Interleukin-16

Interleukin 16 (IL-16) was initially described in 1982 as the first T cell chemoattractant. Through interaction with CD4, IL-16 has now been characterized as a chemoattractant for a variety of CD4+ immune cells. Recent in vivo studies have more fully characterized IL-16 as an immunomodulatory cytokine that contributes to the regulatory process of CD4+ cell recruitment and activation at sites of inflammation in association with asthma and several autoimmune diseases. Since its cloning in 1994, IL-16 structure and function have been studied extensively. This review addresses the current data regarding IL-16 protein and gene structure; the expanding list of cells capable of generating IL-16; the direct interaction of IL-16 with its receptor, CD4; and the functional bioactivities of IL-16 as they relate to inflammation and HIV-1 infection. In addition, potential therapeutic modalities for IL-16 relating to inflammation and immune reconstitution in HIV-1 infection are also discussed.


INTRODUCTION
Interleukin-16 (IL-16) protein and bioactivity were initially identified in 1982 as a T cell chemoattractant factor that was generated from mitogen-or antigen-stimulated human peripheral blood mononuclear cells [1,2]. IL-16 was one of the first characterized cytokines with chemoattractant activity for human T cells and therefore was originally designated as lymphocyte chemoattractant factor (LCF). Since that initial observation the structure, mechanism of synthesis and processing, and other biological activities have been identified and more fully elucidated. Much of that work will be outlined in this review.

PROTEIN STRUCTURE AND BIOCHEMICAL CHARACTERISTICS
In lymphocytes and bronchial epithelial cells IL-16 is generated as a 631-amino-acid precursor molecule, pro-IL-16 [3], which is enzymatically cleaved at a serine residue (S 511 ) [3,4] after stimulation with T cell mitogens or IL-9, respectively, resulting in the secretion of a peptide consisting of the carboxy-terminal 121 amino acids. Pro-IL-16 is enzymatically cleaved by caspase 3 [4], a member of the ICE family of enzymes. Caspase 3 plays an essential role in cell apoptosis. To confer enzymatic activity, caspase 3 must itself be processed from a 40-kDa precursor. It is the process of caspase activation that appears to regulate the rate at which IL-16 bioactivity is released and detected in cell supernatants after antigenic stimulation in CD4 ϩ T cells [5], or IL-9 stimulation of epithelial cells. Although there may be a correlation between cell apoptosis and IL-16 release, as seen for IL-1␤ [6], it is clear that activation of caspase 3 associated with release of bioactive IL-16 can occur without subsequent cell apoptosis [5]. It is interesting that secreted IL-16 does not contain a consensus secretory leader sequence, and the mechanism for its release has not been elucidated. Zhou et al. [7] have reported, however, that amino-terminal deletional mutants of IL-16 demonstrate a reduced capacity for secretion, but did not identify the mechanism by which the amino-terminal 20 amino acids facilitate secretion. Although a potential intracellular role of pro-IL-16 has been postulated based upon its PDZ structure (see below [8]), thus far all of the identified bioactivity ascribed to IL-16 is related to the secreted peptide. Bioactivity is detected only after autoaggregation of the peptide into what are believed to be homotetramers [1,2], although dimers may also impart bioactivity. Resting human CD8 ϩ T cells and mast cells contain processed bioactive IL-16 protein [9][10][11], indicating that at least within these cells, aggregation of IL-16 monomers into bioactive aggregates occurs intracellularly. It is unclear at present whether autoaggregation occurs intracellularly or after secretion from other cell types such as CD4 ϩ T cells [5], eosinophils [12], or epithelial cells [13,14]. High-performance liquid chromatography (HPLC) analysis of native IL-16 from stimulated human T cells and recombinant IL-16 protein generated in COS cells or Escherichia coli has demonstrated the presence of both monomeric and multimeric forms of IL-16 [10,15,16].
The tertiary structure of recombinant E. coli-generated IL-16 monomers has been determined by nuclear magnetic resonance (NMR) spectroscopy, and indicates that the core structure comprises a PDZ (Disc-Large homology repeats) domain [17]. PDZ proteins are classically intracellular proteins that associate with other intracellular proteins via interaction of the conserved G-L-G-F sequence [18]. IL-16 contains a G-L-G-F sequence and therefore may be the first reported example of a secreted PDZ protein. Although this sequence has been associated with protein-protein interaction, it is unclear at present whether this sequence facilitates IL-16 autoaggregation. The IL-16 precursor (pro-IL-16) of lymphoid origin contains two additional PDZ domains. Multiple domains are common in PDZ proteins. Identification of intracellular binding to other proteins has not as yet been reported. The amino and carboxyl termini of secreted IL-16 lie outside of the core PDZ structure [17]. We have localized the binding and active site on IL-16, required to signal via interaction with CD4, to the carboxy terminal. The amino terminal, in addition to function in IL-16 secretion, may also contribute to stabilizing an IL-16/ CD4 interaction.
Biochemically, IL-16 is a basic protein. The isoelectric point for the secreted bioactive protein is 9.1 [15], which is distinct from the predicted pI of 4.5 [17]. There is no required glycosylation because in vitro translation of recombinant IL-16, whether generated under glycosylating or non-glycosylating conditions, maintains all bioactivity identified with native preparations. Furthermore, both native and E. coli-generated IL-16 run at the same molecular mass (17 kDa), as determined by sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) [15].
