Blocking virus infection with soluble cell receptors

poliovirus + receptorWe recently discussed the development of a soluble receptor for HIV-1 that provides broad and effective protection against infection of cells and of nonhuman primates. Twenty-five years ago my laboratory published a paper which concluded that using soluble receptors to block virus infection might not be a good idea. In the first paragraph of that paper we wrote:

…it has been proposed that soluble cell receptors might be effective antiviral therapeutics. It has been suggested that mutants resistant to the antiviral effects of soluble receptors would not arise, because mutations that abrogate binding to receptors would be lethal.

We had previously shown that the cell receptor for poliovirus, CD155, produced in a soluble form, would bind to poliovirus (pictured – the very image from the banner of this blog), blocking viral infection. We then found that it was relatively easy to select for soluble receptor resistant (srr) virus mutants. These viruses still enter cells by binding to CD155, but the affinity of virus for the receptor is reduced. Poliovirus srr mutants replicate normally in cell cultures, and cause paralysis in a mouse model for poliomyelitis. We speculated that receptor binding might not be a rate-limiting step in viral infection, and short of  abolishing binding, the virus can tolerate a wide range of binding capabilities.

The amino acid changes that cause the srr phenotype map to both the exterior and the interior of the viral capsid. The changes on the virion surface are likely to directly interact with the cell receptor. Changes in the interior of the virus particle may be involved in receptor-mediated conformational transitions that are believed to be essential steps in viral entry.

When this work was done, clinical trials of soluble CD4 for HIV-1 infection were under way. We believed that our findings did not support the use of soluble receptors as antivirals, which we clearly stated in the last sentence of the paper:

These findings temper the use of soluble receptors as antiviral compounds.

HIV-1 mutants resistant to neutralization with soluble CD4 were subsequently isolated, and the compound was never approved to treat HIV-1 infection in humans for this and other reasons, including low affinity for the viral glycoprotein, enhancement of infection, and problems associated with using a protein as a therapeutic.

Recently a new soluble CD4 was produced which also includes the viral binding site for a second cell receptor, CCR5. This molecule overcomes many of the issues inherent in the original soluble CD4. It provides broad protection against a wide range of HIV-1 strains, and when delivered via an adenovirus-associated virus vector, protects nonhuman primates from infection. This delivery method circumvents the issues inherent in using a protein as an antiviral drug. Because this protein blocks both receptor binding sites on the viral envelope glycoprotein, it might be more difficult for viruses to emerge that are resistant to neutralization. The authors speculate that such mutants might not be efficiently transmitted among hosts due to defects in cell entry. Given the promising results with this antiviral compound, experiments to test this speculation are certainly welcome.

TWiV 326: Giving HIV a flat tyr

On episode #326 of the science show This Week in Virology, the sternutating TWiVers discuss preventing infection of cells and animals by a soluble CD4-CCR5 molecule that binds to HIV-1 virus particles.

You can find TWiV #326 at

Blocking HIV infection with two soluble receptors

eCD4-FcBecause viruses must bind to cell surface molecules to initiate replication, the use of soluble receptors to block virus infection has long been an attractive therapeutic option. Soluble receptors have been developed that block infection with rhinoviruses and HIV-1, but these have not been licensed due to their suboptimal potency. A newly designed soluble receptor for HIV-1 overcomes this problem and provides broad and effective protection against infection of cells and of nonhuman primates.

Infection with HIV-1 requires two cell surface molecules, CD4 and a chemokine receptor (either CCR5 or CXCR4), which are engaged by the viral glycoprotein gp120 (illustrated). A soluble form of CD4 fused to an antibody molecule can block infection of most viral isolates, and has been shown to be safe in humans, but its affinity for gp120 is low. Similarly, peptide mimics of the CCR5 co-receptor have been shown to block infection, but their affinity for gp120 is also low.

Combining the two gp120-binding molecules solved the problem of low affinity, and in addition provided protection against a wide range of virus isolates. The entry inhibitor, called eCD4-Ig, is a fusion of the first two domains of CD4 to the Fc domain of an antibody molecule, with the CCR5-mimicking peptide at the carboxy-terminus (illustrated). It binds strongly to gp120, and blocks infection with many different isolates of HIV-1, HIV-2, SIV, and HIV-1 resistant to broadly neutralizing monoclonal antibodies. The molecule blocks viral infection at concentrations that might be achieved in humans (1.5 – 5.2 micrograms per milliliter).

