Immunotoxins containing saporin linked to different CD2 monoclonal antibodies: in vitro evaluation

In this study we describe immunotoxins prepared with different CD2 monoclonal antibodies (mAbs) and a ribosome-inactivating protein, saporin. The CD2 immunotoxins were tested on different models. Anti-CD2–saporin conjugates inhibited protein synthesis by a neoplastic CD2+ cell line (SKW-3) and by an interleukin 2 dependent polyclonal CD2+ lymphoid cell culture (T lymphoblasts), with IC₅₀s ranging from 10^-13 m to 10^-11 m (as saporin). Similar results were obtained with proliferation inhibition tests (³H-thymidine incorporation) on phytohaemagglutinin (PHA) driven lymphoid cultures and on mixed lymphocyte culture activated lymphocytes. Moreover a CD2–ricin A chain conjugate was less effective than an analogous immunotoxin containing the same CD2 mAb and saporin in inhibiting lymphocyte proliferation induced by PHA (IC₅₀ approximately 10^-9 m as ricin A chain versus 10^-12 m as saporin). The conjugates were not toxic on bone marrow stem cells. These results suggest that CD2–saporin immunotoxins could represent an effective tool for CD2+ lymphomas or leukaemias, and for T-dependent immune disorders, such as transplanted organ rejection and graft-versus-host disease.     

The primary humoral immune response of European green lizards (Lacerta viridis) to Leishmania agamae

Summary European green lizards, Lacerta viridis, produced relatively thermostable, dithiothreitol-sensitive, non-precipitating, agglutinins and complement-fixing antibodies (CFA) to Leishmania agamae administered subcutaneously (SC), intraperitoneally (IP) or orally (OR). Antibodies were also detected by the immobilization test (IMM) and by enzyme-linked immunosorbent assay (ELISA). The most sensitive method for the detection of stimulated immunoglobulins was ELISA. Antibodies were detected as early as 3 days post-infection with ELISA and between 5 and 7 for CFA, direct agglutination (DA) and indirect haemagglutination (IHA). In the case of IMM, the times of first detection varied from 14 to 28 days. Maximum CFA (2^-8), DA (2^-8), IHA (2^-11) and ELISA (2^-16) titres were reached from 42 to 49 days with significantly higher values occurring in the OR and IP groups. With IMM, maxima occurred after 5 or 6 weeks. Following exposure, two- to five-fold significant increases in serum lysozyme levels were demonstrated but the concentrations in sera following SC, IP or OR routes of antigen administration were not significantly different when the groups were compared with each other. The highest lysozyme values (approximately 12.3–12.5 μgml^-1) were found in the SC and OR groups when compared to the IP(7.40 μgml^-1).     

Effects of all-trans-retinoic acid on superoxide generation in intact neutrophils and a cell-free system

To determine the pathophysiology of the retinoic acid syndrome which occurs during all- trans -retinoic acid (ATRA) treatment of acute promyelocytic leukaemia patients, we investigated the direct effects of ATRA on the function of human neutrophils. We found that ATRA (10–200 μ M ) dose-dependently stimulated superoxide (O⁻₂) generation in intact neutrophils. The maximal activity of ATRA-stimulated O⁻₂ generation was 3.0 nmol/min/10⁶ cells. Adding EGTA to the assay mixture did not affect the activity nor was the intracellular free calcium concentration changed upon stimulation. The treatment of neutrophils with 0.1 μ M staurosporine, an antagonist of protein kinase C, for 10 min, enhanced the activity of ATRA-stimulated O⁻₂ generation up to 186% of that for control samples. Wortmannin (1 μ M ), an inhibitor of phosphatidylinositol 3-kinase (PI 3-kinase), reduced this stimulatory activity by 67%. These results suggest that ATRA activates the signalling pathway related to PI 3-kinase rather than that utilizing calcium and protein kinase C. ATRA enhanced the O⁻₂ generated in a sodium-dodecyl-sulphate (SDS) cell-free system, resulting in rates up to 288% higher than that seen with SDS alone. This enhancement was not affected by pretreatment with staurosporine or wortmannin. ATRA may thus directly activate and/or enhance the function of neutrophils.     

