Sema7A is a Potent Monocyte Stimulator

Scand. J. Immunol. 56, 270±275, 2002

Sema7A is a Potent Monocyte Stimulator S. HOLMES,* A.-M. DOWNS,y A. FOSBERRY,y P. D. HAYES,y D. MICHALOVICH,y P. MURDOCH,y K. MOORES,y J. FOX,y K. DEEN,y G. PETTMAN,y T. WATTAMy & C. LEWISy *Oxford GlycoSciences, Abingdon, Oxon; and yGlaxoSmithKline Pharmaceuticals, Harlow, Essex, UK (Received 5 March 2002; Accepted in revised form 8 May 2002)

Holmes S, Downs A-M, Fosberry A, Hayes PD, Michalovich D, Murdoch P, Moores K, Fox J, Deen K, Pettman G, Wattam T, Lewis C. Sema7A is a Potent Monocyte Stimulator. Scand J Immunol 2002;56:270±275 Sema7A is a recently described member of the semaphorin family that is associated with the cell surface via a glycophosphatidylinositol linkage. This study examined the mRNA expression and biological properties of this protein. Although the expression of Sema7A was demonstrated in lymphoid and myeloid cells, no stimulation of cytokine production or proliferation was evident in B or T cells. In contrast, Sema7A is an extremely potent monocyte activator, stimulating chemotaxis at 0.1 pM and inflammatory cytokine production (interleukin-1 (IL-1b), tumour necrosis factor-a (TNF-a), IL-6 and IL-8) and superoxide release at 1±10 pM. Sema7A is less effective at stimulating neutrophils. Sema7A also significantly increases granulocyte-macrophage colony-stimulating factor (GM-CSF) production from monocytes but has no consistent effect on IL-10, IL-12 or IL-18. Sema7A can also induce monocytes toward a dendritic cell morphology. Sema7A is expressed in monocytes and probably released through proteolysis and acts as a very potent autocrine activator of these cells. Dr S. Holmes, Oxford GlycoSciences, The Forum, 86 Milton Park, Abingdon, Oxon OX14 4RY, UK. E-mail: [email protected]

INTRODUCTION The semaphorin family of transmembrane and secreted proteins was initially implicated in axonal guidance in embryonic development [1]. While previous semaphorins were of neuronal origin, of late several semaphorins have been localized to the immune system. Sema4D (CD100) and SEMA4D (mouse semaphorin G) may play a role in lymphocyte aggregation, germinal centre formation and maturationdependent movement of immature thymocytes [2, 3]. Based on expression patterns, a recently described member of the semaphorin family, Sema7A [4], has also been suggested to have a function in the immune system [5±7]. Sema7A is associated with the cell surface through a glycophosphatidylinositol (GPI) linkage [6] and is highly homologous to the semaphorin encoded by alcelaphine herpesvirus-1 (SEMAVB, 46%) [5, 8]. Homology to all other known semaphorins is 30% or less [5]. The viral semaphorins (vaccinia and alcelaphine herpesvirus) and Sema7A have recently been shown to bind to a plexin-like receptor, # 2002 Blackwell Science Ltd

VESPR (plexin-C1), with dissociation constant, KD, in the low nanomole range [9, 10]. The class 4 semaphorins (Sema4D) bind to a receptor (plexin-B1) which is distinct from Sema7A. VESPR has been characterized as a 200 kDa transmembrane glycoprotein that may signal by associating with a tyrosine kinase. The vaccinia virus semaphorin (SEMAVA) has also been shown to induce CD54 expression and stimulate interleukin-6 (IL-6) and IL-8 production from cultured monocytes. Many viruses encode proteins that function to modulate the immune system, perhaps as a defence mechanism used by the virus to suppress the immune response of the host. Virally encoded proteins have also been found to modulate the immune system by acting as cytokines/cytokine antagonists or by interfering with antigen presentation or the complement cascade [11, 12]. In the present study, we have investigated the biological properties of Sema7A and report on data demonstrating that it is a potent stimulator of monocytes and neutrophils.

