Background and Prior Art

Publications referred to herein are identified in full at the end of this specification.

It is well recognized that macrophages play an essential role in immune and inflammatory processes, both at a cellular level and via the release of mediators such

as interleukin-1 (IL-1), colony stimulating factor (CSF) and tumour necrosis factor (TNF) (Dinarello, 1984; Burgess and Metcalf, 1980; Le and Vilcek, 1987).

The thymus is a key organ in the generation and maturation of lymphocytes in mammals, particularly during foetal and neonatal life. Processing by the thymus is required for the production of T lymphocytes, of which various subsets are required for specific types of immune response. These subsets, such as helper T cells, suppressor T cells, cytotoxic T cells etc. may be identified by the specific antigens present on their cell surfaces.

Thymus ^' cell suspensions may be divided into light and dense cell populations by fractionation in density gradients of Percoll (Salisbury et. al., 1979; Percoll is a trade mark of Pharmacia AB) .

In rats, the dense cell fraction comprises immature thymocytes which are unable to respond to Concanavalin A, and thymic epithelial cells, which produce keratin.

The viability of dense, immature rat thymocytes was found by the present inventors to decrease markedly when a suspension of the cells in complete nutrient medium was incubated in vitro over a four hour period. This was attributed to the phenomenon of programmed cell death, or apoptosis.

It has been estimated that approximately 90% of the cells which are generated within the thymus die in situ (Kinnon et. al, 1986) . This population of thymocytes is derived from the thymic cortex, and the cells are characterized by their high density and susceptibility to cortisone-mediated apoptotic death (Weissman, 1986). The essential role of thymic epithelial cells in thymocyte maturation is well documented by studies in nude mice and rats in which the ectodermally derived epithelial cells have failed to

develop (Douglas-Jones et. al., 1981). In the cortex of the normal thymus, the majority of the stromal cells are epithelial in nature, and have long processes which reticulate among the cortical thymoctyes. In addition, the cortex contains a small number of macrophages which are primarily located in the region adjacent to the medulla (Adkins et. al., 1981).

The epithelial cells in both the cortex and the medulla express both class I and class II histocompatibility antigens (Van Ejwik et. al., 1980) and are thought to induce or select for self class II antigen recognition by thymocytes (Berrih et. al. , 1985). Thymic epithelial cells have also been shown to be the source of thymic hormones which induce the appearance of maturation markers on the thymocytes (Berrih et. al., 1985). More recently, the epithelial cells have also been shown to be a source of IL-1 (Le et. al., 1987).

We have now surprisingly found that a factor released by macrophages in response to stimulation with Iipopolysaccharide is able to protect thymocytes from apoptotic death.

The factor appears to stimulate immature, non-proliferating thymocytes to differentiate to a stage at which they are able to survive. Without wishing to be bound by any proposed mechanism for the observed beneficial effect, it is thought that the macrophage-derived factor binds to a common determinant of the la antigen complex on thymic epithelial cells. This in turn stimulates the thymic epithelial cells to release a factor which promotes survival of thymocytes.

The macrophage-derived factor is clearly different in its biochemical and functional properties from previously known factors such as tumour necrosis factor and interleukin-1.

Summary of the Invention

According to one aspect of the present invention there is provided a monokine released by macrophages in response to stimulation with Iipopolysaccharide, said monokine having the following ^• properties:

Relative molecular weight 36000 kD; Isoelectric point 6.3 - 6.4; Stable at temperatures up to 70°C; Stable at pH 2 to 10;

Activity destroyed by: reduction with

2-mercaptoethanol treatment with proteinases, or treatment with urea; Does not stimulate proliferation of fibroblasts; and Binds to common determinant of la antigen on thymic epithelial cells.

According to a second aspect of the invention, there is provided a method of producing said monokine, comprising the steps of incubating macrophages in nutrient medium in the presence _. of bacterial Iipopolysaccharide or muramyl depeptide for 5 minutes to 2 hours, and recovering the monokine.

Preferably Iipopolysaccharide is used to stimulate monokine production.

Preferably incubation is for 2 hours. Preferably recovering the monokine comprises the steps of: recovering medium from macrophage cultures, removing Iipopolysaccharide or muramyl dipeptide, removing material of Mr less than about 20,000, subjecting the remainder to sequential steps of gel filtration, ion exchange chromatography, and hydrophobic interaction chromatography, and recovering fractions having monokine activity.

Further purification may be effected by methods such as preparative gel electrophoresis and/or high performance liquid chromatography, which are well known to persons skilled in the art. According to a third aspect of the invention there is provided a factor produced by thymic epithelial cells in response to stimulation with said monokine, said factor having the following properties:

Relative molecular weight 320,000; Activity destroyed by trypsin treatment or by boiling;

Does not require active protein synthesis for production;

Binds to immature thymocytes; Protects immature thymocytes from apoptotic death; and

Binding to thymocytes inhibited by preincubation with antibody to CD4 antigen.

According to a fourth aspect of the invention, there is provided a method of bioassay of the monokine comprising the steps of:

(a) adding a sample of fluid containing or suspected of containing monokine to either dense thymic cells, or whole thymus suspension depleted of adherent cells, in nutrient medium in the absence of serum and indicator dye for a period of 3 to 16 hours, preferably 3 to 7 hours,

(b) adding MTT (3-(4, 5-dimethylthiazol-2-yl)- 2,5-diphenyl tetrazolium bromide)

(c ) incubating for 1 hour at 37°C,

(d) removing untransformed MTT,

(e ) adding isopropanol to the samples,

(f ) estimating extinction at 540 or 560 nm, and

(g ) calculating the amount of monokine in the sample .

Preferably the assay is carried out in 96 well miσrotitre plates, and each well contains 0.5 to 5 x 10 6, more preferably 2 x 10 , dense thymic cells in 100 ul medium, suitably RPMl 1640 medium containing no phenol red or serum.

Preferably 100 μl of sample to be assayed is used per 2 x 10 cells.

The preferred incubation time before adding MTT is 3 hours if adherent cells are not first depleted, and 7 hours if adherent cells have been depleted.

