Letter | Published:

Neurotactin, a membrane-anchored chemokine upregulated in brain inflammation

  • An Erratum to this article was published on 04 September 1997

Abstract

Chemokines are small secreted proteins that stimulate the directional migration of leukocytes and mediate inflammation1,2,3,4. During screening of a murine choroid plexus complementary DNA library, we identified a new chemokine, designated neurotactin. Unlike other chemokines, neurotactin has a unique cysteine pattern, Cys-X-X-X-Cys, and is predicted to be a type 1 membrane protein. Full-length recombinant neurotactin is localized on the surface of transfected 293 cells. Recombinant neurotactin containing the chemokine domain is chemotactic for neutrophils both in vitro and in vivo. Neurotactin messenger RNA is predominantly expressed in normal murine brain and its protein expression in activated brain microglia is upregulated in mice with experimental autoimmune encephalomyelitis, as well as in mice treated with lipopolysaccharide. Distinct from all other chemokine genes, the neurotactin gene is localized to human chromosome 16q. Consequently we propose that neurotactin represents a new δ-chemokine family and that it may play a role in brain inflammation processes.

Main

Three classes of chemokines have been categorized to date: α (CXC), β (CC) and γ (C). The definition for the subclass of chemokines is based on the distance between the first two of four characteristic disulphide-forming cysteine residues. The α-chemokines, including interleukin-8 (IL-8), neutrophil-activating protein 2 (NAP-2), GRO-α-γ, epithelial-cell-derived neutrophil activating protein (ENA-78) and granulocyte chemotactic protein 2 (GCP-2), are mainly chemotactic for neutrophils. By contrast, the β- or CC chemokines are chemotactic for monocytes and, in some instances, for eosinophils or basophils, but not for neutrophils. β-Chemokines include monocyte chemotactic protein (MCP) 1-5, the macrophage inflammatory proteins MIP-1α and MIP-1β, eotaxin and RANTES. Lymphotactin is the only γ-chemokine that has only one pair of cysteines and is thought to be chemotactic for lymphocytes5,6,7. Here we report the discovery of a fourth class of chemokine with significant structural and functional differences to other chemokines.

A cDNA library constructed from murine choroid plexus8 was studied by automated high-throughput single-pass sequencing and computer analysis. The predicted amino-acid sequence of one cDNA clone, pT1, exhibited similarity to murine MCP-1. The complete sequence of pT1 was determined and the open reading frame was continuous for 395 amino acids. A full-length cDNA encoding the human homologue of pT1 was further identified, and the predicted open reading frame of this clone (397 amino acids) was 67% identical to that of murine pT1 over its entire length, including a predicted signal peptide of 22 amino acids (Fig. 1). An alignment of the chemokine-like domains from the human homologue of the pT1 gene product (termed neurotactin) with human IL-8, GRO-α, MCP-1, RANTES and lymphotactin is shown in Fig. 2a . Neurotactin contained the four disulphide-forming cysteines present in all α- and β-chemokines. Both murine and human neurotactin were most closely related to MCP-1 from the corresponding species (40% identical at amino acid level).

Figure 1: Amino-acid sequence alignment of human and murine neurotactin (abbreviated as NTN).
figure1

The putative signal peptide is represented by a double underline and the transmembrane domain is underlined. The alignment was generated with the BESTFIT program from the GCG package (Genetics Computing Group, Madison, WI). Identical residues are indicated by asterisks underneath.

Figure 2: Sequence alignment of human neurotactin with α-chemokines IL-8 and GRO-α; β-chemokines MCP-1 and RANTES and γ-chemokine lymphotactin.
figure2

Residues identical to neurotactin are boxed. NTN, neurotactin; LTN, lymphotactin. The alignment was generated with the Pileup program from the GCG package with manual modification. b, Primary structure comparison of human neurotactin with other subclasses of human chemokines. The representatives for the CC, CXC and C chemokine subfamilies are IL-8, MCP-1 and lymphotactin, respectively. Cysteines that are involved in confirmed (IL-8, MCP-1, NTN) or putative (LTN) disulphide-bond formation are linked by straight lines.