Sema7A is a Potent Monocyte Stimulator

MATERIALS AND METHODS Cloning of the human Sema7A cDNA. The complete Sema7A coding sequence (GenBank accession number AF030698) was amplified by reverse-transcription polymerase chain reaction (RT-PCR) as follows. Oligo dT-primed first-strand cDNAs were made from 1 mg of human placental polyA mRNA (Clontech, Palo Alto, USA), using Superscript II reverse transcriptase (Life Technologies, Gaithersburg, MD, USA) according to the manufacturer's instructions. Oligonucleotide primers designed to amplify the coding sequence of human Sema7A (forward 50 -ACC ATG ACG CCT CCT CCG C-30 , reverse 50 GCT CTG AGT GTG AGA CGT TC-30 ) were used for PCR from 1/500th of the RT reaction product. 1/25th of the PCR product was then further amplified in PCR that used the same forward primer and a nested reverse primer (50 -CAG AAG CCT GAG GCA TGC-30 ). Final concentrations of reagents in both PCRs were 0.2 mM dNTP, 40 units/ml of PfuTurbo DNA polymerase (Stratagene, La Jolla, USA), 1 PfuTurbo PCR buffer and 12 pM PCR primers. The thermal cycle used for both PCR reactions began at 94  C for 2 min, followed by eight cycles of 94  C for 10 s, 63  C for 30 s and 72  C for 2 min, 20 cycles of 94  C for 10 s, 58  C for 30 s and 72  C for 3 min, and ended with a further 3 min at 72  C. The PCR products were separated by agarose gel electrophoresis. Those of the size expected for Sema7A were purified by a silica-based method (Qiaquick Kit, Qiagen, Hilden, Germany), cloned into the topoisomerase-activated pcDNA3.1-topo vector (Invitrogen, Groningen, The Netherlands), and transformed into Escherichia coli K12 TOP10 (Invitrogen). PCR colony screening identified the clones containing the plasmid with the Sema7A insert in the correct orientation. Plasmid DNA was prepared from suspension cultures of selected clones by alkaline lysis followed by anionexchange chromatography (Qiagen Miniprep Kit, Qiagen). Clones were sequenced on both strands from standard vector primers and internal gene-specific primers, using an ABI automated sequencer. Sequences were assembled with a software package, SEQMAN (DNASTAR, Madison, USA), and the alignments were optimized manually to give an overall consensus. Vector construction and gene cloning/expression. The C-terminal end of Sema7A contains a hydrophobic region which is cleaved off during the addition of the GPI anchor [6]. The truncated Sema7A gene ( 25AA C0 from terminal end) was PCR amplified using the in-house derived full-length clone as the template. Primers were designed with unique restriction sites at both the N0 - and C0 -terminal ends of the gene to allow efficient subcloning into the expressionvector. PCR amplification was carried out using PfuTurbo DNA polymerase using cycling conditions as recommended by the manufacturer (Stratagene). The addition of 5% dimethyl sulphoxide (DMSO) was required for efficient PCR owing to the high percentage GC ratio of the gene. The PCR product was of expected size and was cloned into the PCR cloning vector TOPO TA as per instructions of the manufacturer (Invitrogen). The expression-vector, pIL4/ FcLinkXa, was altered by replacing the factor Xa cleavage site with the PreScission cleavage site (Pharmacia, Uppsala, Sweden). The IL4 leader sequence was also removed on subcloning, as the Sema7A gene had its own signal sequence. The fidelity of the gene and vector construction was confirmed by DNA sequencing on both strands using an ABI 377 automative DNA Sequencer (Applied Biosystems, Foster City, USA). Generation of stable mass cultures in CHOE1a. Large-scale plasmid DNA isolation of the clone pFcLinkPrec/Sema7A and the empty