Detailed Description of the Invention

The invention will now be described in detail by way of reference only to the following non-limiting examples, and to the accompanying drawings, in which: Fig. 1 shows viability of the dense thymic cells, alone or in the presence of stimulated macrophage supernatant. Percentage cell viability is shown as a function of time. Results are the mean + s.e.m. of 25 experiments. By Student's t-test, *P<0.05 and **P<0.01;

Fig. 2 shows results of molecular weight and isoelectric point measurement of the monokine (a) Molecular weight determination. Sephacryl S200 column run. A column (100 cm x 1.6 cm) calibrated with Combithek II proteins in DPBS with 0.1 mol/1 NaCl was run with a 5 ml sample under identical total conditions. Flow rate 10.8 ml/h, sensitivity x 0.1, chart speed 2 cm/h., run time 17 hours. All of the fractions were tested and MF activity was detected in only one (0.8 ml) fraction. Hatched area indicates activity, (b)

Iso-electric point. LKB broad range ampholine gel (pH 3-9). Initial run, activity localized between 6 and 7 with reference to a standard curve. Second run, samples applied as a 10 cm band with standard proteins at either end. Standard curves were identical and 2 mm strips of

gel were taken between pH 6 and 7 and the activity was localized to ^' one strip: pH 6 .3 - 6.4. Hatched area indicates activity;

Fig. 3 shows the production of the macrophage factor in response to varying levels of two different stimuli. Production was determined by the titre resulting in 100% protection in the thymocyte viability assay. No protection refers to anything less than 100% protection. The supernatant was collected at three different times as shown. Results are the mean of four experiments;

Fig. 4 shows the time course for the production of the macrophage factor. Supernatants from macrophages stimulated with 20 g/ml LPS were collected at various time intervals after stimulation as shown. Production was expressed as the titre of macrophage factor resulting in 100% protection in the thymocyte viability assay. Any protection less than total was referred to as no protection. Representative experiment of two experiments;

Fig. 5 shows the effect of protein synthesis inhibitors on the production of the macrophage factor. Three protein synthesis inhibitors, actinomycin D (1 ug/ml), cycloheximide (10 μg/ml) or puromycin (100 μg/ml) were added to macrophage cultures simultaneously with LPS (20 _j ig/ml) and supernatants were collected at six intervals between five minutes and two hours for testing in the thymocyte viability assay. Protective activity of the macrophage factor at the various times is indicated by the percentage viability of the thymyocytes at three hours. Results are the mean of three experiments;

Fig. 6 shows the effect of control supernatants on the production of the macrophage factor. Three control supernatants, lysed macrophage supernatant, trypsin treated lysed macrophages and unstimulated macrophage supernatants were compared to macrophage

supernatant from macrophages stimulated with 20 μg/ml LPS. The percentage viability of the thymocytes over four hours is shown. Results are the mean of five experiments; Fig. 7 shows the effect of restimulation of macrophage cultures with LPS on the production of the macrophage factor. Supernatants were collected at two hourly intervals and cultures were restimulated with either LPS only (20 μg/ml) or LPS (20 μg/ml) and indomethacin (10 M) in new medium. Production of the factor was expressed as the titre resulting in 100% protection in the thymocyte viability assay for supernatants collected at two, four, six and eight hours after initial stimulation; Fig. 8 shows the role of the la molecule as a receptor for the monokine (representative experiments are shown). In every experiment, the unprotected populations were less than 70% of the control population by 4 hours incubation. (a) Stimulated macrophage supernatant can be mimicked by a monoclonal antibody to la (MRC 0X6) . This effect is also mediated via an epithelial cell. Control DCF; DCF + 2.5 μg/ml MRC 0X6 added (reproduced in four experiments): la-negative DCF with 2.5 μg/ml MRC 0X6 added (reproduced in three experiments); la-negative DCF with supernatant from thymic epithelial cells stimulated with 2.5 μg/ml MRC 0X6; and DCF + 2.5 μg/ml MRC 0X3, the monoclonal antibody to rat la which is strain-specific in its recognition (reproduced in three experiments), are compared.

(b) Removal of MF activity by its absorption on to la-positive cells. MF absorbed on to thymic epithelial cells prior to addition to DCF (reproduced in four experiments) ; MF absorbed on to rat spleen cells (reproduced in three experiments); MF absorbed on to rat spleen cells depleted of la-positive cells (using MRC

OX6/complement lysis) (reproduced in three experiments); and MF absorbed on to rat RBC (reproduced in two experiments) were compared.

(c) The binding site of the MF on spleen cells is blocked by an antibody to la (MRC 0X6) . MF were absorbed on to spleen cells (as in Fig. 3b) but spleen cells were previously incubated with 2.5 μg/ml of MRC 0X6. The possibility of carry over of MRC 0X6 into the assay system was eliminated by labelling the antibody with 125I (reproduced in three experiments). MF absorbed on to spleen cells previously incubated with rat-specific antibody to la (MRC 0X3) (reproduced in three experiments) ;

Fig. 9 shows the role of the Fab portion of MRC 0X6 in stimulating protective activity and in blocking the binding of MF. Populations compared were

DCF + MRC 0X6; DCF + Fab MRC 0X6;

DCF previously incubated with Fab MRC OX 6 + MF; Fig. 10 shows the role of thymic epithelial cells in thymocyte viability. A representative experiment of percentage cell viability as a function of time is shown. Note that in every experiment the unprotected populations were less than 70% of the control population by 4 hours incubation. Populations compared were

Control, DCF depleted of la-positive cells (using MRC 0X6 and complement in a lysis technique) ; la-negative DCF (i.e. DCF after the depletion of la-positive cells) with the addition of MF (reproduced over four experiments) ; la-negative DCF with 1 hour supernatant from unstimulated thymic epithelial cells (reproduced in two experiments) ;

la-negative DCF with supernatant from thymic epithelial cells incubated for 1 hour at 37°C with MF (reproduced in two experiments);

Fig. 11 shows results of molecular weight determination of the thymic epithelial cell factor using Sepharose C1-6B chromatogr phy;

Fig. 12 shows the time course for the production of the thymic epithelial cell factor;

Fig. 13 shows the enhanced proliferation of dense immature thymocytes in response to either whole macrophage supernatant or an antibody to rat la (2.5 μg/ml of clone MRC 0X6 or clone MRC 0X3). Representative experiment of eight experiments showing the mean + standard deviation for quadruplicate cultures; Fig. 14 shows the enhanced proliferation of immature thymocytes in response to partially purified macrophage supernatant. The addition of partially purified macrophage supernatant to dense immature thymocytes results in an increase in proliferation over that of a control population. Pooled data from three preparations of purified Mr 36 000 monokine showing means values + standard error of the mean;

Fig. 15 shows the effect of purified macrophage factor on thymocyte proliferation. Addition of macrophage supernatants which have been treated in one of three different ways restores the proliferative response of the adherent cell depleted population to that of a control population. Representative of two experiments showing mean + standard deviations; and Fig. 16 shows the characteristics of the thymocyte response. The response of divided, recombined and whole thymus to varying levels of Con-A in the presence of 0.5 μg of LPS is shown. Representative experiment of five experiments showing the mean + standard deviations of quadruplicate cultures.

