Controlling the Immune System Through Semaphorins

See allHide authors and affiliations

Science's STKE  16 Apr 2002:
Vol. 2002, Issue 128, pp. re4
DOI: 10.1126/stke.1282002re4


Semaphorins provide crucial attractive and repulsive cues involved in axon guidance. Several semaphorins have also been detected in cells of the immune system. Their influence on cell motility has been reported and is reminiscent of the biological function attributed to nervous system semaphorins. Receptors of the plexin and neuropilin family of proteins, also expressed by some immune cells, may be involved in semaphorin signaling in the immune system. However, semaphorins also affect the functioning of the immune system through receptors regulating lymphocyte activation. An important challenge in the future will be to determine whether, as in the nervous system, semaphorins help immune cells to establish connections with their appropriate targets.


The immune system operates through a highly cooperative network of specialized cell types in which cell migration and cell-to-cell contact play essential roles. For instance, T lymphocytes circulate continuously between the blood, peripheral lymphoid organs, and the tissues, where they may encounter foreign or modified self-antigens to initiate the immune response by generating differentiated effector cells. Moreover, specialized antigen-presenting cells (APCs), such as dendritic cells (DCs), usually encounter foreign antigens at the periphery and migrate to lymphoid organs to present the antigen to specific T cells. Additionally, in lymphoid organs, interactions between T cells, B cells, and APCs are required for simultaneous cell signaling through various receptors. Soluble factors, such as chemokines and cytokines that promote cell movement, cell interactions, and proliferation and differentiation, or both, are also produced in lymphoid organs. In the nervous system, a large number of phylogenetically conserved secreted and transmembrane proteins, the semaphorin family, regulate the navigation of neural growth cones during development (1). These molecules function in axonal guidance by mediating either attractive or repulsive effects through cytoskeletal alterations. The soluble factors and the membrane receptors involved in immune cell migration and arrest in the various lymphoid organs are far from being understood. Thus, the discovery that lymphoid cells express semaphorin molecules is a considerable breakthrough, because it suggests the involvement of the same highly plastic guidance controls used by the nervous system in the orchestration of the immune network.


Structure and expression

The semaphorin CD100 (also designated as Sema4D) was identified in the human immune system with specific monoclonal antibodies (mAbs) (2). The gene was cloned and the expressed protein was identified as a member of the semaphorin and immunoglobulin (Ig) families (3). The murine homolog was also cloned and shows great similarity (88% amino acid identity) with its human counterpart (4). The CD100 protein consists of an NH2-terminal signal sequence followed by a standard 500-amino-acid sema domain (commonly found in semaphorin proteins) with numerous cysteine residues, an Ig domain of the C2 type, a hydrophobic transmembrane region, and a cytoplasmic tail. Both 150-kD monomeric and 300-kD disulfide-linked homodimeric forms of membrane-associated CD100 (mCD100) are expressed; however, the homodimer is usually predominant in lymphocytes. CD100 is highly expressed on T cells and weakly on B cells. Its expression is increased on both cell types after activation. CD100 is also expressed in most hematopoietic cells, with the exception of immature CD34+ stem cells, eosinophils, and erythrocytes. CD100 messenger RNA (mRNA) is also present in nonhematopoietic tissues, including brain, heart, and kidney, but not in liver, colon, or pancreas.

Semaphorins, which are well conserved in many species, have been tentatively classified into seven classes; an eighth class includes viral semaphorins. They mainly differ with respect to membrane anchorage [secreted, transmembrane, and glycosylphosphatidylinositol (GPI)-linked]. CD100 belongs to class 4 of the membrane-associated semaphorins characterized by a Sema domain, an Ig domain, a transmembrane domain, and a cytoplasmic domain (5). However, inclusion of CD100 in the proposed classification is arguable, because CD100 also exists in a soluble form with a structure comparable to that of class 2 and 3 secreted semaphorins.

Long-range signaling through proteolysis

That mCD100 undergoes proteolytic processing at the cell membrane was suspected because the amounts of the molecule are greatly decreased after stimulation of human T cells (2, 6). Proteolytic processing may reveal a crucial role for CD100 as a long-range guidance signal, giving the released molecule the potential to act at a distance. Soluble CD100 (sCD100) is released by an as-yet-unidentified metalloproteinase and consists of two disulfide-linked 120-kD subunits (7, 8). The cleavage site is close to the cell membrane, probably located COOH-terminal to the unique cysteine residue (Cys674) in the basic region that separates the Ig and transmembrane domains. Mutational analyses revealed Cys674 to be necessary for CD100 dimerization, and thus circulating sCD100 would remain dimerized (7). This process is probably highly regulated, because CD100 shedding also increases in activated human T cells and B cells, in human T cells treated with a unique human CD100-specific mAb termed BD16, and after incubating T cells with antibodies to the protein tyrosine phosphatase CD45 (2, 6, 9).

