the elderly �

Regulatory effect of CD25, an activation antigen the alpha subunit of interleukin 2 receptor (IL2R) on the activity of natural killer (NK) cells was studied in fifty elderly (57-70 years old) and fifty young people (19-35 years old). Cytotoxic NK activity was assessed by 51Cr release assay, the levels of interleukin 2 (IL2) and tumour necrosis factors alpha (TNFalpha) were measured using bioassays and expression of CD16 and CD25 proteins by flow cytometry. Low NK activity in the elderly was associated with decline of full health, lowered serum concentration of IL2 and increased production of TNFalpha during NK reaction. Inhibition of TNFalpha activity by anti-TNF monoclonal antibody suppressed exclusively NK activity of low NK responders. Moreover, stimulation in vitro of blood mononuclear cells, with TNFalpha induced in the elderly low NK responders a significantly higher increase of the CD25 expression on the surface of NK cells as compared with that in the elderly high responders. Since the CD25 molecule constitutes a subunit of the high affinity receptor, binding IL2 to immunocompetent cells, its increased expression on NK cells of low NK responders would enable them to bind even low amounts of the endogenous IL2 available in this group of the elderly. Thus, an overproduction of TNFalpha seems to be a mechanism compensating, in the non-fully healthy elderly, for the decreased IL2 production, promoting efficient cytotoxic reaction.

Ageing is associated with impaired specific immune response [1].Deficient immune response can lead to the increased susceptibility of aged subjects to infections, cancers and autoimmune diseases [2].The loss of the ability of T cells of the elderly people to generate an immune response equal to that of T cells of the young is unquestionably an age-related phenomenon.Lower concentration of secreted IL2 [3,4] as well as lower levels of IL2 transcript [5], with simultaneously decreased induction of the transcriptional factors AP-1, NF-AT and NFkB [6,7] -have been detected in the T cell cultures stimulated with mitogens and antigens.Moreover, T cells of the elderly people are characterised by a diminished expression of the alpha chain of IL2 receptor (p55 IL2R, CD25) [8][9][10].
In the healthy old people the decrease in T cell potential is compensated for by an elevated activity of a non-specific arm of the immune response.The most pronounced activation involves NK cells.Unfortunately, such a repolarisation of the immune response happens only in about 10% of the elderly people, while in the remaining population the immune response generally is weakened.Our previous papers showed that the level of NK activity correlates with health status of the individuals selected according to the Senieur Protocol criteria [11].Elderly people who fulfil these criteria (optimally healthy) were characterised by high NK activity, while elderly people who did not meet all the criteria (not fully healthy) had low NK activity.We have also found that low NK activity and lack of a full health in elderly people are connected with low secretion of IL2 and high secretion of TNFa [12,13].
The level of NK cytotoxic activity is regulated by several cytokines.IL2 is considered a main stimulatory factor for NK activity, however other cytokines play also an important (co)stimulatory role [14][15][16].It has been shown that IL2 is necessary, but not sufficient, for optimal proliferation of NK cells [17].Furthermore, although TNFa did not in-duce proliferation by itself, it could augment the IL2-induced proliferation of resting or ionomycin activated NK cells.TNFa and IL2 are known to have agonistic effects on the lymphokine-activated killer cells (LAK).They both are necessary if these cells are to achieve optimal proliferation and exhibit cytotoxic properties [16].The two cytokines are also necessary for induction of expression of the activation antigen CD25 on LAK cells [17].
As TNFa potentiates many functions of NK cells -the intensity of production of this cytokine in the elderly people seems to be of importance.This production by the cells of old subjects is higher in comparison to that of the young, and may occur in the absence of any stimulant.Increased TNFa production has been found in the non-stimulated as well as stimulated cultures of peripheral blood mononuclear cells of aged individuals [18,19].
In the light of these facts it is conceivable that a low NK activity in the elderly may be associated not only with an insufficiency of IL2 but also with an alteration in the production of TNFa.
It seemes of interest to check whether the age-associated overexpression of TNFa has a regulatory effect on the NK cytotoxic activity and whether this effect is realised through the CD25 molecule (activation antigen) on NK cells.

