Toll-Like Receptor 3
The innate immune system of mammals is equipped with various kinds of cells, such as macrophages and dendritic cells (DCs), which provides the first line of defense to the host in recognizing various kinds of pathogens. These cells have developed different classes of protein based receptors for recognizing numerous kinds of pathogen associated molecular patterns (PAMPs) (Miggin and O’Neill 2006). These different classes of pathogen recognition receptors (PRRs) includes, Membrane bound PRRs such as Toll-like receptors (TLRs), Receptor kinases, Mannose receptors and Cytoplasmic PRRs such as Nucleotide oligomerization domain (NOD) receptors, the Retinoic acid inducible gene I (RIG-I)-like receptor (RLR) family, and the recently described AIM2 and DAI cytosolic DNA receptors. All these receptor proteins play a crucial role in “danger” recognition and induction of the innate immune responses against a variety of bacterial and viral infections at the cytoplasmic, endosomal, intercellular, and extracellular level (Kumar et al. 2009; Siednienko et al. 2011).
To date, ten human TLRs have been identified and equivalent forms of many of these have been found in other mammalian species (Kumar et al. 2009). Each member of the TLR family recognizes different kinds of PAMPs. For example, TLR1 in combination with TLR2 or TLR6 recognizes triacyl lipopeptides and diacyl lipopeptides, respectively. TLR4 recognizes lipopolysaccharide (LPS) from gram-negative bacteria and TLR5 recognizes flagellin from bacterial flagellum. In contrast to the above-mentioned plasma-membrane-localized TLRs, the so-called antiviral TLRs, TLR3, TLR7/8, and TLR9 are located on endocytic vesicles, and recognize double-stranded (ds)RNA and single-stranded (ss)RNA, respectively, whereas, murine TLR11 recognizes uropathogenic Escherichia coli (Kumar et al. 2009).
After recognition of PAMPs, activation of TLRs is initiated by the formation of multi-protein complexes containing the TLR and the recruitment of proximal cytoplasmic Toll/IL-1 receptor (TIR) domain-containing adaptor proteins. To date, four activating TLR adaptor proteins have been identified, namely, Myeloid differentiation factor 88 ( MyD88), MyD88 adaptor-like (Mal) (also known as TIR domain containing adaptor protein (TIRAP)), TIR domain-containing adaptor inducing interferon (IFN)-beta (TRIF) (also known as TICAM-1) and TRIF-related adaptor molecule (TRAM). In addition, an inhibitory TLR adaptor protein called sterile alpha and TIR motif-containing protein (SARM) has also been identified (Vercammen et al. 2008). Upon recruitment of the activationary adaptor proteins to the TLRs via TIR domain interactions, a series of TLR signaling cascades are elicited which culminates in the production of proinflammatory cytokines such as tumor necrosis factor α (TNF-α), interleukin (IL)-6, IL-1β, and IL-12. In addition, activation of TLR3, -7, -8, and -9 induces the production of antiviral type I interferons (IFN-β and IFN-α). Moreover, TLRs induce DC maturation, which is essential for the activation of pathogen-specific adaptive immune responses (Miggin and O’Neill 2006; Kumar et al. 2009).
In spite of various similarities between all TLRs in terms of signal transduction pathways and the production of proinflammatory cytokines, there is rigid specificity with respect to their adaptor usage (Miggin and O’Neill 2006; Kumar et al. 2009). MyD88 is the common downstream adaptor that is recruited by all TLRs except TLR3 and leads to activation of nuclear factor (NF)-κB. Mal is required for signaling by TLR4 and TLR2 (Miggin and O’Neill 2006; Kenny et al. 2009). TRIF mediates TLR3 and TLR4 signaling and induces the activation of transcription factors IRF3 and IRF7 which leads to the expression of Type I IFNs. Finally TRAM regulates TLR4 signaling exclusively acting as a bridging adaptor to recruit TRIF to the TLR4 complex (Kumar et al. 2009).