Mercury compounds induce a rapid increase in procoagulant activity of monocyte-like U937 cells

. When monocytic leukaemia line U937 cells were incubated in the presence of HgCl₂ there was a rapid increase in tissue factor (TF)-dependent procoagulant activity, reaching a maximum (equivalent to the total TF activity observed when cells had been subjected to a freeze/thaw cycle) after 15 min at 50 μ m HgCl₂ and after 30 min at 10 μ m HgCl₂. Two other heavy metal compounds, AgNO₃ and phenylmercuric acetate, caused a similar increase in TF activity. The increase was independent of protein synthesis. Other reagents tested, CdCl₂, ZnCl₂, NiCl₂, ADP, FMLP and monocyte chemotactic factor (MCF-1), did not cause a rapid increase in functional activity, when tested under the same experimental conditions. The addition of HgCl₂ to the cells causes, in a concentration-dependent manner, a 10-12-fold increase in intracellular calcium (Ca_i) which coinicides with increase in TF activity. Calcium ionophore also caused an increase in TF activity of the U937 cells. Upon treatment with HgCl₂ the cells surface of U937 cells showed a large increase in the level of phosphatidylserine (PS) on the cell surface (as measured by potentiation of the rate of activation of prothrombin by factor Xa-factor Va) but with no change in the level of TF antigen on the cell surface. We consider that the TF is present on the cell surface of the monocyte but relatively inactive towards the physiological substrate, factor X (FX), until HgCl₂ causes a change in the polarity of the cell membrane exposing PS on the outer leaflet by a mechanism likely to be enhanced by the increase in intracellular calcium.     

Effect of Ebrotidine on Gastric Mucosal Calcium Channel Activity

Ebrotidine is a new H2-receptor antagonist also known for its gastroprotective effect against ethanol-induced mucosal injury. In this study, we investigated the effect of ebrotidine on the activity of the gastric mucosal calcium channels. The channel complex was isolated from the solubilized gastric epithelial cell membranes by affinity chromatography on wheat germ agglutinin. After being labeled with [³H]PN200–110, the complex was reconstituted into phosphatidylcholine vesicles which exhibited active ⁴⁵Ca²⁺ uptake into intravesicular space and responded in a concentration-dependent manner to calcium channel activator, BAY K8644, as well as to calcium channel antagonist, PN200–110. The ⁴⁵Ca²⁺ uptake was inhibited by ebrotidine. Maximum inhibitory effect was attained at 50 μg/ml ebrotidine, at which point a 54.9% decrease in uptake occurred. The gastric mucosal calcium channels, on epidermal growth factor binding (EGF) in the presence of ATP, responded by an increase in tyrosine phosphorylation of 55 and 170 kDa proteins, and the vesicles containing the phosphorylated channels displayed a 48% greater ⁴⁵Ca²⁺ uptake. This phosphorylation process was inhibited by ebrotidine. Furthermore, ebrotidine also interfered with the binding of EGF to calcium channel protein. The results point toward the importance of KGF in the maintenance of gastric mucosal calcium homeostasis, and suggest that ebrotidine has the ability to protect the cellular integrity from calcium imbalance by modulating the ECF-stimulated gastric mucosal calcium channel phosphorylation.     

