HIV restriction factor blocks respiratory viruses - but not how we thought

Measles virus - is it targeted by an HIV restriction factor?

Cellular organisms have evolved multiple defences to keep viruses and other genetic parasites at bay. One such shield is the development of an advanced adaptive immune system seen in vertebrates while another is the more evolutionary widespread 'innate' system made up of various expressed proteins and small RNAs. These molecules prevent particular stages of the viral replication cycle like entry, replication or exit. 

Another mechanism is to mutate the virus out of existence, i.e. change nucleotides throughout the viral genome so much so as to effectively knock-out that protein's function during infection and slow down replication so much so to allow your bodies' other defences to clear it. Since 2002 we have had tantalising clues that this is functioning during a viral infection, particular HIV and other retroviruses:

Isolation of a human gene that inhibits HIV-1 infection and is suppressed by the viral Vif protein.

Influenza - putting the Trojan into the horse but should you open it?

A trojan horse (dendritic cell) filled with virus

Inspired by a recent journal club article:

A number of pathogens infect via one organ but are able to move to another. Think of the likes of the initial HIV infection of cells within the reproductive tract followed by its transfer to the immune system (see my earlier post here). This transfer of viruses, bacteria or other parasites is responsible for the induction of an effective immune response but also it can lead to some of the more serious disease symptoms during infection, but how is it controlled?

Revisiting the origins of contagion - the measles story

The mechanisms behind the incredible infectiousness of measles are poorly understood - that is, until now, where two studies have now come forward investigating the molecular biology of measles person-person transmission. Two groups have independently identified the protein, nectin-4 (a cell adhesion molecule) as the receptor allowing measles to infect and emerge from the respiratory tract and spread from person to person, potentially filling in a major gap in our understanding of this important human pathogen.

The measles virus is one of - if not the most - infectious agents currently circulating in human populations. It is also responsible for considerable disease and death worldwide, particularly across the developing world. But luckily, the introduction of a live-attenuated vaccine has significantly reduced it's incidence - this is in spite of a number of recent outbreaks associated with reduced vaccination rates. 

What's in a cough? Viruses. Lots of viruses.

A cough - Tang et al 2008

Respiratory pathogens are pro's when it comes to person-person transmission; easy in and easy out. Viral replication - and also I guess bacterial - followed by the subsequent immune response, induces a range of behaviors in the host which results in the expulsion of doses of infectious particles. These include your runny nose, coughs and sneezes -  you've all experienced them. Although even talking and breathing may also be spreading virus.

But, how exactly does the virus cause this? Which mechanism is potentially the most potent transmitter of virus? How many particles can be found in your average cough? And, how do these behaviors physically transfer infection? Perhaps most important - what can we do to stop it?

Closing the book on measles infection - do we know it all?

Measles infection. (http://biowiki.org)

One of most important factors in establishing a viral infection is the presence or absence of particular receptor molecules. These proteins/sugar molecules (or whatever) must be expressed on the surface of the cell, allowing the virus to bind and initiate entry. 

The expression of a receptor therefore governs the behavior of an infection by allowing only certain cells to permit virus replication. This thus controls how a virus enters your body, spreads throughout it, causes disease and finally escapes to continue the infection in a new host. The identity of the receptor that a virus uses allows a better understanding of infection and pathogenesis and may facilitate the development of new antiviral treatments.

A paper published in PLoS Pathogens (see here), this week documents the discovery that a protein - Nectin-4 - acts as the receptor that may allow measles virus (MeV) to escape your body in the ends-stages of the disease, through the infection of the epithelial cells lining your respiratory tract. However, this molecule differs from that used bythe virus to initially enter your body.

This Nectin-4 study finally sheds light on what has been an elusive molecule to find and goes a long way in explaining the basic biology of this historic human pathogen. Also, the findings that Nectin-4 is highly expressed on the surface of a number of common cancers further adds weight to the use of MeV as an anti-cancer agent. But does Nectin-4 fully explain MeV infection? And, can we finally close the book on measles infection?