271

vector control pFcLinkPrec was carried out using maxi-preps (Qiagen) as per instructions of the manufacturer. Concentration of the plasmid DNA was determined by optical density (OD) 260/280 nm and the fidelity of the clones determined by restriction endonuclease digestion. Transfection of pFcLinkPrec/Sema7A and also the control vector pFcLinkPrec into CHOE1A cells was performed using electroporation (3  107 cells, 0.4 cm gap cuvette, 380 V 25 mF). Cell lines were maintained after transfection in MR1-4 medium ‡ 1 X nucleosides. Selection was commenced 48 h post-transfection in MR1-4 medium (no nucleosides). Media was changed every 48±72 h, until stable mass cultures were obtained (4 weeks). The supernatant of the stable mass of both pFcLinkPrec/Sema7A and the control pFcLinkPrec was evaluated for secreted protein expression after 72 h of growth. Culture and purification of Sema7A-Fc. The CHO_Sema7A-Fc cells were routinely maintained in shake flask cultures which were shaken at 100±120 rpm, incubated at 37  C in an atmosphere of 5% CO2 in air. The cells were grown in a proprietary serum-free medium and were maintained at a cell concentration between 2 and 30  106 cells/ml. For the production of recombinant protein, the cultures were scaled up into 1 l shake flasks and grown to a cell concentration of approximately 8  105 cells/ml, at which point the cultures were transferred to 34  C for the duration of the production. The cultures were harvested for product by centrifugation (2000  g, 15 min) before the cell viability dropped below 85%. Routinely, the cell viability at harvest was in excess of 90%. The supernatant containing the recombinant protein was stored sterile at 4  C prior to purification. Purification was accomplished by capture of conditioned media on protein A (ProsepA, BioProcessing, Consett, UK), washing with phosphate-buffered saline (PBS), followed by elution with 25 mM citrate pH 3. Eluted fractions were immediately neutralized with 500 mM Tris pH 9.0 (by titration to pH 7.5). The yield of Sema7A-Fc was in the range of 20±40 mg/l of conditioned media. Cleavage of Fc was accomplished by buffer exchange into 50 mM Tris pH 7.0, 150 mM NaCl, 1 mM ethylenediaminetetraacetic acid (EDTA), 1 mM dithiothreitol DTT and addition of 16 EU PreScission protease per 25 mg of Sema7A-Fc and incubation for 16 h at 4  C. The cleaved material was then passed over a second ProsepA column and the non-adsorbed material collected and concentrated by stirred cell UF. TaqMan mRNA profiles. Total RNA from multiple cell lines and polyA‡ RNA from tissues of four different individuals were isolated and cDNA was prepared. TaqMan PCR was performed to detect either Sema7A, VESPR or housekeeping genes [13]. Cell culture. Human monocytes were prepared from healthy donors by an indirect or direct purification scheme using kits purchased from Miltenyi Biotec (Bergisch Gladbach, Germany) (similar results being obtained from both procedures). The monocytes were >90% pure by flow cytometric analysis. The cells (400 K/well) were cultured in 96-well plates in RPMI-1640 with 10% heat inactivated autologous serum, and the media were harvested as indicated for cytokine analysis. Human neutrophils were prepared by the method of Merritt et al. [14], and B- and T-cell preparations were purified using kits supplied by Miltenyi Biotec. Monocyte superoxide release assay. To each well was added 50 ml of monocytes (5  106/ml) in Dulbecco's phosphate buffered saline (DPBS) containing 0.5 mM CaCl2, 0.2% bovine serum albumin (BSA), 25 ml of 0.6 mM cytochrome C and 25 ml of test compound. The plate was then read at 550 nm, 37  C. Absorbance at 550 nm was converted to nanomoles O2±, based on the extinction coefficient of (reduced minus oxidized) cytochrome C, DE550 nm ˆ 21  103/M/cm.