The primary structure of neurotactin is notably different from all other chemokines (Fig. 2b). First, it differs in the arrangement of the first two cysteines, by the inclusion of a three amino-acid spacer: CXXXC. Second, the neurotactin precursors are substantially larger than other chemokine precursors, which are usually less than 150 amino acids long. The region of homology to other chemokines is located at the amino terminus approximately from residues 23 to 92. Third, neurotactin contains a stretch of 19 conserved hydrophobic residues towards the carboxy terminus, suggesting that it is a membrane protein. Three additional cysteines are present at the C-terminal portion of the neurotactin, one in the transmembrane domain and two in the cytoplasmic domain.

The expression profile of pT1 was determined by RNA (northern) blot analysis of murine and human tissues. Expression of a 3.5-kilobase (kb) mRNA was detected predominantly in murine brain and to a lesser extent in kidney, heart and lung (Fig. 3), as well as uterus (data not shown). Three minor mRNA species were also observed in the brain. This pattern of expression is unusual in the chemokine family. Human neurotactin gene expression was most abundant in the brain and heart, but was also present in all other tissues tested except peripheral blood leukocytes (data not shown). Murine neurotactin mRNA was upregulated in cells of endothelial and fibroblastic origin (data not shown). In EOMA (endothelial cells) and STO (embryonic fibroblasts) the expression of neurotactin mRNA was constitutive and inducible by 12-O-tetradecanoylphorbol-13-acetate (TPA) and lipopolysacchride (LPS), whereas in BMS-12 (bone marrow stromal) cells the neurotactin mRNA was detected only after stimulation with TPA or LPS. Neurotactin mRNA was not detected in several cell lines of haematopoietic origin with or without stimulation by TPA and LPS. The tested cell lines were WEHI-3 and Pu5-1.8 (myelomonocytic), P388D1 and IC-21 (monocytic/macrophage), AKR.G.2 (thymoma), BaF3 (pro B cell), EL-4 (lymphoma), NFS-1.0 (B-cell lymphoma) and BCL (B-cell leukaemia).

Figure 3: Murine tissue northern blot probed with a full-length murine neurotactin cDNA probe.
figure3

Hybridization of the same blot with actin probe is shown at the bottom to demonstrated equal loading RNA in each lane.

In order to determine the association of neurotactin with the cell membrane, a construct containing the full-length neurotactin coding region was made in mammalian expression vector pN8e vector (kindly provided by J. Morgenstern and K. Theriault). A FLAG epitope tag was added to the C terminus. As shown in Fig. 4 , strong staining was detected on the surface of 293 EBNA cells transiently transfected with this construct, suggesting that full-length neurotactin is indeed membrane-anchored. The signal was competable by FLAG peptide.

Figure 4: Immunofluorescent labelling of transfected 293 cells.
figure4

a, b, Fluorescent staining and light field image from transfected 293 cells. c, d, Fluorescent staining and light field image from transfected 293 cells after competition with FLAG peptide. e, f, Fluorescent staining and light field image in mock transfection.

Two soluble versions of murine neurotactin, corresponding to the chemokine-like domain (P1, from residue 22 to 105) and entire extracellular domain (P3, from residue 22 to 337) were constructed and expressed as glutathione S-transferase (GST) fusion proteins in Escherichia coli. Purified P1 was analysed by matrix-assisted laser desorption ionization mass spectrometry after a Staphylococcus aureus V8 protease digestion to examine its disulphide formation (data not shown). Under non-reduced conditions two proteolytic fragments were recovered which represent the covalent attachment of peptides (F-1)-E33 to S57-R66 and I34-E56 to T67-E79, respectively. These peptides encompass the C32-C58 and C36-C74 disulphide pairings, respectively. Under reducing conditions in the presence of DTT, the consistent peptides of the pairs were observed along with complete absence of the disulphide-bonded peptides substantiating the assignment. No additional ions were detected that would support alternative disulphide pairing and thus indicate proper folding of the domain with respect to homologous chemokines.

