British /ournal of Haernatology. 199 1, 77, 2 5 - 3 1 AWNIS 0007104891000042

In vivo effects of macrophage colony-stimulating factor on human monocyte function

ASIM K H W A J A , BERYL JOHNSON, I A N E. A D D I S O N , K W E E Y O N G , KAREN RUTHVEN.* STEPHEN ABRAMSON* A N D D A V I D C . IJNCH Department of Haematology. University College and Middlesex School of Medicine, London, and *Alpha Therapeutic Ltd, Thetford. Norfolk

Received 20 July 1990; accepted for publication 20 September 1990

Summary. Macrophage colony-stimulating factor (M-CSF) is reported to enhance a variety of functions of mature monocyte/macrophages in vitro. We have examined the effects of a 2 h intravenous infusion of M-CSF obtained from human urine (hM-CSF) on haematological parameters and selected monocyte functions. There was a rapid, small, but consistent reduction in Hb concentration (mean 6 - 5 f 2.3%. P<0.0005 by paired t test) by the completion of the hM-CSF infusion and small, transient falls in platelet, monocyte and neutrophil counts were noted in the 2 h following the end of the infusion. No effect on monocyte or neutrophil CDl lb cellular adhesion molecule expression was detected. Expo- sure to hM-CSF in vivo did not directly stimulate the monocyte

respiratory burst, but increased the percentage of monocytes responding to f-met-leu-phe from 9 . 8 f 2 . 5 to 1 6 . 6 f 4 . 2 ( P < 0.0 1 ). The number of candida ingested and degraded per 100 monocytes increased from 101 f 14 pre-infusion to 160 f 22 post-infusion (P < 0.0 1 ). There was a rapid increase in the numbers of monocytes entering a skin window membrane from a mean of 226f 71 pre-infusion to 1 0 6 4 f 4 0 4 at the end of the infusion, with no effect on neutrophil migration. These data show that the administra- tion of hM-CSF enhances several of the functions of periph- eral blood monocytes in vivo, and this may be of benefit in the treatment of selected infections.

The proliferation and differentiation of haemopoietic progeni- tors is controlled by a family of glycoprotein growth factors, termed the colony-stimulating factors (CSFs). Macrophage CSF (M-CSF) stimulates the development of monocyte/ macrophage colonies in vitro (Clark & Kamen, 1987). Human M-CSF has a homodimeric structure, and two major forms, produced by differential mRNA splicing events of the M-CSF gene, have been characterized (Kawasaki et al, 1985: Wong et al, 1987). Additional processing and N-glycosylation result in 70-90 kD and 40-50 kD glycoproteins. Human urinary CSF contains the entire amino acid sequence of the larger form of M-CSF. and has a n apparent molecular weight of 8 5 0 0 0 (Ishizaka et al. 1986).

In addition to its haemopoietic effects, M-CSF enhances various effector functions of mature mononuclear phago- cytes in vitro. These include stimulation of direct and antibody dependent cytotoxicity (Sampson-Johannes & Car- lino. 1988: Mufson et al. 1989), chemotaxis (Wang et al, 1988). respiratory burst activity (Wing et al. 1985), Fc

Correspondence: Professor D. C. Linch. 98 Chenies Mews, London WClE 6HX.

receptor expression (Magee et al, 1987) and killing of Candida albicans (Wang et al. 1989). Exposure of monocyte/macro- phages to M-CSF causes production of plasminogen activator (Chen & Lin. 1984). and stimulates the synthesis and secretion of cytokines such as interleukin-1 , tumour necrosis factor (TNF) and granulocyte CSF (G-CSF) (Warren & Ralph, 1986: Motoyoshi et al, 1989).

M-CSF is now available for clinical use in both purified human urinary (hM-CSF) and recombinant (rhM-CSF) forms. hM-CSF has been reported to ameliorate chemotherapy- induced neutropenia in early clinical trials (Masaoka et al. 1988: Motoyoshi et al. 1986), but the effects of this factor on monocyte functions following clinical administration have not as yet been characterized. In this study we have examined the effects of a single intravenous infusion of hM-CSF on haematological variables, and on selected monocyte and neutrophil functions in haematologically normal patients with malignant lymphoma. These include phagocyte migra- tion to a sterile inflammatory site ('skin window'), expression of cellular adhesion molecules (CAMS). respiratory burst activity, and candida phagocytosis and killing. We show that hM-CSF enhances monocyte function in vivo and suggest that this has potentially beneficial effects in the treatment of

