (T^;^) (Table 2) In the eczema group, the range for T^;^ was 2OO-3,469/mm^ Six patients, all with severe eczema, were T lymphopaenic: S.P. (455/mm3); D.C. (227/mm3); B.H. (2oo/mm3); R.B. (24o/mm3); J.S. (5ii/mm^)j A.B. (325/mm^). Only five patients had a T-lymphocyte count higher than the normal mean value of I,74o/mm^ The overall mean T-cell level in the eczema group was much lower than that in the control group (P = 0-0009) (Table 2). The mean 'null' (non-T, non-B) lymphocyte level for the eczema group was raised in comparison with the control group mean, although a significant difference could not be shown. None of the normal controls was T lymphopaenic (normal range for T^i^ = 470-1,8oo/mm-'). T^•,„ levels failed to correlate with other parameters tested (see Table i). TABLE I. Grid showing correlations between some parameters of a population of thirty children with atopic eczema. Results obtained by Spearman rank correlation coefficient method (r = correlation coefficient) Log plasma IgE Log plasma IgE
Eosinophils/mm ^ Eczema severity No correlation
P< 00002
r = 064 (Fig. 3) Eosinophils/mm'
No
P = 0 01 r = 0 44 (Fig. 2)
correlation T cells in medium (T^in/mm^) T cells with thymosin (T ma,/mm^) Thymosin-inducible T cells (Ti/mm^)
No
No
correlation
correlation
correlation
No
No
No
correlation P = 004 r = 0 38 (Fig. I)
correlation
correlation
No
No
correlation
correlation
No
'Null' cells in atopic eczema
503
3000
m
2000
3
1000
Eczema severity grade
FIGURE 2. Plot of absolute eosinophil level and disease severity grade in a population of thirty children with atopic eczema.
T-lymphocyte numbers in medium with thymosin (T^^,) {Table 2) Thymosin, in vitro, abolished the apparent T-lymphocyte deficit in the eczema patient population, i.e. it increased the levels of E-rosette-forming cells to normal. In the eczema group, the range for T ^ ^ was 1,048-4,11 i/mm^. All six of the T-lymphopaenic patients had normal levels of E rosettes after in vitro lymphocyte induction with thymosin, and sixteen patients had a T-cell level higher than the normal mean for T^,,,. Correspondingly, the difierence in null cell levels between patient and control groups was abolished with in vitro thymosin. The entire deficit of E-rosette-forming cells could be corrected by thymosin. Although the controls had a small number of thymosin-inducible T lymphocytes (T,), the mean level for the patients was more than ten times higher. In vitro incubation of lymphocytes with thymosin at low concentration (mean 0-02 /ig/ml) was sufficient to increase the mean T-lymphocyte level of the Tlymphopaenic patients into the normal range. No single thymosin concentration produced maximal rosette formation for all of the patients or control subjects. At above-optimum concentrations, thymosin actually inhibited E rosetting. Tn,3, levels failed to correlate with other parameters tested (see Table i). Thymosin-inducible T lymphocytes {T-^ = T^.^^ — T^-^^ (Table 2) All children with eczema had Tj cells (Tindu 480/mm^), and there was a correlation with eczema severity (P = o-oi) (Fig. 2). No other correlations with eosinophil levels were found (Table i). B lymphocytes (Table 2) Ofthe B lymphocytes, only IgE-bearing lymphocytes were increased in eczema (P = 0002; range 1-21%; mean 7-5%). Plasma IgE (Table 3) The mean plasma (total) IgE level was significantly elevated in the patient group (P = o-oi). Plasma IgE levels (range 20-64,000 iu/ml) exceeded 1,000 iu/ml in eighteen (60%) and 10,000 iu/ml in six (20%) ofthe patients. Eleven (37%) had normal IgE levels (
