Autologous Enhancement by Interferon of Natural Killer Activity of Human Peripheral Blood Lymphocytes

The in vitro action of interferon (IFN)-α and IFN-γ from six healthy donors and ten patients with multiple sclerosis (MS) on natural killer (INK) activity of peripheral blood lymphocytes (PBL) was studied in an autologous system. The NK activity of PBL was detected by a cytotoxic test using 3H-uridine human erythromyeloblast K562 cells. Autologous IFN-α and IFN-γ did not augment NK activity of PBL from healthy donors in vitro, whereas in samples from MS patients the IFNs strongly stimulated NK cell cytotoxic function. This stimulation suggests the existence of an inhibitor of regulatory IFN action, that is produced in healthy donors simultaneously with IFN in response to IFN induction, but which is lacking in commercial IFN preparations. The factor-containing supernatants from healthy donors reduced the stimulatory action of autologous IFNs in patients with MS almost until complete blockade. Because this inhibitor was absent in patients with MS, deficiency of an inhibitor of IFN regulatory action in MS could open the way to treatment of this compartment of the immune system.


Introduction
The ability of interferons (IFNs) to activate natural killer (NK) cells is an important property of these mediators. By inducing differentiation of pre-NK to active NK forms, enhancing the lytic potency of a single killer cell, and accelerating the recycling of NK' cells after lytic interaction with their targets ', 3 IFNs serve factors playing a key role in the function of this lymphocyte population. 1," The spectrum of immunomodulatory action of IFNs, as well as their strong antiproliferative activity, has led to the use of IFN preparations in the therapy of oncological diseases. 5,6 However, in spite of proven clinical efficiency, IFN action does not correspond to expectation in patients with idiopathic autoimmune haemolytic anaemia in vitro, in a course of IFN therapy in chronic granulocytic leukaemia or in myelodisplasia patients.
Multiple sclerosis (MS) is a disorder where the effects of IFN are puzzling. 1-1" If the use of IFN-ot and IFN-[3 is characterized by a reduction in the number 11 12 and severity of exacerbations then IFN-,treatment leads, at least in some patients, to exacerbation of the pathological process, ha3 MS is associated not only with a strong deficiency of cytotoxic NK cell activity 1-7 and a functional defect of the IFN system, 5,6,18a9 but also anomalous responses of NK cells to regulatory IFN action. 14,7,' NK cells produce multiple cytokines including IFN, ,. and the reduced IFN production by peripheral blood lymphocytes (PBL) possessing NK activity, 5 in 1994 Rapid Communications of Oxford Ltd patients with MS, is probably related to a defect in monocyte helper."" There also appears to be a dissociation between the cytotoxic and IFN-producing activities of PBL exhibiting NK properties in this disease, ,. as well as maintenance of an excess of IFN in the serum of blood and cerebrospinal fluid in these patients. 9, '4 In the light of these relationships, the aim of the present work was to study the autologous (auto-) IFN-{z and IFN-7 actions on the cytotoxic NK activity of PBL in patients with MS and in healthy donors.

Materials and Methods
Six healthy donors (two men and four women) aged between 18 and 48 years, and ten patients (four men and six women) aged between 18 and 42 years with evident MS ,.5 with remitting course of the process at the stage of exacerbation, cerebrospinal form and various duration of the disease (from 3 to 8 years), and IV-V degrees of invalidization (Kourtske scale) were investigated.
Determination of the action of auto-IFN on NK activity of PBL was carried out in two steps. Initially, peripheral venous blood of patients was collected for IFN induction while PBL were used to evaluate NK cytotoxicity. Then, after IFN induction (on the 8th day of investigation), PBL were isolated from peripheral blood of the same healthy donors and patients with MS (MS patients received a placebo triaD, and the action of auto-IFN on NK activity of these PBL was determined. PBL were isolated from patients' blood on Ficoll-Paque (Pharmacia Fine Chemicals) gradients of a density of 1.077 g/cm using the method of B6yum. 26 The cytotoxic NK activity of PBL was measured by the radiometric assay of Hashimoto and Sudo, 27 as modified by Rykova et al., 28 using target cells of the standard human erythromyeloblast line K562 labelled with H-uridine at a dose of 3 btCi/ml. PBL and K562 cells were inoculated into the U-shape wells of 96-well microplates followed by co-incubation for 14 h at 37C in a humid atmosphere containing 5% CO,.. The incubation was carried out in 0.2 ml RPMI-1640 medium (Amimed) together with 12% v/v foetal calf serum (NF Gamaleya Institute for Epidemiology and Microbiology, Moscow), 10 mM HEPES (Serva), 2 mM glutamine, and 40 l.tg/ml gentamicin (Pharmachim). NK activity was tested over a wide range of effector:target cell (E:T) ratios 100:1, 50:1, 25:1, 12:1 and 6:1. Each E:T ratio was prepared in two or three parallel wells. Initial suspensions of PBL contained 10 PBL/ml, and the suspension of K562 targets contained 105 cells/ml. After stopping co-incubation the PBL and K562 cells were harvested by a 12-channel harvester (Dynatech) on glass fibre filters with 2.5 l-tm pores (Whitman). Radioactivity remaining on the filters was counted for 1 min for each probe using I]-counters (Packard Mark II, USA). Efficiency of IFN action was determined by regulatory parameters (RPs) presenting the ratios of absolute changes of the IC to initial NK activities, counted as mean values for all tested E:T ratios and expressed with corresponding signs, i Total RP (TRP) for a group of observations was counted as mean value of RPs in this group. Student's t-test was used for evaluation of the statistical significance of the results.