25

26 Asim Khwaja et a1 selected severe infections, particularly in the immunocom- promised host.

PATIENTS

The infusions were carried out on 1 3 haematologically normal (as defined by normal peripheral full blood counts and iliac crest bone marrow aspirate and trephine) adult patients (12 male and one female) with relapsed malignant lym- phoma (six with Hodgkin's disease and seven with high- grade non-Hodgkin's lymphoma). prior to treatment with high-dose chemotherapy and autologous bone marrow transplantation (BMT). One group of five patients received hM-CSF at 4 million units (mU)/mz each, and two groups of four patients each received 8 mU/mz and 16 mU/mz, The patients were taking part in a trial examining the effects of hM-CSF on haemopoietic recovery following BMT, this infusion being a test dose to assess adverse effects. changes in vital signs, and monocyte function. Local ethical committee approval and full informed written consent from each patient were obtained.

MATERIALS

Human M-CSF with a specific activity of 10i U / p g , as measured in a mouse colony-forming assay, was purified from human urine and provided by Alpha Therapeutic Ltd/ Green Cross Corporation (Osaka, Japan). The endotoxin content of the preparation was less than 0.6 pg/pg of hM-CSF as measured by the limulus lysate assay. The lyophilysed product was reconstituted with sterile water for injection, made up to a volume of 200 ml with 0.9% NaCI, and administered over 2 h via an indwelling central venous catheter.

Dichlorofluorescein diacetate (DCF-DA, Molecular Probes, Oregon, U.S.A.) and f-met-leu-phe (fmlp) (Sigma, Poole) were dissolved in dimethyl sulphoxide (DMSO) at stock concentra- tions of 100 mM and 50 mg/ml respectively. stored at - 20OC. and diluted in RPMI (Gibco, Paisley) on the day of use. Monoclonal CD1 Ib antibody (antibody 44) was a kind gift from Dr Nancy Hogg (ICRF Laboratories, London) Fluorescein isothiocyanate (FITC) conjugated rabbit anti- mouse antibody was obtained from Dako (Bucks.).

METHODS

Huematological variables. Full blood counts were done on either a Coulter STKR or a Technicon H1. All differential WBC counts were done by a single observer on May- Grunwald-Giemsa (MGG) stained smears; a t 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 30 min with a saturating concentration of CDI l b antibody. Following three washes the cells were further incubated with fluorescein conjugated rabbit anti- mouse antibody for 30 min at 4 O C . After three further washes at 4OC. samples were processed using the Coulter Imrnuno- prep Workstation (Coulter, Hialeah, Florida. 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 CDI 1 b expression was based on mean cell fluorescence using a linear scale.

Measurement 01 respiratory burst activity. This was mea- sured using a modification of the method of Bass et al ( I 98 3 ) for assessing intracellular H20z production as previously described (Jaswon et al. 1990). In brief, freshly drawn whole blood was incubated with 100 PM DCF-DA for 1 5 min at 37OC. Where the effect of hM-CSF was being examined in vitro, this was followed by incubation with hM-CSF at 20000 U/ml (an optimal concentration determined by prior in vitro experiments), or with diluent control, for 2 h at 37°C. Fmlp at l o - " M was then added for 1 5 min and the samples kept at 3 7OC. 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.

Skin window migration. Monocyte and neutrophil migration into a n inflammatory site was assessed using a previously described technique (Addison et al. 1982). In brief, superficial forearm skin abrasions were made in duplicate and covered with 8 mm micropore membranes (Sartorius 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 hM-CSF infusion, continuing during the infusion and for 20 min following its completion. A similar protocol, but without any hM-CSF administration, was followed in the controls. The 20 min 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 mem- brane by the leading front of neutrophils. and the cellularity a t 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 could be calculated.

Monocyte phagocytosis and degradation o j candida. Phago- cytosis and killing were measured immediately before the start of the hM-CSF infusion, immediately after its end, and 24 h later. This was done by a modification of the method as previously described (El Maalem & Fletcher, 1976). Mono- nuclear cells were separated from whole blood by density gradient separation (Lyrnphoprep, Nycomed, Oslo. Norway). washed twice in RPMI. and resuspended to a count of 5 x IOh/ml in 30% autologous plasma. The percentage of monocytes in the mononuclear fractions was consistent within patients at sequential measurements. Candida guiller- mondii (stock- 10X/ml) at three concentrations (1/27, 1/81 and 1 /243 dilutions of stock) were added and incubated in an end over end mixer a t 3 7°C for 60 min. Cytospins were then prepared and stained by MGG and NSE. An assessment of phagocytosis was made by counting the percentage of

M-CSF and Monocyte Function 27

120 e J - g 110 l2O1

p 100 = 90 n

c

6 60

..