Serum 'Uracil+Uridine'Levels Before and After Vitamin B12 Therapy in Pernicious Anaemia

S ummary . The serum 'uracil+uridine'levels, expressed as uracil, have been measured in 10 cases of pernicious anaemia both before and after treatment, and compared with the levels in 97 normal subjects. The mean pre-treatment value (8.82 μmol/l., range 6.0–12.0 μmol/l.) differed significantly from that of the normal controls (15.7 μmol/l., range 5.7–40.5 μmol/1., t = 8.8, P <0.001). This confirms the low serum uracil level previously reported in pernicious anaemia in relapse. The level rose progressively after treatment, reaching a maximum on the fourth day (mean 17.85 μmol/1., range 9.3–23.4 μmol/l.). This was not significantly different from the mean of the normal control group. The difference between the pre- and post-treatment levels was significant on days 3, 4 and 5 ( P <0.005, P <0.001 and P <0.005 respectively) and the rise preceded the reticulocyte response by 24 h. A further case was treated with physiological doses of vitamin B₁₂ (2 μg daily for 6 d) and a similar rise in the serum uracil level noted.
These results are not explained by any of the known functions of vitamin B₁₂. They are, however, similar to the changes in the serum methionine levels previously reported in pernicious anaemia. The latter were readily explained by the known action of vitamin B₁₂ on ' de novo 'methionine synthesis and it is suggested that the synthesis of uracil, like that of methionine, might be influenced by vitamin B₁₂ in man.     

Mechanisms involved in the antiplatelet activity of Escherichia coli lipopolysaccharide in human platelets

In this study, Escherichia coli lipopolysaccharide (LPS) dose-dependently (100–300 μg/ml) and time-dependently (10–60 min) inhibited platelet aggregation in human platelets stimulated by agonists. LPS also dose-dependently inhibited the phosphoinositide breakdown and the intracellular Ca⁺² mobilization in human platelets stimulated by collagen. LPS (300 μg/ml) also significantly inhibited the thromboxane A₂formation stimulated by collagen in human platelets. Moreover, LPS (100–300 μg/ml) dose-dependently decreased the fluorescence of platelet membranes tagged with diphenylhexatrience. In addition, LPS (200 and 300 μg/ml) significantly increased the formation of cyclic GMP but not cyclic AMP in platelets. LPS (200 μg/ml) also significantly increased the production of nitrate within a 30 min incubation period. Rapid phosphorylation of a platelet protein of M _r 47 000, a marker of protein kinase C activation, was triggered by phorbol-12-13-dibutyrate (PDBu, 50 n M ). This phosphorylation was markedly inhibited by LPS (200 μg/ml) within a 30 min incubation period.
These results indicate that the antiplatelet activity of LPS may be involved in two important pathways. (1) LPS may induce conformational changes in the platelet membrane, leading to change in the activity of phospholipase C. (2) LPS also activated the formation of nitric oxide (NO)/cyclic GMP in human platelets, resulting in inhibition of platelet aggregation. Therefore, LPS-mediated alteration of platelet function may contribute to bleeding diathesis in septicaemic and endotoxaemic patients.     

Ca2+-CaM activation of AMP deaminase contributes to adenine nucleotide dysregulation and phosphatidylserine externalization in human sickle erythrocytes

Ca²⁺-calmodulin (Ca²⁺-CaM) activates erythrocyte adenosine monophosphate deaminase (AMPD) in conditions of disturbed calcium homeostasis, prompting us to investigate adenine nucleotide metabolic dysregulation in sickle cell disease (SCD). However, higher ATP concentrations in reticulocytes, compared to erythrocytes, confound a comparative evaluation of SCD and normal RBCs. Therefore, a combination of centrifugation and antiCD71-labelled magnetic bead selection was used to prepare reticulocyte-poor fractions (reticulocytes <4% of total RBCs) of SCD RBCs. ATP and total adenine nucleotide concentrations were 12% lower in sickle erythrocytes compared to normal erythrocytes and inosine monophosphate (IMP) concentrations were threefold elevated (all P  < 0·05). Furthermore, preincubation with a diffusible CaM antagonist slowed IMP accumulation in sickle erythrocytes during an experimental period of energy imbalance, thus showing that Ca²⁺-CaM activates AMPD in SCD. Finally, adenine treatment (100 μmol/l) of ex vivo SCD RBCs significantly expanded ATP levels (16% higher) and reduced phosphatidylserine (PS)-exposure, specifically those cells with the highest levels of PS externalization (46% fewer events) (both P -values <0·05 compared to untreated samples). We conclude that Ca²⁺-CaM activation of AMPD contributes to increased turnover of the adenine nucleotide pool in sickle erythrocytes and that this metabolic dysregulation promotes PS exposure that may contribute to the pathogenesis of SCD.     