MATERIALS AND METHODS
Older volunteers (50 people aged 57-70) were recruited from The Geriatric Outpatient's Department in Gdañsk.The younger (50 people aged 19-35) ones were students and staff of our Department.All the volunteers were informed about the goal of the study and gave consent to it.Older volunteers had had during the last three years a full medical examination every 12 months by the physicians from the Geriatric Outpatient's Department.The laboratory tests were carried out in the Department of Biochemistry of the Medi-cal University of Gdañsk.All volunteers were qualified into the study according to the Senieur Protocol criteria [11].According to these criteria older volunteers were divided into the group of fully healthy people (called "healthy"), fulfilling all the criteria of the protocol (n = 8; all females, mean age 63.2 ± 6.5) and those not fulfilling all these criteria (called "almost-healthy") (n = 42; 30 females and 12 males, mean age 64.1 ± 4.2).The "almost-healthy" people were generally in a good physical and mental health.They led an independent and active life and regarded themselves as "healthy".They had minor symptoms: moderately elevated and controlled blood pressure, degenerative changes of the skeleton, without pain and without limitation of the movement.They did not suffer from acute and chronic inflammatory diseases, autoimmune or neoplastic diseases.They did not receive medicaments known to affect the immune system.Fifty young people who remained under medical care were divided in the same way as the older volunteers into two groups: "healthy" (n = 12; all females, mean age (30.6 ± 5.5) and those not meeting all the protocol criteria "almost -healthy" (n = 38; 20 females and 18 males, mean age 27.8 ± 8.8).As the next step, K562 cells were mixed with PBMC (effector cells) at a ratio of 1 : 25.To the selected samples of mixed cells, monoclonal antibodies (mAbs) against TNFa and anti-IL2 (Genzyme, Cambridge, MA) were added at the beginning of the reaction.The cell mixture was incubated for 4 h at 37°C in a humidified atmosphere containing 5% CO 2 .