0 40 80 120 160 200 240 280 time (minutes)

= 130 3 8 120 f 110 g 100 a g 90

O 80

a. c

- I! 70 c

8 50 605 0 40 80 120 160 200 240 280

time (minutes)

0

8 60

50 ,i: 0 40 80 120 160 200 240 280

time (minutes)

c 9 130

8 120 Q 110 g 100

L o

0

80

70

0

0

c

* * 6 60:. # s o 1 . 1 . 1 . 1 . 1 . 1 . ~

0 40 80 120 160 200 240 280 time (minutes)

Fig 1. The effect of a 2 h hM-CSF infusion on haematological variables. Results are expressed as mean ( f SE)% of time 0 value ( r l = 1 3 ) . ( a ) Mean platelet counts. * f’<O.Ol. (b) Mean haemoglobin concentration. * P < O ~ O O O 5 . (c) Mean neutrophil count. * P 1 0 . 0 2 5 . (d) Mean monocyte count. * P < O . O 2 5 .

monocytes that had one or more intracellular candida, and the total number of intracellular candida per 100 monocytes. and of killing by counting the number of visibly degraded intracellular organisms and expressing this as a %of the total number ingested. Data was collected for all three dilutions of candida but the data presented is for that dilution where there were - 1 - 3 candida per monocyte. as we have found this to be the most reproducible and least liable to inter-observer error.

KES U LTS

Eflerts of hM-CSF irfusion on blood cell counts During the 2 h infusion of hM-CSF. there was a small but consistent fall in Hb concentration. This was apparent within 30-60 min and at the end of the infusion the mean Hb concentration ( +SE) had fallen from a pre-infusion value of 11.3f0.5 g/dl to 1 1 . 5 f 0 . 5 (P<0.0005 by paired t test), representing a mean fall of 6.5% (Fig 1). Parallel falls in haematocrit and red cell numbers were observed. During the infusion there were no significant changes in leucocyte or platelet numbers.

In the 2 h period following the infusion of hM-CSF there were small. transient, but significant falls in platelet (mean of 8%. P<O.Ol). neutrophil (16%. P<O.O15). and monocyte counts (24%. P < O 4 2 5 ) . with no further fall in the haema- tocrit (Fig 1 ).The lymphocyte count did not vary significantly throughout the infusion, or during the immediate post- infusion period. 2 4 h after the infusion the Hb was a mean of 94 f 3% of the pre-treatment value, and leucocyte and platelet numbers were similar to pre-infusion values. The changes in the blood count were observed at all doses of hM- CSF used with no discernible dose effect. No clinical adverse effects were detected during or after any of the infusions.

Phagocyte cellular adhesion niolerule expression Neutrophil and monocyte surface expression of the CDI Ib antigen (Mac I ) was measured before, immediately after, and at 120 min after the end of the hM-CSF infusion. CDI 1 b expression, expressed as a percentage of the pre-infusion value, was 1 15 f 34% and 1 3 1 f 52% for neutrophils and monocytes respectively at the end of the infusion (n= 9. N.S. by paired t test), and 89 f 22% and 84 f 16% at 1 2 0 min after the end of the infusion ( n = 6. N.S.).

28 Asim Khwaja et al

1400

g 1200-

2

Table 1. In vitro monocyte HzOz production before and after the administration of an infusion of hM-CSF using a flow cytometric assay. Figures refer to the mean (fSE) percentage of responding monocytes from seven separate infusions.

- M-CSF INFUSION

Pre Post

Medium 4 .h f0 .2 4 . 3 f 0 . 4 M-CSF alone 4 . 4 f 0 . 4 -

fmlp 9 . 8 f 2 . 5 16.hf4.2' M-CSF+fmlp l h . X f 3 . 5 1 7 . 2 f 4 . 2

'Pc0.01 by paired I test.