Purified P1 and P3 were subsequently assayed for their chemotactic specificity in vitro. As shown in Fig. 5 , recombinant neurotactin P1 was found to be chemotactic for human neutrophils in vitro with a maximal activity at 10−8M (n = 3), whereas P3 has no significant effect on all three types of human leukocyte tested. The same chemotaxis index for IL-8 has been documented9. P1 was also chemotactic for human T lymphocytes. Further testing showed that neither P1 nor P3 had chemotactic effect on human monocytes, human myelomonocytic THP1 cells and murine P388D1 monocytic cells (data not shown). P1 and P3 were subsequently injected into mice peritoneally to determine their chemotactic efficacy in vivo. Bacterial lysate containing only GST was purified in the same fashion as P1 and P3, to serve as a control for the in vivo assay. Two hours after injection, peritoneal exudate was collected and assayed for leukocyte infiltration (Fig. 6). Consistent with the in vitro results, P1 was chemotactic for neutrophils and lymphocytes invivo. However, P1 in vivo was also chemotactic for monocytes, but not for eosinophils. The activity of P1 towards monocytes is probably due to a secondary effect in vivo and therefore may not reflect its chemotactic specificity. The inability of P3 to chemoattract monocytes and lymphocytes in vivo was also consistent with the in vitro data, but in contrast to the observations in vitro, P3 was chemotactic for neutrophils, suggesting differential activation of neurotactin in vivo.

Figure 5: Chemotaxis assay of recombinant murine neurotactin on purified human neutrophils (a), monocytes (b) and lymphocytes (c).
figure5

The chemotaxis index is calculated as the ratio of the number of leukocytes migrated onto the sample filter to those on the control filter (n = 3). The percentage migration of neutrophils and lymphocytes induced by P1 (at the optimal concentrations) is 12% plus or minus 3% and 6.5% plus or minus 0.3%, respectively. The background migration for neutrophils and lymphocytes is 4.5% plus or minus 2% and 2.3% plus or minus 0.5%.

Figure 6: Neutrotactin-induced recruitment of leukocytes to the peritoneum.
figure6

Peritoneal exudate was collected from C57BL/6J mice 2h after injection of either P1 or P3 neurotactin (fitted symbols), PBS or GST (empty symbols). Each dot represents one individual mouse analysed (6 and 4 mice per experimental and control group, respectively). The bar in each panel represents the mean of the total number of cells of the leukocyte type indicated.

Neurotactin expression in normal and LPS-treated murine brain was determined by immunohistochemistry. Frozen brain sections from normal, LPS-treated mice were stained with an anti-neurotactin peptide antibody with or without peptide competition (Fig. 7a–f). In normal brain, staining was localized to capillary vessels and resident microglia. Two hours after LPS treatment, increased labelling intensity on the same cell types was observed. An increasing number of the activated microglial cells stained positive for the anti-neurotactin antibody. In both normal and LPS-treated brain, staining of larger vessels was restricted to the apical region of the endothelium. It is known that, in addition to resident microglial cells, two other subtypes of microglial cell are present in the central nervous system at the blood–brain barrier: perivascular (completely surrounded by basal lamina), and juxtavascular microglia (directly in contact with basal lamina)10. The anti-neurotactin antibody staining associated with micro-vessels was consistent with staining of those microglial cells.

Figure 7: Expression of murine neurotactin in normal, LPS-treated and EAE brain.
figure7

Normal murine brain sections stained with anti-neurotactin polyclonal antibody, with (a) or without (b) peptide competition (original magnification ×250). c, Higher magnification of an area containing vascular endothelium (original magnification ×1,000). LPS-treated murine brain sections stained with anti-neurotactin polyclonal antibody, with (d) or without (e) peptide competition. f, Higher magnification of a microglial cell. Paraffin sections from control (g) and EAE (h) mice stained with anti-neurotactin polyclonal antibody (original magnification ×250).