When administered to mice, eCD4-Ig protected the animals from HIV-1. Rhesus macaques inoculated with an adenovirus-associated virus (AAV) recombinant containing the gene for eCD4-Ig were protected from infection with large amounts of virus for up to 34 weeks after immunization. Levels of eCD4-Ig in the sera of these animals ranged from 17 – 77 micrograms per milliliter.

These results show that eCD4-Ig blocks HIV infection with a wide range of isolates more effectively than previously studied broadly neutralizing antibodies. Emergence of HIV variants resistant to neutralization with eCD4-Ig would likely produce viruses that infect cells less efficiently, reducing their transmission. eCD4-Ig is therefore an attractive candidate for therapy of HIV-1 infections. Whether sustained production of the protein in humans will cause disease remains to be determined. Because expression of the AAV genome persists for long periods, it might be advantageous to include a kill-switch in the vector: a way of turning it off if something should go wrong.

How influenza virus infection might lead to gastrointestinal symptoms

influenza virusHuman influenza viruses replicate almost exclusively in the respiratory tract, yet infected individuals may also have gastrointestinal symptoms such as vomiting and diarrhea. In mice, intestinal injury occurs in the absence of viral replication, and is a consequence of viral depletion of the gut microbiota.

Intranasal inoculation of mice with the PR8 strain of influenza virus leads to injury of both the lung and the intestinal tract, the latter accompanied by mild diarrhea. While influenza virus clearly replicates in the lung of infected mice, no replication was observed in the intestinal tract. Therefore injury of the gut takes place in the absence of viral replication.

Replication of influenza virus in the lung of mice was associated with alteration in the populations of bacteria in the intestine. The numbers of segmented filamentous bacteria (SFB) and Lactobacillus/Lactococcus decreased, while numbers of Enterobacteriaceae increased, including E. coli. Depletion of gut bacteria by antibiotic treatment had no effect on virus-induced lung injury, but protected the intestine from damage. Transferring Enterobacteriaceae from virus-infected mice to uninfected animals lead to intestinal injury, as did inoculating mice intragastrically with E. coli.

To understand why influenza virus infection in the lung can alter the gut microbiota, the authors examined immune cells in the gut. They found that Mice lacking the cytokine IL-17A, which is produced by Th17 helper T cells, did not develop intestinal injury after influenza virus infection. However these animals did develop lung injury.

Th17 cells are a type of helper T cells (others include Th1 and Th2 helper T cells) that are important for microbial defenses at epithelial barriers. They achieve this function in part by producing cytokines, including IL-17A. Th17 cells appear to play a role in intestinal injury caused by influenza virus infection of the lung. The number of Th17 cells in the intestine of mice increased after influenza virus infection, but not in the liver or kidney. In addition, giving mice antibody to IL-17A reduced intestinal injury.

There is a relationship between the intestinal microbiome and Th17 cells. In mice treated with antibiotics, there was no increase in the number of Th17 cells in the intestine following influenza virus infection. When gut bacteria from influenza virus-infected mice were transferred into uninfected animals, IL-17A levels increased. This effect was not observed if recipient animals were treated with antibiotics.

A key question is how influenza virus infection in the lung affects the gut microbiota. The chemokine CCL25, produced by intestinal epithelial cells, attracts lymphocytes from the lung to the gut. Production of CCL25 in the intestine increased in influenza virus infected mice, and treating mice with an antibody to this cytokine reduced intestinal injury and blocked the changes in the gut microbiome.

The helper T lymphocytes that are recruited to the intestine by the CCL25 chemokine produce the chemokine receptor called CCR9. These CCR9 positive Th cells increased in number in the lung and intestine of influenza virus infected mice. When helper T cells from virus infected mice were transferred into uninfected animals, they homed to the lung; after virus infection, they were also found in the intestine.

How do CCR9 positive Th cells from the lung influence the gut microbiota? The culprit appears to be interferon gamma, produced by the lung derived Th cells. In mice lacking interferon gamma, virus infection leads to reduced intestinal injury and normal levels of IL-17A. The lung derived CCR9 positive Th cells are responsible for increased numbers of Th17 cells in the gut through the cytokine IL-15.