Modulation of fmlp stimulated respiratory burst activity The generation of superoxide and H202 as products of the respiratory burst plays a central role in the cytotoxic potential of the phagocyte. We therefore examined the effect of hM-CSF infusion on monocyte and neutrophil H r 0 2 production using a fluorescent flow cytometric technique as described: experi- ments were carried out using a whole blood technique, as it is possible that separation procedures may themselves modu- late function, and thus mask significant changes caused by cytokine exposure by increasing baseline activity (Fearon & Collins, 198 3) . In vitro incubation of whole blood with hM- CSF was carried out prior to starting the infusion and the respiratory burst stimulated by the bacterial peptide f-met- leu-phe (fmlp). Blood sampled 2 h after the completion of the infusion was stimulated directly with fmlp. and also subjected to a further in vitro incubation with hM-CSF prior to stimulation.

llnprimed neutrophils in whole blood show little increase in fluorescence when stimulated with fmlp. with the propor- tions of cells scored as positive rising from a baseline value of 5 . 4 f 0 . 5 % to 13 .5 f2 .4% (n=22) after addition ofoptimal concentrations of fmlp. Pre-incubation of the pre-infusion sample with M-CSF did not enhance this response

(4.3 f0 .9% in unstimulated and 13.9 f 3.7% in fmlp stimu- lated neutrophils). Similarly, exposure to hM-CSF in vivo did not significantly modify the neutrophil responses ( 3 . 1 f 0.5% in unstimulated and 1 5 . 6 f 5 .3% in stimulated cells). In control experiments, where whole blood was incubated in vitro with optimal concentrations of recombinant human granulocyte-macrophage colony-stimulating factor (rhGM- CSF). the percentage of neutrophils responding to fmlp rose to

Monocytes are similarly hypo-responsive to fmlp stimula- tion in whole blood unless previously primed. It is seen from Table I that in vitro incubation with hM-CSF significantly increases the proportion of monocytes responding to fmlp from 9.8 f 2 . 5 to 16.8 f 3.5% (P<O.01). In vitro incubation with rhGM-CSF similarly increases the number of responding monocytesfrom 7 f l to21.3f 3.6x(n=7).Invivoadminis- tration of hM-CSF produced priming similar to that seen in vitro, and this effect was maximal with no further increase seen on immediate cx vivo pre-incubation of blood with optimal concentrations of hM-CSF (Table I). These effects were seen at all infusion doses, and there was no detectable dose effect.

46 .7 f6 .3 (n=22).

Skin window studies The influence of hM-CSF 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 hM-CSF infusion. The leading front distance (IFD) (mm). the cellularity per high-power field at 50% of the LFD and the product of the two values were 65 .2 f3 .3 . 36 .5 f4 .6 and 2396f339 pre- infusion, and 61 .4 f4 .9 . 38 .4f4 .1 and 2280f 360 after 60 min of the infusion. Similar values were obtained at all other time points during the infusion, and in the period immediately following its completion.

The mean number of monocytes entering each membrane prior to starting the infusion was 226f 71 . which was similar to normal controls (241 f 9 8 , n= 7). During the first 20 min of the infusion, this increased to 544 f 1 55 and reached a peak value of 1064 f 404 between 100 and 120

M-CSF and Monocyte Function 29

Candida ingestion and killing by monocytes The ability of separated mononuclear cells to ingest and degrade Candida guillermondii was measured and four para- meters were used to assess monocyte activity: the percentage of NSE positive monocytes that had ingested one or more organisms, the number of candida ingested per 100 mono- cytes. and the number of degraded intracellular candida expressed as a percentage of the total ingested. Finally, the product of the last two values was derived as an index of the total number of candida killed in 60 min by 100 monocytes. There was no significant change in the number of monocytes phagocytosing candida following the test infusion of hM-CSF. with values pre-infusion of 6 7 f 4 % (mean fSEj , immedi- ately post of 7 0 f 5% and 2 4 h post of 6 4 f 5%. However, the mean number of candida ingested by 100 monocytes increased significantly from a baseline level of 188 f 30 to 2 5 8 f 4 0 at 2 h (P<O.O25 by paired t test), returning towards pre-infusion levels by 2 4 h (208 f 52) (Fig 3aj. and there was significant enhancement of killing ability, expressed as the percentage of ingested organisms degraded, and the total number of organisms killed by 100 monocytes (Figs 3b and 3c). The data suggest that exposure to hM-CSF in vivo augments the phagocytosis and killing activity of peripheral blood monocytes.