Two receptors have been identified for CD100: plexin-B1 and CD72. Secreted semaphorins require plexin-neuropilin receptor complexes to trigger guidance effects in neurons and nonneuronal cells. Plexin-B1 appears to be a high-affinity receptor [dissociation constant (KD)~ 10–9] for CD100, on the basis of binding assays performed with recombinant sCD100 (10). Plexin-B1 is expressed in a broad range of fetal and adult tissues where, associated with neuropilin-2, it can also bind another soluble semaphorin, Sema3C, with a high affinity. However, plexin-B1 is apparently not expressed in T cells or B cells and so cannot be the receptor for CD100 in these cell types. Nevertheless, in the mouse, sCD100 does bind to resting and activated B cells, as well as to activated T cells (11), suggesting that a receptor distinct from plexin-B1 may be present on the surface of these cells.

Using an expression cloning strategy, Kumanogoh et al. found that CD72, a homodimeric type II integral membrane glycoprotein, is a low-affinity receptor (KD ~ 3 × 10–7) for recombinant murine sCD100 (11). In mice, CD72 is expressed on both T and B cells and also on DCs and monocytes. In humans, CD72 is a typical B cell marker, also weakly expressed in macrophages and DCs, but not in T cells. CD72 is a B cell inhibitory receptor with two copies of a conserved cytoplasmic amino acid sequence, the so-called immunoreceptor tyrosine-based inhibitory motif (ITIM), characterized by the prototypic sequence (Ile or Val or Leu or Ser)-X-Tyr-X-X-(Leu or Val) (where X is any amino acid) (12). Upon B cell receptor (BCR)-dependent phosphorylation of the ITIMs by a Src family protein tyrosine kinase such as Lyn, these ITIMs recruit the cytosolic Src homology 2 (SH2) domain-bearing tyrosine phosphatase SHP-1 (13). SHP-1 dephosphorylates various key signaling proteins required for B cell activation, such as the tyrosine kinase Syk, the B-linker adapter protein BLNK, or phospholipase-Cγ enzymes (14, 15). Thus, CD72 can attenuate or extinguish signaling. The sCD100 protein blocks the tyrosine phosphorylation of CD72 and its association with SHP-1, and enhances B cell responses (11). CD100-deficient mice have immune defects, mainly in their B cell responses to T cell-dependent antigens (16). Moreover, the humoral (antibody) response is increased in transgenic mice expressing sCD100 (17). Thus, sCD100 appears to function as a positive modulator of immune responses in the mouse. Such an effect on B cells mediated by CD72 has not yet been reported in humans.

In human, sCD100 inhibits the spontaneous or chemokine-induced migration of monocytes (18). Another member of the semaphorin family, Sema3A, also inhibits migration, whereas Sema3F, whose sequence is more similar to that of Sema3A than to that of CD100, is ineffective (18). This cell migration arrest may be mediated through binding of sCD100 or Sema3A to the same receptor, because there is no additive or synergistic effect of the two semaphorins in combination (18). CD72 is not detected on human monocytes, indicating that a different receptor is involved in this phenomenon, possibly plexin-B1. Similar effects of CD100 and Sema3A on the migration of DCs have not yet been documented. However, human DCs do express neuropilins(19) that may assemble with plexins into functional receptors for these soluble semaphorins.

Short-range signaling of CD100

Whether the CD100 expressed by murine or human lymphoid cells operates as a transmembrane receptor is still unclear. However, several observations with mAbs that recognize discrete epitopes of CD100 suggest that it might have signaling functions in human lymphocytes. For example, a CD100-specific mAb called BB18 induces T cell division in the presence of autologous APCs and phorbol ester; another mAb, BD16, recognizes a different epitope on CD100 and can activate T cells cooperatively with signals delivered through the T cell antigen receptor or through the CD2 activation molecule (2, 6). In addition, CD100 associates with CD45 in human T and B cells in vivo (9). One functional consequence of this association is CD45-dependent increased T cell homotypic adhesion in the presence of CD100 mAbs. Finally, the cytoplasmic domain of human CD100 appears to associate with a serine kinase that may regulate the shedding process of CD100 itself (7, 20). So far, no such results have been observed with murine mCD100, and it is not known whether CD72 can trigger signals in lymphocytes expressing CD100.

