British Journal of Haematology, 1994, 88, 515-519

Interleukin-3 administration enhances human monocyte function in vivo

ASIM KHVVAJA, IAN E. ADDISON, BERYL JOHNSON, KWEE YONG* AND DAVID C. LINCH Departments of Haematology, Universitg College London Medical School, and *Royal Free Hospital Medical School, London

Received 1 1 April 1994; accepted for publication 18 July 1994

Summary. In addition to its haemopoietic effects, inter- leukin-3 (IL-3) enhances leucocyte function in vitro. In this study we examined the effects on haematological variables and monclcyte function of a single IL-3 infusion in five haematologically normal individuals. There was a rapid fall in circulating monocyte (to 24 f 6% of pre-infusion value) and eosinctphil numbers (to 3 * 2%) with a nadir at 30 min and gradual return to baseline over 6h . No significant changes in monocyte expression of the adhesion molecules CD1 l b or ],-selectin or of monocyte respiratory burst activity were detected. There was a significant increase in monocyte phagocytosis and killing of Candida after IL-3 infusion: the

percentage of monocytes which had ingested Candida increased from 39 * 10% to 62 f 12% and the total number of Candida killed per 100 monocytes increased from 6 3 & 3 4 to 2 1 0 & 5 9 (P<0.05 and P c 0 . 0 1 respectively). There was no inhibition of neutrophil migration into a ‘skin window’ site and monocyte migration was moderately enhanced (peak increase of 260 f 47%). These results show that IL-3 has significant effects on monocyte function in vivo and could be of use in augmenting host defence mechanisms in immunocompromised patients.

Keywords IL-3, monocyte function, in viva

The proliferation and differentiation of haemopoietic progenitors is regulated by a family of glycoprotein growth factors which includes erythropoietin, granulocyte- macrophage (GM), granulocyte (G) and macrophage (M) colony-stimulating factors (CSFs) and interleukin-3 (IL-3) (Metcalf, 1991). A feature common to the majority of these cytokines is their additional ability to enhance various functions of mature blood cells. These effects have been demonstrated in vitro and in cells taken from individuals receiving growth factors in clinical trials (Devereux et al, 1989; Sulllivan et al, 1989; Jaswon et al, 1990; Khwaja et al, 1991. 1992).

IL-3 has been shown to enhance monocyte (see below), eosinophil (Takafuji et al. 1991) and basophil (Lopez et al, 1990) function in vitro. No significant effect on neutrophil activation has been reported. In monocytes, IL-3 stimulates antitumour cytotoxicity (Cannistra et al, 1988), increases production of G-CSF (Oster et al, 1989). primes the production of superoxide (Yuo et ul, 1992), enhances anti- Candida activity (Wang et al. 1989) and may prime the release of several cytokines (Takahashi et al, 1993; Maurer et al. 1993). In this study we have examined the effects of a single intravenous infusion of IL-3 on haemato-

Correspondence: Dr A. Khwaja. UCL Medical School, 98 Chenies Mews, London WClE 6HX.

logical variables and on selected monocyte functions in haematologically normal patients with malignant lymph- oma. In addition we have measured plasma Ievels of IL-3 following infusion. We show that IL-3 is able to enhance certain monocyte functions that may be relevant to its application in clinical practice.

PATIENTS AND METHODS

Patients. The infusions were carried out in five haemato- logically normal (no significant cytopenia and normal iliac crest bone marrow biopsy) adult patients with relapsed malignant lymphoma (three with Hodgkin’s disease and two with high-grade non-Hodgkin’s lymphoma). IL-3 was given prior to treatment with high-dose chemotherapy and autologous bone marrow transplantation using the BEAM protocol. Local ethics committee approval and full informed written consent were obtained.

IL-3 infusion. Recombinant human IL-3 was provided by Sandoz Ltd, U.K. The lyophilied product at a dose of 5 pg/kg was reconstituted with sterile water for injection, and added to 2 50 ml of 0.9% NaCl containing 0.1 mg/ml human serum albumin. The IL-3 was administered over 2 h via an indwelling central venous catheter.