Functionally, bioactivity of IL-16 has been localized to the hydrophilic region located in the carboxyl end of the secreted molecule. Synthetic peptides generated based on the sequence of residues Arg 106 to Ser 109 for the human sequence [19], in the carboxy terminal are capable of blocking all bioactivity, and monoclonal antibodies specific for this sequence neutralize all IL-16 bioactivities [19]. Along these lines, expression of CD4 antigen is an absolute requirement for all cellular responses to IL-16. We have mapped the CD4 binding site for IL-16 to the D4 domain, and the active signaling site to the sequence Trp 345 to Ser 350 contained within the D4 domain (see below).
Recent studies by Kurschner et al. [20] have also demonstrated a neuronal form of IL-16 (NIL-16). NIL-16 message is detected in neurons localized in the cerebellum and hippocampus. The protein consists of 1322 amino acids where the carboxy-terminal 624 amino acids are identical to IL-16, however, amino acids 1-698 represent a novel sequence. The amino terminus of NIL-16 contains four PDZ domains and interacts selectively with a variety of neuronal ion channels. Similar to IL-16, NIL-16 is processed by caspase 3 enzymatic activity [20]. The potential role of IL-16 in the brain is unclear, however, neuronal cells do express CD4 and incubation of cerebellar granule neurons with IL-16 results in induction of the immediate-early gene, c-fos [20]. Therefore it is hypothesized that in the nervous system NIL-16 may have two separate roles, serving as a signaling molecule after processing by caspase 3, and as a scaffolding protein possibly involved in anchoring ion channels in the membrane.
The sequence, structure, and function of IL-16 are highly conserved in all species examined. Simian IL-16 has been cloned from Rhesus and Pig-tailed Macaque and Sooty Mangabey [21], as well as from African Green monkeys [22]. The simian IL-16 intraspecies homology for the whole IL-16 molecule is approximately 98%, whereas the homology compared with the human sequence is approximately 96%. Sequence homology for the secreted molecule comparing simian to human is Ͼ98%. Functionally, all simian-derived IL-16 has demonstrated the same bioactivities on human or simian T cells (chemotaxis, effects on a mixed lymphocyte reaction [21], and inhibition of HIV or SIV viral replication [21,22]), as observed with human IL-16. Murine IL-16 has also been isolated and sequenced [23]. The predicted amino acid sequence is Ͼ85% homologous with the human sequence in the region of the secreted protein. Recent studies in our lab have indicated that the murine pro-IL-16 as well as the secreted molecule have high sequence homology and approximately the same molecular mass as found for human IL-16. Native murine IL-16 protein has been isolated from splenocytes stimulated with mitogen, and from bronchoalveolar lavage (BAL) of ovalbumin-sensitized mice challenged with aerosolized antigen [24]. As the conservation of protein sequence would predict, murine IL-16, purified by anti-human IL-16 antibody affinity chromatography, induces cell migration of human, murine, or rat CD4 ϩ T cells with approximately the same dose response and magnitude of induced migration as seen with human IL-16 [23]. Murine IL-16 is also inhibited by neutralizing monoclonal antibodies generated to human IL-16.

IL-16 GENE STRUCTURE AND mRNA
The IL-16 gene is a single copy gene located on chromosome 15q26.1-3 for human [25] and chromosome 7 D2-D3 in the mouse [26]. The neuronal and lymphoid forms are created from a single gene by alternate splicing. The human lymphoid pro-IL-16 gene consists of seven exons and six introns [3]. Thus far sequence analysis for the human, mouse, and feline IL-16 gene have shown high (Ͼ84%) sequence homology [3,23,27]. The IL-16 promoter lacks a TATA box; however, it does contain two CAAT box-like motifs and three putative binding sites for GA-binding protein (GABP) transcription factor [28]. Regulation of the promoter activity in T cells has been shown to occur after binding of the co-activator CREB complexed with GABP␣␤ [28].
Northern blot analysis of T lymphocytes [3,23] reveals a single IL-16 mRNA species of approximately 2.6 kb. The message is constitutively expressed in greater than 95% of all lymphocytes and has a half life of 2 h [3,23]. All T cells appear to contain substantial amounts of pro-IL-16. Stimulation of resting T cells with either antigen, anti-CD3 plus anti-CD28 antibody, or by phorbol myristate acetate (PMA) results in little, if any, additional mRNA synthesis [5]. Taken all together, these results suggest that regulation of bioactive IL-16 synthesis and secretion occurs at the level of the required posttranslational caspase 3 cleavage. An open reading frame (ORF) in the IL-16 mRNA codes for a predicted precursor molecule of approximately 70-80 kDa [3,29], which is confirmed by Western blot analysis indicating the presence of an 80-kDa molecule obtained from resting T lymphocytes [29]. This band is present in a variety of cell types from both human and murine [23] cell lysates and is detected using antibodies directed to the amino-terminal of pro-IL-16 or to the secreted 17-kDa peptide. The 3' untranslated region contains three AUUUA sequences [15], however, message stabilization has not been observed in lymphocytes thus far, following either TCR/CD28 co-stimulation or by histamine. The structure of the full IL-16 gene inclusive of the neuronal coding regions is unknown, but appears to be approximately 24 kb long. There is no current information available about the neuronal promoter.