Isolation of peripheral blood mononuclear cells (PBMC
After incubation the samples were centrifuged and the supernatants was collected.Radioactivity of the supernatants was counted in a LKB ultragamma counter.The following formula was used for calculations: percent of chromium release (% of cytotoxicity) = (experimental release -spontaneous release)/(maximal release -spontaneous release) ´100.
Incubation of PBMC stimulated with K562 cells.K562 cells, used in this test as stimulatory cells, were cultured in the RPMI 1640-FCS medium containing 100 mg streptomycin, 100 U/ml penicillin, 2 mM L-glutamate and 100 mM/ml of pyruvate (Gibco Life Technologies Inc., Gaithesburg, MD).PBMC (responders to the stimulation with K562 cells) were suspended in the same medium at a concentration of 1 ´10 6 cells/ml.Stimulatory and responder cells (K562 and PBMC) were mixed at a ratio of 1 : 1 (2 ´10 5 cells /100 ml), centrifuged and incubated for 4 h in a humidified atmosphere containing 5% CO 2 at 37°C.After incubation the supernatants were harvested and kept at 80°C until use.Sera.Sera obtained from venous blood of the subjects tested were frozen and kept at 80°C until use.
Bioassay for TNFa.TNF-sensitive fibrosarcoma cell line WEHI 164, cell clone 13 (these cells undergo a dose-dependent apoptotic death under the influence of TNFa) (from Dr T. Espevik, University of Trondheim, Norway) was cultured for 24 h on 96-well plastic plates (Corning Science Products, Rochester, NY) at a concentration of 20 ´10 3 cells/well in the medium RPMI 1640-FCS supplemented with 2 mM L-glu- tamine, gentamycin and actinomycin-D (Sigma, St. Louis, MO) at a concentration of 1.0 mg/ml.A 10 ml portion of the supernatant (from above the mixture of K562 cells + PBMC) was added, in triplicate, to each well of the plate.Increasing concentrations of rTNFa (Genzyme, Cambridge, MA) were added, in duplicate, instead of the supernatants to the plate wells for obtaining a standard curve.Neutralising rabbit anti-TNFa antibody (Genzyme, Cambridge, MA) (1 : 10, 1 : 20, and 1 : 50) was added to the set of samples to check specificity of the test.To parallel control samples normal rabbit serum was added (1 : 50).After 24 h of incubation, 20 ml of MTT 3-(4,5dimethylthiazol-2-yl)-2,5-diphenyl-tetrazolium bromide (Sigma, St. Louis, MO) solution was added to each well of the plate.The plate was then incubated for another 4 h (37°C, 5% CO 2 ), 100 ml of isopropanol was added.Absorbance read at 570 nm on an automated plate reader (Multiscan MCC/340, Labsystems, Helsinki, Finland) fitted with titration standards of rTNFa.The anti-TNF antibody completely blocked the cytotoxic effect of the supernatants on WEHI cells.This assay had a detection limit of 1 pg/ml.The intra-assay coefficient of variation ranged from 5.5 to 13.8%, while the inter-assay coefficient of variation ranged from 15.5 to 22.3%.
Bioassay for interleukin 2 (IL2).IL2-dependent CTLL cells (from Institute of Immunology, Wroc³aw, Poland) (proliferating in a dose-dependent fashion under the influence of IL2) were cultured for 48 h on 96-well plastic plates (Corning Science Products, Rochester, NY) at a concentration of 20 ´10 3 cells/well in the RPMI 1640-FCS medium, 2 mM L-gluta- mate and gentamycin.A portion of 10 ml of the sera from the examined subject was added, in triplicate, to each well of the plate.Increasing concentrations of rIL2 were added, in duplicate, instead of sera to the wells for obtaining a standard curve.Neutralising rabbit anti-IL2 antibody (1 : 10, 1 : 20, and 1 : 50) was added to the set of samples to check specificity of the test.To parallel control samples normal rabbit serum was added (1 : 50).After 48 h of incubation, 20 ml of MTT solution was added to each well of the plate to penetrate into living cells.The plate was then incubated for another 4 h (37°C, 5% CO 2 ) and the 100 ml of isopropanol was added.Absorbance read at 570 nm on an automated plate reader from experimental wells fitted with titration standards of rIL2.The anti-IL2 antibody completely blocked the proliferative effect of the sera on the CTLL cells.The assay had a detection limit of 12.5 pg/ml.The intra-assay coefficient of variation ranged between 8.5 and 12.8%.The inter-assay coefficient of variation ranged from 16 to 25.%.
Stimulation of PBMC with TNFa.PBMC were suspended in the culture RPMI 1640-FCS medium at a concentration of 1 10 6 cells/ml.The cells were stimulated with human rTNFa (26 U/ml) (Genzyme, Cambridge, MA).Non-stimulated PBMC served as controls.Only sera negative for endotoxin were used in the experiments.Cells were incubated for 48 h in a humidified atmosphere containing 5% CO 2 at 37°C.Then cells were washed twice in cold PBS, kept on ice, and divided into samples of 30 ´10 4 cells for cytometry analysis.
Flow cytometry for surface staining: CD16 and CD25 molecules on PBMC.Surface staining of antigens was done on PBMC isolated from peripheral blood and for PBMC after stimulation in vitro with rTNFa.Monoclonal antibodies: anti-CD16 conjugated with phycoerythrin and anti-CD25 conjugated with FITC (Immunotech S.A., Marseille, France) were used for surface staining of PBMC cells.Cells (30 ´10 4 ) were suspended after washing in 50 ml of PBS followed by 10 ml of anti-CD16-PE and 10 ml of anti-CD25-FITC monoclonal antibodies.For control of each set a separate sample was stained with 10 ml of IgG1-PE and 10 ml of IgG2-FITC.Samples were incubated in the dark, at 4°C for 30 min, then washed twice and suspended in 1 ml of cold PBS for immediate flow cytometry on an EPICS XL flow cytometer (Coulter, Hialeah, FL).Data from at least 10 thousand cells was collected from each sample and stored as listmodes; off-line analysis of these data was done using the WinMDI software (by Dr.Joseph Trotter, Scripps Institute, La Jolla, CA).Statistical analysis.Statistical analysis was performed using the Excel 5.0 software (Microsoft), the statistical significance of differences was quantified by the unpaired T test (for comparing the groups) and paired T test (for assessing the effect of monoclonal antibodies on the NK cytotoxic activity).