Abbreviations used hereinafter are defined as follows

Con-A Concanavalin-A

DCF Dense cell fraction of thymus

5 DPBS Dulbecco's phosphate-buffered saline

DTEC Dense thymic epithelial cells

FCS Foetal calf serum

IL-I Interleukin-1 0 LPS Lipolysaccharide

MDP Muramyl dipeptide

MF Macrophage factor (monokine)

Mr Relative molecular weight

RBC Red blood cells 5 TECF Macrophage-induced thymic epithelial cell factor

TNF Tumour necrosis factor

Example 1 Preparation of macrophage supernatant

Macrophage cultures were prepared from the 0 peritoneal washings of 6-8 week old Wistar-Furth rats. The cells in RPMI-1640 medium containing 10% FCS were seeded into tissue culture flasks at a density of 1 x 10 6 cells/ml (5 x 10 per flask), and allowed to adhere for 1 hour at 37°C. The monolayers were then washed five times

25 with DPBS, 5 ml serum-free medium was added per flask, and the cells were stimulated with 20 μg/ml Iipopolysaccharide from Salmonella enteritidis (Sigma) . The supernatant was harvested at 2 hours, filtered through an XM300 filter (relative molecular weight (Mr)

30 300,000 cut off) to remove high Mr components, dialysed against DPBS and concentrated over a PM10 membrane filter (Mr 10,000 cut off). Both XM300 and PM10 ultra-filtration membranes were from Amicon Ltd. By titration it was shown that the described MF activity was

35 increased at least 500 times over that in control.

unsti ulated macrophage cultures. The cultured peritoneal cells were identified as macrophages by the criteria of being adherent cells, with 99.6% showing a positive reaction for lysozyme.

Example 2 Effect of Macrophage Supernatant on the Viability of Thymus Cells Thymus glands were removed from 4-8 week old male and female rats, taking care to dissect the thymus free from surrounding tissues including lymph nodes. Thymuses were washed in RPMI-1640 containing 10% FCS, sliced and pressed through a 0.5 mm pore size stainless steel wire mesh. The resulting thymic suspension was fractionated on a gradient of Percoll according to a described method (Salisbury et. al., 1979) The dense fraction was then washed three times in RPMI-1640 with 10% FCS.

Purified dense thymic epithelial cells and dense thymocytes were prepared by treating whole thymus suspensions with antibody and complement prior to fractionation on Percoll. Thymus cell suspensions

Q containing 1 x 10 cells in 5 ml RPMI-1640 were treated with monoclonal antibody (MRC 0X52, a rat pan T cell and thymocyte marker; Robinson et. al., 1986 for thymocytes, or MRC 0X6, directed against a common determinant on rat la, obtained from Sera Labs, for la-positive cells) at 2.5 μg/ml final concentration, together with complement at 2% final concentration. The preparations were incubated for 1 hour at 37°C and then layered on to Percoll. The isolated populations were washed and characterized by indirect immunofluorescence and by enzyme histochemistry. Results for 10 thymus preparations are summarized in Table 1.

Table 1. Identification and preparation of cell populations.

Populations Preparation Identification

DCF of thymus (immature Fraciionation of whole thymus on (1) 12-8 ± 1-9% keratin +ve thymocytes and dense epithelial Percoll colloidal silica density (2) 12-6 ± 2-4% remain after cells) gradient M RC OX52-complement mediated lysis of rhymocyres

(3) Non-responsive to lecrins

(4) All cells — ve for lysozyme

(5) All cells — ve for acid phosphatase

DCF depleted of la - ve cells MRCOXό/complement mediated (I) Remaining cells lysed with lysis of DCF MRC OX52 and complement

DTEC MRC OX52/comρlement mediated (I) 96-9 ± 3-7% -l-ve for keratin lysisofDCF (2)98-9 ± I -6% -Hve for MRC 0X

(3)99-0 ± 1-5% +ve for MRC 0X

(4) All cells -ve for lysozyme

(5) All cells -ve for acid phosphatase

(6) Homogeneous c population-ultrasrructure

09

H

C H m 1,2 Immunof luorescence with affinity-purified antibody to rat keratin and antibody to rat lysozyme respectively. Both antibodies were raised in rabbits. m -1

Aliquots of DCF were counted, and the number of viable cells (excluding trypan blue) was expressed as a percentage of the starting population. The results shown in Fig. 1 indicate a rapid loss of viability. This was typically in the range of 40-70% of the starting population after 4 hours incubation.

The addition of 50 ul/ml of supernatant from macrophage cultures which had been stimulated with 20 μg Iipopolysaccharide/ml 2 hours previously protected the dense cell fraction from loss of viability. Neither Iipopolysaccharide (from 1-20 μg/ml) alone, nor supernatant from unstimulated macrophages protected the DCF. The supernatant from stimulated macrophages was titrated, and was found to be protective at an end-point dilution of 1/500. Approximately half of the peritoneal cells which were seeded into culture flasks became adherent macrophages. This gave 2.5 x 10 cells per flask in 5 ml medium. Therefore 50 μl of supernatant

4 represents the release from 2.5 x 10 macrophages. This equates to this number of macrophages releasing enough

7 activity to protect 500 x 10 dense thymic cells, or one macrophage protecting 2 x 10 dense thymic cells.

The loss of viability in vitro was confined to the dense, immature population and did not affect the buoyant fraction or the thymic epithelial cells alone. The protective effect was found in whole supernatant, and in the fraction retained above the PM10 membrane. Although these membranes have a nominal cut off of Mr 10,000, it was found that under the conditions used, soybean trypsin inhibitor (Mr 23,000) was filtered out while bovine serum albumin (Mr 67,000) was retained. Repeated assessment of macrophage cultures by examining at least 200 cells per culture showed that -99.6 + 0.6% of the adherent cells were positive for lysozyme by indirect immunofluorescence staining. The specificity of the anti-rat lysozyme serum was demonstrated by a

single arc in immunoelectrophoresis against a rat neutrophil lysate and against the purified protein used for immunization. A control or second stage antibody alone was negative. The cells in the cultures were 100% positive for acid phosphatase, the reaction being ^■ characterized by diffuse cytoplasmic staining of variable intensity. For further studies a batch of fractionated macrophage supernatant was prepared from the peritoneal washings of 30 Wistar-Furth rats. This preparation (MF) was used at a concentration equivalent to a 1/250 dilution of whole supernatant in subsequent protection studies.

Example 3 Characterization of Monokine (MF) (a) Physical properties. The Mr of the monokine was determined by gel filtration fractionation on Sephacryl S200. The results are shown in Fig. 2a. MF eluted at Mr 36,000.

The isoelectric point of the MF was determined by isoelectric focusing. Standard proteins and 100 μl of MF were loaded on to a broad range (pH 3-9) LKB ampholine gel and run to equilibrium. A standard curve was prepared from the reference proteins as shown in Fig. 2b and strips of the gel were cut at known intervals in relation to this. The strips were homogenized, dialysed against DPBS and tested. A second run was used to narrow the range further. The activity was localized between pH 6.3 and 6.4.

The effect of heating was assessed by treating aliquots of MF for 30 minutes at 40, 56, 60, 70 and 80°C. Activity was destroyed at 80°C, but was stable at lower temperatures. The effect of pH on the stability was tested by adding MF to appropriate 0.1 mol/1 buffers at pH 2, 4, 6, 8 and 10 for 60 min at 37°C. Samples were

then dialysed against DPBS and tested for activity. MF was stable over this pH range. MF activity was lost following reduction with 5% 2-mercaptoethanol.