Results
The data in Fig. 1A show that auto-IFNs did not augment the cytotoxic NK activity of PBL obtained from blood of healthy donors. The TRP of auto-IFNwas equal to -8_+16% and of auto-IFN-7 to 6 _+ 10%. At the same time auto-IFNs strongly activated PBL possessing NK activity in patients with MS (Fig. 1B). The TRPs were 28 _+ 11% (p< 0.05) for auto-IFN-0t and 56 + 26% (p < 0.05) for auto-IFN-,.
Comparison of the doses of auto-IFNs in the study of IFN action on NK activity of PBL shows that the same doses of auto-IFNs were used in normal individuals as in MS patients (Table 1). Moreover, these doses of auto-IFNs were similar to those adopted in previous work concerning the regulatory action of commercial IFN preparations on PBL NK cytotoxicity, z where the difference was not more than one dilution of IFN-containing supernatant.  The distribution of regulatory influences of auto-IFNs on PBL NK activity was compared in MS patients and in healthy donors. As can be seen from Table 2, the number of observations when stimulation or inhibition of NK activity of PBL, or absence of an effect were seen, using all E:T ratios, were similar in  Table 2).
The modulatory action of auto-IFNs on PBL possessing NK cytotoxicity in MS patients was the same as in healthy donors (Table 3), and these data may indicate that the stimulatory effects of auto-IFNs detected in MS patients were not related to changes in the properties of these IFNs themselves or in the sensitivity of the cells to regulatory IFN influence, thus suggesting the existence and special role of some additional factor controlling the regulatory properties of IFNs. Evaluation of NK activity of the blood cells obtained from healthy donors and MS patients on the 1st and 8th days of study showed that NK activity of PBL, expressed in both mean values of the IC and comparison of the spectrum of its variability, did not change over I week (Table 4). Thus, a change in NK activity of PBL in analysing IFN regulatory action on the cytotoxic function of the cells can be excluded.
Moreover, the data in the Table 4 show that PBL NK activity in MS patients was lower than in healthy donors but was sufficient for the evaluation of IFN action, whether enhancing or reducing.
A possible influence of PHA, which may be present in IFN-7-containing supernatants, could be also excluded. Control tests were performed, including a study of in vitro PHA P action on PBL NK activity in three healthy subjects and four MS patients. PHA P was used at doses equal to those inducing IFN-, or auto-IFN-7 production. PHA P did not affect in vitro PBL NK cytotoxicity in either healthy donors or MS patients (data not shown).
The IFN-containing supernatants obtained from PBL of healthy donors, acting on IFN-enhanced NK activity of MS patients' PBL completely abolished the stimulatory effect of auto-IFNs on PBL NK cytotoxicity in MS patients (Fig. 2). The action of the suggested inhibitor of IFN regulatory action on NK (n is the number of observations) activity was evaluated by adding it, at doses corresponding to those used for induced auto-IFNs, to a PBL suspension stimulated with auto-IFN, and incubating the mixture as described above for auto-IFNs.
After this procedure TRPs became equal to 6 + 5% (p < 0.05) for IFN-and to 9 + 7% (p < 0.05) for IFN-y. The difference was statistically significant for both these auto-IFNs (Fig. 2), and the values of TRPs corresponded to those obtained for PBL from MS patients possessing NK activity without any treatment. This abolition of the IFN stimulatory effect on PBL NK cytotoxicity by allogeneic factors was not surprising because a negative regulation was found to be efficient not only for autologous but also for allogeneic systems in healthy donors (data not shown).  It is interesting that the presence of the IIRA in IFNcontaining supernatants, and its 'antiregulatory' action on the properties of IFNs do not change the antiviral activity of preparations, which can be evaluated by a standard method that tests the protective effect of IFN on monolayer human fibroblast cultures infected with VSV. 31 These supernatant preparations abolished the auto-IFN-induced stimulation of PBL possessing NK activity in MS patients up to the values of TRPs characteristic of the initial, untreated suspension of PBL. Thus, the blocking properties of the IIRA can be exerted in both an autologous and an allogeneic system.
MS represents a multifactorial immunological abnormality associated with a strong deficiency of NK activity [14][15][16][17] probably caused by a functional defect of the IFN system. 18,35 This IFN-dependent NK deficiency is linked with a reduction in IFN production by practically all types of cells involved in the process. 18'19'22 It is also associated with anomalous in vitro responses of NK themselves as PBL possessing NK activity to the IFN regulatory action, 14,17,2 and is probably located at the stage of NK differentiation controlled by IFN. An interleukin (IL)-dependent step of maturation may also be involved, because IL-2 was found to augment NK activity of PBL in MS patients more slowly than in healthy donors, and induced less IFN-,. 6 On the other hand, some authors describe activatory changes in MS patients. Maintenance of excess IFN in serum and cerebrospinal fluid of MS patients was reported by Degr et al. 24 and Hertzog et aL, 19 accompanied by dissociation between the cytotoxic and IFN-producing functions of PBL possessing NK activity. 2