Materials. Dichlorofluorescein diacetate (DCF-DA, Mole- cular Probes, Oregon, U.S.A.) and f-met-leu-phe (f-mlp)

515

5 16 (Sigma, Poole, U.K.) were dissolved in DMSO at stock concentrations of 100 m~ and 50mgfml respectively, stored at -20°C. and diluted in RPMI (Gibco, Paisley, U.K.) on the day of use. Monoclonal CDllb antibody (antibody 44) was a kind gift from Dr Nancy Hogg (ICRF Laboratories, London) and antibody to LAM-1 (Leu-8) was obtained from Becton Dickinson (Oxford, U.K.). FITC-conjugated rabbit anti-mouse antibody was obtained from Dako (Bucks., U.K.). IL-3 ELISA kit was purchased from British Biotechnology Ltd (Cambridge, U.K.).

Asim Khwaja et a1

METHODS

IL-3 ELISA. This was carried out on plasma samples according to the manufacturer's instructions. Blood was collected into EDTA, centrifuged and frozen at -70°C within 60 min of collection. Samples were batched and run on the same day.

Haematological variables. Full blood counts were done on a Technicon H1. All differential WBC counts obtained were confirmed by visual inspection of May-Grunwald-Giemsa and non-specific esterase stained smears: at least 200 cells per specimen were counted.

CAM expression. All patient samples were taken into preservative-free heparin, immediately placed on ice and incubated for 30min with a saturating concentration of CD1 l b antibody. Following three washes the cells were further incubated with fluorescein-conjugated rabbit anti- mouse antibody for 30min at 4°C. After three further washes at 4"C, samples were processed using the Coulter Immunoprep Workstation (Coulter, Hialeah, Fla., U.S.A.) which lyses red blood cells and fixes white cells. Analysis was carried out by flow cytometry using a Coulter Epics-C machine, monocytes and neutrophils being selectively gated by virtue of their light-scattering properties. Assessment of CDllb expression was based on mean cell fluorescence using a linear scale.

Measurement of respiratory burst activity. This was measured using a whole blood technique for assessing intracellular H20z production as previously described (Jaswon et al, 1990). In brief, freshly drawn whole blood was incubated with 1 0 0 ~ ~ DCF-DA for 15min at 37°C. f-mlp at M was then added for 15 min and the samples kept at 37°C. The reaction was stopped by placing the samples on ice and they were then prepared for analysis by flow cytometry as described above. Samples were taken immediately before the start of the infusion, and at 2 h after its completion.

Monocyte phagocytosis and degradation of Candida. This was carried out as previously described (Khwaja et al, 1991). Mononuclear cells were separated from whole blood by density gradient separation (Lymphoprep. Nycomed, Oslo. Norway), washed twice in RPMI, and resuspended to a count of 5 x 106/ml in 30% autologous plasma. Candida guillermondiae at three concentrations were added and incubated in an end-over-end mixer at 37°C for 60min. Cytospins were then prepared and stained by MGG and NSE. An assessment of phagocytosis was made by counting the percentage of monocytes which had one or more

intracellular Candida and by the total number of intracellular Candida per 100 monocytes; Candida killing was measured by counting the number of dead and visibly degraded intracellular organisms and expressing this as a percentage of the total number ingested. Phagocytosis and killing were measured immediately before the start of the IL-3 infusion, at the end of infusion, and 24 h later.

Skin window migration. Monocyte and neutrophil migra- tion into an inflammatory site was assessed using a previously described technique (Addison et al, 1989). Superficial forearm skin abrasions were made in duplicate and covered with 8 mm micropore membranes (Sartorious Ltd), and the whole area covered with a moist pad (sterile pyrogen-free 0.9% saline) and an occlusive dressing. Healthy laboratory staff acted as control subjects. The membranes were left undisturbed for 3 h to allow phagocyte migration to be established. In patients, serial membranes were applied at 20 min intervals from immediately preceding the start of the IL-3 infusion, continuing during the infusion and for 20 min following its completion. The identical protocol, but without any IL3 administration, was followed in the controls. The 20min membranes were fixed and stained with non-specific esterase (NSE) and counterstained with methyl green. For monocyte migration, the total number of cells entering the membrane were counted. Neutrophil migration was assessed by measuring the distance travelled into the membrane by the leading front of neutrophils, and the cellularity at 50% of the leading front distance per high power field. Using the product of these two figures, an index of the total cells entering the membrane was calculated

Statistical analysis. Comparison between pre-infusion and post-infusion values for white cell counts and monocyte function tests was made using the paired t test.