CELLS OF ORIGIN
IL-16 is synthesized by a variety of immune (T cells, eosinophils, and dendritic cells) and non-immune (fibroblasts, epithelial, and neuronal) cells ( Table 1). It was first identified as the major T cell chemoattractant factor produced by CD8 ϩ lymphocytes [9,10]. CD8 ϩ T cells released IL-16 in response to stimulation by mitogens, antigens, or vasoactive amines such as histamine or serotonin. It is unclear at present how IL-16 is secreted from the cells. The predicted amino acid sequence for processed, secreted IL-16 does not contain a secretory signal peptide [3,15], suggesting that IL-16 secretion does not involve the endoplasmic reticulum. Studies by Zhou et al. indicate that deletion of the amino-terminal 30 residues significantly reduces secretion of IL-16 in transfected Jurkat cells [7], suggesting that the amino terminal may be involved in the secretory process.
The mechanism of IL-16 processing is regulated separately from synthesis in CD8 ϩ T cells, eosinophils, and mast cells. Cell lysates generated from these cells without stimulation contain preformed IL-16 bioactivity [9,10,30]. This finding is consistent with the reports that these cells express constitutive IL-16 message as well as contain large amounts of precursor IL-16 molecule [9][10][11]. In CD8 ϩ T cells, this pool of processed IL-16 appears to result from cleavage of constitutively synthesized pro-IL-16 by (low levels of) constitutively activated caspase 3 enzyme [4]. Preformed bioactive IL-16 is not detected in CD4 ϩ T cells [5,9,10], fibroblasts, or dendritic cells [31] despite containing constitutive IL-16 message and pro-molecule. Constitutively active caspase 3 has not been detected in these cells and may explain the absence of preformed IL-16. CD4 ϩ T cells are capable of generating and secreting IL-16 after stimulation with either mitogens or specific antigen [5]. The time course for release of IL-16 is 12-24 h and the synthesis is blocked by either transcription or translation inhibitors. Co-stimulation through CD28 does not further increase the amount of IL-16 message, however, it does result in more rapid caspase 3 activation and consequently a more rapid generation and release of IL-16 protein [5].
The kinetics for release of IL-16 from CD8 ϩ T cells is dependent on the type of stimulation. When stimulated with either mitogen or antigen, 12-24 h is required to detect IL-16 bioactivity in the cell supernatants [1,2,5,32], and no IL-16 is detected after treatment with either transcription or translation inhibitors. It is conceivable that synthesis and/or activation of a required enzyme or associated protein in addition to caspase 3 activation must also occur for proper cleavage and secretion. In contrast, the release of IL-16 after stimulation of CD8 ϩ T cells with either histamine, acting through interaction with the H2 receptor [10,30,33], or serotonin, acting through the S2 receptor [9], is 1-4 h. Under these conditions, release of IL-16 is not affected by inhibitors of transcription or translation inhibitors, suggesting that in this time frame the vasoactive amines are functioning primarily as secretagogues. However, in CD8 ϩ T cells the time course for IL-16 release does not appear to be consistent with granule extrusion. Whether IL-16 is secreted from the suppressor or cytotoxic CD8 ϩ T cell subset or from both subsets is unclear at present. Because IL-16 has immunomodulatory bioactivities (see below), a functional classification would be that IL-16 is generated at least in part by suppressor T cells.
Three other immune cell types have also been shown to generate IL-16. Eosinophils obtained from either normal or hypereosinophilic donors express IL-16 message and protein when cultured in the presence of granulocyte-macrophage colony-stimulating factor (GM-CSF) [12]. Primary mast cells, as well as several mast cell lines, have been shown to produce IL-16 after stimulation with either C5a or PMA plus calcium ionophore [11]. Most recently, human dendritic cells, both primary cells and cell lines, have been shown to be capable of generating as well as responding chemotactically to IL-16 [31].
Several non-immune cell types are also capable of generating IL-16. Initially, Bellini et al. [13] identified IL-16 in cell supernatants of cultured primary airway epithelial cells obtained from asthmatic individuals, but not from non-asthmatic individuals, stimulated in culture with histamine. Consistent with this, Arima et al. [34] reported an increase in IL-16 message in an epithelial cell line stimulated with histamine, suggesting that histamine is inducing de novo IL-16 protein production. Both IL-16 protein and message have been detected in airway epithelium obtained from chronic asthmatics [35] where the majority of the IL-16 staining was observed in the epithelium, suggesting that it is a primary source of IL-16 in asthmatic inflammation. The amount of IL-16 message and protein detected in the epithelium demonstrated a positive correlation with the numbers of infiltrating CD4 ϩ T lymphocytes [35]. Generation of IL-16 by the epithelial cells appears to be somewhat selective because normal or atopic non-asthmatic individuals showed only infrequent isolated pockets of immunoreactive IL-16 and IL-16 mRNA [35].