p r o p o r t 24 hr

T

6

U

u

E E 60

B t z 6 20

" 40

- z

0 -1

p r o p o r t 24 hr

1

' T

p r o p o r t 24 hr

Fig 3. The effect of hM-CSF infusion on the ability of monocytes to ingest and degrade candida (see Methods). (a) Mean ( f SEj number of candida ingested by 100 monocytes. (b) Mean % of ingested candida that were killed. (c) Mean number of candida killed by 100 monocytes. ' P i O . 0 2 5 .

min. This contrasts markedly with the findings in healthy control volunteers undergoing skin windows over the same time frame, but without a hM-CSF infusion (Fig 2). There is thus a marked and specific increase in monocyte migration into the inflammatory focus during the hM-CSF infusion. This effect was observed at all three doses of hM-CSF infused.

DISCUSSION

The monocyte/macrophage plays a n essential part in a co- ordinated host defence system, and in this study we have characterized the effects of exposure to hM-CSF on haemato- logical variables and on monocyte functions both in vitro and ex vivo. The administration of hM-CSF leads to a rapid, small fall in Hb. haematocrit and RBC numbers, which is detectable by 30 min following the onset of the infusion. Similar effects have been reported in toxicological studies of rhM-CSF in cynomolgus monkeys and were stated to be transient (Garnick et al. 1989). The Hb had not returned to pre- infusion levels by 24 h after the completion of the infusion in our study, but as patients receiving hM-CSF had between 200 and 300 ml of blood venesected over the 24 h period, it is difficult to reach any conclusion as to the significance of this late effect on Hb concentration. A small and consistent reduction in circulating platelets to a nadir at 9 0 min after the end of the infusion and recovery by 120 min was seen. A reduction in platelets was also seen in primates receiving rhM-CSF, but this effect was stated to be maximal a t 3-6 d after the start of treatment (Munn et al, 1990). No data for the comparable time period under study here were given. Other workers have reported transient reductions in platelet counts of a similar magnitude in five of six patients receiving rhGM- CSF infusions (Lieschke et al, 1989). More prolonged therapy with GM-CSF has also been associated with unmasking of immune thrombocytopenia (Lieschke et al, 1989). and it is possible that the small reductions in platelet counts observed in this study are secondary to increased phagocytic clearance of platelets, with non-specifically associated Ig, by an acti- vated reticuloendothelial system.

It has been reported that the bolus intravenous injection of rhM-CSF to Lewis rats a t doses of 200-400 pg/kg leads to a

3 0 Asim Khwuja et ul transient monocytopenia maximal at 1 5 min after the injection, followed by a monocytosis apparent by 24 h (Ulich et al. 1990). In our study we have noted some fluctuation in monocyte numbers followed by a small reduction from a mean of 0.773f0 .177 x 10y/l pre-infusion to 0 .587f0 .136 x 10y/l a t 240 min after the start of the infusion. with a return toward pre-treatment levels by 6 h. This effect was not clearly dose-dependent and no subsequent monocytosis was detected up to 24 h. This small reduction in monocyte numbers may be related to monocyte activation and consequent margination, although we were unable to demonstrate any increase in CD1 l b expression. This con- trasts with findings in patients receiving rhGM-CSF infusions. where there is a rapid upregulation of CAMS (Devereux et al. 1989), and almost complete disappearance of circulating neutrophils and monocytes due to margination in the pulmonary vasculature, followed by a gradual return to the circulation over the next 2 h (Devereux et al. 1987). It may be possible that larger amounts of hM-CSF would result in a greater fall in monocyte counts similar to that described with rhGM-CSF, although the doses of hM-CSF used here are at least equal to that reported to be sufficient to ameliorate post- chemotherapy neutropenia (Masaoka et al, 1988).

We have shown that exposure to hM-CSF in vivo enhances the ability of peripheral blood monocytes to both ingest and degrade candida, and that this effect is still apparent at 24 h. M-CSF has previously been shown to enhance the expression of both FcRI and FcRII on mature mouse macrophages (Magee et al. 1987). and has been reported as augmenting both human monocyte and mouse macrophage activity against Candida albicans in vitro (Wang et al. 1989: Karbassi et nl. 1987).