Despite the widespread use of a highly effective - and safe - vaccine, MeV still infects an estimated 20 million people a year, mostly young children and, in 2008, the WHO documented that measles caused the death of 164,000 of these children that year. Measles is characterised by fever, photophobia, cough, runny nose, sickness, and a nasty rash over most of the body (see picture above). Every so often the virus will enter your brain and can cause serious inflammation there. And, even more rarely will the virus persist in your body only to re-emerge years later as an incurable degenerative brain disease known as sub-acute sclerosing pan-encephalitis (SSPE). The molecular basis of how this virus enters our body, replicates, spreads inside you exits your body are only now being realised with modern recombinant viruses, like those expressing fluorescent proteins (see below). For examples, see here, here, here and here. 

Disclaimer: these references are from work that my supervisor and the group I am currently in are heavily involved in but are the only work documenting MeV in the entire organism. Read here when I previously wrote about the discovery of how MeV enters our body through dendritic cells.

GFP-expressing MeV infection in a Rhesus macaque model - many organs afflicted and recapitulates human disease. Virus replication in, A) Outer skin, corresponding to rash, B) oral mucosa, C) tongue and tonsil, D) lymph node, E) Lung, F) stomach and intestine (gut-associated lymphoid tissue - GALT), G) intestine and spleen, H) and I) both close-up of 'koplik spot' skin rash. (de Swart, et al 2007)

It was always assumed that following breathing in virus-laden aerosols MeV initially infected your epithelial cells lining the upper respiratory tract. From here it could spread throughout your body, replicating in lymphocytes and finally be released back out via the epithelial cells. But, researchers using a green-fluorescent protein (GFP)-expressing MeV showed that it was in fact dendritic cells (sentinal immune cells) found within your respiratory tract that were the first cells to be infected. These cells could then easily transmit the virus to your immune cells. These cell types can be justified through the identify of the receptor for measles, SLAM that is only expressed on these immune cells. 

GFP-expressing MeV infection of A) polarised epithelial cells, B) and C) lymphoid like cells in mucosa. (de Swart et al, 2007)

However, based on animal models and autopsy reports, we know that the virus can - and does - productively infect epithelial cells lining your airway (see figure above) yet no receptor was known. That is, until Chris Richardson's team at Dalhousie University, Canada discovered that Nectin-4 could function as a MeV epithelial receptor. 

PVRL4 (Nectin-4) allows MeV infection

The group used microarray analysis to determine what genes were expressed in cells that could, and those that could not, be infected by a wild-type, pathogenic MeV. 

They were then able to bioinformatically pull out proteins that were found on the plasma membrane surface (ones that were likely to be receptors) and expressed them in cells that couldn't be infected. The gene that allowed MeV to infect was the receptor molecule and this was confirmed to be Nectin-4 through siRNA knockdown and antibody binding assays.

So does Nectin-4 explain how MeV infects, causes disease and escapes the human body? Based on the Human Protein Atlas entry for Nectin-4, this protein is on most normal cells weakly - including the brain and lung, which are both targets for MeV and it is not really expressed on lymphoid cells, although these cells already express a highly-active receptor for the virus. This study may explain how the virus can infect your respiratory tract epithelium and be secreted into your airways to spread the infection. The role of Nectin-4 should be established in further model systems like the macaque and even in vitro differentiated epithelial cells.

de Swart RL, Ludlow M, de Witte L, Yanagi Y, van Amerongen G, McQuaid S, Yüksel S, Geijtenbeek TB, Duprex WP, & Osterhaus AD (2007). Predominant infection of CD150+ lymphocytes and dendritic cells during measles virus infection of macaques. PLoS pathogens, 3 (11) PMID: 18020706 


Lemon K, de Vries RD, Mesman AW, McQuaid S, van Amerongen G, Yüksel S, Ludlow M, Rennick LJ, Kuiken T, Rima BK, Geijtenbeek TB, Osterhaus AD, Duprex WP, & de Swart RL (2011). Early target cells of measles virus after aerosol infection of non-human primates. PLoS pathogens, 7 (1) PMID: 21304593

Noyce, R., Bondre, D., Ha, M., Lin, L., Sisson, G., Tsao, M., & Richardson, C. (2011). Tumor Cell Marker PVRL4 (Nectin 4) Is an Epithelial Cell Receptor for Measles Virus PLoS Pathogens, 7 (8) DOI: 10.1371/journal.ppat.1002240

When you say flu, what do you really mean?