In recent years, researchers have gained a more in depth understanding of the signaling pathways that are modulated upon TLR engagement. This essay focuses on Toll-like receptor 3, a key molecule that recognizes viral double-stranded (ds) RNA originating from dsRNA viruses such as reovirus. TLR3 also recognizes dsRNA produced during the replication of single-stranded (ss) RNA viruses, such as West Nile virus (WNV), respiratory syncytial virus (RSV), and encephalomyocarditis virus (EMCV). In addition, TLR3 also recognizes a synthetic analogue of dsRNA known as polyriboinosinic:polyribocytidylic acid (poly(I:C)) (Kumar et al. 2009). Herein, Toll-like receptor 3 will be described in terms of its structure, expression, subcellular localization, and its ligands. In addition, the signaling pathways that are activated by TLR3 and its curtailment by various molecules thus limiting proinflammatory cytokines and Type I IFN production will also be described. Also, major perspective applications of Toll-like receptor 3 in the context of antiviral and cancer immunotherapies will also be described.
Properties of TLR3
Discovery of the first human Toll-like receptor by Nomura and colleagues in 1994 (Nomura et al. 1994) has paved the way for a breakthrough in Immunobiology. Since then, a family of ten human TLRs with detailed crystal structures has been identified and they have been proved to be pivotal elements in the human innate immune system.
Structure of TLR3
Human TLR3 consists of an extracellular domain containing 23 leucine-rich repeats (LRRs), N- and C-terminal flanking regions, a transmembrane domain, and an intracellular TIR domain (Bell et al. 2006). Analysis of the TLR3-ectodomain (ECD) revealed that the LRRs form a large horseshoe shaped structure which has an inner diameter of approximately 42 Å, an outer diameter of 90 Å, and a thickness of 35 Å. The TLR3-ECD is highly glycosylated with 15 predicted N-glycosylation sites, while the lateral surface is completely carbohydrate free and offers a large surface area for interaction with other molecules. It has been demonstrated that amino acid residues H539 and N541 located on the glycosylation-free lateral surface of the LRR domain of TLR3 are critical for ligand binding (Bell et al. 2006). TLR3-ECD bind as dimeric units to dsRNA oligonucleotides of at least 45 bp in length, the minimal length required for the transduction of signal through the receptor (Botos et al. 2009). A LRR-deletion study on TLR3 has shown that the C-terminal LRRs control receptor signaling and dimerization; the intermolecular contacts between the C-terminal domains of two TLR3-ECD stabilize the dimer and position the C-terminal residues closely, which is considered to be essential for signal transduction across plasma membrane in intact TLR3 molecules (Botos et al. 2009).
Expression and Subcellular Localization of TLR3
The expression and subcellular localization of TLR3 is regulated in a cell-type specific manner and is modulated by the activation status of the cell. TLR3 expression has been detected in immune cells including macrophages, DCs, γδ T cells, NK cells, T cells, and mast cells (Siednienko and Miggin 2009; Gauzzi et al. 2010). TLR3 expression has also been detected in nonimmune cells, including fibroblasts, epithelial cells, keratinocytes, glial cells, and neuronal cells (Gauzzi et al. 2010; Siednienko et al. 2010).
Unlike the other viral RNA sensor proteins whose localization is exclusively cytoplasmic, TLR3 has been found localized intracellularly within endosomal vesicles and also localized at the cell surfaces (Siednienko et al. 2010). For example, in DCs, TLR3 is predominantly detected in the intracellular endocytic vesicles (Siednienko et al. 2010). In contrast, while fibroblasts and epithelial cells express intracellular TLR3, its expression predominates on the cell surface. Likewise, peritoneal and bone marrow-derived murine mast cells express TLR3 both in the cytoplasm and on the cell surface. In keratinocytes, high levels of intra-cytoplasmic TLR3 expression and low levels of surface TLR3 are detected (Gauzzi et al. 2010). Interestingly, Arg740 and Val741 are involved in the intracellular localization and expression of TLR3 in humans, while in murine TLR3, no crucial structural motif residues have been identified to date (Funami et al. 2004). The expression of TLR3 in both DCs and macrophages is upregulated by viral infection and exogenous addition of poly(I:C) or type I IFN (Gauzzi et al. 2010; Siednienko et al. 2010).