The effect of proteinases on MF activity was assessed by adding MF to 10 μg/ml solutions of enzyme for ^' 60 minutes at 37°C. The proteinases were papain, pronase, thermolysin and trypsin, which were used in RPMI-1640, pH 7.4, and pepsin in RPMI-1640 adjusted to pH 4.0 with 1 mol/1 HC1. The enzyme-MF mixtures or enzyme controls were then added to DCF in RPMI-1640 containing 10% FCS as a source of proteinase inhibitors . Proteinases alone had no effect on thymocyte survival, whereas all five proteinases destroyed MF activity. Thus the monokine is at least partly protein in nature. The results of these studies are summarized in Table 2.

TABLE 2 Physical Characteristics of Monokine.

Parameter Monokine

Molecular weight Mr 36,000 Isoelectric point 6.3 - 6.4 Heat stability Stable at 70°C, activity lost at 80°C. pH stability Stable between pH 2 and 10 Proteinases Activity destroyed by papain, pepsin, pronase, thermolysin and trypsin

5% 2-Mercaptoethanol Destroys activity 8 mol/1 urea Destroys activity

(b) Functional Properties.

MF was tested for its ability to stimulate 3T3 fibroblasts to proliferate both in the presence and the absence of FCS. Proliferation of fibroblasts was expressed as the mean of quadruplicates of counts per 3 minute of [ H]-thymidine uptake by the cells. As controls, fibroblasts alone and fibroblasts with whole macrophage supernatant were tested. The results given in

Table 3 indicate that while the whole supernatant has stimulating activity both in serum-containing and in serum-free conditions, MF has none under either condition.

TABLE 3 Effect of Macrophage Culture Supernatants on Fibroblast Proliferation

Sample Serum-free 10% FCS

(cpm) (cpm)

Control (fibroblasts alone) 1010 + 99 1880 + 315 Whole macrophage supernatant 5075 + 434 4305 + 1300 Partially purified macrophage 903 + 125 1586 + 588 supernatant (MF)

Example 4 Thymocyte Viability Assay

Dense thymus cell fractions prepared as described in Example 2 were divided into one ml aliquots containing 1 x 10 cells. Test supernatants were added as 50 μl samples to these cells and viability was assessed at hourly intervals by the exclusion of trypan blue (0.4% solution). Viability was expressed as a percentage of the original cell number.

The titre of the activity was defined as the reciprocal of the maximum dilution of unfractionated macrophage supernatant which gave complete protection at four hours in the thymocyte viability assay. This

7 applied to 100 μl of supernatant added to 1 x 10 cells in a total volume of 1 ml.

Unfractionated thymus cell suspensions show no decrease in cell viability with time. However, depletion of macrophages, e.g. by adherence to plastic, results in apoptotic death of immature thymocytes in the suspension.

This was prevented by addition of the Mr 36,000 monokine. Surprisingly, if the thymocytes were first treated with corticosterone, cell death was increased.

The monokine evidently induces di ferentiation in a population of la prothymocytes, which thereby become corticosterone-sensitive. This is supported by results described below with Con A.

Example 5 Further Characterisation of Synthesis and Release of the Monokine. The time course of production for the factor, together with the effect of different stimuli at various doses and at various times was examined. In addition, the nature of the production of the macrophage factor was studied using three different protein synthesis inhibitors and trypsin treated and untreated lysed macrophage preparations.

(a) Macrophages were prepared as in Example 1, and were stimulated with either LPS (1-20 μg/ml) or MDP (0.5-2.5 μg/ml) . Supernatants were collected at various time intervals between five minutes and forty-eight hours according to the experiment. In the case of the unstimulated macrophage supernatant, an identical procedure was followed without the addition of the stimulant, and the supernatant was collected two hours after the replacement of serum-free medium.

Varying concentrations of LPS and MDP were added to freshly established peritoneal macrophage cultures, and the supernatant was collected after two, eight or twenty-four hours incubation and tested for activity in the thymocyte viability assay. The results are shown in Figure 3. For LPS, there was a proportionate increase in titre from 1 to 10 ug LPS and then an abrupt increase from ten units at 10 jug LPS to 500 units at 20 μg LPS. There was no demonstrable activity at eight or at twenty-four hours after the addition of LPS. The response to MDP was quite different in both the titre achieved and in the persistence of the response, so that for 2.5 μg MDP, although the titre was only one unit, it persisted for at least twenty-four hours. Controls were set up to test the effect of the stimulants alone or in combination with the partially purified macrophage factor. Neither stimulant protected the thymocytes or affected the activity of the macrophage factor. (b) The time course for the release of the macrophage factor was examined using 20 μg/ml LPS as the stimulant. As shown in Figure 4, low level activity was released from the macrophages within five minutes after the addition of the stimulant. After one hour the activity had increased ten-fold, with the peak response of 500 units being reached at two hours. The activity decreased to 100 units at three hours, while from four hours to forty-eight hours there was no demonstrable activity. The yield as assessed by the number of units of protective activity in the thymocyte viability assay is greatest in response to 20 μg/ml LPS when the supernatant is collected two hours after the cells are challenged. Increasing the culture time beyond two hours led to a decrease in the number of units of activity at three hours and a loss of protective activity beyond that time and up to forty-eight hours.

(c) Effect of Inhibitors of Protein Synthesis.

Macrophage supernatants produced in the presence of inhibitors of protein synthesis were prepared as follows: the macrophages were allowed to adhere and were then washed to remove any traces of serum. The protein synthesis inhibitors were used in the following concentrations: actinomycin D 1 μg/ml, puromycin 100 μg/ml, cycloheximide 10 μg/ml.

Inhibitors were added to macrophage cultures at the same time as 20 μg LPS and the supernatants were collected at time intervals up to two hours. They were dialysed to remove low molecular weight inhibitors and then tested for protective activity in the thymocyte viability assay. The results shown in Figure 5 show that in the presence of each of the three protein synthesis inhibitors, protective activity was demonstrable in the culture supernatants at five and ten minutes after LPS challenge. By fifteen minutes, however, activity had been lost in the puromycin and cycloheximide treated cultures, whereas it was still present in the cultures which had been incubated with actinomycin D. From thirty minutes to two hours, activity was lost in all of the treated cultures.

The early release of low level activity from five minutes onwards suggested that the factor was already present within the macrophages and was released. In addition it was shown that lysates of unstimulated cells also contained protective activity, although this activity was not titratable. Addition of the protein synthesis inhibitors, actinomycin, puromycin and cycloheximide to cultures at the same time as the LPS demonstrated that de novo protein synthesis was necessary for production of the factor. In contrast to cultures with LPS alone, protective activity could not be detected beyond ten minutes with puromycin and cycloheximide or after fifteen minutes with actinomycin. Hydrolysis of

surface proteins by trypsinization prior to washing or by lysis of the cells failed to remove the protective activity, indicating that the factor was not a surface protein on the macrophage, but was probably cytosolic. These results are shown in Figure 6.