Health status and NK activity in the young and elderly
An analysis of the NK activity and the health status of the elderly volunteers revealed that those characterised by full health (n = 8) had simultaneously higher NK (cytotoxicity 68 ± 22%) activity in relation to the "almost healthy" (n = 42) (cytotoxicity 32 ± 18% ).Similar results were obtained with the young.The "healthy" (n = 12) young subjects had significantly (P < 0.05) higher NK activity (cytotoxicity 81 ± 16%) than those "almost healthy" (n = 38) (cytotoxicity 38 ± 22%).Basing on these results, in the further considerations the individuals with higher NK activity were called "high NK responders" and those with low NK activity "low NK responders".

Secretion of TNFa during the NK cytotoxic action in the groups differing in age and NK activity
Secretion of TNFa into the supernatants of the cultures containing mixtures of the stimulating cells (K562) with peripheral blood mononuclear cells as effectors, is shown in Table 1.The comparison of TNFa secretion during the NK action revealed that the levels of that cytokine released by the cells of the elderly with low NK activity were significantly higher than those of the elderly with high NK activity.There was no significant difference in the TNFa levels between the low and the high NK responders in the young group.
The statistical analysis confirmed a negative correlation between the level of TNFa released during the NK cytotoxic action and the intensity of NK activity in the elderly people (r = -0.735;P < 0.05).We did not find such a correlation in the young group.

Blood levels of interleukin 2 in the groups differing in age and NK activity
Both the young and the elderly people with low NK activity had significantly lower levels of IL2 in their sera in comparison with those of the age-matched subjects with high NK activity (Fig. 1).There was a positive correlation Table 1.Level of TNFa (U/ml) in the cultures of PBMC stimulated with K562 cells.
The young and elderly subjects were divided into groups with low and high NK activity.The effector to-target cell ratio was 1 : 1, and the reaction was carried out for 4 h.Activity of TNFa was measured in the supernatants and expressed as U/ml.(r = 0.766; P < 0.05) between the serum level of IL2 and the intensity of the NK reaction in the elderly.

Effect of anti-TNFa and anti-IL2 antibodies on NK activity
To examine the role of TNFa and IL2 in the regulation of NK activity, anti-TNFa and anti-IL2 monoclonal antibodies were applied.An addition of anti-TNFa antibody to the NK reaction mixture significantly decreased the NK activity in the elderly people with low NK activity, while it had no effect on the intensity of the NK potential in the elderly with high NK activity.Monoclonal anti-IL2 antibody significantly decreased the NK activity both in the low and high NK responders (Fig. 2) (almost healthy and healthy subjects, respectively).
The results of this part of the study suggest that in the elderly people with low NK activity, a high production of TNFa during the NK reaction may be necessary for this reaction to reach the maximum possible intensity.The NK activity in those subjects remained, however, below that of the elderly with high NK activity.

Resting and activated NK cells in the young and elderly subjects
The percentages of resting (CD16 + cells) and activated NK cells (CD16 + CD25 + cells) were measured by flow cytometry in freshly obtained, non-stimulated PBMC of the young and elderly volunteers.There were no significant differences between groups of the young and elderly (young: CD16 + =13 ± 4.4% versus elderly: 19.3 ± 5% and young: CD16 + CD25 + = 0.7 ± 0.5% versus elderly: 1.7 ± 0.7).Similarly, the values for the high and low NK responders did not differ significantly within the age groups.