The Anti-D Content of IgG Preparations for Use in the Prevention of Rh Haemolytic Disease

Summary. Anti-D concentrations in IgG preparations for use in the prevention of haemolytic disease of the newborn have been estimated by direct labelling of the IgG with ¹²⁵I and by an indirect method using ¹²⁵I-labelled antiglobulin.
The concentrations ranged from 50–2000 μg anti-D/ml for total protein concentrations of 10–16 g/100 ml in the IgG preparations.     

I-, MN- and Pr1/Pr2-Activity of Human Erythrocyte Glycoprotein Fractions Obtained by Ficin Treatment

Abstract. Human erythrocyte glycoproteins obtained by the phenol/saline extraction procedure are split using ficin. Ficin treatment results in diffusable and non-diffusable glycopeptides. The nondialysable material is separated on sephadex G 50 columns in 4 fractions with different antigen activities. The main I activity is found in fraction 1 (neuraminic acid (NANA) content 21%). Quantitative differences concerning the MN activity are observed in the fractions 1–3. While fraction 2 (NANA content 22%) is Pr₁active but Pr₂-inactive, fraction 3 (NANA content 20%) is Pr₁-inactive but Pr₂-active. Therefore, the NANA determinating Pr₁ is not involved in the Pr₂ antigenicity. NANA-free fraction 4 is serologically inactive.     

Carrageenan-induced activation of human platelets is dependent on the phospholipase C Pathway

Stimulation of washed human platelets by the pro-inflammatory polysaccharide carrageenan is accompanied by shape change, aggregation and release of granule contents and unaccompanied by thromboxane A₂ synthesis. Carrageenan triggers platelet activation through a prostaglandin synthetase-independent mechanism. The phospholipase A₂ (PLA₂) inhibitor, p-bromophenacyl bromide suppresses platelet responses to carrageenan (Vargaftig et al , 1980) probably by mechanism (s) other than those which involve PLA₂ activity. Exposure of platelets to carrageenan (2–25 μg/ml) induced inositol phosphate formation in a time- and concentration-dependent manner, the level of inositol phosphate formation correllating with the intensity of aggregation. Neomycin, an aminoglycoside antibiotic which inhibits the phospholipase C -mediated phosphatidylinositol 4,5-bisphosphate breakdown, suppressed both platelet activation and inositol phosphate formation. Inhibition was concentration-dependent with an IC₅₀ value of about 180 μM. Platelet-activating factor (PAF) is not responsible for carrageenan-induced platelet activation and inositol phosphate formation, since exposure of platelets to carrageenan (25 μg/ ml) in the presence of compound WEB 2086 (100 μM). a PAF antagonist, failed to inhibit carrageenan responses. However, compound Ro 19-3704, a structurally related antagonist of PAF reported to be also an inhibitor of phospholipases A₂ and C, inhibited concentration-dependently (0.1–10 μg/ml). These findings indicate that carrageenan activates human platelets through a phospholipase C-dependent mechanism and show that neomycin, at low concentrations, can be a selective inhibitor of phospholipase C-mediated PIP2-breakdown.     