Paraffin-embedded brain sections from mice with severe experimental autoimmune encephalomyelitis (EAE) were examined to evaluate neurotactin expression further. EAE develops spontaneously in immunodeficient α-myelin basic protein T-cell receptor transgenic mice11 or in mice that have been actively immunized with proteolytic peptide (PLP)30, and these have been used as a model of chronic inflammatory diseases. Neurotactin was found to be upregulated in activated microglia from such EAE mice, in a similar manner to LPS-treated mice (Fig. 7g–h).

Both human and murine neurotactin genes were located on different chromosomes from other chemokines. The genes for the α-chemokines have been mapped to human chromosome 4 and 10 (ref. 31), and the β-chemokines to human chromosome 17 and mouse chromosome 11 (ref. 3). Lymphotactin was located to chromosome 1 in both human and mouse5,6. Using a panel of backcross progeny of C57B1/6J Mus musculus and Mus spretus mice the murine neurotactin gene was mapped to the long arm of chromosome 8 (data not shown). The human neurotactin gene was localized by radiation hybrid mapping to the syntenic region on chromosome 16q (data not shown). This result further supports the notion that neurotactin represents a new class of chemokine.

The unusual pattern of cysteine residues (CXXXC) and membrane association suggest that neurotactin represents a new class of chemokine and may have new functions through membrane association that contrast with soluble chemokines. Several chemokines, such as IL-8 and GRO-α, have been shown to act locally because of their association with the extracellular matrix12, and by doing so to establish a gradient in the blood flow. Although it is unknown whether neurotactin functions through a membrane-bound or soluble form, or both, the membrane association of neurotactin represents a new mechanism for chemokine presentation, localization and juxtacrine signalling. Juxtacrine stimulation is found in a wide range of species including insects and mammals, and is thought to be important in developmental processes that have spatial restrictions13. There are many examples of other cytokine classes that can be membrane associated or secreted, such as steel factor (SLF), macrophage colony-stimulating factor, epidermal growth factor (EGF) and members of the tumour necrosis factor (TNF) family13. Lack of the membrane-associated form of SLF in sl/sldmice causes pathology despite the presence of a soluble form14,15.

By sequence comparison, neurotactin is most closely related to MCP-1, although functionally it seems to be different from the β-chemokines based on chemotaxis data with recombinant proteins. Both in vitro and in vivo, recombinant neurotactin P1, which contains the chemokine domain only is chemotactic for neutrophils, resembling the α-chemokines, such as IL-8. However, from the limited tissues and cells tested, neurotactin gene expression is overlapping but different from IL-8. For example, they are both expressed in fibroblastic and endothelial cells; however, by northern analysis, neurotactin RNA was not detected in nine murine B, T and monocyte/macrophage cell lines or human peripheral blood leukocytes. It is also interesting to note that neurotactin does not contain the ‘ELR’ motif present in other neutrophil-chemoattractic chemokines16,17,18.

Our results have demonstrated that the chemokine-like domain (P1) and complete extracellular domain (P3) of neurotactin are differentially active in vitro and in vivo. Principally the entire extracellular domain construct (P3) is incapable of inducing chemotaxis of leukocytes in vitro, yet it is chemotactic for neutrophils in vivo. One explanation for the difference is that P3 undergoes further post-transitional processing in vivo to generate a form more analogous to P1 that is able to chemoattract neutrophils. Our data suggest the possibility that multiple forms of neurotactin exist with different ranges of activity controlled by post-translational proteolysis.

Primary sequence analysis of neurotactin indicates the possibility for the generation of secreted or cleaved forms of neurotactin. Both human and murine neurotactin contain a conserved dibasic potential protease cleavage site that lies towards the N terminus of the predicted transmembrane domain (R339–340 in human and R337–338 in mouse). Pairs of basic residues have been shown to serve as the signals for precursor cleavage by mammalian proteases such as furin and PC3 (ref. 19) and might also enable the release of a soluble version of neurotactin. The C-terminal residue of both human and murine neurotactins is valine. A C-terminal valine residue in membrane-anchored proTGF-α and steel factor is essential for their extracellular proteolytic cleavage20. Alternatively spliced forms of neurotactin mRNA, such as those observed in the brain (Fig. 3), might also result in shorter soluble forms, although the molecular nature of these have yet to be defined.