Ligands and its Delivery to TLR3
The primary ligand for TLR3 is dsRNA. As previously stated, TLR3 senses dsRNA viruses such as reovirus and dsRNA produced during the replication of ssRNA viruses, such as WNV, RSV, and EMCV. In addition to dsRNA from viral origin, endogenous dsRNA released from dying cells also activates TLR3 (Gauzzi et al. 2010). Interestingly, poly(I:C) in a cell-associated form is reported to be more efficient in triggering TLR3 than soluble dsRNA, suggesting that dsRNA derived from dying cells is a more potent and physiol ogically relevant TLR3 ligand than dsRNA from live cells (Gauzzi et al. 2010). In the laboratory, the stable synthetic dsRNA analogue, poly(I:C) is extensively used as an artificial activator of TLR3. It must be noted that beyond TLR3, dsRNA is also recognized by several other cytosolic sensors, such as protein kinase R, 20–50-oligoadenylate synthetases, and the more recently identified RLR RNA helicases, RIG-I and melanoma differentiation-associated gene5 (Mda5) (Gauzzi et al. 2010). In certain cell types, TLR3 is predominantly expressed intracellularly, thus the TLR3 ligand is transfected into the cells with cationic lipososmes such as lipofectin and DOTAP and these ligand-liposomes complexes are believed to be delivered to the endosome to activate TLR3 (Gauzzi et al. 2010). It has been found that TLR3 provides several sites to accommodate ligand binding depending on the glycosylation status of the extracellular domain and the structural characteristics of the ligand (example: modification and ligand length). Notably, TLR3 signaling has been shown to be more potently induced by longer duplex length of poly(I:C) rather than shorter ones (de Bouteiller et al. 2005).
Regarding the delivery mechanism of dsRNA to the TLR3 containing endocytic vesicle, it has been demonstrated that CD14 directly binds to extracellular poly(I:C) and enhances dsRNA-mediated TLR3 activation (Gauzzi et al. 2010) by direct internalization and co-localization of CD14 with TLR3 which then facilitates the transfer of the dsRNA to the intracellular TLR3 (Gauzzi et al. 2010). Supporting this concept is the fact that intracellular co-localization of TLR3 and CD14 in multiple compartments, including the endoplasmic reticulum and the Golgi apparatus has been demonstrated (Gauzzi et al. 2010). Interestingly, human DCs, which do not express CD14 and do not express TLR3 on their surface, produce type I IFN and IL-12p70 in response to exogenously added dsRNA and antibodies against TLR3 do not inhibit dsRNA-induced IFN-β production by DCs (Gauzzi et al. 2010). This suggests that the extracellular dsRNA must be internalized via a putative uptake receptor (e.g., clathrin-mediated endocytosis) or by the uptake of the apoptotic bodies derived from virally infected cells containing viral dsRNA in order to activate intracellularly localized TLR3. Thus, the dsRNA may originate from many sources including dying cells, apoptotic bodies, and viruses and the mechanisms that facilitate association of the dsRNA with TLR3 largely depends on the cell type and concomitant subcellular distribution of TLR3.
TRIF-Mediated TLR3 Signaling
Regarding TRIF, the proline-rich N-terminal domain of TRIF mediates IRF3 activation by facilitating the phosphorylation of IRF3 at critical residues via a number of auxiliary molecules. NAK-associated protein 1 (NAP-1) forms part of the active kinase complex for IRF3 and facilitates the association of TRIF with TANK-binding kinase 1 (TBK1; also called NAK or T2K) and IκB kinase related kinase ɛ (IKKɛ; also called IKKι) (Gauzzi et al. 2010). TBK1 and IKKɛ also interact with TANK (TRAF family member-associated NF-κB activator) which also facilitates the activation of IRF3 (Kumar et al. 2009). Furthermore, TNF receptor-associated factor 3 (TRAF3) plays a crucial role in TLR3 signaling as various independent studies show that TRAF3 forms a complex with NAP-1 and TRIF and TRAF3-deficient mice exhibit impaired IFN-β in response to poly(I:C) (Fig. 1) (Oganesyan et al. 2006). It must be noted that TRAF3 and NAP-1 are also critical adaptor proteins in RIG-I/Mda5-mediated IRF3 activation (Oganesyan et al. 2006). TRAF6 also interacts with TRIF, though its role in IFN-β production is largely non-critical (Yamamoto and Takeda 2010). The RNA helicase, DEAD box protein 3 (DDX3) has also been shown to play a role in TLR3 signaling (Schröder et al. 2008). Upon TKB1/IKKɛ-mediated phosphorylation and homo/heterodimerization of IRF3/7, the complex subsequently translocates into the nucleus and binds to the IFN-β enhanceosome (Schröder et al. 2010) and IFN-stimulated response elements (ISREs) to induce the transcription of responsive genes including the Type I IFN-β and CCL5 genes (Siednienko et al. 2011) (Fig. 1). Thus, TRAF3, NAP-1, TBK1/IKKɛ, and the N-terminal region of TRIF play an important role in TLR3-mediated IRF3/7 activation (Oganesyan et al. 2006).