# 2002 Blackwell Science Ltd, Scandinavian Journal of Immunology, 56, 270±275

272 S. Holmes et al. Chemotaxic assay. Cell migration was evaluated using a 48-well modified Boyden microchemotaxis chamber as described by Berkhout et al. [15]. Cytokine immunoassays. 96-well plates were coated with 1 mg/ml of anticytokine monoclonal antibody (MoAb) in PBS overnight at 4  C. Following blocking and washing, the wells were incubated for 60 min shaking at room temperature with 50 ml of standard/ unknown plus 50 ml of 0.5 mg/ml biotin anticytokine MoAb in assay buffer. After washing, the wells were incubated for 30 min shaking at room temperature with Eu3‡-labelled streptavidin (Perkin Elmer, Shetton, USA), wells washed, and enhancer (Perkin Elmer) was added and the time-resolved fluorescence measured. Anticytokine-paired MoAbs were purchased from Pharmingen (San Diego, CA, USA) or prepared in-house (IL-4, IL-5 and IL-18).

RESULTS Sema7A cloning cDNA encoding human Sema7A [5, 6] was cloned from placental mRNA by RT-PCR. The PCR reaction produced a band of expected 2 kb size for the full-length Sema7A coding sequence, and this was cloned into the mammalian expression-vector pcDNA3.1. Complete sequencing of the full-length clone revealed three single nucleotide differences compared with that reported by Lange et al. [5]. However, these did not change the encoded protein, and agreed with the partial sequence reported by Xu et al. [6]. The purified Sema7A-Fc and Sema7A had molecular weights of 100 and 76 kDa, respectively, from SDS-PAGE gel electrophoresis.

TaqMan analysis The highest expression of Sema7A was detected in monocytes and C13 cells (immortalized immature microglial cells) and to a lesser extent in peripheral blood mononuclear cell (PBMC), lung and brain (Fig. 1). The expression in monocytes remains unchanged after 1 day in culture with human serum; however, after 4 days, the expression increases to four times the naive levels (data not shown). Sema7A (or VESPR) was not present in THP1 cells ± an immature monocytic cell line. Cell lines derived from vessel wall, bladder, liver, bone, cartilage, astrocytoma, neuroblastoma, prostate and kidney were tested, and only those positive are depicted in Fig. 1. Low-level expression of Sema7A was detected in freshly isolated neutrophils but was tripled upon attachment of the cells to plastic, and was further increased with 20 nM granulocyte-macrophage colony-stimulating factor (GM-CSF) treatment, up to 10-fold at 1 h (data not shown). Levels of VESPR mRNA were 50-fold higher in macrophages than Sema7A, but highest levels were found in neutrophils (Fig. 1). In general, the expression of VESPR correlated with Sema7A, the exceptions being the presence of VESPR in CCF-STTG1 cells (astrocytoma), negative for Sema7A and high levels of Sema7A in the C13 cells but low levels of VESPR. Effect of Sema7A on cytokine production Sema7A had no effect on inducing proliferation or cytokine production from T cells (interferon-g (IFN-g), IL-2, IL-4, IL-5)

VESPR

Sema7A 1500

80 000

1000 750

40 000

500 20 000

Copies Sema7A mRNA

Copies VESPR mRNA

1250 60 000

+

IMR32 NT-2 C13 CCF-STTG1

0

Brain Pituitary Hypothalamus Spinal cord Heart Lung Liver Foetal liver Kidney Muscle Stomach Intestine Spleen Lymphocytes Adipose Pancreas Prostate Cartilage Bone Bone marrow Colon Macrophages Platelets THP1 Neutrophils Hep-G2

250 0

Fig. 1. Tissue and cell profile of Sema7A and VESPR mRNA by TaqMan analysis. Data have been presented as copies of gene's mRNA detected per nanogram of polyA‡ RNA for tissues and copies of mRNA detected per 50 ng of total RNA for cells. These units are equivalent if assuming the mRNA pool represents 2% of total RNA. Cell lines ± Hep-G2, hepatocyte carcinoma; IMR32‡, neuroblastomas; NT-2, neuronal precursor; C13, immature microglial; CCF-STTG1, astrocytoma. # 2002 Blackwell Science Ltd, Scandinavian Journal of Immunology, 56, 270±275