Everyone knows the 'flu when they see it. Its like the cold but only worse, right? But often the public are very quick to brand something the flu when it clearly isn't and this even extends to doctors in the clinic. As highlighted by this recent paper, published in BMC Infectious Diseases.

BioMed Central | Full text | During the summer 2009 outbreak of "swine flu" in Scotland what respiratory pathogens were diagnosed as H1N1/2009?

Turns out a lot of viral infections look like influenza, clinically. But what are doctors misdiagnosing as flu? Based on this study, which paired molecular diagnostics with what the clinic thought each patient had  based on well-defined guidelines (mostly flu at the early stages of the 2009 swine-influenza pandemic - April - July).

Are these symptoms clinically specific to influenza infections? http://symptomswineflu.com 

What they did was:

We examined the results from 3247 samples which were sent to the laboratory during April-July 2009

And,

Real-time RT-PCR was carried out in order to detect influenza A (a generic assay and a H1N1/2009 specific assay [7]), B and C, RSV, rhinovirus, parainfluenza 1-4, human metapneumovirus, coronavirus (229E, NL63, HKU1 and OC43), adenovirus, and Mycoplasma pneumoniae.

 Out of the 3247 samples analysed, 27.9% were diagnosed as being infected with a respiratory pathogen they looked for, what is causing their symptoms??. While human rhinovirus infections were the most common, H1N1 influenza infection was one of the least common at 0.8%. A whole heft of viral and bacterial pathogens were detected at levels intermediate to these.


So, whats going on here? How come it's so difficult to diagnose these guys and after-all, does it really matter?


Well, this work highlights the difficulty in accurately determining the causative pathogen from basic symptoms. The viruses looked at in this study all share replicative and epidemiological characteristics: respiratory pathogens, infect the upper/lower respiratory tract and, as viruses, induce a general interferon/inflammatory response (sore throat, headache, malaise, fever, muscle pains etc.) and are transmitted among the winter/spring months. It is not until you look at the molecular level, that you realise they are all at least slightly different.


And, yes it does really matter. Many of these people may have developed life-threatening symptoms as a consequence of an infection that was not influenza and would have been missed by the clinicians. Also, at a great cost to society and the patient (financial, medical and emotiona;), some people were wrongly treated with antiviral medication after being diagnosed swine flu.


What studies like this show, is the importance of molecular diagnostics in times of pandemic infections. But the problem here is that, despite their high specificity, they can be expensive and difficult to run, especially in 'field' locations across the world. 

Can we target antivirals to the host cell?

Two papers published recently explore the idea of targeting antiviral compounds to the host cell using high-throughput organic synthesis and an in vitro screen. Both groups identify a single novel inhibitor of virus replication and both are host specific.  One paper goes further and identifies the target and also the in vivo potential of this compound. These papers highlight the hopes and pitfalls of targeting the host cell against viruses while shedding light on the basic biology of these viruses.

   We can protect ourselves from the harmful effects of viral infection through the use of vaccination, antivirals or indeed behavioral changes. Yet not all viruses have proved to be amenable to these and some, like measles and mumps, are quite easily combated with mass immunization campaigns while the likes of HIV have shown to be quite sensitive to antiviral compounds.

Respiratory infections are a problem! hastings.gov.uk

   For a majority of viruses though, we have neither effective vaccines nor antivirals. We need these now more than ever. And, even those which we do have vaccines and/or antivirals (Influenza A), resistance is quickly built up over time rendering them clinically useless. How then are we able to generate new, more effective antiviral molecules against a whole range of viral pathogens, especially given the cost of taking a single drug to patients?; we want more bang for our buck with antivirals. One way maybe through the targeting not of a viral gene or protein but a cellular one. The virus life cycle is so intimately connected to that of the host cell that removal of cellular pathways may have a dramatic effect on virus functioning and applications like these have the beauty of possibly targeting both established and emerging pathogens.

   Krumm et al, from the Emory University School of Medicine, Atlanta, U.S.A, demonstrate the feasibility of this approach using viruses from the paramyxo- and the orthomyxoviruses (generally called the 'myxoviruses') as their model of choice. Read the paper here. Using this, they discover a novel compound that demonstrates host-cell specificity while greatly inhibiting virus replication. Even at nanomolar concentrations (potent!).