Heterogeneity of Human Platelets III. GLYCOGEN METABOLISM IN PLATELETS OF DIFFERENT SIZES

S ummary Glycogenolysis, glycogenesis, glyconeogenesis and the enzymes phosphorylase, glycogen synthetase, and fructose-1–6-diphosphatase were investigated in heavy and light platelet populations in order to determine the presence or extent of metabolic heterogeneity. The mean volume of the heavy platelets was approximately 2.4-fold greater than that of the light platelets. Heavy platelets contained 3.4-fold more glycogen and had a 5.9-fold greater rate of glycogenolysis per g wet weight, than light platelets. Heavy platelets incorporated 5.9-fold more [¹⁴C]glucose into platelet glycogen when data were expressed per g wet weight, and 1.7-fold more [¹⁴C]glucose when data were expressed per μmole of glycogen. This difference in glycogen synthesis was paralleled by a 1.8-fold greater glycogen synthetase activity. Heavy platelets incorporated 3.0-fold more [¹⁴C]citrate into platelet glycogen when data were expressed per g wet weight, and 1.4-fold more [¹⁴C]citrate when data were expressed per μmol of glycogen. Similar results were obtained utilizing [¹⁴C]pyruvate as precursor. This difference in glyconeogenesis was paralleled by a 2.5-fold greater fructose-1–6-diphosphatase activity for heavy platelets.     

Two calcium/calmodulin-dependent protein kinases, which are highly concentrated in brain, phosphorylate protein I at distinct sites.

Two calcium-stimulated protein kinase activities (ATP:protein phosphotransferase, EC 2.7.1.37) that phosphorylate protein I, a specific synaptic protein, have been identified in homogenates of rat brain. One of these is found in both the particulate and cytosolic fractions and phosphorylates a region of protein I that is phosphorylated in intact synaptosomes in response to calcium but not to cyclic AMP. The stimulation by calcium of the particulate enzyme and of the partially purified cytosolic enzyme requires the addition of calmodulin. It is not yet known whether the particulate and cytosolic enzymes are related. A second calcium-stimulated protein I kinase is found only in the cytosol and phosphorylates a region of protein I that is phosphorylated in intact synaptosomes in response to either calcium or cyclic AMP. The calcium stimulation of this latter kinase is probably mediated by calmodulin, judging from its inhibition by low concentrations of trifluoperazine. Both of the calcium-stimulated protein I kinases are more highly concentrated in brain than in other tissues. The two cytosolic kinases are distinguishable from each other and from myosin light chain kinase and phosphorylase b kinase by their substrate specificities and their chromatographic behavior on DEAE-cellulose.     

Phosphatidylserine surface expression and integrin αIIbβ3 activity on thrombin/convulxin stimulated platelets/particles of different sizes

Platelets stimulated by a combination of thrombin/convulxin have been shown to develop two to three populations characterized by different phosphatidylserine (PS) surface expression and integrin αIIbβ3 activity. To determine how these markers are distributed on the surface of platelets/particles, we studied Annexin V and PAC-1 binding to platelets/particles of different sizes by flow cytometry analysis and evaluated influences of calpain and caspase inhibitors on thrombin/convulxin-activated platelets. Analysed platelets/particles were divided by their sizes, according to the standard size beads, into seven populations from 0·37 to 4·8 μm. PAC-1 binding/μm² was almost equal in platelets/particles ranging from 1·2 to 4·8 μm and was significantly lower on smaller-sized particles sizes (0·37–0·7 μm). PS surface exposure/μm² was high in the particles of 0·37–1·2 μm and very low in platelets (2·6–4·8 μm). Upon thrombin/convulxin stimulation caspase inhibitors prevented microparticle (MP) formation, while a calpain inhibitor stimulated MP formation. It was also shown that stimulated platelets are heterogeneous not only in their ability to activate αIIbβ3 integrin complex and expose PS on their surface, but also in the distribution of activation markers, which strongly depends on platelet/particle size and that platelets/particles of different sizes provide different responses to the same stimulus.     