Franz et al. [36] have demonstrated that human synovial fibroblasts obtained from patients with rheumatoid arthritis, but not from osteoarthritis, were immunoreactive for IL- 16

IL-16 ASSOCIATION AND SIGNALING THROUGH CD4 ANTIGEN
IL-16 requires cell surface expression of CD4 for induction of all its bioactivities. The most compelling evidence for this was demonstrated in studies in which murine T cell hybridoma cells were transfected with the cDNA for human CD4. CD4 expression imparted IL-16 responsiveness, as determined by induction of second messenger signaling and cell migration, in otherwise unresponsive cells [37,38]. A physical association between IL-16 and CD4 has been demonstrated in studies where IL-16 is purified from biological fluids through the use of recombinant soluble CD4 (rsCD4) affinity chromatography [15]. In addition, IL-16 physically binds to rsCD4 in solution as indicated by immunoprecipitation and Western blot analysis [15]. Using peptide inhibition, the binding on CD4 for IL-16 has been localized to the D4 domain [39]. Specifically, IL-16 signaling mediated through CD4 requires the amino acid sequence W 345 to S 350 , located in the proximal end of the D4 domain. At present it does not appear that IL-16 requires a co-receptor for CD4 to elicit cell signaling because IL-16 induces bioactivity on a variety of CD4 ϩ cells, including lymphocytes [1,32,40,41], monocytes [41], and eosinophils [42], and that cross-reactivity for IL-16 is observed across several different species [23]. However, recent findings that IL-16 stimulation cross-desensitizes CCR5 signaling [43], combined with the report that CD4 and CCR5 are constitutively associated [44], presents the possibility that IL-16, under certain circumstances, may be associating with, and in part signaling through, the co-receptor CCR5. This type of co-receptor relationship has already been reported for HIV-1 gp120 binding and signaling [45,46].
There is high sequence homology and functional crossreactivity between human, simian [21,22], murine [23], feline [27], and rat IL-16, perhaps suggesting that its conservation has imparted evolutionary advantages. Thus far, all species of IL-16 tested induce similar bioactivities on CD4 ϩ T cells. In addition, neutralizing monoclonal antibodies generated against human IL-16 are also capable of neutralizing simian [21], murine [23], or rat IL-16 bioactivities. It appears that the highly conserved carboxyl end of IL-16, and in particular the sequence Arg 107 -Arg-Lys-Ser 109 , is essential to confer bioactivity because peptides derived containing this sequence and antibodies directed to this region are capable of neutralization [19]. Although there is high sequence homology throughout IL-16 from various species, the greatest homology usually resides in the carboxyl terminal, centered around the R-R-K-S sequence [21][22][23]. Our mutational analysis has confirmed that both arginines are essential, whereas the lysine can be substituted. Similarly, although overall sequence homology for CD4 from different species is less than 55%, homology in the proximal region of the CD4 domain, corresponding to murine CD4 344 WQCLLS 349 likely where IL-16 associates with CD4, is almost absolute. Our mutational analysis suggests that leucines 347, 348 are essential for IL-16 binding and activation of CD4 ϩ T cells. The significance of the evolutionary conservation of CD4 in this region has not been determined, but may also relate to the ability of CD4 to dimerize in this region facilitating major histocompatibility complex (MHC)-dependent oligomerization of CD4s in providing accessory function to TcR signaling.
The existence of a soluble natural ligand for CD4 expands the potential role of CD4 on T cells beyond that of just co-receptor for the TCR/CD3 complex during cell-cell interaction, and would now include cell-cell interaction-independent induction of cell motility and cell cycle progression in lymphocytes. It would also help to define a role for CD4 expressed on non-lymphoid cells such as eosinophils, monocyte/macrophages, dendritic cells, and neuronal cells. These cells lack TCR/CD3 and therefore a different role for CD4 must exist on these cells. As such, IL-16 stimulation of eosinophils [12], monocytes [41], and dendritic cells [31] results in cell motility; increased eosinophil adhesion to matrix proteins [47], and up regulation of HLA-DR expression in monocytes [41]. It is interesting that both of these cell types lack the CD4-associated src tyrosine kinase p56lck found in lymphocytes. The mechanism by which CD4 in these cells transmits the migratory signal has not been determined. Similarly, a role for CD4 expressed on neuronal cells has not been clearly defined. The potential of an autocrine function for IL-16 in the brain opens the possibility of non-immune bioactivities for CD4 and IL-16.
As noted above, surface expression of CD4 is an absolute requirement for IL-16-induced bioactivities. IL-16 is biologically active only while in the multimeric form [16,48]. This suggests that specific cross-linking of CD4 molecules is required to elicit IL-16-induced cell signaling (not observed with uncross-linked ligands such as dimeric anti-CD4 antibodies). This paradigm is consistent with studies by Sakihama et al. [49] and Konig et al. [50] demonstrating that oligomerization of CD4 is required for optimization of MHC class II-dependent cell activation. Because CD4 dimerization occurs via autoaggregation of the proximal portion of the D4 domain [51] in a region that overlaps the sequence necessary to transmit an IL-16 signal it is possible that binding of multimeric IL-16 may facilitate CD4 cross-linking by interaction with the site of CD4 dimerization.