Sema7A is a Potent Monocyte Stimulator Table 1. Effect of Sema7A on cytokine production from monocytes*

Table 2. Effect of Sema7A-Fc on proinflammatory cytokine production from neutrophils*

Cytokine (ng/ml)

Donor 2 Control Sema7A-Fc (10 nM) LPS (200 ng/ml) Sema7A-Fc ‡ LPS

IL-8

TNF-a

0.0 0.1 1.3 17.5 236.0

0.1 0.1 0.2 1.3 22.6

28.0 32.0 41.0 116.0 903.0

0.0 0.1 0.5 6.0 25.0

4.5 182.0 189.0 194.0

0.3 24.6 14.8 25.8

295.0 828.0 794.0 703.0

0.1 9.2 7.0 11.1

*Monocytes were exposed to Sema7A for 24 h and the cytokines measured by immunoassay. Results are the mean of triplicate determinations (standard deviation (SD) < 5%). Sema7A at 1 pM (interleukin-6 (IL-6) and tumour necrosis factor-a (TNF-a)) and 10 pM (IL-1b, IL-8) was significantly different from control (P < 0.05).

with or without T-cell activators, and Sema7A also has no effect on B-cell cytokine release (IL-6, tumour necrosis factor-a (TNF-a), TNF-b, IL-10) (data not shown). Dependent on the donor cell preparation, Sema7A is active at 1±10 pM in stimulating proinflammatory cytokines (IL-1b, TNF-a, IL-6, IL-8) from monocytes (Table 1). Sema7A and Sema7A-Fc are equally effective at stimulating proinflammatory cytokines from monocytes, maximum activity being approximately at 10 nM. The stimulation of monocyte cytokine release by Sema7A is equivalent to a maximal dose of lipopolysaccharide (LPS), and the two stimulators do not produce an additive effect (Table 1). Sema7A also consistently stimulated the production of GM-CSF from monocyte cultures, and this cytokine together with IL-6 and IL-1b and to some extent IL-8 was refractory to Sema7A stimulation after 4 days in culture, whereas TNF-a levels were unaffected ± this is a similar profile to LPS stimulation (data not shown). In all experiments, controls of heat-inactivated Sema7A/ Sema7A-Fc (to test for the presence of low amounts of endotoxin) were included, and no protein preparations accommodated any stimulation of cytokine. Monocytes were exposed to immobilized CD80-Fc, used as a control for Fc-receptor-mediated effects, and this produced the same level of cytokines as untreated cells. Sema7A also has no consistent effect on IL-10, IL-12 or IL-18 secretion from monocytes (data not shown). TNF-a-induced secretion from monocytes could be detected after 1 h of stimulation with Sema7A; other cytokines being undetectable after this time.

Cytokine (ng/ml)

Control Sema7A-Fc (1 nM) Sema7A-Fc (10 nM)

IL-1b

IL-6

IL-8

TNF-a

0.0 0.0 0.3y

0.0 0.1y 11.2y

0.2 0.5y 15.1y

0.0 0.0 0.4y

*Cells were cultured for 24 h, and results are the mean of triplicate determinations (standard deviation (SD) < 5%). IL, interleukin; TNF, tumour necrosis factor. ySignificantly different from control (P < 0.05).

The secretion of TNF-a did not stimulate other proinflammatory cytokines, as the addition of excess neutralizing antibodies to TNF-a did not inhibit Sema7A-stimulated IL-1b, IL-6 or IL-8 expression (data not shown). It was also noted that monocyte cultures in the presence of Sema7A contained large number of cellular aggregates, whereas control cultures did not, suggesting that Sema7A induces cell-surface adhesion/activating proteins. Sema7A was also less effective, approximately 20-fold, at stimulating proinflammatory cytokines (IL-6, IL-8, IL-1b and TNF-a) from neutrophils compared with monocytes (Table 2). Effect of Sema7A on monocyte superoxide release and chemotaxis Sema7A stimulates superoxide release at picomole concentrations (Fig. 2), and is similar in potency to fMLP. Sema7A has distinct differences compared with MCP-1 which has no effect on superoxide release and stimulates Ca transients (EC50 ˆ 1.5  0.2 nM), whereas Sema7A had no effect on Ca transients in monocytes (data not shown). The effect of Sema7A on chemotaxis of freshly isolated human peripheral 2.5