Serum markers of bone metabolism in multiple myeloma: prognostic value of the carboxy-terminal telopeptide of type I collagen (ICTP)

This study was performed to evaluate the prognostic significance of serum markers of bone and collagen metabolism in multiple myeloma. Serum C-terminal telopeptide of type I collagen (ICTP) reflects degradation of bone, whereas serum osteocalcin, together with serum C-terminal propeptide of procollagen type I (PICP) and serum bone-specific alkaline phosphatase (bAP) reflect synthesis of bone matrix. The N-terminal propeptide of procollagen type III (PIIINP) in serum reflects synthesis of type III collagen. We analysed frozen sera from 109 patients with newly diagnosed multiple myeloma. Serum ICTP was elevated (>5.0μg/l) in most patients (median 6.6 μg/l, range 1.4–29.4 μg/l). Serum PIIINP was elevated (>4.2μg/l) in 46% (median 4.0 μg/l, range 1.4–20.1 μg/l). Serum PICP was generally within the reference limits, whereas serum osteocalcin and serum bAP were elevated in 19% and 37%, respectively. Serum ICTP correlated with serum PIIINP, serum β₂-microglobulin (β₂m), serum calcium, performance status, and stage. In univariate analysis, the test variables serum ICTP ( P =0.026) and serum osteocalcin ( P =0.036) were found to be of prognostic value, but PIIINP, PICP, or bAP in serum were not. Serum ICTP and serum β₂m had a similar prognostic value. In multivariate analysis, serum calcium showed the highest prognostic significance, and serum β₂m was the only other variable of independent prognostic value. However, in normocalcaemic patients, serum ICTP showed the highest prognostic significance, followed by serum osteocalcin. Thus, the serum levels of ICTP and osteocalcin seem related to bone turnover and calcium metabolism, and provide further information about myeloma activity, particularly in normocalcaemic patients.     

Lack of Conformity in the Behaviour of Platelets during Normal Storage Conditions at 22°C

Abstract. Platelets prepared from citrate-phosphate-dextrose blood under normal routine conditions and concentrated to 1,000–1,600 · 10⁹ · 1^-1 varied considerably in their capacity to acidify the medium. Serotonin uptake was well maintained for 5 days provided that adenosine 5'-triphosphate was within 50% of normal and pH was above 6.0. A strong release of platelet factor 4 was occasionally seen in relation to preparation. In some preparations the platelet factor 4 was well maintained for 5 days, in others a sudden or gradual release was seen during storage. The investigation indicates that differences in the quality of routinely prepared platelets occur during preparation and storage which are insufficiently well explained and controlled.     

Effect of Testosterone and Oestradiol on Erythroprotein Production in Vitro

The effects of testosterone and oestradiol on erythropoietin production by mouse fetal liver in primary culture was studied. No stimulatory effect on in vitro erythropoietin production was observed by any of the oestradiol concentrations employed (10^-10M to 10^-6M); whereas testosterone, only over a dose range of 10^-12 M to 10^-10 M, stimulated erythropoietin production by mouse fetal liver in culture. In addition, a stimulatory effect of testosterone on fetal liver tissue was obtained by injecting testosterone (20 mg/mouse) intramuscularly into pregnant mice 4 d prior to obtaining the fetal liver tissue for culture. Therefore, it is concluded that testosterone can exert a direct stimulatory effect on erythropoietin production by mouse fetal liver.     

Observations of the Number of Available c, D, and E Antigen Sites on Red Cells

Abstract. The number of available antigen sites within the Rh system were estimated using trace-labelled antibodies. The results were as follows, given as the number of sites per red cell: c-antigen: on cc cells, 70,000–85,000: on Cc, 37,000–53,000; D sites on cells of phenotype -D-: 110,000–202,000; E-antigen sites showed considerable heterogeneity depending on phenotype as well as source of anti-E, and estimates varied between 450 and 25,600; e-antigen: on eE cells, 13,400 and 14,500: on ee cells, 18,200 and 24,400.
The average equilibrium constants of the antibodies were: anti-E, 4 × 10⁸ 1/mol; anti-e, 2.5 × 10⁸ 1/mol; anti-c, 3.2 × 10⁷ and 5.6 × 10⁷ 1/mol.     