IL-16-INDUCED CELL SIGNALING
Cross-linking of CD4 by multimeric IL-16 results in the generation of several second messengers. In lymphocytes and monocytes detectable increases in intracellular Ca 2ϩ , inositol (1,4,5)-trisphosphate (IP 3 ), and phosphorylation of CD4 are observed within minutes after stimulation [41]. In lymphocytes, IL-16 stimulation also results in autophosphorylation of p56lck [37]. Similar signaling occurs in murine T hybridoma cells after transfection and surface expression of human CD4. In that system, IL-16-induced signals are dependent on the amount of expressed CD4 and are not detectable in cells transfected with mutated constructs lacking the cytoplasmic tail, essential for its interaction with p56lck [37]. Cells expressing chimeric constructs of CD4/p56lck that lack the SH1 (kinase) domain of lck respond normally to IL-16-induced chemotaxis [37]. Thus, the transmission of a migratory signal through CD4/lck does not require the catalytic activity of lck, and may reside in the SH2/SH3 domains that mediate recruitment and association to other signal-transducing molecules. The migratory signal is not observed in constructs that lack the SH3 domain, indicating a requirement for the SH2/SH3 recruitment domains for other intracellular molecules, such as phosphatidylinositol 3-kinase (PI3-kinase). The migratory response in the wild-type CD4-p56lck hybridoma cells is sensitive to the PI3-kinase-specific inhibitor, wortmannin, as well as to selective protein kinase C (PKC) inhibitors [T. Ryan, unpublished observation, and 52]. Although either wortmannin or PKC inhibitors completely block an IL-16-induced migratory signal, the integration between these two pathways is unclear at present.
In addition to the induction of migration, IL-16 stimulation results in cell cycle progression in a percentage of human CD4 ϩ T cells [53], or more uniformly in CD4 ϩ cell lines. The murine T cell hybridoma cells expressing CD4 also demonstrate increased activation and up regulation of the IL-2 receptor alpha chain after IL-16 stimulation. There is no detectable IL-2 production after IL-16 stimulation in either primary T cells or T cell lines. Unlike the migratory response, induced cell cycle progression does appear to require the catalytic domain of p56lck because SH1-deficient mutants do not demonstrate increased IL-2R␣ [A. Lee, unpublished observation].
IL-16-induced signaling has also been elucidated in a CD4 ϩ murine macrophage cell line. Krautwald [54] reported that human IL-16 stimulation of the cell line BAC-1.2F5 resulted in phosphorylation of SEK-1, subsequently inducing activation of stress-activated protein kinases (SAPKs) p46 and p54. Stimulation with IL-16 also led to phosphorylation of c-Jun and p38 MAPK (mitogen-activated protein kinase), but did not induce activation of MAPK family members ERK-1 and ERK-2. It is interesting that IL-16 stimulation differed from that of other proinflammatory cytokines, such as TNF-␣ and IL-1␤, in that, despite SEK-1 activation, there was no detectable induction of cellular apoptosis [54].

FUNCTIONAL BIOACTIVITIES OF IL-16
A variety of CD4 ϩ target cells for IL-16 stimulation have been identified (see Table 2). Although initially characterized as a chemoattractant specifically for CD4 ϩ T cells [40], it was later determined that IL-16 is also a potent chemoattractant for all peripheral immune cells expressing CD4, including CD4 ϩ monocytes [41], eosinophils [42], and dendritic cells [31]. In vitro studies have indicated that the ED 50 (half-maximal effective dose) for recombinant IL-16 is 10 Ϫ11 M. For lymphocytes, IL-16 demonstrates both chemotactic and chemokinetic activity and, unlike most of the chemokines, does not require prior activation of the T cells for induction of responsiveness [15,38,41,55].
In addition to induced cell migration, IL-16 is a competence growth factor. Stimulation with IL-16 results in cell cycle progression in CD4 ϩ T lymphocytes [41,53]. Twenty-four to forty-eight hours after stimulation with IL-16, 15-35% of normal human CD4 ϩ T cells increase surface expression of IL-2R␣ and ␤ [53]. Although IL-16 alone is insufficient to induce cell proliferation, the addition of either IL-2 or IL-15 to IL-16-primed cells results in an increase in thymidine uptake [53]. Thus, IL-16 stimulation can induce a G o to G 1 transition but it is not sufficient to induce production of IL-2. Stimulation of normal human peripheral blood mononuclear cells with IL-16 in combination with IL-2 results in a 1000-fold increase in CD4 ϩ T cells, observed over an 8-to 10-week period. The resultant cell population is homogeneously CD4 ϩ CD25 ϩ CD29 ϩ CD45RO [53]. Similar but less dramatic results are observed when IL-16 and IL-2 are used to stimulate HIV-1infected cells [N. Parada, unpublished observation]. IL-16 stimulation of peripheral CD4 ϩ T cells for up to 5 days does not appear to induce preferentially either Th1 or Th2 cytokines because only GM-CSF and IL-3 are detected [53]. However, with subsequent stimulation by either IL-2 or TCR ligation preferential generation of interferon-␥ (IFN-␥), compared with IL-4 or IL-5, is detected, suggesting that in combination with Certain CD4 ϩ T lymphoma cells increase their growth rate in response to IL-16 stimulation alone without additional IL-2 [56]. The CD4 ϩ monocytoid cell line THP1 has been shown to synthesize as well as respond to IL-16. The addition of anti-IL-16 antibodies reduces baseline cell proliferation, and the addition of exogenous IL-16 results in enhanced baseline proliferation. The CD4 ϩ CD8 ϩ TCR/CD3 Ϫ lymphocytic cell line, SUPT1, do not synthesize IL-16, however, their growth rate, as indicated by increased RNA synthesis, thymidine uptake, and cell numbers, is increased by IL-16 stimulation. Along those lines, IL-16 mRNA and protein expression has been associated with lesions of mycosis fungoides, the most common cutaneous T cell lymphoma, characterized by accumulation and activation of CD4 ϩ T cells [57].