Activated NK cells in the elderly after stimulation with TNFa
To compare the effect of TNFa on NK numbers and phenotype in the elderly high and low NK responders, the cytometric tests were performed after 48 h incubation of PMBC with or without TNFa.In both groups examined after incubation of cells with TNFa there appeared cells with elevated values of the forward scatter (size) and slightly elevated side scatter (granularity) parameters (Fig. 3b) in relation to those for the non-stimulated population (Fig. 3a).Simultaneously, the double positive CD16 + CD25 + cells (Fig. 3d) appeared which were not seen after incubation without TNFa (Fig. 3c).
It was therefore interesting to establish whether the double positive CD16 + CD25 + cells remained within the major population of small lymphocytes with low values of both FSC and SSC, or whether they belonged to the bigger blastoid subpopulation.For this purpose, the graph displaying a plot of the FSC and SSC parameters was subdivided into two regions ("gates"): gate A -for the resting lymphocytes as in peripheral blood analysis and gate B -for lymphoblasts (Fig. 4a, 4b).Comparison of the presence of CD16 + CD25 + population in gates A and B showed that these cells were localised almost exclusively in gate B, i.e., among the lymphoblasts (Fig. 4d) while they were absent in gate A (Fig 4c).

Effect of TNFa on CD16+CD25+ cells in elderly people differing in NK activity
After TNFa stimulation the percentage of CD16 + CD25 + cells increased both in the elderly with low and high NK activity.In the cultures from the high NK responders the percentage of activated NK cells increased 3.7 times in relation to the values before stimula-Vol.47 Low NK activity in the elderly is compensated by TNFa 307 tion.In the cultures from low NK responders an increase of activated NK cells was higher, reaching a value 20.85 times as high as the values before stimulation.The differences between the two groups were statistically significant (P < 0.05).The summary of these experiments is presented in Table 2.These data indicate that the NK of the elderly low NK responders were more responsive to the stimulating effect of TNFa than the cells of the high NK responders.

DISCUSSION
In this study we have examined whether the age-associated overexpression of TNFa may have a regulatory effect on the NK cytotoxic activity and whether this effect on NK cells is realised through an activation antigen -the CD25 molecule.The results obtained showed that low NK activity in the elderly was associ-ated with a lack of full health, a lower serum concentration of interleukin 2 and a higher production of TNFa during the NK reaction [12,13].Moreover, stimulation in vitro of blood mononuclear cells, with TNFa induced in the elderly low NK responders a significantly higher increase of the CD25 expression on the surface of NK cells as compared with the elderly high responders.
There is a consensus in the immunogerontological literature that production of TNFa is increased on ageing.An increased secretion of TNFa by cells of old individuals in the absence of any stimulant [19] as well as after a mitogenic stimulation of PBMC have been described [18].This phenomenon has also been observed in the sera of elderly people [3,20].We have not come across any data analysing the changes of TNFa level during the NK cytotoxic reaction or its effect on that activity.It has been demonstrated, however, that TNFa may be secreted by NK cells after stim- ulation with carcinoma cells, transformed viral cells etc. [21,22].We had also earlier observed a negative correlation between the TNFa secretion and the intensity of NK activity in a group of elderly people, and lack of such a correlation in young individuals [12].
The results presented showed that the anti-TNFa monoclonal antibodies exerted an inhibitory effect on NK activity exclusively in the elderly low NK responders.This may indicate that an excessive secretion of TNFa during NK cytotoxic action in this group is necessary to maintain the highest possible level of NK activity.
The correlation between the level of NK activity and serum level of IL2 found in our study confirmed the role of IL2 in the stimulation of NK cells.The production of IL2 decreases during the ageing process.This phenomenon has been observed at different levels: at the level of secreted IL2, of IL2 mRNA and of IL2 receptors on the surface of T cells [4][5][6][7][8][9][10].In our previous paper [13], we have found that not all elderly people are characterised by a low IL2 production.The level of IL2 production in vivo in the fully healthy elderly was similar as that in the healthy young people.Thus, the lower NK cytotoxic activity found in the majority of the elderly is primarily due to the IL2 insufficiency.
The coexistence of high levels of TNFa with an insufficiency of IL2 at a low NK cytotoxic response in the elderly leads us to hypothesize about the role of TNFa.If the endogenous IL2 production is low, high amounts of TNF are secreted to up-regulate the expression of CD25 molecule on the surface of CD16 + NK cells.CD25 (p55, IL2R), the subunit of the IL2 receptor, can form, together with the beta and gamma chains, a high affinity receptor for IL2 on the surface of NK cells.In this way, this receptor might be able to bind low amounts of IL2, available in the elderly individuals.
In this paper we have shown that TNFa really contributes to up-regulation of the CD25 molecule on CD16 + cells of the elderly and that this effect is more pronounced in the low NK responders.
Summarising, the results of our paper suggest that the age-related overexpression of TNFa, though associated with low NK activity, does not contribute to its suppression.This overproduction seems to be a mechanism compensating for the decreased IL2 production so as to achieve a possibly maximal activation for the cytotoxic effect of NK cells.