Platelet-activating factor induces protein kinase activity in the particulate fraction of human neutrophils

Platelet-activating factor (PAF) is a proinflammatory lipid that has both platelet- and phagocyte-stimulating properties. Because several known activators of calcium-, phospholipid-dependent protein kinase (protein kinase c, PKC) also stimulate neutrophil responses and because neutrophil stimuli such as phorbol diesters and the chemotactic peptide f-Met-Leu-Phe are reported to increase protein kinase activity in neutrophil (PMN) particulate fractions, we investigated the effect of PAF on neutrophil protein kinase activities. In neutrophils exposed to 10(-6) mol/L PAF, cytosolic PKC activity was 521 +/- 38 pmol 32P/10(7) PMN/min (mean +/- SEM), which was not significantly lower than cystolic activity in buffer-treated controls (558 +/- 32 pmol 32P/10(7) PMN/min, n = 14). PAF-exposed cells exhibited a concomitant rise in protein kinase activity associated with the particulate fraction with 53 +/- 4 pmol 32P/10(7) PMN/min compared with 32 +/- 2 pmol in control cells (n = 14). Particulate protein kinase activity was independent of the presence of calcium and phospholipid in the assay medium. The specific PKC inhibitor H-7 inhibited particulate protein kinase activity, however, which suggested that the enzyme activity assayed in this fraction may be PKC in a constitutively activated form. The increase in particulate protein kinase activity induced by PAF required the presence of cytochalasin B, was detectable within 5 seconds of exposure to PAF, and was not reversed by washing the cells free of extracellular PAF after initial exposure. Although PAF did not have a direct effect on PKC activity from cytosolic fractions from resting cells, the increase in particulate protein kinase activity induced by PAF was inhibited when the cells were first depleted of calcium by incubation with Quin 2. These results suggest that PAF induces an increase in particulate protein kinase activity in neutrophils by a calcium- dependent mechanism and that the induction of membrane-associated protein kinase activity may be involved in neutrophil-stimulating actions such as superoxide production, which occur at higher concentrations of PAF.     

Granulocyte colony-stimulating factor given in addition to interferon-α to mobilize peripheral blood stem cells for autologous transplantation in chronic myeloid leukaemia

In order to potentially mobilize and harvest the Ph⁻ cells observed in most patients with chronic myeloid leukaemia (CML) during interferon-α (IF-α) therapy, G-CSF (filgrastim), 5 μg/kg/d, was administered subcutaneously together with IF-α to 30 CML patients in haematological remission but with various degrees of cytogenetic remission, after IF-α therapy. Peripheral blood stem cells (PBSC ) were harvested using standard aphereses from day 5 of G-CSF. Patients underwent one to four (median three) aphereses. Median total yields/kg were 7.6 (range 3.8–25) × 10⁸ MNC, 3.4 (0–140) × 10⁶ CD34⁺ cells, and 17 (1.1–107) × 10⁴ CFU-GM. No patient had a significant increase in the percentage of Ph⁺ cells in the bone marrow under G-CSF therapy. The percentage of Ph⁺ cells in apheresis products tended to decrease between the first and the last apheresis ( P  = 0.05). 14 patients who were not responsive to IF-α were transplanted after conditioning with busulphan 16 mg/kg and melphalan 140 mg/m². Median time to neutrophils > 0.5 × 10⁹/l was 20 d (16–114 d) and to platelets > 50 × 10⁹/l 18 d (12–149 d). Nine patients had a major cytogenetic response post graft, which correlated with the amount of Ph⁺ cells reinfused with the graft ( P  = 0.02). We conclude that this procedure is feasible, allowing the harvest of enough PBSC, some of them Ph⁻ in patients who responded to IF-α, to allow autologous transplantation.