While the N-terminal region of TRIF mediates IRF3/7 activation, the leucine-rich C-terminal region of TRIF is involved in NF-κB activation via receptor-interacting protein 1 (RIP1) (Fig. 1). RIP1 interacts with TRIF through the RIP homolytic interaction motif (RHIM) domain of TRIF, followed by polyubiquitination (U) of RIP1, which recruits the ubiquitin receptor protein transforming growth factor β-activating kinase (TAK) binding proteins 2 and 3 (TAB2 & TAB3) and TAK1 to form a TAK1 complex (Fig. 1) (Gauzzi et al. 2010; Yamamoto and Takeda 2010). TAK1 then phosphorylates IKKα and IKKβ, which consequently phosphorylates the NF-κB inhibitor IκB, concomitantly leading to its degradation and nuclear translocation of NF-κB and induced expression of IFN-β (Fig. 1) (Gauzzi et al. 2010). TAK1 also activates two other classes of kinase, JNK and p38 and these activate the transcription factor, AP-1 (Jin et al. 2010). In addition, another protein, TRAF6 also mediates NF-κB activation, though this involves binding of TRAF6 to the N-terminal region of TRIF (Fig. 1). Studies have suggested that the participation of TRAF6 in TRIF-mediated NF-κB induction is cell-type specific as TRAF6 is essential for NF-κB activation in mouse embryonic fibroblasts, whereas poly(I:C) induced NF-κB activation is not impaired in TRAF6 deficient macrophages (Gauzzi et al. 2010; Sasai et al. 2010). The TAK1 complex also activates the mitogen-activated protein kinases (MAPKs) including c-jun N-terminal kinase, p38, and extracellular signal related kinase (ERK) to control mRNA expression and the stability of mRNA for inflammatory genes by mediating phosphorylation of the AP-1 transcription family proteins (Yamamoto and Takeda 2010).
Of all the TLR adaptor proteins, TRIF is the only adaptor that is able to engage mammalian cell death signaling pathways whereby TRIF interacts with Fas-associated cell death domain (FADD) through RIP1 and activates procaspase-8 to initiate cell apoptosis (Vercammen et al. 2008; Jin et al. 2010); this is an important host defense mechanism for limiting the spread of viral infection. Moreover, TRIF also plays a role in TLR3 induced proliferative responses in B cells (Vercammen et al. 2008).
PI3K and c-src in TLR3 Signaling
While the TIR domain of TLR3 serves to recruit the adaptor molecule, TRIF, additional molecules are also capable of interacting with TLR3 and modulating its signaling. For example, studies have shown that phosphatidylinositol 3-kinase (PI3K) is capable of interacting with phosphorylated tyrosine residues Tyr759 and Tyr858 located within the TIR domain of TLR3. This interaction serves to activate the TLR3 pathway toward IRF3 and concomitant induction of IFN-stimulated genes while impairing NF-κB-dependent proinflammatory signaling (Fig. 1) (Vercammen et al. 2008). Another tyrosine kinase molecule known as c-Src has also been shown to interact with endosomally localized TLR3 and has been shown to play a crucial role in dsRNA-mediated TLR3 signaling (Vercammen et al. 2008).