Example 6 Effect of Restimulation of Macrophage Cultures Restimulation of cultures with LPS was examined to see whether yield of the monokine could be increased. In addition, the abrupt decrease in titre from a peak at two hours after stimulation to no detectable protective effect at four hours was studied by examining the effects of restimulation with LPS at two, four and six hours of culture. The abrupt decrease in titre of protective activity after 2 hours could have been due to prostaglandin production by the macrophages. The spontaneous production of prostaglandin E _? by thymic phagocytes in culture is greatly enhanced by adding LPS to the cultures; the prostaglandin response to LPS is reduced 10-fold by indomethacin (Papiernik and Homo-Delarche, 1983).

We found a further peak of protective activity when 10 -5M indomethacin was added to cultures at 4 hours together with LPS and new medium, but not when cultures were restimulated in the absence of indomethacin.

Supernatants were taken after two hours from cultures stimulated with 20 μg/ml LPS. At this time, the cultures were restimulated with 20 μg/ml LPS alone or in combination with 10~ M indomethacin. The supernatants were collected at four hours and the cultures were restimulated. This procedure was repeated at six hours and the final supernatants were collected at eight hours.

The results are shown in Figure 7, and indicate that macrophages were anergic to further stimulation. In this context, it was noted that 10 nM of prostaglandin

E-, added together with LPS at 20 μg/ml to fresh macrophage cultures, totally removed the activity at two

_5 hours. Indomethacin (10 M) , added together with LPS to fresh cultures, did not alter the peak titre at two hours; however, in the presence of this inhibitor of prostaglandin synthesis macrophages which were restimulated at four hours did release low titre activity at six hours of incubation. There was no effect of indomethacin, however, on restimulation at two hours or at six hours.

Example 7 Partial Purification of the Monokine Macrophage supernatant was prepared as in

Example 1, and filtered through an XM 300 membrane, to remove the bulk of the Iipopolysaccharide, and then through a PM 10 membrane. The fraction above the membrane was retained and sterile filtered as described. The molecular weight was determined by gel filtration fractionation on Sephacryl S200, and the activity was localized to an Mr 36,000 fraction (Yield = 80% of loaded sample). The macrophage factor was further purified by ion exchange chromatography on DEAE Sephacel (Pharmacia) . The active fraction (5ml) from Sephacryl S200 was loaded onto a DEAE Sephacel column (1.6 x 60 cm), and the activity eluted from this column in 2 x 2 ml fractions at 120-124 ml after a gradient of 200 mis of 0.05 M Tris (pH 8.0) + 200 mis of the same buffer with 0.5M NaCl was applied (Yield = 50% of loaded sample). The active fractions were pooled and sterile filtered, and half of the active fraction was dialysed against 1M (NH.J-SO. at pH 7.0 in preparation for chromatography on Phenyl-Sepharose CL-4B (Pharmacia) . A 7 ml sample was applied to 10 mis of Phenyl-Sepharose in a 0.9 cm

diameter column with adaptor, the matrix having been equilibrated with 1M (NH ₄ ) ₂ S0 ₄ , (pH 7.0). The flow rate was 20 ml/hour and a 100 ml gradient of 1M ( H ₄ ) ₂ S0 ₄ , (pH 7.0) + 100 mis H ₂ 0 was applied immediately after the addition of the sample and 2 ml fractions were collected. ^■ The activity was recovered in 2 fractions, which represented 2% of the total gradient (Yield = 27% of loaded sample).

Example 8 Receptor Site for the Monokine. During the course of optimizing procedures for the depletion of la-positive cells it was observed that MRC 0X6, in the absence of complement, promoted thymocyte survival (Figure 8a), thus mimicking the effect of MF. As expected, this antibody had no effect on the survival of the la-negative DCF. When incubated with isolated thymic epithelial cells at 1.2 x 10 per ml, it promoted the release of thymocyte survival activity. In contrast, MRC 0X3, which recognizes a rat strain-specific epitope on la, failed to protect. The possibility that MF also bound la was probed initially by attempting to deplete MF activity by absorption with different types of cells, as outlined earlier. MF was absorbed at 0°C, using 3 cycles of incubation for 1 hour with 1 x 10 cells. As shown in Figure 8b, the DTEC effectively removed the MF activity. Spleen cells also removed the activity, whereas spleen cells depleted of la-positive cells did not. RBC were also ineffective. Similarly, MF was absorbed on to spleen cells which had previously been incubated at 0°C with MRC 0X6 or MRC 0X3, and the supernatant was tested for protective activity. In order to negate the possibility of carry over of MRC 0X6 into the viability

125 assay, the antibody was labelled with I. Thhe specific activity of the dialysed product was 2.09 x 10 5 disintegrations/min per μg protein. No carry-over of

radioactivity was detected. The results shown in Figure 8c indicate that prior binding of MRC 0X3 did not prevent absorption of MF, whereas prior binding of MRC 0X6 did so. This provided evidence that MF bound to surface la at a site which was the same as or nearby the site recognized by MRC 0X6. The finding that MRC 0X3 did not block absorption of MF mitigated against an intermolecular steric blocking of MF by MRC 0X6. The binding site of the MF was further examined by studying the competitive effect of univalent Fab MRC 0X6 antibodies prepared as described. Fab 0X6 at 2.5 μg/ml failed to protect the DCF in the thymocyte viability assay. Pre-incubation of the DCF with this Fab antibody preparation for 15 min. at 37 C did, however, effectively block the protective action of MF, indicating a close relationship between the binding site of the antibody and the macrophage product (Figure 9).

Example 9 Mediation of the Protective Effect Via Thymic Epithelial Cells The DCF was depleted of la-positive cells by treatment with MRC 0X6 and complement as described earlier. Similarly, DTEC were prepared by lysis of thymocytes using MRC 0X52 and complement. Cell populations were characterized as described in Table 1. There was good agreement between the number of cells positive for keratin and those remaining after lysis of thymocytes in the DCF. The presence of macrophages and non-epithelial dendritic cells in this fraction was excluded on the basis of the buoyant nature of these cells and that all of the dense cells were accounted for by either anti-keratin staining of epithelial cells (12.8%) or MRC 0X52 staining of thymocytes (87.2%). In addition, none of the cells in this fraction gave a positive reaction for lysozyme or for acid phosphatase. Furthermore, the purified

epithelial cells appeared to be a uniform population by electron microscopy. The DTEC maintained viability over

24 hours in culture while the dense la-negative thymocytes showed 100% mortality over this time period.

7 Aliquots of MF (50 μl) failed to protect 1 x 10 thymocytes depleted of la-positive cells (Figure 2).,

The incubation of 50 μl MF with 1.2 x 10 ⁶ DTEC (in proportion to the percentage in the DCF) did, however, result in the release of protective activity by these cells (Figure 10).

Thus removal of the la-positive cells in the DCF left a population which displayed the described loss of viability over 4 hours. The addition of macrophage supernatant failed to prevent cell death, as did the addition of supernatant from unstimulated thymic epithelial cells. The addition of supernatant from thymic epithelial cells which had previously been incubated with MF was protective, suggesting that a cascade was occurring in which macrophage products were acting on thymic epithelial cells, resulting in the release of a factor responsible for the survival ^" of thymocytes.