RESULTS

Pharmacokinetics Plasma IL3 levels were measured before and after the completion of the 5pg/kg IL-3 infusion and the data are shown in Fig 1. Only one patient had detectable levels of IL-3 prior to the infusion (31 pg/ml). Peak values at the completion of the infusion were remarkably consistent with a mean (+ SEM) of 8-63 + 0.2 ng/ml. There was a rapid clearance of IL-3 from the plasma with a mean value at 2 h after infusion of 0.62 f 0.26 ng/ml. Values for initial and clearance tl12 were 2.5 f 0.3 and 35 * 5 min respectively.

Effects of lL3 infusion on blood cell counts Fig 2 shows the effect of the 2 h infusion on the component parts of the white cell count. There were rapid and significant falls in both monocyte and eosinophil numbers reaching a nadir at 30min following the start of the infusion (24 f 6% [P < 0.00051 and 3 f 2% [P < 0.00051 of pre-infusion values respectively). The striking effect of IL-3 on circulating eosinophils was demonstrated by a patient with a pre-infusion Hodgkin's associated eosinophilia (1.3 x 109/1) who had no detectable circulating eosinophils after 15 min of IL-3 and did not recover his pre-infusion count for a further 8 h. There was also a slower and less marked fall in circulating lymphocytes

IL-3 Enhances Monocyte Function in Vivo 5 17

'O- -1 I

8000 - k ; b

- E n

6000 - _

0 6 0 1 2 0 1 8 0 2 4 0 3 0 0

t ime [minl

Fig 1. Serial plasma IG3 levels measured by ELISA in four patients after the completion of a 2 h infusion of IL-3 at 5 pg/kg body weight. Peak values at the completion of the infusion were mean +SEM 8.63 f 0.2 ng/ml. There was a rapid clearance of ILL3 from the plasma with a mean value at 2 h after infusion of 0.62 f 0.26 ng/ml.

reaching a nadir of 54 & 8% (P < 0.01) by the end of the infusion. There were no significant changes in neutrophil numbers during or after the LL3 infusion.

Phagocgte cellular adhesion molecute expression Neutrophil and monocyte surface expression of the CD1 l b and LAM-1 antigens was measured before, during, and at 60min after the end of the IL-3 infusion. No significant changes were seen in the expression of either antigen at any stage.

IL-3 infusion 150 11-

125 c C 3

0

0

0 100

a9 75 E .- c)

50

8 25

Modulation of f-mlp stimulated respiratory burst activity Measurement of monocyte and neutrophil respiratory burst activity was made before, at the end of the IL-3 infusion, and 2 h later. No change in baseline or f-mlp stimulated intracellular H2O2 production was detected following the IL-3 infusion.

Candida ingestion and killing bg monocytes The ability of separated mononuclear cells to ingest and degrade Candida guillermondae was measured and three parameters were used to assess monocyte activity: the percentage of NSE positive monocytes which had ingested one or more organisms, the number of Candida ingested per 100 monocytes, the number of killed intracellular Candida expressed as a percentage of the total ingested. These measurements were used to derive data for the number of Candida ingested and killed per 100 monocytes. There was a significant increase in the percentage of monocytes which had ingested Candida following the test infusion of IL-3 from 39 f 10% (mean &SE) to 62 f 12% immediately post infusion (P < 0.05). The mean number of Candida ingested by 100 monocytes increased significantly from a baseline level of 109 f 51 to 294 77 at 2 h (P < 0.005) and there was a significant increase in the total number of organisms killed by 100 monocytes from 63 & 34 to 210 & 59 after the infusion (P < 0.01) (Table I).