In summary, TLR3 is distributed on the cell surface and the endosome in a cell-type-dependent manner. The adaptor protein TRIF is localized in the cytoplasm. Upon stimulation of TLR3 with synthetic dsRNA or upon recognition of an invading virus, TRIF is recruited to the endosomal/cell surface TLR3 via TIR domain interactions and thereafter moves away from the receptor to speckle-like structures that also include downstream signaling molecules. The interaction of TRIF with RIP1 results in polyubiquitination (U) of RIP1 and its binding to the ubiquitin receptors TAB2 and TAB3 which activates TAK1. Activated TAK1 induces phosphorylation of IKKα and IKKβ leading to phosphorylation and degradation of IκB and release and translocation of NF-κB to the nucleus. TAK1 also activates two other classes of kinase, JNK and p38 and these activate the transcription factor, AP-1. Also, TRIF activates TBK1 and IKKɛ through NAP1 and this results in phosphorylation and nuclear translocation of IRF3 and IRF7 resulting in IFN-β production. TRAF3 also binds with the TBK1/IKKɛ complex inducing IRF3 activation. PI3K is also recruited to the TIR domain of TLR3 and activates kinase Akt for full phosphorylation and activation of IRF3 in the nucleus. Tyrosine kinase c-Src also plays a role in TLR3 signaling. TRIF interacts with FADD through RIP1 and activates procaspase-8 to initiate cell apoptosis. Many of the mechanisms by which TLR3/TRIF is activated and transmits signals remain largely unknown. Phosphorylation and activation of the various transcription factors such as IRF3, IRF7, NF-κB, and AP-1 molecules leads to IFN-β and proinflammatory cytokine production and consequent DC maturation and complete activation of the innate and adaptive immune system.
Negative Regulation of TLR3 Signaling
Excessive or sustained activation of TLR3 signaling, leading to elevated levels of proinflammatory cytokines and IFN-β is potentially harmful or even fatal for the host cell. Clinically, this may present as systemic inflammatory response syndromes such as inflammation-associated myopathies, WNV-driven CNS inflammation and viral or autoimmune liver diseases. Given this, the body has evolved several mechanisms for modulating and curtailing TLR3-mediated cellular responses. For example, protein inhibitor of activated signal transducers and activators of transcription (PIASy), TRAF1, and TRAF4 have been shown to inhibit TLR3 ligand induced activation of IRF3 and NF-κB (Vercammen et al. 2008). Interestingly, PI3K has been shown to inhibit TLR3-mediated NF-κB activation while activating IRF3. Similarly, another kinase known as RIP3 has been shown to inhibit TLR3 induced NF-κB activation by competing with RIP1 for TRIF binding (Vercammen et al. 2008). In contrast, suppressor of IKKɛ (SIKE) has been shown to inhibit TLR3-mediated IRF3 activation, without affecting NF-κB activation (Vercammen et al. 2008).
The TLR adaptors Mal and MyD88 were not thought to be involved in TLR3 signaling until a recent study demonstrated that MyD88 negatively regulates TLR3/TRIF-induced corneal inflammation through a mechanism involving JNK phosphorylation, but not p38, IRF3, or NF-κB (Johnson et al. 2008; Siednienko et al. 2011). MyD88 has also been shown to inhibit TLR3-dependent phosphorylation of IRF3 and thus curtail TLR3-mediated IFN-β and RANTES production (Siednienko et al. 2011). MyD88 also inhibits TLR3-dependent IL-6 induction, but not IκB degradation nor p38 activation in murine macrophages (Kenny et al. 2009). The TLR adaptor Mal has also been shown to inhibit TLR3-dependent IFN-β production through a mechanism distinct from MyD88 whereby Mal inhibits TLR3 ligand-mediated IRF7 activation (Siednienko et al. 2010). Mal has also been shown to inhibit TLR3-dependent IL-6 induction (Kenny et al. 2009). In addition, the fifth TLR adaptor protein, SARM which contains a sterile α-(SAM) and a HEAT/armadillo (ARM) motif has been shown to inhibit TLR3 signaling. While SARM does not induce IRF3 or NF-κB activation itself, it inhibits TRIF-mediated TLR3 signaling (Carty et al. 2006) with both the TIR and SAM domains proving vital for SARM functionality.
Thus, it is clear that TLR3 inhibitory molecules have the ability to curtail global TLR3 signaling and/or to inhibit a specific pathway on the TLR3 signaling cascade in a cell-type specific manner. These molecules may serve to alter the balance between NF-κB and IRF3/7 following TLR3 activation, thus providing an immune response that it tailored to a specific cell-type/ligand/biological milieu. Moreover, the vast array of molecules and mechanisms that have evolved to curtail TLR3 signaling highlights the importance of TLR3 immunomodulation.
Therapeutic Manipulation of TLR3
Once microbes breach the physical barriers of the body such as skin or the intestinal tract mucosa, they are recognized by TLRs which activates the innate immune cell response.
It is well established that dsRNA-induced activation of TLR3 instigates a series of signaling cascade events which leads to the induction of IFN-ß gene transcription and the production of antiviral IFN-β, IL-12, and various inflammatory cytokines (Vercammen et al. 2008; Gauzzi et al. 2010). Thus, TLR3 plays a key role in antiviral immune responses.