These results show that influenza virus infection of the lung leads to production of CCR9 positive Th cells, which migrate to the gut. These cells produce interferon gamma, which alters the gut microbiome. Numbers of Th17 cells in the gut increase, leading to intestinal injury. The altered gut microbiome also stimulates IL-15 production which in turn increases Th17 cell numbers.

It has been proposed that all mucosal surfaces are linked by a common, interconnected mucosal immune system. The results presented in this study are consistent with communication between the lung and gut mucosa. Other examples of a common mucosal immune system include the prevention of asthma in mice by the bacterium Helicobacter pylori in the stomach, and vaginal protection against herpes simplex virus type 2 infection conferred by intransal immunization.

Do these results explain the gastrointestinal symptoms that may accompany influenza in humans? The answer is not clear, because influenza PR8 infection of mice is a highly artificial model of infection. It should be possible to sample human intestinal contents and determine if alterations observed in mice in the gut microbiome, Th17 cells, and interferon gamma production are also observed during influenza infection of the lung.

TWIV 266: A pathogenic vicious cycle

On episode #266 of the science show This Week in Virology, Vincent, Alan, and Kathy discuss finding viruses in outer space, varying results obtained from personal genetic testing, and depletion of CD4 cells during HIV infection by pyroptosis.

You can find TWiV #266 at

Antimicrobial peptides induced by herpesvirus enhance HIV-1 infection

Langerhans cellsThe risk of being infected with human immunodeficiency virus type 1 (HIV-1) is substantially enhanced in individuals with other sexually transmitted diseases. For example, infection with herpes simplex virus type 2 (HSV-2) increases the risk ratio of acquiring HIV from 2 to 4. Explanations for this increased risk include direct inoculation of HIV-1 into the blood through genital ulcers, and the induction of inflammatory cells by HSV-2 which act as sites of replication for HIV-1. The results of infections carried out in cell culture suggest a biological mechanism for the enhancement of HIV-1 infection by HSV-2.

Langerhans cells (LC) are believed to one of the first cells in which HIV-1 replicates after sexual exposure. LCs are dendritic cells which patrol the mucosal epithelium, taking up and processing antigens and presenting them to T cells in the lymph nodes. These cells express the HIV-1 receptors CD4 and CCR5, but not CXCR4, and can therefore be infected with CCR5-tropic* but not CXCR4-tropic HIV-1. Individuals who do not express CCR5 are resistant to HIV infection. For these and other reasons CCR5-tropic HIV-1 viruses are believed to be ones that transmit infection from one individual to another.

In human skin explant cultures, which contain LCs, co-infection with HSV-2 substantially increased the number of HIV-1 cells. This observation could not be explained by co-infection of individual cells because very few of these were observed in the cultures. When applied to fresh cells, the supernatant of cultures infected with HSV-2 also stimulated the number of HIV-1 infected LCs. These observations suggested that HSV-2 infection stimulates the production of one or more substances from infected cells which in turn improve HIV-1 infection.

Human epithelial and epidermal cells are known to produce antimicrobial peptides such as defensins and cathelicidin. These are short, evolutionarily conserved peptides that inhibit the growth of bacteria, viruses, and fungi. HSV-2 infected keratinocytes were found to produce a number of antimicrobial peptides, but the most important one is called LL-37. This peptide enhanced the expression of HIV-1 receptors CD4 and CCR5 on LCs, leading to increased susceptibility of the cells to HIV-1. Removing LL-37 from the supernatant of HSV-2 infected cells reduces the ability of the medium to stimulate susceptibility to HIV-1.

These findings provide a plausible mechanism by which HIV-1 infection is enhanced by HSV-2. When HSV-2 infects the genital mucosa, the epithelial cells produce LL-37. This antimicrobial peptide enhances the production of CD4 and CCR5 on LCs, allowing more efficient infection by HIV-1. This mechanism is supported by the observation that elevated levels of LL-37 correlate with HIV-1 infection in sex workers.

I wonder why antimicrobial peptides up-regulate CD4 and CCR5. In addition to their antimicrobial properties, the cathelicidins possess chemotactic, immunostimulatory, and immunomodulatory effects, and the upregulation of CD4 and CCR5 are likely part of these activities.

These are exciting findings, and if they are further correlated in humans, they might lead to novel ways of interfering with HIV-1 infection, such as by antagonizing LL-37.

*CCR5 and CXCR4-tropic refer to HIV-1 virions that bind to chemokine receptors CCR5 or CXCR4, respectively, in addition to CD4, to initiate infection.