The induction of neurotactin by inflammation and the ability of recombinant neurotactin to cause leukocyte chemotaxis suggests a role in inflammatory responses. Neutrophil demargination and infiltration in the brain is characteristic of acute bacterial infections such as sepsis. It also occurs early in cerebral ischaemia, stroke and brain trauma21. The increased expression of neurotactin in brain microglia 2h after LPS administration suggests rapid upregulation following inflammation. Furthermore, the high levels of neurotactin expression in EAE brain by activated microglia indicates that neurotactin might also be involved in the generation and progression of other brain inflammatory diseases, such as multiple sclerosis.

Given the many unique properties of neurotactin we propose that it is the first of a new class of chemokines, and by convention we propose that this new class of chemokines be termed the δ-family of chemokines.

Methods

High-throughput sequencing and computer analysis. DNA was prepared from individual transformants using the AGTC (Gaithersburg, MD) system in a 96-well format. Samples were sequenced by dye-primer fluorescent chemistry on automated ABI373 and ABI377 machines. Expressed sequence tag (EST) sequences were automatically processed after filtering out E. coli, ribosomal or mitochondrial contaminants. Next, vector and repetitive elements were masked and/or removed from each sequence. Sequences were then searched against the Genbank nucleotide database using the BLASTN program22, and through a non-redundant protein database using the BLASTX program22,23. More detailed analysis of individual DNA sequences was done using the GCG package of programs. Contig assembly of sequences was done using Sequencher (Gene Codes Corp., Ann Arbor, MI). The putative signal peptide was predicted by a signal-peptide prediction program24. The human homologue was initially identified as a partial sequence entry in dbEST. The clone was obtained from Genome Systems Inc. (St Louis, MO) and the complete sequence was determined as described above.

Northern blot analysis. Northern blots were probed using standard techniques25 with a 32P-labelled DNA fragment encoding the full-length pT1. Northern blots from multiple murine tissues and control actin probe were obtained from Clontech (Palo Alto, CA).

Detection of membrane-associated neurotactin. Full-length murine neurotactin cDNA was modified for expression in mammalian system by polymerase chain reaction (PCR) with the following primers: 5′-GGGAAAGAATTCATGGCTCCCTCGCCGCTCGCGTGG-3′ and 5′-GGGAAACTCGAGTCATTTATCATCATCATCTTTATAATCCACTGGCACCAGGACGTATGA-3′. Nucleotides coding for DYKDDDDK (FLAG epitope) were incorporated into the 3′ reverse primer for subsequent detection by the M2 anti-FLAG antibody. The PCR products were cloned into the pN8e vector bearing the EBV origin of replication. The construct DNA was prepared with the Qiagen Maxiprep kit (Qiagen, Chatsworthy, CA) and transfected into 293 EBNA cells with lipofectamine (Gibco, Gaithersburg, MD) in 8-well chamber slides. 48 hours after transfection, the cells were fixed with 50% methanol and 50% acetone for 1min at room temperature. After washing with 2.5ml Tris-buffered saline four times, the cells were first incubated with 10µgml−1 of M2 anti-FLAG monoclonal antibody (Eastman Kodak, New Haven, CT) and then with fluorescein isotricyanate (FITC)-conjugated goat anti-mouse antibody at 1:1,000 dilution (Jackson Immuno Research, West Grove, PA) for 1h each. The immunofluorescent staining was visualized with a Zeiss Axioskop 20 with ×200 magnification.