Example 10 Production of the Thymic Epithelial Cell Factor The thymic epithelial cell factor was induced with either the macrophage factor or an antibody to a common determinant on la, as follows: Whole thymus cell suspensions were treated with MRC 0X52 and complement for thirty minutes at 37°C. The remaining cells were then fractionated on Percoll and the dense cell fraction which contained keratin-positive thymic epithelial cells was collected. These cells were divided into three aliquots and were cultured in tissue culture flasks at a density of 1 x 10 ⁸ cells/ml in RPMI 1640 medium. The first aliquot served as a control, the second aliquot was

stimulated with 2.5 μg/ml of MRC 0X6 and the remaining suspension was stimulated with the macrophage factor (at dilution equivalent to 1/250 of the original macrophage supernatant) . The cultures were incubated for three hours, after which the supernatants were collected. The "resultant supernatants were tested in the thymocyte viability assay using a dense cell fraction depleted of la positive cells. The macrophage-induced thymic epithelial cell factor (MF-TECF) was active at a concentration of 1/5000 and the antibody-induced epithelial cell factor (MRC OX6-TECF) protected the thymocytes at a dilution of 1/750 in the thymocyte viability assay. The control supernatant from unstimulated thymic epithelial cells was inactive in the thymocyte viability assay.

Example 11 Properties of the Thymic Epithelial Cell Factor The protective activity of the macrophage factor-thymic epithelial cell factor was destroyed by incubating the factor (at a 1/1000 dilution) with 1 mg/ml of trypsin. Activity was also destroyed by boiling the factor prior to testing it in the thymocyte viability assay, suggesting that it was probably protein.

The molecular weights of both MF-TECF and MRC OX6-TECF were determined by gel filtration fractionation on Sepharose CL-6B (Pharmacia) . The results are shown in Figure 11. A column (100 cm x 2.6 cm) was calibrated with ferritin (Mr 440,000), aldolase (Mr 158,000), albumin (Mr 68,000) and ovalbumin (MR 43,000) in PBS with 0.1 M NaCl and was run with a 1 ml sample under total identical conditions. The MF-TECF eluted at a molecular weight of 320,000 and the MRC 0X6-TECF eluted at 760,000.

A possible binding site for the MF-TECF was suggested by experiments using a monoclonal antibody to rat T helper cells which bound at CD4 (Sera Lab, Clone

W3/25). Whole thymus suspensions were incubated with this antibody, la positive cells were removed and the remaining cells were fractionated on Percoll for use in the thymocyte viability assay. The addition of MF-TECF to this dense cell fraction failed to protect the ^• thymocytes, suggesting that the availability of CD4 molecules was important in this process.

Example 12 Characteristics of the Release of the Thymic Epithelial Cell Factor. The MF-TECF was prepared as previously described but supernatants were collected from cultures at various intervals between five, minutes and three hours and tested in the thymocyte viability assay. As shown in Figure 12, low titres of activity are produced in the first fifteen minutes following which the titre increases steadily up to three hours. The addition of one of three protein synthesis inhibitors, actinomycin D, cycloheximide or puromycin had no effect on the production of MF-TECF, indicating that protein synthesis was not required.

These experiments suggest that the thymic epithelial cell factor, when induced by the macrophage factor, has a molecular weight of around 320,000. Induction of the thymic epithelial cell factor with MRC 0X6 (the antibody to a common determinant on la), results in the release of a protein of molecular weight 760,000. The large molecular weight of the protein together with the fact that the molecular weight of the product varied according to the method by which it was induced, suggested that the thymic epithelial cell factor might in fact be an Ia-ligand complex that is shed from the thymic epithelial cell. If this does occur, then the selection of thymocytes for which apoptosis is arrested could be based on their ability to bind the Ia-ligand complex that is shed. This appears to be related to the expression of

CD4 by the thymocytes since pre-incubation of immature thymocytes with anti-CD4 antibodies inhibits the observed protective effect of the monokine-induced thymic epithelial cell factor. It was not possible to absorb the monokine with the monoclonal antibody W3/25 to CD4, indicating that it is unlikely that this macrophage product is soluble CD4.

Example 13 Stimulation of Thymocyte Maturation by the Monokine. The immature thymocytes in the dense cell fraction do not proliferate in response to Con A. The effect of the monokine on their ability to respond to Con A was therefore tested.

The dense thymic fraction was incubated at 1 x 10 ⁶ cells per well with 2.5 μg of MRC 0X6 or MRC 0X3. The proliferative response to Con A was measured. While MRC 0X3 had only a slight effect, there was a substantial increase in proliferation with both MRC 0X6 and 2 hours whole macrophage supernatant, and an additive effect when the two treatments were combined, as shown in Figure 13. This trend was reproduced over eight experiments.

Incubation of the dense thymic cell fraction with partially purified macrophage supernatant (Mr 36,000 fraction from chromatography on Sephacryl S200) also resulted in an increase in the proliferative response to Con A. This fraction was used at a concentration which demonstrated protective activity in the thymocyte viability assay, and three different preparations were tested. All preparations showed a similar increase in proliferation, as shown in Figure 14, and the effect was reproduced over five experiments.

The depletion of adherent cells from whole thymus suspensions resulted in a decrease in thymocyte proliferation in response to varying levels of Con A and

0.5 μg of LPS compared with a control population. The mean of the ratios of the control population to the depleted population was 2.27 + 0.6. The addition to each well of 25 μl of whole macrophage supernatant (prepared from 1 x 10 cells stimulated with 20 μg/ml LPS and harvested at 2 hours) restored the response to that of the control. The purity of the macrophage cultures was confirmed by staining for lysozyme and acid phosphatase. The means of the ratio of the control response to the restored response was 1.0 + 0.07.

The restoration of the response with macrophage supernatant did not appear to be due to IL-1. Supernatant treated with 10 mmol/1 phenylglyoxal to remove IL-1 activity was still capable of restoring the response to control levels. Similarly, the response was restored by the addition of Mr 36,000 factor from chromatography on Sephacryl S200, at a concentration which was active in the thymocyte viability assay. These results are shown in Figure 15. In other experiments it was found that 0.5 μg

LPS increased thymidine uptake in response to a range of concentrations of Con A, but that the shape of the dose-response profile was unchanged. Similar results were obtained with MDP (1 μg/10 cells) or streptococcal cell wall fragments (0.5 μg/10 cells). Conversely, in the presence of a constant amount of Con A (0.25 μg/well), 1 pg of LPS sharply stimulated thymidine uptake, with gradual further increase up to 1 mg LPS. Again, similar results were obtained using MDP or streptococcal cell wall fragments.

Fractionation of the thymus cell suspension on a density gradient of Percoll confirmed that the buoyant fraction which contained mature thymocytes, macrophages, epithelial cells and non-epithelial dendritic cells proliferated in response to Con A and 0.5 μg of LPS. The dense fraction containing immature thymocytes and dense

epithelial cells responded only weakly to Con A and LPS. Recombination of the two fractions restored the response to a level equivalent to that of whole thymus, showing that cells had not been lost in the separation procedure. Figure 16 shows the response of the buoyant, dense, recombined and whole fractions to Con A and LPS.