–

(1 pM) (10 pM) (100 pM) (1 nM)

IL-1b

nmol O2 /250 K cells

Donor 1 Control Sema7A Sema7A Sema7A Sema7A

IL-6

273

Control FMLP100 nm Sema7A 0.1 nm

2.0

Sema7A 1 nm Sema7A 10 nm 1.5

1.0 0

10

20

30 40 Time (min)

50

60

Fig. 2. Effect of Sema7A on monocyte superoxide release. Results are the mean of triplicate determinations ± standard deviation (SD) < 5%. Significant stimulation with Sema7A (P < 0.05) at all concentrations was evident after 15 min of incubation.

# 2002 Blackwell Science Ltd, Scandinavian Journal of Immunology, 56, 270±275

274 S. Holmes et al. 4.0

Control (5 days)

3.5

26.1

C.I.

3.0

60.1

Sema7A-Fc (5 days) 89.3

2.1

CD11b-PE

2.5 3.2

0.0

2.0 CD14-FITC

1.5 82.4

2.1

84.7

4.8

1.0 0.1

1

10

CD86-PE

100

pM

0.4

Fig. 3. Effect of Sema7A on monocyte chemotaxis. Results are the mean of triplicate determinations ± standard deviation (SD) < 5%. Significant (P < 0.05) increase in chemotactic index was observed at 0.1±30 pM Sema7A. Representative of four similar experiments. EC50 for Sema7A was 0.4 pM.

blood monocytes was tested in a 48-well Boyden chamber, and a representative dose±response curve of these studies is shown in Fig. 3. Sema7A gave the expected bell-shaped response curve with the inhibition of chemotaxis at high concentrations owing to receptor desensitization. In all experiments, Sema7A consistently induced monocyte chemotaxis at femtomole concentrations and has an EC50 of 0.4 pM which is 1000 times more potent than MCP-1 (EC50 value of MCP-1stimulated human monocyte chemotaxis being 0.37 nM). Sema7A was less effective in stimulating neutrophil chemotaxis (EC50 ˆ 20 pM). Flow cytometric analysis of Sema7A stimulated monocytes After 5 days in culture in the presence of immobilized Sema7A-Fc, the monocytes are showing a trend towards becoming CD11b‡/CD14±/CD1a±/CD83‡/CD40‡/CD86‡ dendritic cells (Fig. 4). Sema7A probably increases monocyte cell survival, as well as inducing maturation into dendritic cells through stimulation of M-CSF and GM-CSF. The Sema7A-Fc data was compared with immobilized CD80-Fc as a control for Fc-mediated effects. DISCUSSION Lange et al. [5] were the first to elucidate the sequence of Sema7A, with a subsequent publication of Xu et al. [6] demonstrating that Sema7A was GPI-linked. There are a number of GPI-anchored leucocyte surface proteins that play a diverse role in immunoregulation [16], and it has been demonstrated that these GPI-anchored proteins occur in microdomains or `rafts' at the surface of cells [17]. GPI proteins, for example CD14, can also be shed from the surface of monocytes [18], and specific proteases responsible

0.1

CD1a-FITC 2.7

1.0

36.2

21.8

CD83-PE 0.2

12.9

CD40-FITC

Fig. 4. Flow cytometric analysis of monocytes cultured in the presence of Sema7A-Fc. Monocytes were cultured with plate-bound Sema7A-Fc (10 nM) or CD80-Fc (control) for 5 days and then the cells analysed by flow cytometry. Specificity of staining of the anti-CD monoclonal antibodies (MoAbs) was demonstrated with phycoerythrin (PE)- and fluorescein isothiocyanate (FITC)-labelled immunoglobulin G (IgG). One representative of three similar experiments. Percentage of cells in each quadrant is depicted.