Skin window studies The influence of IL-3 on the ability of neutrophils and monocytes to migrate into a sterile inflammatory site in vivo was examined using a skin window technique. No effect on neutrophil migration was seen during the IL-3 infusion. The mean number of monocytes entering each membrane prior to starting the infusion was 851 & 197, which was higher than in normal controls (241 4 98, n = 7) and in a similar

......... * ....... _.-. *.-.

*- lymph

0 6 0 1 2 0 1 8 0 2 4 0 3 0 0 3 6 0 4 2 0 4 8 0 time [min]

Fig 2. Mean white blood cell differential values during and after the completion of a 2 h infusion of IL-3 (5pg/kg) in five separate patients.

518 Asim Khwaja et a1

- m 0 c

c

Table I. The effect of IL-3 infusion on monocyte phagocytosis and killing of Candida. Results (mean f SEM) are from inksions in five separate patients.

CONTROLS I T T I I I -f t P

Monocytes Mean no. No. Candida No. Candida phagocytosing Candidal ingested/100 killed/100 Candida (%) monocyte monocytes monocytes

Pre-infusion 39 f 10 2.1 f 0.6 109 f 51 63 f 3 4 Post-infusion 62 f 12" 4 .1 f 0.9' 294 f 77b 210 f 59' 24 h post infusion 4 5 5 6 3.3 f 0.7 171 f 47 134 & 39

group of patients we have examined in a previous study (Khwaja et al, 1991). Following the infusion of IL-3 there was no fall in neutrophil or monocyte migration. An increase in monocyte migration to a peak value of 2 1 8 2 f 5 0 1 was reached between 100 and 120min (Fig 3). In healthy control volunteers undergoing skin windows over the same time frame, but without an IL-3 infusion, there was no change in monocyte migration (Fig 3).

IL-3 infusion

m 30001

DISCUSSION

In this study we have described the effects of an infusion of IG3 on haematological parameters and on monocyte function. IL3 infusion led to a rapid reduction (nadir 30 min) of circulating monocytes and the virtual disappear- ance of circulating eosinophils. This was followed by their reappearance over the following 4-6h. This transient cytopenia is similar to the effects of GM-CSF infusion on

circulating neutrophils and monocytes (Devereux et al, 1987) and to the effect of G-CSF on neutrophil numbers (Yong & Linch, 1992). The reduction in circulating phagocytes following GM-CSF infusion is due to transient margination in the pulmonary vasculature and it is possible that IL-3 has a similar effect on its target cells. It has been proposed that the mechanism of phagocyte margination after GM-CSF administration is related to quantitative and/ or qualitative changes in the surface expression of cell adhesion molecules including those recognized by C D l 1 b and LAM-1 antibodies. We found no quantitative change in the expression of either of these molecules on monocytes obtained at the end of the IL-3 infusion. However, it is possible that changes occur maximally in the cells that are slow to demarginate and thus escape examination or that such changes that do occur are of a transient nature.

We have also examined the effect of IL-3 administration on priming of the monocyte respiratory burst using a whole blood flow cytometric assay of intracellular oxidant produc- tion. We were unable to detect any priming of the monocyte respiratory burst stimulated by f-mIp either at the end of the infusion or 2 h later. In addition, we did not find any significant effects of IL-3 incubation in vitro in samples taken from the same group of patients prior to infusion. This contrasts with the modest priming effect of M-CSF infusion in a similar group of patients (Khwaja et al, 1991). Recently, Yuo et al (1992) have described a small but significant priming of monocyte superoxide release by IL-3 in vitro. The reasons for the discrepancy between their results and those presented here are not clear, but may relate to different methods of cell preparation (whole blood versus purified cells) or the different assays employed (intracellular oxidant production versus extracellular superoxide release). In addition, we have been unable to detect IL3 primiig of the monocyte respiratory burst using the DCF assay in four different normal individuals (unpublished observation). Maurer et al (1993) have shown a significant increase in monocyte oxygen radical release in two of three patients receiving IL-3 therapy for 14-42 d, a much more prolonged period than examined in our study.