These results indicate that macrophage products are responsible for the increased proliferative responses of thymocytes to co-stimulation with Con A and microbial products. The data support the efficient uptake of picogram quantities of a variety of bacterial components by macrophages, which comprised less than 0.1% of the thymic cell suspension. The evidence for the role for macrophages in the augmentation of the proliferative response was provided by a diminished response following depletion of adherent cells, and by the restoration of the activity by products from highly purified peritoneal macrophages. The active macrophage component was shown to be the Mr 36,000 protein, which did not have IL-1 activity as determined by a fibroblast proliferation assay.

Example 14 Effects of the Monokine in vivo.

It is known that the cytokines IL-1, TNF, and interleukin-6 can induce liver epithelial cells to secrete acute-phase plasma proteins. Macrophages are a major source of these cytokines.

A single intravenous injection of the partially-purified Mr 36,000 protein into rats induced a significant elevation of plasma fibrinogen and <tf ₂ ~πi _c icroglobulin levels.An injection of heated protein caused no response.

The amount of protein injected was equivalent to that released by 1 x 10 macrophages. This small amount of monokine produces a substantial effect in the

animal by a cascade phenomenon whereby responding macrophages secrete one or more cytokines, which in turn activate the acute phase response.

Example 15 An Automated Bioassay for the Monokine. ^' In order to monitor large numbers of samples conveniently, an assay which can be carried out in microtitre plates and in which many samples can read simultaneously is desirable.

We have developed an assay based on the conversion of a soluble, yellow compound (MTT) to an insoluble blue formazan by enzymes in living cells. The insoluble product is dissolved in isopropanol, and estimated spectrophotometrically. The assay can be carried out in 96 well microtitre plates of the type used for ELISA assays.

Two complementary variations in technique were used. The assay is based on the cleavage of the tetrazolium salt MTT (3-(4, 5-dimethylthiazol-2- yl)-2,5-diphenyl tetrazolium bromide) into a blue coloured product (formazan) by the mitochondrial enzyme succinate dehydrogenase (Denizot and Lang, 1986). The conversion takes place only in living cells, and the amount of formazan produced is proportional to the number of cells present. Thymic suspensions were either fractionated on

Percoll as described earlier, or depleted of adherent g cells by incubating 1 x 10 thymic cells in 5ml RPMI 1640

2 medium in a 75 cm tissue culture flask for one hour at

37°C. This depletes the suspension of thymic macrophages and destabilizes the population so that the immature, dense thymocytes undergo apoptotic death.

Using the dense cell fraction from Percoll gradient separation, cells were washed into RPMI 1640 medium, without serum or phenol red, at a density of 2 x 7 10 per 0.9 ml. Dilutions of macrophage supernatant or

of the fractions containing monokine activity were added

7 at 100 μl/2 x 10 cells. The cells were then aliquoted in replicates of eight into flat bottom 96 well culture plates at 2 x 10 cells/100 μl. The effective range is 0.5 x 10 ⁶ to 5 x 10 ⁶ cells/well but 2 x 10 ⁶ cells are preferred. The final row of cells received cycloheximide at 10 μg/ml instead of macrophage products. This arrests the protein synthesis necessary for apoptotic death, and provides a positive control at 100% viability. The plates were incubated at 37° for three hours and then MTT was added as 50 μl of a 1 mg/ml solution in RPMI 1640 medium. After one hour of incubation at 37°C, the plates were centrifuged at 800 x g for 10 minutes and the untransformed MTT was removed by inverting, flicking and blotting the tray. Two hundred microlitres of isopropanol was then added to each well, the plates briefly shaken and the formazan read at 540 or 560 nm (preferred) test wavelength against a 690 nm reference wavelength in an ELISA reader. Formazan generation in test samples was read with reference to both cycloheximide-treated positive controls and unprotected controls. The conditions for using adherent cell-depleted whole thymic populations were identical, except that the incubation step was for seven hours or sixteen hours (preferably seven) before adding the MTT.

The assay using fractionated dense thymic cells involves rapid apoptotic death and therefore a short incubation period while the assay using depleted adherent cells involves a slower rate of death and longer incubation time. It does, however, have the advantage that it is easy to establish, and is applicable to screening large number of samples.

Example 16 Distinction Between the Monokine and Previously known Macrophage-derived Factors. The characteristics of IL-1 and TNF in the rat have been described ("Lovett et al, 1983; Schmitt et al, 1986; Vilcek et al. 1986; Rupp et al. 1986).

The failure of the monokine of this invention to stimulate fibroblast proliferation in either the presence or absence of serum (Example 3(b)) suggests that it is neither IL-1 nor TNF. This is supported by the homogeneity of the factor with respect to both Mr and isoelectric point; both IL-1 and TNF have a number of molecular forms (Smith et al., 1986; Wood et al. 1985). Phenylglyoxal is an inhibitor of IL-1 activity (Klamfeldt, 1985; Krakauer, 1985). Whole macrophage supernatant from cultures incubated for 2 hours with 20 μg LPS/ml were treated with 10 mmol phenylglyoxal/ml. This treated supernatant was able to restore the response of adherent cell-depleted thymus cell cultures to Con A, as shown in Figure 15.

The time course for the production of this factor differs from that of IL-1 or lymphocyte activating factor in mice, humans and rats. Mizel (1981) showed that mouse peritoneal macrophages stimulated with phorbol myristate acetate reach a peak level of production of IL-1 around twenty-four hours, and that this then plateaus. Similarly, Wood et al, (1983) collected supernatants from a Balb/c macrophage line, 2-3 days after the cells were stimulated with LPS from Escherichia coli in order to harvest lymphocyte activating factor for testing in the thymocyte proliferation assay. Human monocytes and macrophages stimulated with LPS show a maximum yield of lymphocyte activating factor at twenty-four hours with very little production prior to two hours as assessed by the ability of the supernatants to augment proliferation of mouse thymocytes in response

to phytohaemagglutinin (Treves et al. 1983). Bird et al. (1985) prepared IL-1 from Sprague-Dawley rat peritoneal macrophages by culturing the cells for forty-eight hours in the presence of 10 μg/ml LPS. Examination of the time requirement for TNF production by LPS induction of human monocytes showed that cytotoxicity reached plateau values within two hours (Kornbluth and Edgington, 1986). The monocytes, however, need to be primed two to three days prior to challenging with LPS, with an agent such as tubercle bacilli (Carswell, 1975; Mestan, 1986). This was not necessary for the early production of the factor of the present invention.

While it is recognized that the macrophages will produce a large number of soluble factors in response to challenge with LPS, the early release together with the abrupt decrease in production at two hours of this factor would appear to be different from other well recognised monokines.