Vol. 47
Low NK activity in the elderly is compensated by TNFa 309

Figure 1 .
Figure 1.The level of IL2 in the sera of young and elderly people.Bars represent arithmetical means of values obtained from the number of subjects (n) indicated, ± S.E.M. Asterisks show statistically significant differences at P = 0.043 (*) and P = 0.03 (**).

Figure 2 .
Figure 2. The effects of anti-TNFa and anti-IL2 monoclonal antibodies on NK cytotoxic activity in the elderly people.The anti-TNFa and anti-IL2 mAb were added to the cellular mixtures containing PBMC + K562 cells at the start of the 4 h NK cytotoxic reaction.The spontaneous NK activity (mean ±S.D.) of the group with low NK response (n = 8) was assumed as the cytotoxic index = 1.0.All other values were calculated in relation to that value.The mouse anti-human TNFa mAb decreased NK activity at the concentration of 50 ng/ml while the mouse anti-human IL2 mAb had its maximal effect at the concentration of 100 ng/ml.The mean ±S.D. values of the cytotoxic indexes were as follows: low spontaneous NK = 0.69 ± 0.31; low spontaneous NK + anti-TNFa = 0.43 ± 0.17; low spontaneous NK + isotypic control mouse IgG1 (PharMingen, A Beckton Dickinson Comp., Warsaw) = 0.77 ± 0.28.Bars represent arithmetical means of values obtained from the number of subjects (n) indicated, ±S.E.M. Differences significant at P = 0.014 (*), P = 0.007 (**) and P = 0.0002 (***).

Figure 3 .
Figure 3.The effect of TNFa on resting and activated NK cells.Panel (a) represents the forward and side scatter parameters of the non-stimulated, and panel (b) of the TNFa-stimulated PBMC.Activated NK (CD16 + CD25 + ) are visible only after incubation of PBMC with TNFa (d), but not in a control, non-stimulated, sample (c).The panels show typical result of the flow cytometric analysis in the form of a two-dimensional density plot.

308 J. Myoeliwska and others 2000 Figure 4 .
Figure 4. Characteristics of the TNFa activated NK cells.Panel (a) shows "small cells" (Gate A) and "large cells" (Gate B).Panel (b) shows expression of CD25 versus CD16 in the total PBMC population.The activated NK (CD16 + -CD25 + ) are visible only after incubation of PBMC with TNFa (d) but not in a control, non-stimulated, sample (c).The panels are a typical result of the flow cytometric analysis in the form of a two-dimensional density plot.