IMMUNOLOGY: After the Toll Rush

See allHide authors and affiliations

Science  05 Mar 2004:
Vol. 303, Issue 5663, pp. 1481-1482
DOI: 10.1126/science.1096113

This article has a correction. Please see:

You're not feeling well because you have an infection. It might be a urinary tract infection or, perhaps, influenza. How does your immune system detect and eliminate the microscopic intruders that are responsible for infectious diseases? Three papers in this issue by Zhang et al. (1) on page 1522, Heil et al. (2) on page 1526, and Diebold et al. (3) on page 1529 bring us one step closer to answering this question. They describe new information about Toll-like receptors (TLRs). These transmembrane proteins expressed by cells of the innate immune system recognize invading microbes and activate signaling pathways that launch immune and inflammatory responses to destroy the invaders. Zhang et al. (1) describe the newest TLR on the block, termed TLR11, which senses bacteria that cause infections of the bladder and kidney. Heil et al. (2) and Diebold et al. (3) demonstrate that TLR7 and TLR8 are required for recognition of the single-stranded RNAs (ssRNAs) found in many viruses. All three reports provide important new information about host defense and point the way toward new therapies to combat infectious diseases.

Many immunologists have joined the Toll rush in the past 5 years as the importance of TLRs for sensing microbial products, such as lipopolysaccharides, has become clear (see the figure) (4). Notable TLRs include TLR4, which senses bacterial lipopolysaccharides; TLR3, which detects the double-stranded RNAs (dsRNAs) of viruses; and TLR9, which senses CpG motifs common to both bacterial and viral DNA. Using TLR4 as a guide, Zhang and colleagues searched for similar gene sequences in the databases of the National Center for Biotechnology Information. They discovered TLR11 in a mouse liver expressed sequence tag database. Further analysis revealed that TLR11 is particularly abundant in the kidney and bladder. When infected with so-called uropathogenic bacteria, mice deficient in TLR11 harbored 10,000 times as many bacteria in their kidneys as normal mice. Mouse TLR11 senses bacteria that infect the kidney and provokes an inflammatory response that leads to bacterial clearance. What about humans? It turns out that humans have a truncated form of TLR11 that is probably inactive. The authors speculate that this might be why humans are susceptible to urinary tract infections.

Going down the Toll mine.

Toll-like receptors (TLRs) provide a repertoire for sensing pathogen-derived molecules during the innate immune response. TLRs in dendritic cell membranes sense proteins or lipids from bacteria or viruses. TLRs in endosomal membranes detect bacterial and viral nucleic acids. The relative contribution of each TLR to the innate immune response is not yet known because pathogens contain multiple ligands specific for several different TLRs. The signaling pathways associated with each TLR are different although they share common components. Specific signals may emanate from the adapter proteins recruited by each TLR (MyD88, Mal, Trif, and Tram). An important question concerns how the immune response is tailored to each pathogen according to activation of specific signaling pathways triggered by different pathogen products (9). HSV-1, herpes simplex virus 1; LPS, lipopolysaccharide; RSV, Roux sarcoma virus; MMTV, mouse mammary tumor virus.


The Heil et al. (2) and Diebold et al. (3) studies investigated antiviral TLRs. Both groups report the same result, that mouse TLR7 senses ssRNA from viruses. It is well established that TLR7 (and its close relative TLR8) are receptors for a family of antiviral compounds including imiquimod (5) that resemble the purine base adenine or that contain guanosine. Hence, the prediction would be that TLR7 and TLR8 would bind to viral nucleic acids that resemble the antiviral compounds. Heil et al. (2) demonstrate that a sequence from the U5 region of HIV-1 RNA, which is single-stranded and rich in guanosine and uracil (U), boosted production of the antiviral cytokine interferon-α by dendritic cells of the immune system. These authors showed that a TLR was the likely receptor involved because of the requirement for the MyD88 adapter protein used by many TLRs for signal transduction (6). Cells from mice deficient in TLR7 were unresponsive to viral ssRNA, implying that TLR7 is the receptor that detects ssRNA. Mice lacking TLR8 responded normally to viral ssRNA. In contrast, analysis of human TLR8 suggested that this TLR may sense viral ssRNA, pointing to species differences among the TLRs.

Essentially, Diebold and colleagues (3) come to the same conclusion about mouse TLR7. They found that genomic ssRNA from the influenza virus drives interferon-α production by dendritic cells. DNA containing CpG motifs, dsRNA, or indeed the antivirals that act via TLR7 also activate interferon-α production. In order to activate interferon-α production, influenza virus must be taken up by dendritic cells (endocytosis) and the resulting endocytic vesicles made acidic (as indicated by the inhibitory effect of chloroquine). Mice lacking TLR7 proved to be unresponsive to influenza virus. Because influenza RNA is U-rich, Diebold et al. then tested a synthetic polyU RNA on cells from TLR7-deficient animals and observed the same lack of response. The conclusion to be drawn from the Diebold and Heil studies is that mouse TLR7 detects viral U-rich ssRNAs.

Together, the three new reports make an important addition to the TLR literature. A lack of functional TLR11 might indeed predispose humans to urinary tract infections, although how this observation might be used therapeutically is not clear. Another question centers on what the actual ligand for TLR11 might be. The similarity between TLR11 and TLR5 suggests that a flagellin-like protein, such as the one found in the pili of uropathogenic bacteria, might bind to TLR11. The studies on TLR7 and TLR8 identify the natural viral ligands for these TLRs, and emphasize the importance of TLRs for mobilizing the innate immune system against viral infection. Whether mice deficient in TLR7 (or indeed TLR3) are more prone to viral infection than normal mice has yet to be determined.

The studies also highlight an emerging theme of the TLR field: the response of TLRs to host-derived molecules. It is well established that, in the absence of infection, damaged host tissues provoke inflammation and that this might be mediated by TLRs that sense products released by damaged cells. Once endocytosed by dendritic cells or other scavenger cells, host nucleic acids could trigger an inflammatory response via endosomal membrane TLRs, such as TLR3 (7), TLR7, or TLR9 (8). TLR9 senses host DNA bound to histones or anti-histone autoantibodies (8). Host RNA released by damaged tissues and bound to autoantibodies (or possibly host nucleic acids in apoptotic vesicles) when endocytosed by dendritic cells could be recognized by endosomal TLR7, resulting in an inflammatory response. These pathways may present new opportunities for the treatment of autoimmune diseases. Another possibility stemming from the new work is that ligands for TLR7 or TLR8 may be powerful adjuvants that could be used in antiviral or antitumor vaccines to stimulate dendritic cell activity.

The fight against pathogenic bacteria and viruses has witnessed many triumphs. The three new studies and other findings from the TLR research field will provide us with more options in the effort to save people from the damaging and sometimes fatal effects of infectious and autoimmune diseases.


  1. 1.
  2. 2.
  3. 3.
  4. 4.
  5. 5.
  6. 6.
  7. 7.
  8. 8.
  9. 9.
View Abstract

Navigate This Article