In addition to IL-16's potential effects on the cell cycle of normal and abnormal CD4 ϩ T cells, Szabo et al. [58] have demonstrated that IL-16 stimulation of murine bone marrow cells results in the differentiation of CD4 ϩ pro-B cells into pre-B cells. This transition is facilitated by the ability of IL-16 to induce activation of both RAG-1 and RAG-2 gene expression. In vivo treatment of nude mice with recombinant human IL-16 resulted in expansion of pre-B cells detected in the bone marrow [58]. The effects of IL-16 on CD4 ϩ T cell differentiation in the thymus is not known but is currently under investigation.
It has been well established that there are pleiotrophic effects induced by CD4 after interaction with multivalent ligands such as HIV-1 virus or cross-linked anti-CD4 antibody. Specifically, stimulation of CD4 by aggregated HIV-1 gp120, or cross-linked divalent anti-CD4 antibodies induces second messenger generation [59][60][61][62][63] and cellular responses [62][63][64][65] such as chemotaxis [62,66]. However, stimulation by these ligands has also been shown to be capable of inhibiting cell activation induced through TCR/CD3 [67,68]. In fact one of the first described functions for CD4 was to modulate TCR signaling [69]. As expected, IL-16 stimulation is sufficient to inhibit TCR signaling, as indicated by its inhibitory effect on a mixed lymphocyte reaction (MLR) [70]. Similarly, IL-16 stimulation inhibits anti-CD3 or specific antigen-induced activation in a dose-dependent fashion when added before TCR activation [71]. IL-16 stimulation prevents TCR-induced IL-2R␣ expression [71] and IL-2 production [72], which may be related to observed inhibitory effect of IL-16 on TCR-or PMA-induced NF-B activity [N. Parada, unpublished observation]. In addition we have determined that there is a direct positive correlation between the inhibitory activity of IL-16 on subsequent TCR stimulation, with cells that demonstrate IL-16induced migratory activity. It is interesting that point-mutated IL-16, which does not induce signaling, does inhibit an MLR to the same extent as wild-type IL-16 [19]. This suggests that there are potentially two mechanisms by which CD4 ligation can modulate TCR signaling; by negative signaling as well as by disruption of CD4's involvement in the TCR complex. One activation-dependent molecule of note that is not induced after IL-16 pre-treatment with subsequent antigenic stimulation is Fas (CD95) [71]. The significance of this finding is unclear, however, prevention of CD95 expression may serve to limit CD95-mediated activation-induced cell death (AICD), seen at sites of inflammation [73].
Thus it appears that CD4 can function to regulate T cell responses by augmenting antigen-induced activation if bound by MHC/peptide complexes via cell-cell interaction, or by binding a soluble ligand, IL-16. The role of IL-16 in vivo remains to be elucidated, however, one hypothesis is that it is capable of contributing at least in part to a general antigenindependent non-clonal recruitment and priming of CD4 ϩ cells in an inflammatory process. Although the recruited cells would be responsive to cytokine stimulation, they would be refractory to antigen-specific activation. The effect would be to increase the number of cells recruited to an inflammatory focus and to further increase the number of viable cells by simultaneously reducing the susceptibility of those cells to antigen-specific induced cell death.
Although functions for IL-16 have largely been attributable only to the secreted form, the vast majority of detectable IL-16 protein exists as unprocessed pro-IL-16 located intracellularly. Even after maximal mitogenic stimulation only 20-25% of total intracellular IL-16 is secreted from the cell [H. Yamasaki, unpublished observation]. Therefore the possibility exists for some intracellular role for pro-IL-16. As mentioned above, Kurschner et al. [20] have identified that the amino terminus of NIL-16 binds to the class C ␣1 subunit of a mouse brain calcium channel, and they hypothesize that after caspase cleavage not only would secreted IL-16 have autocrine effects, but pro-NIL-16 could serve as a cytosolic scaffolding protein that anchors ion channels in the membrane [20]. In lymphoid cell lines pro-IL-16 may reside in both cytoplasm and nucleus; and nuclear localization appears to be augmented after caspase cleavage of IL-16 [20]. An intracellular role for pro-IL-16 in lymphocytes has yet to be fully elucidated, however, some effects have been seen on cell growth [Y. Zhang, unpublished observations].