The effects of HGF administration on phagocyte migration into skin windows has varied significantly between cytokines. GM-CSF infusion causes inhibition of neutrophil migration (Addison et al, 1989; Peters et d, 1988b). G-CSF is reported to not affect migration (Peters et al, 1988a) and

ZL-3 Enhances Monocyte Function in Vivo 519 M-CSF causes a marked increase in monocyte migration (Khwaja et al, 1991). In this study we show that IL-3 infusion (does not inhibit either neutrophil or monocyte migration. The rise in monocyte migration with IL-3 infusion (peak increase of 260*47%) is modest when compared with a near 5-fold increase seen during M-CSF infusion in a similar group of patients (Khwaja et al, 1991). Furthermore, as the five patients studied had unusually high baseline levels of monocyte migration we are reluctant to draw any firm conclusions regarding the possible migration enhancing effects of IL-3.

We have also shown a clear effect of IL-3 infusion on both phagocytosis and killing of Candida by purified monocytes. Wang et 111 (1989) showed that incubation with IL-3 could significantly enhance monocyte anticandidal activity in vitro. Our data show that IL-3 infusion augments monocyte phagocytosis and killing of Candida by approximately 3-fold. The mechanism for this enhancement is not clear, and in particular. as stated above, we have been unable to show any concurrent priming of respiratory burst activity.

The role of IL-3 in haemopoiesis has been extensively explored and its pleiotropic effects on the development of multiple cell lineages have been documented in vitro and in vivo (HIDelzer et al, 199 1). The data presented in this paper show that IL-3 also has significant effects on the function of mature nnonocytes in vivo, in particular enhancing anti- candidal activity. We were unable to examine the effects of IL-3 on eosinophil function ex vivo as infusion was followed by a prolonged fall in circulating cells. These data suggest that treatment with L 3 for chemotherapy-related or other cytopeniais is unlikely to have any significant adverse effect on phagocyte function and that IL-3 administration may be of use in augmenting monocyte function in immuno- compromised patients.

ACKNOWLEDGMENTS

We thank Julia Carver for technical assistance. This work was partly supported by the Kay Kendall Leukaemia Fund. A.K. is supported by the Wellcome Trust.

REFERENCES

Addison, I.E., Johnson, B.. Devereux, S.. Goldstone, A.H. & Linch, D.C. (1989) Granulocyte-macrophage colony-stimulating factor may inhibit neutrophil migration in viva Clinical and Experimental Immunology. 76, 149-153.

Cannistra, S.A.. Vellenga, E.. Groshek. P.. Rambaldi. A. & Gritfin, J.D. (1988) 13uman GM-CSF and IG3 stimulate monocyte cytotoxicity through a tumor necrosis factordependent mechanism. Blood,

Devereux. S.. Bull. H.A.. Campos-Costa. D.. Saib, R. & Linch, D.C. (1989) Granulocyte-macrophage colony-stimulating factor induced changes in cellular adhesion molecule expression and adhesion to endothelium: in vitro and in vivo studies in man. British journal of Haernatology. 71, 323-330.

Devereux, S., Linch, D.C., Campos-Costa, D., Spittle, M.F. & Jelliffe, A.M. (11987) Transient leucopenia induced by granulocyte- macrophage colony-stimulating factor. (Letter). Lancet, ii, 1523.

Hoelzer. D.. Seipelt, G. & Ganser. A. (1991) Interleukin 3 alone and

71,672-676.

in combination with GM-CSF in the treatment of patients with neoplastic disease. Seminars in Hematology, 28, (Suppl. 2).

Jaswon. M.S.. Khwaja, A., Jones. H.M., Roberts, P.J. & Linch. D.C. (1990) The effect of rhGM-CSF on the neutrophil respiratory burst in whole blood. British Journal of Haematology. 75, 181-187.