TLR3 is involved in the sensing of dsRNA derived from reovirus (Vercammen et al. 2008). TLR3 also recognizes dsRNA produced during the replication of many viruses including single-stranded (ss)RNA viruses, such as WNV, RSV and EMCV (Vercammen et al. 2008). It must be noted that TLR3 is dispensable for the detection of certain viruses (Vercammen et al. 2008). Regarding the therapeutic value of TLR3 manipulation, it has been shown that activation of TLR3 using poly(I:C) inhibits human immunodeficiency virus (HIV) infection by increasing the expression of Type I IFN antiviral factors thus restricting HIV expression and replication (Zhou et al. 2010). Also, poly(I:C) has proven beneficial as a mucosal adjuvant for an influenza virus vaccine in a murine infection model (Vercammen et al. 2008). Interestingly, another TLR3 agonist, Poly(I:C12U), has been shown to have antiviral activity against hepatitis B virus, several flaviviruses, coxsackie B3 virus, and Punta Toro virus. Given that excessive TLR3 signaling that is associated with many viral based infections, the use of TLR3 antagonists to curtail excessive TLR3 signaling would also be of therapeutic value.
TLR agonists have the capacity to prime and shape adaptive immunity and so have aroused significant interest in the development of cancer immunotherapies (Cheng and Xu 2010). Regarding TLR3, its activation by dsRNA induces dual antitumoral activity, either an anticancer immune response or apoptosis of the cancer cells via TLR3 which is expressed by the cancer cell (Jin et al. 2010); these two modes of action work synergistically against cancer. Given that cancer cells are malignantly transformed cells of the host, they express antigens that are either not expressed or only expressed in trace amounts by the healthy host and so are referred to as tumor associated antigens (TAAs). Thus, induction of TLR3-mediated apoptosis of the cancer cells presents the immune system with a new repertoire of TAAs in a TLR3 activation context that is favorable to the development of long-term anticancer immune responses (Jin et al. 2010). In addition, activation of TLR3 using agonists leads to production of type I IFN and other inflammatory mediators, which aid in the phenotypic and functional maturation of DCs. For example, immunization with the melanoma peptide trp2 and adjuvants consisting of cationic liposomes complexed with TLR3/9 agonists has been shown to control the growth of established B16 melanoma tumors in a therapeutic tumor vaccine model (Vercammen et al. 2008). Despite increased expression of TLR3 on many tumors and the immunostimulatory and protective properties of dsRNA, many questions remain to be answered regarding TLR3 in the context of its antitumor activity.
Much information has come to light in the last decade providing a convincing picture of the important, yet complex, interplay between TLR3, the TLR adaptor proteins, the signaling pathways and the role of TLR3 in antiviral and antitumor therapies. It must be noted that although it is clear that several viruses stimulate TLR3-dependent signaling, the importance of TLR3 signaling in the antiviral immune response is not definitive. For example, an absence of TLR3 does not affect the outcome of a viral infection in some instances; this may be attributed to redundancy with other dsRNA sensors, such as the RLRs, RIG-I, and Mda5. Another complication in understanding TLR3 functionality is the absence of a physiological ligand for TLR3. Most studies use poly(I:C), a synthetic agonist, but the identity of the viral RNA sequences that actually activate TLR3 are poorly understood. Moreover, the potential of endogenous RNA (e.g., from dying cells) to mediate TLR3 signaling requires further study. Regarding the cellular localization of TLR3, it is important to consider that the subcellular localization of TLR3 may affect subcellular signaling events. Lastly, to fulfill the therapeutic potential of TLR3, studies must be performed to further elucidate the signaling pathways and proteins that can positively and/or negatively regulate TLR3 signaling. These molecules could potentially be used to modulate inappropriate TLR3 signaling in a number of clinical contexts and to develop enhanced vaccine adjuvant therapy against tumor and viral infectious diseases.
- Nomura N, Nagase T, Miyajima N, Sazuka T, Tanaka A, Sato S, et al. Prediction of the coding sequences of unidentified human genes. II. The coding sequences of 40 new genes (KIAA0041-KIAA0080) deduced by analysis of cDNA clones from human cell line KG-1 (supplement). DNA Res. 1994;1:251–62.PubMedCrossRefGoogle Scholar