We have also shown that hM-CSF does not itself stimulate the monocyte respiratory burst but primes it in response to the bacterial peptide fmlp. Even after hM-CSF priming, less than 20% of fmlp stimulated monocytes appear to produce HLOL in whole blood: similar values were obtained with rhGM-CSF, suggesting that circulating monocytes are gener- ally less responsive to fmlp than circulating neutrophils. Increased respiratory burst activity in M-CSF treated murine macrophages in response to stimulation by phorbol ester has been reported (Wing et al. 1985); on the other hand, no increase in H20L production in human monocyte derived macrophages exposed to M-CSF was seen (Nathan et al. 1984). In contrast to the methods used in our study, these experiments were carried out after separation procedures and incubation with M-CSF for 3 d. followed by stimulation of the respiratory burst by phorbol ester, and this may account for the observed differences in results. The stimulation of the respiratory burst seen in the present study could also explain, at least in part, the enhanced candidacidal activity observed.

It has been shown that rhM-CSF has both chemokinetic and chemotactic effects on human monocyte migration in vitro (Wang et al, 1988). The response was relatively specific for mononuclear cells. although some increase in neutrophil migration was seen at high rhM-CSF concentrations. We have shown in the present study that the systemic adminis- tration of hM-CSF caused a marked increase in the number of monocytes entering an artificially created inflammatory site.

The effect was seen at all dose levels examined. had a rapid onset within 20 min of starting a n infusion, lasted for the duration of the infusion, with a return toward pre-treatment levels after its conclusion. No effect on neutrophil migration was seen at the doses studied. We and others have previously reported that the administration of GM-CSF leads to a reduction in neutrophil migration into skin windows (Addison et al. 1989; Peters et al. 1988a). Treatment with G- CSF has been reported not to affect neutrophil migration (Peters et al, 1988b). Stimulation of monocyte entry into inflammatory sites has potential benefit in the treatment of selected infections in the immune compromised patient, for example in the management of disseminated fungal infec- tion. In conjunction with our observation that hM-CSF increases monocyte phagocytosis and killing of candida. and those of others that M-CSF enhances monocyte survival and maturation (Becker et al. 1987). treatment with M-CSF may represent valuable adjunctive therapy in the treatment of such infections.

In conclusion. we have shown that the clinical administra- tion of hM-CSF to haematologically normal patients with malignant lymphoma augments several mononuclear phagocyte functions, including migration into skin windows. priming of the respiratory burst, and phagocytosis and killing ofcandida. Theseeffects were seen at all dose levels used, with no clear dose relationship. Although these data need to be confirmed in multiple dosing studies, enhancement of mono- cyte function by M-CSF has potential value in the treatment of selected infections in the immune compromised patient as an adjunct to conventional anti-microbial therapy.

ACKNOWLEDGMENTS

This work was supported in part by the Kay Kendall Leukaemia Fund. A.K. is supported by Hoechst U.K.. K.Y. by the MRC, U.K.. and D.C.L. 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 vivo. Clinical and Experimental Immunologg, 76. 149-1 5 3 .

Addison, I.E.. Johnson. B. & Shaw. M. (1982) A human skin window technique using micropore membranes. /ournal of Immunologitwl Methods. 54, 129-1 39.

Bass, D.A.. Parce. J.W.. DeChatelet. L.R.. Szejda. P.. Seeds, M.C. & Thomas, M. (1983) Flow cytometric studies of oxidative product formation by neutrophils: a graded response to membrane stimu- lation. /ournal of Immunology. 1 30, 19 10- 19 1 7.

Becker. S.. Warren. M.K. & Haskill. S. (1987) Colony-stimulating factor induced monocyte survival and differentiation into macro- phages in serum free cultures. Iournal oflmrnunology. 139. 3703- 3709.

Chen. B.D.-M. & Lin. H.4 . (1984) Colony stimulating factor (CSF-I ): its enhancement of plasminogen activator production and inhibi- tion of cell growth in a mouse macrophage cell line. /ournu1 of Immunology. 132, 2955-2960.

Clark, S.C. & Kamen. R. ( 1 987) The human hematopoietic colony- stimulating factors. Science. 236. 1229-1 2 37.

Devereux, S.. Bull. H.A.. Camps-Costa. D.. Saib. R. & Linch. D.C. (1 989) Granulocyte-macrophage colony-stimulating factor

M-CSF and Monocyte Function 31 Motoyoshi. K.. Yoshida. K.. Hatake. K.. Saito, M.. Miura. Y.. Yanai.