Khwaja, A., Carver, J.E. & Linch, D.C. (1992) Interactions of granulocyte-macrophage colony-stimulating factor (CSF), granu- locyte CSF, and tumor necrosis factor (Y in the priming of the neutrophil respiratory burst. Blood. 79, 745-753.

Khwaja. A.. Johnson, B.. Addison, I.E.. Yong. K.. Ruthven. K., Abramson. S. & Linch. D.C. (1991) In vivo effects of macrophage colony-stimulating factor on human monocyte function. British Journal of Haematology, 77, 25-31.

Lopez. A.F., Eglinton, J.M., Lyons. A.B., Tapley. P.M., To, L.B., Park, L.S., Clark, S.C. & Vadas, M.A. (1990) Human IL-3 inhibits the binding of GM-CSF and IL5 to basophils and strongly enhances their functional activity. Journal of Cellular Physiology. 145, 69-77.

Maurer. A.B., Ganser, A., Buhl, R., Seipelt, G.. Ottmann. O.G., Mentzel. U.. Geissler. R.G. & Hoelzer. D. (1993) Restoration of impaired cytokine secretion from monocytes of patients with myelodysplastic syndromes after in vivo treatment with GM-CSF or IL-3. Leukemia, 7, 1728-1733.

Metcalf, D. (1991) Control of granulocytes and macrophages: molecular, cellular, and clinical aspects. Science, 254. 529-533.

Oster. W., Lindemann. A,, Mertelsmann. R. & Hermann. F. (1989) GM-CSF and multilineage CSF recruit human monocytes to express granulocyte CSF. Blood, 73, 64-67.

Peters, W.P., Kurtzberg, J., Atwater. S.. Borowitz, M., Gilbert, C.. Rao, M., Currie, M., Shogan. J., Jones, R.B., Shpall, E.J. & Souza. L. (1988a) Comparative effects of rHuGM-CSF and rHuG-CSF on hematopoietic reconstitution and granulocyte function following high-dose chemotherapy and ABMT. Blood. 72, 130a.

Peters, W.P., Stuart, A,, Atironti, M.L., Kim, C.S. & Coleman, R.E. (1988b) Neutrophil migration is defective during recombinant human GM-CSF infusion after autologous bone marrow trans- plantation in humans. B h ~ d , 72, 1310-1315.

Sullivan, R., Fredette. J.P., Socinski, M., Elias. A.. Antman, K., Schnipper, L. & G&n. J.D. (1989) Enhancement of superoxide anion release by granulocytes harvested from patients receiving GM-CSF. British Journal of Haematology, 71,475.

Takafuji. S.. Bischoff, S.C., De, W.A. & Dahinden. C.A. (1991) IL3 and IL5 prime normal human eosinophils to produce leukotriene C4 in response to soluble agonists. Journal of Immunology. 147,

Takahashi, G.W., Andrews, D.F., Lilly, M.B., Singer, J.W. & Alderson, M.R. (1993) Etrect of granulocyte-macrophage colony-stimulating factor and interleukin-3 on interleukin-8 production by human neutrophds and monocytes. Blood, 81,

Wang, M.. Friedman, H. &Djeu. J.Y. (1989) Enhancement of human monocyte function against Candida dbicans by the colony- stimulating factors (CSF): IG3, GM-CSF and M-CSF. Journal of Immunology, 143, 6 7 1-6 77.

Yong, K.L. & Linch, D.C. (1992) Differential effects of granulocyte and granulocyte-macrophage colony-stimulating factors on neutrophil adhesion in vitro and in vivo. European Journal of Haematology. 49, 251-259.

Yuo. A., Kitagawa, S., Motoyoshi, K., Azuma, E.. Saito. M. & Takaku, F. (1992) Rapid priming of human monocytes by human hematopoietic growth-factors-granulocyte-macrophage colony- stimulating factor (CSF), macrophage-CSF. and interleukin-3 selectively enhance superoxide release triggered by receptor- mediated agonists. Blood. 79, 1553-1557.

17-24.

3855-3861.

357-364.