for GPI cleavage and protein secretion have been identified [19]. There is evidence that Sema7A can be released from cells through proteolysis [20], which would implicate Sema7A in having an autocrine function. Although semaphorins are widely known for their role in neuronal function, immunological properties have also been demonstrated for Sema4D [2] and Sema3C [21], as well as the vaccinia virusA39R protein SEMVA [7], which shares around 30% homology with Sema7A. This protein has a biological function on monocytes, producing modest increases in IL-6 and IL-8 (when compared with Sema7A). Sema7A is an extremely potent stimulator of monocytes and neutrophils. We have demonstrated that Sema7A is active at 1 pM, stimulating monocyte cytokine production, and significantly induces monocyte chemotaxis at 100 fM. EC50 for monocyte chemotaxis is 1000 times more potent than for MCP-1 and in neutrophils approximately 100 times more potent than for IL-8. The estimated KD of Sema7A binding to its receptor VESPR (plexin-C1) in transfected cell lines has been reported as 2.1 nM by Tamagnone et al. [10]. It is therefore surprising that Sema7A is biologically active at 0.1±1 pM, some three±four orders of magnitude less than the KD. We have not been able to confirm

# 2002 Blackwell Science Ltd, Scandinavian Journal of Immunology, 56, 270±275

Sema7A is a Potent Monocyte Stimulator

Sema7A binding to monocytes by various methods, probably owing to a very low receptor density. It is also possible that a coreceptor/cofactor is also involved in Sema7A binding to VESPR which may explain the enhanced biological activity. TaqMan expression data confirmed that Sema7A and VESPR, besides being present in monocytes and neutrophils, are also present in other lymphoid cells. Previous reports have localized VESPR to T and B cells [7], and antibodies to surface-bound Sema7A have been developed and demonstrate the presence of Sema7A on most T and natural killer (NK) cells and all B cells, following activation [20]. However, Sema7A does not stimulate B or T cells. As mentioned above, this could be owing to a cofactor missing in T and B cells to prevent Sema7A stimulation or a defect in the signalling cascade in these cells. However, this does demonstrate a divergent role for Sema7A in myeloid versus lymphoid cells. Monocytes cultured in the presence of serum tend towards a macrophage morphology, and although Sema7A expression is higher in macrophages than monocytes, macrophages are less responsive to Sema7A stimulation of GM-CSF, IL-1b, IL-6 and IL-8 but not TNF-a. Neutralizing antibodies to TNF-a added to monocyte cultures rule out the possibility that Sema7A cytokine stimulation is mediated by TNF-a. It has been suggested that the monocyte is a relatively immature precursor cell that can differentiate into macrophage or dendritic cell requisite on the stimuli supplied [22]. Sema7A induces the maturation of monocytes towards a dendritic cell morphology in vitro and may play a role in this process, although future experiments will need to delineate whether this is wholly owing to GM-CSF stimulation. High levels of Sema7A expression are also present in the brain and especially microglial cells. Microglial cells are postulated to arise from blood monocytes, which infiltrate the brain during embryonic development and the second and third week postnatal. Besides an immune regulation, Sema7A may also contribute to a neural immune interaction. This study has demonstrated that Sema7A is a potent autocrine stimulator of monocytes, but further investigations are required to delineate the exact biological role of Sema7A and its function within the immune system with respect to myeloid cells.

REFERENCES 1 Kolodkin AL, Matthes DJ, Goodman CS. The semaphorin genes encode a family of transmembrane and secreted growth cone guidance molecules. Cell 1993;75:1389±99. 2 Hall KT, Boumsell L, Schultze JL et al. Human CD100, a novel leukocyte semaphorin that promotes B cell aggregation and differentiation. Proc Natl Acad Sci USA 1996;93:11780±5. 3 Furuyama T, Inagaki S, Kosugi A et al. Identification of a novel transmembrane semaphorin expressed on lymphocytes. J Biol Chem 1996;271:33376±81.