N.. Yamada, M.. Kawashima, T.. Wong. C.C.. Temple, P.A.. Leary. A.C.. Witek. Giannotti. IS.. Fujisawa. M.. Yuo. A.. Okabe, T. & Takaku. F. (1989) Recombinant and native urinary colony- stimulating factor directly augments granulocytic and granulo- cyte-macrophage colony-stimulating factor production of human peripheral blood monocytes. Experimental Haematology. 1 7,

Mufson. R.A.. Aghajanian. J., Wong. G.G.. Woodhouse. C.&Morgan. A.C. (1 989) Macrophage colony-stimulating factor enhances monocyte and macrophage antibody-dependent cell-mediated cytotoxicity. Ce/lular Immunology. 119, 182-1 92.

Munn. D.H.. Garnick. M.B. & Cheung. N.-K.V. (1990) UTects of parenteral recombinant human macrophage colony-stimulating factor on monocyte number. phenotype. and antiturnour cytotoxi- city in non-human primates. Blood. 75, 2042-2048.

Nathan. C.F.. Prendergast.T.1.. Wiebe. M.E.. Stanley, E.R.. Platzer. E.. Remold, H.G.. Welte. K.. Rubin. B.Y. & Murray, H.W. (1984) Activation of human macrophages: comparison of other cytokines with interferon p . /ourno/ ofExperirnenml Medicine, 160, 600-605.

Peters, W.P.. Stuart, A.. Affronti, M.L.. Kim, C.S. & Coleman, R.E. ( 1988a) Neutrophil migration is defective during recombinant granulocyte-macrophage colony-stimulating factor infusion after autologous bone marrow transplantation in man. Blood. 72.

Peters. W.P., Kurtzberg. 1.. Atwater. S.. Borowitz. M.. Gilbert. C.. Rao. M.. Currie. M.. Shogan. I., ]ones. R.B.. Shpall. E.J. & Souza. L. (1988b) Comparative effects of rHuG-CSF and rHuGM-CSF on hematopoietic reconstitution and granulocyte function following high-dose chemotherapy and autologous bone marrow transplan- tation. (Abstract). Blood. 72, 13Oa.

Sampson-Johannes, A. & Carlino. J.A. (1988) Enhancement of human monocyte tumoricidal activity by recombinant M-CSF. /ourna/ of Immunology. 141, 3680-3686.

Ulich, T.R.. del Castillo. 1.. Watson. L.R.. Yin. S. & Garnick. M.B. ( 1990) In vivo hematologic effects of recombinant human macro- phage colony-stimulating factor. Blood. 75. 846-8 50.

Wang. I.M.. Griffin. J.D.. Rambaldi. A.. Chen. Z.G. & Mantovani. A. ( 1988) Induction of monocyte migration by recombinant macro- phage colony-stimulating factor. journal of Immunology. 141, 575-579.

Wang, M.. Friedman. H. & Djeu. J.Y. (1989) Enhancement of human monocyte function against candida albicans by the colony- stimulating factors (CSF): IL-3. granulocyte-macrophage CSF. and macrophage CSF. journal of Immunology. 143, 671-677.

Warren, M.K. & Ralph. P. ( 1 986) Macrophage growth factor CSF-I stimulates human monocyte production of interferon. tumor necrosis factor, and colony stimulating activity. Iournal oflmmuno-

Wing, E.J.. Ampel. N.M.. Waheed. A. & Shadduck. R.K. (1985) Macrophage colony-stimulating factor (M-CSF) enhances the capacity of murine macrophages to secrete oxygen reduction products. Iournal oflmmunology. 135, 2052-2056.

Wong. G.C.. Temple. P.A.. Leary. A.C.. Witek-Giannotti. J.S.. Yang. Y.-C., Ciarletta. A.B.. Chung. M.. Murtha. P.. Kriz. R., Kaufman. R.J.. Ferenz. C.R.. Sibley, B.S.. Turner. K.J.. Hewick. R.M.. Clark, S.C.. Yanai. N., Yokota. H.. Yamada. M.. Saito. M.. Motoyoshi. K. & Takaku. F. (1987) Human CSF-1: molecular cloning and ex- pression of 4-kb cDNA encoding the human urinary protein. Science, 235, 1504-1 508.

68-71.

1310-1315.

logy. 137, 2281-2285.

induced changes in cellular adhesion molecule expression and adhesion to endothelium: in vitro and in vivo studies in man. British /ournu/ of Haematology, 71, 323-3 3 0 .