RELATIONSHIP OF IL-16 TO HIV-1 INFECTION
One of the more intriguing functions identified for IL-16 is as a suppressor of human immunodeficiency virus (HIV-1) and simian immunodeficiency virus (SIV) infection. Although clearly not a physiological role for IL-16, Baier et al. [22] and later Mackewicz et al. [74] reported that IL-16, at a concentration of 1-5 µg/mL, could suppress approximately 40% of viral infection. This activity of IL-16 may be distinct from the CD8 ϩ cell-derived CAF activity initially described by Levy et al. [61]. Because IL-16 binds to CD4 at an epitope distinct from HIV-1 and there is no steric inhibition of viral binding to T cells, the inhibitory effect of IL-16 is also distinguished from the mechanism of HIV-1 inhibition induced by RANTES, macrophage inflammatory protein (MIP)-1␣, and MIP-1␤ [39]. The inhibitory effect of IL-16 appears to be at the level of transcriptional regulation. Maciaszek et al. [75] have reported that in transient transfection studies with HIV-1 LTR-reporter gene constructs, IL-16 pretreatment repressed either PMA-or Tat-stimulated HIV-1 promoter activity by 60-fold. This effect of IL-16 required sequences within the core enhancer, but was not simply due to down-regulation of the binding activity of transcription factors such as NF-B. Data thus far suggest that IL-16 stimulation results in activation of a transcriptional repressor that functions through sequences within or immediately adjacent to the core enhancer. Zhou et al. [7,76] confirmed this finding and further demonstrated that cells transfected to express the bioactive portion of IL-16 were resistant to HIV-1 infection. In dendritic cells, IL-16 reportedly not only inhibits viral replication, but also prevents viral entry when added to cell cultures during the infection period [77]. Studies by Idziorek et al. [48] and Lee et al. [21] suggest that IL-16 is capable of inhibiting both T tropic and M tropic isolates of HIV or SIV and that some antiviral effects are observed even if IL-16 is added post-infection [21].
In addition to the described natural anti-viral activity of IL-16, a possible therapeutic role for IL-16 in HIV-1 infection would be as an adjunct to IL-2-based therapy. The ability of IL-16 to up-regulate IL-2R␣ and impart IL-2 responsiveness to CD4 ϩ lymphocytes would predict that IL-16 could be used for immune reconstitution of CD4 ϩ T cells. Preliminary studies have indicated that peripheral blood mononuclear cells obtained from HIV-1 ϩ individuals, cultured with IL-16 and IL-2 for up to 10 weeks, results in an increase in total cell numbers, which is comprised of a homogenous CD4 ϩ T cell population. In addition, cells obtained from some of the patients demonstrated renewed antigen responsiveness after expansion by IL-16/IL-2 co-treatment [N. Parada, unpublished observation]. Because IL-2 clinical trials have yielded some encouraging results, it is feasible that IL-16 treatment would increase the IL-2R ϩ population and likely decrease the amount of IL-2 required to reconstitute CD4 ϩ cell counts, thus reducing the risk of IL-2 toxicity. An additional benefit would be that IL-2 stimulation results in maintenance of Bcl-2 protein levels, thus decreasing induced apoptosis [78]. Increasing IL-2-responsive cells might not only facilitate cell proliferation but reduce cellular apoptosis as well. Therefore, it appears that in vitro IL-16 stimulation of CD4 ϩ lymphocytes obtained from HIV-1infected individuals can result in priming the cell for IL-2 responsiveness without directly activating HIV-1 viral replication, and in fact may have some antiviral properties. Coincidently, serum IL-16 levels have been shown to remain close to normal levels in HIV long-term non-progressors compared with AIDS-positive patients [79] in which levels are low. Also, IL-16 levels rise dramatically in HIV-1-infected individuals after antiviral treatment with indinavir [80].

IL-16 IN INFLAMMATION
IL-16 is detected in organ-specific secretions in a number of inflammatory processes. However, the role it plays in these disease states has yet to be resolved. Initial studies described the presence of IL-16 at sites of inflammation and, based on in vitro bioactivities, such as the ability to induce cell migration, prime T cells for proliferation, and potentially protect T cells from AICD, classified IL-16 as a pro-inflammatory cytokine. A murine model of asthma reported that neutralization of IL-16 by administration of anti-IL-16 monoclonal antibodies reduced airway hyper-reactivity and IgE antibody levels [24]. This finding supported the concept that IL-16 can function as a pro-inflammatory cytokine in vivo. Recent data suggest that this interpretation is not so clear-cut. As mentioned above, in vitro data have indicated that IL-16 stimulation results in inhibition of TCR stimulation [71]. Several animal studies have supported this concept and further propose that IL-16 is an immunomodulatory cytokine. Klimiuk et al. have identified IL-16 in association with rheumatoid arthritis (RA) and report that administration of IL-16 in their murine model of RA significantly reduced IFN-␥, IL-1␤, and TNF-␣ [81]. In addition, IL-16-transfected squamous cells inhibit T cell activation in a model of tolerogenic skin allograft [Fujita, unpublished observation]. As a ligand for CD4, IL-16's bioactivities are defined by the role of CD4 in cell activation. CD4 appears to function as a sentinel molecule for TCR-mediated cell activation. Its involvement in the TCR/MHC complex augments cellular activation, whereas ligation by soluble factors (antibody, HIV-1 gp120, gp17, and IL-16) results in cell activation for some parameters (chemotaxis) but inhibition of parameters dependent on TCRmediated activation. It appears that CD4 can regulate T cell activation by its ability to switch from an immune phenotype when bound by MHC class II and TCR, to an inflammatory phenotype when bound separately from the TCR complex by IL-16, antibody, or gp120. This concept would be consistent with the observations that only a small percent of T cells found at sites of inflammation are specific for the inciting antigen. This concept would also allow for a particular T cell to function either as an immune or as an inflammatory cell, however, not as both simultaneously. Therefore, although IL-16 has been detected at sites of inflammation in association with a variety of diseases, its particular role in that disease may be dictated by the presence and sequence of other cell types and cytokines. A definitive understanding of the role of IL-16 for a given inflammatory situation may have to await generation of an IL-16 knockout animal.