Recombinant murine neurotactin expression in E. coli. Soluble murine neurotactin was constructed by PCR using modified oligonucleotides. The forward primer for P1 construct was 5′-GGGAAAGAATTCCTGCCGGGTCAGCACCTCGGCATG-3′ and the reverse primer was 5′-GGGAAACTCGAGTCACTTCTCAAACTTGCCACCATTTTT-3′. The reverse primer for the P3 construct was 5′-GGGAAACTCGAGTCACCTTGTGGCTGCCTGGGTGTCGGG-3′. pfu Taq polymerase was purchased from Stratagene (La Jolla, CA). The PCR products were ligated to pGEX-4T (Pharmacia, Piscataway, NJ) by EcoRI and XhoI sites and transformed into E. coli DH5α (LifeTechnology, Gaithersburg, MD). Recombinant protein was induced, purified with glutathione-Sepharose and cleaved from GST according to the manufacturer's instructions. The products P1 and P3 contain five additional amino acids (GSPEF) at the N terminus resulting from thrombin digestion of GST fusion proteins. As a control, GST was prepared along with P1 and P3 at the same time. The same amount of thrombin was added to the control GST before binding to an endotoxin BX column. Endotoxin was removed from the preparation using an endotoxin BX column (Cape Cod Associates, Falmouth, MA) and was determined to be <0.01EUml−1.

Chemotaxis assay. The ability of neurotactin to elicit chemotaxis was tested in an in vitro assay as described for PMN26, monocytes26 and lymphocytes27. Neutrophils, monocytes and T lymphocytes were isolated from human peripheral blood as described27,28,29. Preparations were greater than 90% pure by Wright–Geimsa staining. 50,000 neutrophils were added to each well of a 48-well micro chemotaxis chamber containing a 5-µm pore-size filter. Human IL-8 and MCP-1 were purchased from Biosource International (Camarillo, CA). Varying concentrations of neurotactin and other chemokines were added to the lower chamber of each appropriate well. The chamber was incubated at 37°C, 5% CO2 for 30min. Stained and fixed migrated cells were quantified by counting three high-power fields (×400 magnification) per well. The number of cells that migrated to the buffer control were subtracted as background.

The in vivo chemotactic migration of leukocytes to neurotactin was determined by injection of either P1 fraction or P3 fraction of neurotactin protein into the peritoneum of C57BL/6J mice. Two hours after neurotactin (0.5µg in 400µl PBS per mouse) or control protein (GST, 0.5µg in 400µl PBS per mouse) or PBS administration (400µl per mouse), peritoneal exudate was collected and the number and subtype of leukocytes recovered in this fluid determined. Number and type of leukocytes was determined for four high-power fields (×40 magnification; total area 0.5mm2) per section and area. Number of neutrophils and eosinophils were determined by staining with Wright–Giemsa. Macrophages and monocytes were assayed by Moma-2 antibody. Lymphocytes were asessed by Thy 1.2 (53-2.1) and IgM (II/41) immunostaining.

Neurotactin expression in normal, LPS-induced and EAE murine brains by immunohistochemistry. Polyclonal anti-neurotactin antibody was raised in rabbit against peptide LPGQHLGMTKCEIM at the N terminus of the protein (Research Genetics, Huntsville, AL). Antibody was affinity-purified from 12-week bleeds. Eight-week-old CD1 mice were injected intravenously with 40µg LPS and then killed by cervical dislocation after 2h. The brain was excised, bissected traversely before each piece was rolled in Tissue Tek OCT compound (Cryoform, IEC, MA) snap frozen in isopentane/dry ice and stored at −70°C. Sections (3µm) were cut onto microscope slides, air-dried, fixed in 2% paraformaldehyde (5min, 4°C) and methanol (10min, −20°C). Brain from mice suffering active EAE were fixed in 10% neutral buffered formalin before paraffin embedding. Sections (4µm) were microwaved twice for 5min in 0.01M sodium citrate (pH 6) before staining. Fixed sections were stained with antibody using an avidin/biotin staining method. All incubations were done under humidified conditions and slides were washed twice between steps for 5min each in 0.1M PBS supplemented with 0.2% gelatin (PBSG). Sections were overlaid with 20% fetal calf serum in PBS for 15min and then incubated overnight at 4°C with polyclonal anti-neurotactin or normal rabbit serum (both diluted to 1/200 in PBS supplemented with 0.1% BSA). Endogenous peroxide was blocked by incubation for 20min in methanol containing 0.3% hydrogen peroxide. Nonspecific staining due to cross-reaction with endogenous avidin or biotin was blocked by incubation with avidin solution followed by biotin solution, both for 20min. Bound monoclonal antibody was visualized by incubation with biotinylated swine anti-rabbit immunoglobulin (Dako, CA) and then streptavidin peroxidase complex both diluted in 10% normal mouse serum PBS, and incubated for 1h. Finally, slides were flooded with peroxidase substrate solution (400µg diaminobenzidine in 10ml PBS, containing 0.01% hydrogen peroxide) for 10min before counter-staining with haematoxylin. Control sections were included where monoclonal antibody, biotinylated anti-rabbit immunoglobulin or streptavidin complex were selectively omitted. In addition competitive inhibition of the antibody was accomplished by preincubation of antibody with the peptide (25µgml−1) for 45min at 37°C before incubation with tissue sections.