275

4 Goodman CS, Kolodkin AL, Luo Y et al. Unified nomenclature for the semaphorins/collapsins. Cell 1999;97:551±2. 5 Lange C, Liehr T, Goen M, Gebhart E, Fleckenstein B, Ensser A. New eukaryotic semaphorins with close homology to semaphorins of DNA viruses. Genomics 1998;51:340±50. 6 Xu X, Ng S, Wu ZL et al. Human semaphorin K1 Is glycosylphosphatidylinositol-linked and defines a new subfamily of viral-related semaphorins. J Biol Chem 1998;273:22428±34. 7 Comeau MR, Johnson R, DuBose RF et al. A poxvirus encoded semaphorin induces cytokine production from monocytes and binds to a novel cellular semaphorin receptor, VESPR. Immunity 1998;8:473±82. 8 Ensser A, Fleckstein B. Alcelaphine herpesvirus type 1 has a semaphorin-like gene. J Gen Virol 1995;76:1063±7. 9 Winberg ML, Noordermeer JN, Tamagnone L et al. Plexin A is a neuronal semaphorn receptor that controls axon guidance. Cell 1998;95:903±16. 10 Tamagnone L, Artigiani S, Chen H et al. Plexins are a large family of receptors for transmembrane, secreted and GPIanchored semaphorins in vertebrates. Cell 1999;99:71±80. 11 Spriggs MK. Shared resources between the neural and immune systems: semaphorins join the ranks. Curr Opin Immunol 1999;11:387±91. 12 Smith GL, Symons JA, Khanna A et al. Vaccinia virus immune evasion. Immunol Rev 1997;159:137±54. 13 Sarau HM, Ames RS, Chambers J et al. Identification, molecular cloning, expression, and characterization of a cysteinyl leukotriene receptor. Mol Pharmacol 1999;56:657±63. 14 Merritt JE, Jacob R, Hallam TJ. Use of manganese to discriminate between calcium influx and mobilization from internal stores in stimulated human neutrophils. J Biol Chem 1998;264:1522±7. 15 Berkhout TA, Sarau HM, Moores K et al. Cloning, in vitro expression, and functional characterization of a novel human CC chemokine of the monocyte chemotactic protein (MCP) family (MCP-4) that binds and signals through the CC chemokine receptor 2B. J Biol Chem 1997;272:16404±13. 16 Horejsi V, Drbal K, Cebecauer M et al. GPI microdomains: a role in signalling via immunoreceptors. Immunol Today 1999;20:356±61. 17 Varma R, Mayor S. Microdomains of GPI anchored proteins in living cells revealed by cross-linking. Nature 1998;294:802±4. 18 Bazil V, Strominger JL. Shedding as a mechanism of down modulation of CD14 on stimulated human monocytes. J Immunol 1991;147:1567±74. 19 Fritz BA, Lowe AW. Polarized GP2 secretion in MDCK cells via GPI targeting and apical membrane restricted proteolysis. Am J Physiol 1996;270:176±83. 20 Angelisova P. Characterisation of the human leukocyte GPI anchored glycoprotein CDw108 and its relation to other similar molecules. Immunobiology 1999;200:234±45. 21 Mangasser-Stephan K, Dooley S, Welter C et al. Identification of human semaphorin E gene expression in rheumatoid synovial cells by mRNA differential display. Biochem Biophys Res Commun 1997;234:153±6. 22 Peters JH, Gieseler R, Thiele B, Steinbach F. Dendritic cells: from ontogenic orphans to myelomonocytic descendants. Immunol Today 1996;17:273±7.

# 2002 Blackwell Science Ltd, Scandinavian Journal of Immunology, 56, 270±275