Devereux. S.. Linch. D.C.. Canipos Costa. D.. Spittle, M.F. & Jellilk A.M. ( 1987) Transient leucopenia induced by granulocyte-macro- phage colony-stimulating factor. Lancet. ii, 1523.

Donahue. R.E., Wong. G.G.. Metzger. M.. Seghal. P.K.. Morris, J.P., Turner. K.J.. Morin. S.H.. Sibley. S.B.. Stoudemire. 1.. Clark, S.C. & Garnick. M. (1988) In vivo effects of recombinant human macrophage colony-stimulating factor in primates. (Abstract). Blood, 72, 114a.

El Maalem. H. & Fletcher. 1. (1976) Defective neutrophil function in chronic granulocytic leukaemia. British /ournal of Haematology.

Fearon. D.T.. & Collins. L.A. (1983) Increased expression of C3b receptor on polymorphonuclear leukocytes induced by chemotac- tic factor and by purification procedures. /ourria/ of Immunology. 130. 370-375.

Garnick, M.B.. Stoudemire. J., Donahue. R.E.. Metzger. M.. Bree. A,. Timothy. G.. Wong. C.. Munn. D. & Cheung. N.K.V. ( 1989) In vitro and in vivo biological effects of recombinant human macrophage colony-stimulating factor. ASCO Proceedings, 8. 8 14.

Ishizaka, Y.. Motoyoshi. K.. Hatake, K.. Saito. M.. Takaku. F. &Miura. Y. ( 1986) Mode of action of human urinary colony-stimulating factor. Experimental Haematology. 14, 1-8.

Jaswon. M.S.. Khwaja, A.. Roberts. P.J., Jones. H.M. & Linch. D.C. ( 1 990) The effects of rhGM-CSF on the neutrophil respiratory burst when studied in whole blood. British Iorrrnal of Haematologg. 75,

Karbassi, A,. Becker. J.M.. Foster, J.S. & Moore, R.N. (1987) Enhanced killing of candida albicans by murine macrophages treated with macrophage colony-stimulating factor: evidence for augmented expression of mannose receptors. /ournu/ of Immuno-

Kawasaki, E.S.. Ladner. M.B.. Wang. A.M.. Van Arsdell. 1.. Warren. M.K.. Coyne. M.Y.. Schweickart. V.L.. Lee. M.T.. Wilson, K.J.. Boosman. A, . Stanley. E.R.. Ralph. P. & Mark. D.F. (1985) Molecular cloning of a complementary DNA encoding human macrophage-specific colony-stimulating factor (CSF-I ). Science.

lieschke, GJ., Maher. D.. Cebon. J., O'Connor. M.. Green. M.. Sheridan. W.. Boyd. A.. Rallings. M.. Bonnem. E.. Metcalf. D., Burgess, A.W.. McGrath. K.. Fox, R.M. & Morstyn. G. (1989) Effects of bacterially synthesised recombinant human granulo- cyte-macrophage colony-stimulating factor in patients with advanced malignancy. Annals of Internal Medicine. 1 10, 35 7- 364.

Magee. D.M.. Wing, E.J.. Ampel, N.M.. Waheed. A. &Shadduck. R.K. ( 1987) Macrophage colony-stimulating factor enhances the ex- pression of Fc receptors on rnurine peritoneal macrophages. Immunology. 62, 373-378.

Masaoka. T.. Motoyoshi. K.. Takaku. F.. Kato. S.. Harada. M.. Kodera. Y.. Kanamaru. A.. Moriyama. Y.. Ohno. R.. Ohira. M.. Shibata. H. & Inoue. T. ( 1988) Administration of human urinary colony- stimulating factor after bone marrow transplantation. Bone Marrow Transplantation. 3 , 121-127.

Motoyoshi. K.. Takaku. F.. Maekawa. T.. Miura. Y.. Furusawa. S.. Hattori. M., Nomura. T., Mizoguchi. H.. Ogawa, M.. Kinugasa. K.. Tominaga. T.. Shirnoyama, M., Deura. K.. Ohta, K.. Taguchi. T.. Masaoka. T. & Kimura. 1. ( 1 986) Protective effect of partially purified human urinary colony-stimulating factor on granulocyto- penia after antitumor chemotherapy. Experimental Haernatologu,

34. 95-103.

181 -187.

logy. 1 39. 41 7-42 1 .

230. 291-29h.

14, 1069-1075.