References

  1. 1

    Oppenheim, J. J., Zachariae, C. O. C., Mukaida, N. & Mursushima, K. Properties of the novel proinflammatory supergene “intercrine” cytokine family. Annu. Rev. Immunol. 9, 617–648 (1991).

  2. 2

    Baggiolini, M. & Dahinden, C. A. CC chemokines in allergic inflammation Immunol. Today 15, 127–133 (1994).

  3. 3

    Baggiolini, M., Dewald, B. & Moser, B. Interleukin-8 and related chemotactic cytokines-CXC and CC chemokines. Adv. Immunol. 55, 97–197 (1994).

  4. 4

    Schall, T. J. & Bacon, K. B. Chemokines, leukocyte trafficking, and inflammation. Curr. Opin. Immunol. 6, 865–873 (1994).

  5. 5

    Kelner, G. S. et al. Lymphotactin: a cytokine that represents a new class of chemokine Science 266, 1395–1399 (1994).

  6. 6

    Kennedy, J. et al. Molecular cloning and functional characterization of human lymphotactin. J. Immunol. 155, 203–209 (1995).

  7. 7

    Yoshida, T., Imai, T., Kakizaki, M., Nishimura, M. & Yoshie, O. Molecular cloning of a novel C or g type chemokine, SCM-1. FEBS Lett. 360, 155–159 (1995).

  8. 8

    Tartaglia, L. A. et al. Identification and expression cloning of a leptin receptor, OB-R. Cell 83, 1263–1271 (1995).

  9. 9

    Moser, B., Clark-Lewis, I., Awahlen, R. & Baggiolini, M. Neutrophil-activating properties of the melanoma growth-stimulatory activity. J. Exp. Med. 171, 1797–1802 (1990).

  10. 10

    Gehrmann, J., Matsumoto, Y. & Kreutxberg, G. W. Microglia: intrinsic immunoeffector cell of the brain. Brain Res. Rev. 20, 269–278 (1995).

  11. 11

    Lafaille, J. J., Nagashima, K., Katsuki, M. & Tonegawa, A. High incidence of spontaneous autoimmune encephalomyelitis in immunodeficient auto-myelin basic protein T cell receptor transgenic mice. Cell 78, 399–408 (1994).

  12. 12

    Witt, D. P. & Lander, A. D. Differential binding of chemokines to glycosaminoglycan subpopulations. Curr. Biol. 4, 394–400 (1994).

  13. 13

    Massague, J. & Pandiella, A. Membrane-anchored growth factors. Annu. Rev. Biochem. 62, 515–541 (1993).

  14. 14

    Flanagan, J. G., Chan, D. C. & Leder, P. Transmembrane form of the kit ligand growth factor is determined by alternative splicing and is missing in the Sld mutant. Cell 64, 1025–1035 (1991).

  15. 15

    Brannan, C. I. et al. Steel-Dickie mutation encodes a c-kit ligand lacking transmembrane and cytoplasmic domains. Proc. Natl Acad. Sci. USA 88, 4671–4674 (1991).

  16. 16

    Baldwin, E. T. et al. Crystal structure of interleukin 8: symbiosis of NMR and crystallography. Proc. Natl Acad. Sci. USA 88, 502–506 (1991).

  17. 17

    Hebert, C. A., Vitangcol, R. V. & Baker, J. B. Scanning mutagenesis of interleukin-8 identifies a cluster of residues required for receptor binding. J. Biol. Chem. 266, 18989–18994 (1991).