Work directed at identifying a role for IL-16 in inflammation has focused on diseases characterized by CD4 ϩ cellular infiltrates; specifically, asthma and granulomatous diseases. Asthma was the first disease to be directly associated with IL-16 production [13]. IL-16 bioactivity was identified in cultures of primary epithelial cells, obtained from asthmatics but not from normals, stimulated with histamine. This association was confirmed by studies that identified IL-16 in the BAL fluid obtained 4 h after antigen challenge of asthmatic subjects [16]. IL-16 was not detected in the BAL fluid obtained from either normal or atopic non-asthmatic individuals. At this 4-h time point IL-16 represented the major chemoattractant activity, approximately 80% of total activity, with the balance of the activity attributable to MIP-1␣ bioactivity [16]. Furthermore, direct subsegmental histamine challenge of asthmatic subjects resulted in the elaboration of IL-16 protein detected in the BAL fluid [33]; whereas atopic non-asthmatics and normals did not secrete IL-16 into their airways. In contrast to antigen challenge, histamine challenge resulted in release of only IL-16 into the BAL fluid. Detection of IL-16 after airway challenge with either histamine or antigen from asthmatics but not from normals or atopic non-asthmatics suggested the existence of a phenotypic difference between asthmatics and non-asthmatics. The difference between asthmatics and non-asthmatics was identified by immunohistochemical staining and in situ hybridization for IL-16 protein and mRNA. Analysis of biopsies from asthmatics revealed readily detectable and uniformly distributed IL-16 protein and message in their airway epithelium and infiltrating CD4 ϩ cells [14]. There was a high correlation between the amount of detectable IL-16 protein and mRNA in the airway epithelium with the number of infiltrating CD4 ϩ mononuclear cells. In contrast, non-asthmatics had little detectable IL-16 protein and message. These studies suggest that a phenotypic change occurs associated with asthma, such that the airway epithelium is induced to synthesize IL-16. It is our current hypothesis that in asthma IL-16 is released from the epithelium after stimulation with histamine, which has been secreted from activated mast cells in response to antigen inhalation.
Despite all these descriptive studies, the in vivo role of IL-16 in human asthma has not been clearly identified. Recent studies have indicated that 48 h after histamine challenge of asthmatics, a stimulus that induces release of only a single detectable T cell chemoattractant, IL-16, a 2.5-to 3-fold increase in CD4 ϩ T lymphocytes is observed in the BAL fluid [M. Vallen-Mashikian, unpublished observation]. Using a murine model of allergic asthma, treating ovalbumin-sensitized mice with neutralizing anti-IL-16 antibodies before ovalbumin challenge significantly reduced the hyper-airway reactivity and IgE antibody production observed in animals treated with control antibodies [24]. Similar findings were obtained using IL-16-neutralizing peptides derived from the putative bioactive site of IL-16 [D. Center, unpublished observation]. These findings suggest that IL-16 may contribute to the accumulation of CD4 ϩ T cells and overall pathophysiology seen in asthmatic inflammation. IL-16's role in modulating the Th2-dependent immune response is currently under investigation by a number of laboratories.
There are also data indicating that IL-16 may play a role in the development of granulomatous inflammation. In murine models of delayed hypersensitivity granuloma formation it has been established that release of histamine and serotonin are pivotal mediators for the full development of the granuloma, characterized by CD4 ϩ T cell infiltrates [82]. Consistent with the potential role of IL-16 in CD4 ϩ T cell recruitment, immunohistochemical staining of granuloma, associated with sarcoidosis, from the lymph node and lung reveals high levels of IL-16 staining. The staining was most abundant in areas associated with perivascular accumulation of lymphocytes. IL-16 was also detected in high levels in the BAL obtained from individuals with lung-involved sarcoidosis. There is a similar IL-16 immunohistochemical staining pattern in airway tissue and bioactive protein contained in BAL of granulomas of infectious origin, such as Mycobacterium tuberculosis [J. Berman, unpublished observation].
The potential role of IL-16 has also been examined in the inflammation associated with inflammatory bowel disease. Analysis of colonic tissue sections from patients with Crohn's disease demonstrate increases in both IL-16 message and protein when compared with either uninvolved colonic tissue from the same patient or with tissue from normal individuals. In addition, using a murine model of inflammatory bowel disease, animals treated with neutralizing anti-IL-16 antibodies demonstrated significantly less weight loss, mucosal ulcerations, and myeloperoxidase activity as compared with animals receiving control antibodies [A. Keates, unpublished observation]. Taken together, these studies indicate that IL-16 is present at sites of inflammation and neutralization of IL-16 bioactivity may significantly alter the inflammatory process.