In summary, the monokine of this invention acts on thymic epithelial cells to cause the release of a second factor, thought to be a complex between the monokine and la antigen, which in turn promotes the survival and differentiation of immature thymocytes. The monokine would be useful in: (a) Stimulating the maturation of the immune system e.g. in children with congenital or acquired immune deficiency; AIDS patients; patients being treated with radiotherapy or cytotoxic drugs;

(b) Selective stimulation of immunity; (c) Selective suppression of inflammation or autoimmunity.

It will be clearly understood that the invention in its general aspects is not limited to the specific details referred to hereinabove.

REFERENCES

1. Adkins, B., Mueller, C, Okada, C.Y. and Reichert, R.A. (1987). Early events in T-cell maturation. Ann. Rev. Immunol. 5: 325-365.

2. Berrih, S., Savino, W. and Cohen S. (1985). Extracellular matrix of human thymus; Immunofluorescence studies on frozen sections and cultured epithelial cells. J. Histochem. Cytochem. 33: 655-664.

3. Bird, J. , Sheng, Y.J., Florentin, I., and Giroud, J.P. (1985). Release of interleukin and low-molecular-weight lymphocyte-activating factors by rat peritoneal macrophages and its enhancement by acute non-specific inflammatory processes. Br. J. Exper. Path., 66: 271-277.

4. Burgess, A.W. and Metcalf, D. (1980). The nature and action of granulocyte-macrophage colony stimulating factors. Blood 56: 947-958.

5. Carswell, E.A., Old, L.J., Kassell, R.L. , Green, S., Fiore, N. , and Williamson, B. (1975). An endotoxin-induced serum factor that causes necrosis of tumours. Proc. Natl. Acad. Sci. USA, 72: 3666-3670.

6. Denizot, F. and Lang, R. (1986). Rapid colorimetric assay for cell growth and survival. Modifications to the tetrazolium dye procedure giving improved sensitivity and reliability. J. Immunol. Methods. 89:271-277.

7. Dinarello, CA. (1984). Interleukin-1. Rev. Infect. Dis. 6: 51-95.

8. Douglas-Jones, A., Nelson, J. , Jansen, V. and Miller, T. (1981). Characterization of the (rnu ^nz ) nude rat. Morphological characteristics of the lymphoid system. Aust. J. Exp. Biol. Med. Sci. 53: 277-286.

■9. Kinnon, C, Diamond, R.A. and Rothenberg, E.V. (1986). Activation of T-cell antigen receptor oζ- and β, -chain genes in the thymus; implications for the lineages of developing cortical thymocytes. J. Immunol. 137: 4010-4015.

10. Klamfeldt, A. (1985). Enhanced breakdown in vitro of bovine articular cartilage proteoglycans by conditioned synovial medium. The effect of phenylglyoxal. Scand. J. Rheumatol. 14: 217-223.

11. Kornbluth, R.S. and Edgington, T.S. (1986). Tumour necrosis factor production by human monocytes is a regulated event: induction of TNF-a-mediated cellular cytotoxicity by endotoxin. J. Immunol., 137(8): 2585-2591.

12. Krakauer, T. (1985). Biochemical characterization of interleukin 1 from a human monocytic cell line. J. Leuk. Biol. 37: 511-518.

13. Le, J. and Vilcek, J. (1987). Biology of disease. Tumour necrosis factor and interleukin 1: cytokines with multiple overlapping biological activities. Lab. Invest. 56: 234-248.

14. Le. P.T., Tuck, D.T., Dinarello, C.A., Haynes, B.F., and Singer, K.H. (1987). Human thymic epithelial cells produce interleukin 1. J. Immunol. 138: 2520-2526.

15. Lo, D. and Sprent, J. (1986). Exogenous control of I-a expression in foetal thymus explants. J. Immunol. 137: 1772-1775.

16. Lovett, D.H., Ryan, J.L. and Sterzel, R.B.

^■ (1983) Stimulation of rat mesangial cell proliferation by macrophage interleukin 1. J. Immunol. 131: 2830-2836.

17. Mestan, J., Digel, W., Mittnacht, S., Hillen, H., Blohm, D., Mδller, A., Jacobsen, and Kirchner, H. (1986). Antiviral effects of recombinant tumour necrosis factor in vitro. Nature, 323: 816-819.

18. Mizel, S.B. (1981). Production and quantitation of lymphocyte-activating factor (interleukin 1), in Manual of Macrophage Methodology (H.B. Herscowitz, H.T. Holden, J.A. Bellanti and A. Ghaffor, eds. ) Marcel Dekker Inc., New York.

19. Papiernik, M. and Ho o-Delarche, F. (1983). Thymic reticulum in mice III. Phagoc tic cells of the thymic reticulum in culture secrete both prostaglandin E- and interleukin 1 which regulate thymocyte proliferation. Eur. J. Immmunol., 13: 689-692.

20. Robinson, A.P., Puklavec, M. and Mason, D.W. (1986) MRC OX-52; a rat T-cell antigen. Immunology 57: 527-531.

21. Rupp, E.A., Cameron, P.M., Ranawat, C.S., Schmidt, J.A., and Bayne, E.K. (1986). Speific bioactivites of monocyte derived interleukin 1 alpha and interleukin 1 beta are similar to each other on cultured human connective tissue cells. J. Clin. Invest. 78: 836-839.

22. Salisbury, J.G., Graham, J.M. and Pasternak,

*

C.A., (1979). A rapid method for the separation of large and small thymocytes from rats and mice. J. Biochem. Biophys. Meth. 1:341-347.

^■ 23. Schmitt, A., Hauser, C, Jaunin, F., Dayer, J.M., and Saurat, J.H. (1986). Normal epidermis contains high amounts of natural tissues II-l, biochemical analysis by HPLC identifies a Mw approximately 17Kd with a Pi 5.7 and an Mw 30 Kd form. Lymph. Res. 5: 105-118.

24. Smith, R.A. , Kirstein, M. , Fiers, W. and Boglioni, C. (1986). Species specificity of human and murine tumour necrosis factor. A comparative study of tumour necrosis factor receptors. J. Biol. Chem. 261: 14 871-14 876.

25. Treves, A.J., Barak, V., Tal, T., and Fuks, Z. (1983). Constitutive secretion of interleukin 1 by human monocytes. Europ. J. Immmunol., 13: 647-651.

26. Van Ewijk, W., Rouse, R.V. and Weissman, I.L. (1980). Distribution of the H-2 microenvironments in the mouse thymus. Immunoelectron microscopic identification of the I-A and H-2K bearing cells. J. Histochem. Cytochem. 28: 1098-1099.

27. Vilcek, J. , Palombella, V.J., Henriksen-Di Stefano, D., Swenson, C, Feinman, R., Hirai, M. and Tsujimoto, M. (1986). Fibroblast growth enhancing activity of tumour necrosis factor and its relationship to other polypeptide growth factors. J. Exp. Med. 163:632-643.

28. Weissman, I.L. (1986). Nursing the thymus (editorial) Lab. Invest. 55: 1-4.

29. Wood, P.R. Andrus, L. and Clark I.A. (1983). Production of lymphocyte activating factor in vivo. Immunol 50: 637-644.