  18. 18

    Clark-Lewis, I., Schumacher, C., Baggiolini, M. & Moser, B. Structure-activity relationships of interleukin-8 determined using chemically synthesized analogs. Critical role of NH2-terminal residues and evidence for uncoupling of neutrophil chemotaxis, exocytosis, and receptor binding activities. J. Biol. Chem. 266, 23128–23134 (1991).

  19. 19

    Hosaka, M. et al. Arg-X-Lys/Arg-Arg motif as a signal for precursor cleavage catalyzed by furin within the constitutive secretory pathway. J. Biol. Chem. 266, 12127–12130 (1991).

  20. 20

    Bosenberg, M., Pandiella, A. & Massague, J. The cytoplasmic carboxy-terminal amino acid specifies cleage of membrane TGFα into soluble growth factor. Cell 71, 1157–1165 (1992).

  21. 21

    Kochanek, P. M. & Hallenbeck, J. M. Polymorphonuclear leukocytes and monocytes/Macrophages in the pathogenesis of cerebral ischemia and stroke. Stroke 23, 1367–1379 (1992).

  22. 22

    Altschul, S. F., Gish, W., Miller, W., Myers, E. W. & Lipman, D. J. Basic local alignment search tool. J. Mol. Biol. 215, 403–410 (1990).

  23. 23

    Gish, W. & States, D. J. Identification of protein coding regions by database similarity search. Nature Genet. 3, 266–272 (1993).

  24. 24

    von Heijne, G. Anew method for predicting signal sequence cleavage sites. Nucleic Acids Res. 14, 4683–4690 (1986).

  25. 25

    Chirgwin, J. M., Przbyla, A. E., MacDonald, R. J. & Rutter, W. J. Isolation of biologically active ribonucleic acid from sources enriched in ribonuclease. Biochemistry 5294–5299 (1979).

  26. 26

    Falk, W., Goodwin, R. H. J. & Leonard, E. J. A48-well micro chemotaxis assembly for rapid and accurate measurement of leukocyte migration. J. Immunol. Methods 33, ((1980)).

  27. 27

    Larson, C. G., Anderson, A. O., Appella, E., Oppenheim, J. J. & Matsushima, K. The neutrophil-activating protein (NAP-1) is also chemotactic for T lymphocytes. Science 243, 1464–1466 (1989).

  28. 28

    Nelson, R. M., Dolich, S., Aruffo, A., Cecconi, O. & Devilacqua, M. P. Higher-affinity oligosaccharide ligands for E-selectin. J. Clin. Invest. 19, 1157–1166 (1993).

  29. 29

    Cowen, D. S., Lazarus, H. M., Shurin, S. B., Stoll, S. E. & Dubyak, G. R. Extracellular adenosine triphosphate activates calcium mobilization in human phagocytic leukocytes and neutrophil/monocyte progenitor cells. J. Clin. Invest. 83, 1651–1660 (1989).

  30. 30

    Kuchroo, V. K. et al. B7-1 and B7-2 costimulatory molecules activate differentially the Th1/Th2 developmental pathways: application to autoimmune disease therapy. Cell 80, 707–718 (1995).

  31. 31

    Shirozu, M. et al. Structure and chromosomal localization of the human stromal cell-derived factor 1 (SDF1) gene. Genomics 28, 495–500 (1995).

Download references

Acknowledgements

We thank F. Lee, G. Duyk, B. Tepper, M. P. Das, J. Lafaille, D. Levinson, S. Lin, P.Stroobant, G. Jia, D. Holtzman, S. McCarthy, D. Michnick and S. Busfield for advice and support. We are grateful for the expert technical assistance of the Millennium sequencing and bioinformatics group.

Author information

Correspondence to Yang Pan.

Rights and permissions

Reprints and Permissions

About this article

Further reading

Comments

By submitting a comment you agree to abide by our Terms and Community Guidelines. If you find something abusive or that does not comply with our terms or guidelines please flag it as inappropriate.