Other Lymphoid Semaphorins

In addition to CD100 and Sema3A, two other semaphorins, A39R and CD108 (also known as Sema7A), have a role in the immune system. A39R is not a lymphoid semaphorin, but a viral semaphorin produced by vaccinia virus. It belongs to the V semaphorin subclass composed of secreted viral semaphorins with a single sema domain (21). However, the A39R receptor, VESPR (viral-encoded semaphorin protein receptor, also designated plexin-C1 or CD232), a member of the plexin family of molecules, is expressed by various hematopoietic cells, including lymphocytes, monocytes, and human DCs, and has strong biological effects as a viral immune modulator (22). In particular, stimulation of A39R increases the expression of the proinflammatory cytokines interleukin-6 (IL-6), IL-8, and tumor necrosis factor-α (TNF-α) on human monocytes. A39R also triggers the expression of molecules like ICAM-1 (a key adhesion molecule in cell interactions) that binds to LFA-1 (CD54) expressed by T and B cells. A39R, like CD100, appears to inhibit the migration of human monocytes (23).

Human CD108 (identical to the murine semaphorin K1) is a 75- to 80-kD GPI-anchored membrane glycoprotein (24, 25). In the immune system, the molecule is preferentially expressed on T cells. CD108 is also expressed on erythrocytes, where it defined the high-frequency John-Milton-Hagen human blood group sometimes involved in clinically benign autoimmune disorder. No biological effects for CD108 have been reported, but its extracellular domain is very similar to that of the semaphorin encoded by alcelaphine herpesvirus-1 (24). A recombinant CD108 molecule can bind various immune cell lines (24), and VESPR, but not neuropilins, might be a receptor for the molecule (10, 24). Nevertheless, it is not known whether a soluble form of CD108 is produced and whether it could have biological effects on lymphoid cells.


The study of semaphorins in the immune system is yet in its infancy, but a tentative proposal on the role of the most studied immune semaphorin, CD100, raises many questions (Fig. 1). First, what is the mechanism of CD72 dephosphorylation by sCD100? Does decreased phosphorylation reduce the lateral mobility of CD72 and disrupt its localization in the vicinity of activated BCRs and Src family kinases? Does mCD100 trigger the same antagonistic effect after cell-to-cell contact? Another major issue will be to identify whether sCD100 has identical biological effects in human B cells, where the same negative function is assigned to CD72. Further studies on the reported inhibitory action of sCD100 on human monocyte migration would also be of interest, because it recalls the collapsing effect described for semaphorins in the nervous system (26). In human monocytes, a plexin receptor alone may be the sCD100 receptor because neither CD72 nor neuropilin-1 or -2 is expressed by these cells (18). However, this question is still open. An important challenge is to determine whether soluble semaphorins can significantly alter the migration of DCs and other DC functions, such as antigen presentation, and to document the presence of neuropilin-plexin functional semaphorin receptors in this cell type. Finally, other work may focus on Rho-related small guanosine triphosphatases (GTPases) and cytoskeletal alterations that are involved in plexin-based semaphorin signaling in the nervous system, especially because CD100 is reported to stimulate the interaction between plexin-B1 and active Rac (27). Substantiation of a role for these pathways might then indicate that immune semaphorins are bona fide guidance signals that direct the recruitment of the immune cells in fighting pathogens.

Fig. 1.

A proposed role for soluble CD100 (sCD100) in immune responses. In lymphoid organs, T lymphocytes are activated by antigen processed by APCs. The expression of membrane-bound CD100 (mCD100) dimers greatly increases in the activated cells. sCD100 is presumably produced after proteolysis by a metalloproteinase. sCD100 binding to CD72 releases B cells from the inhibitory signal mediated by CD72 and its associated cytosolic protein tyrosine phosphatase SHP-1, which results in stronger signaling and more potent responses to cell stimuli. This release from inhibition may further enhance sCD100 production by the activated B cells. In parallel, the binding of sCD100 to an unidentified receptor on monocytes (possibly a plexin) inhibits cell migration. This could improve monocyte recruitment and possibly promote differentiation into functional APCs, which could then amplify the immune response. This scheme may also apply in the periphery at sites of infection in the presence of activated antigen-specific T cells, increasing the enrolment of phagocytic cells.


  1. 1.
  2. 2.
  3. 3.
  4. 4.
  5. 5.
  6. 6.
  7. 7.
  8. 8.
  9. 9.
  10. 10.
  11. 11.
  12. 12.
  13. 13.
  14. 14.
  15. 15.
  16. 16.
  17. 17.
  18. 18.
  19. 19.
  20. 20.
  21. 21.
  22. 22.
  23. 23.
  24. 24.
  25. 25.
  26. 26.
  27. 27.
  28. 28.
View Abstract

Stay Connected to Science Signaling

Navigate This Article