   This work falls on the back of a series of experiments they reported back in 2008 where they screened a 137,500 compound library against in vitro infection with the related measles, canine distemper and human parainfluenza 3 viruses. This was to determine if they could uncover broad range, potent inhibitors of a large number of viruses. Moving away from the 'one-bug, one-drug' paradigm, this work identified a class of 11 compounds that did just that: they significantly inhibited replication and cytopathic effect of all the three viruses. They use the data from this series of experiments to find a host-targeted compound.

They state that their 'ideal' inhibitor would be:

In search of candidates with a host-directed antiviral profile, we anticipated three distinct features of desirable compounds: a) potent inhibition of virus replication at the screening concentration (3.3 µM); b) a primary screening score, representative of the selectivity index (CC₅₀/EC₅₀), close to the cut-off value for hit candidates due to some anticipated host-cell interference ( = 1.9); and c) a broadened viral target spectrum in counter screening assays that extends to other pathogens of the myxovirus families

initial compound A) and B), the synthetically optimised JMN3-003

   A single compound was taken forward and then synthetically optimized (high-throughput synthetic chemistry anybody?) to achieve even better inhibition. No longer do we have to rely on the natural diversity of compounds out there. This new and improved compound (JMN3-003) could inhibit not just myxo-virus replication but also that of clinically relevant RNA (Sindbis virus) and to a lesser extent DNA viruses, such as vaccinia virus. The broad activity of the molecule suggests that it may target a more general aspect of viral replication than others had previously. The activity of this drug was also host-cell specific, suggesting that it did not target a virus structure but rather one found within the host cell, whose activity would change depending on the species assayed.

JMN3-003 antiviral activity

   This molecule had little or no effect on the metabolic activity of the immortalized cell lines used or of other primary cell cultures when administered at >7,000X the active dose, possibly hinting at the safety profile of such a drug. And, under liver cell extract stability tests, the compound showed a desirable extrapolated half-life of around 200 minutes. Host transcription and translation weren't affected by drug administration.To be safe, the authors put forward that this compound may only be preferentially used in acute respiratory infections where the treatment time would be substantially shorter. This group did not carry out any in vivo work so the real safety of this compound cannot really be assessed. See below for how this should be done.

Activity changes depending on the cell used. Human PBMCs or primate Veros

   So, we have demonstrated that the compound inhibits virus replication. But just how exactly does it do this? At what stage of the virus replication cycle does it act? Entry? Transcription? replication? assembly? exit? The group addressed these points through the use of assays that measure virus envelope fusion with the cell membrane (virus entry) and virus gene expression and replication (post entry). This work showed that fusion was not inhibited when the drug was added suggesting a post entry step was being targeted. When compared to a known inhibitor of virus polymerase activity, JMN3-003 looked very similar. Other experiments, assessing levels of mRNA and genomes produced from virus replication showed a significant inhibition in one or both processes during measles or influenza virus infection. Some component associated with virus polymerase reactions is being targeted. What it is, sadly this paper did not address, yet further experiments I am sure will uncover the mechanism of action of this compound and maybe through this identify an important host-virus association that has been previously overlooked. Recently, evidence has suggested that these viruses require the activity of host kinases for proper, regulated replication. Could it be these?

Little risk of resistance?

   Their experiments also suggest that it is extremely difficult for these viruses to adapt to the inhibition from JMN3-003. After 90 days of replication on cells treated with the compound, little resistance had emerged. Again, the reasons for this were not addressed. When compared to previous problems of drug resistance seen in Influenza A and HIV-1, this activity looks very promising.

RSV - the biggest cause of bronchiolotis in young children. http://www.empowher.com/

   Interestingly, this approach has been used for another medically important virus (Respiratory syncitial virus, RSV - also a paramyxovirus) and just published recently in the journal PNAS (Bonavia et al 2011). This group used effectively the same strategy as described above, except using a lot more compounds (1.7 million). They picked up a number of molecules that also targeted a post-entry step in virus replication, possibly virus replication ( RdRp activity); it also shows broad spectrum activity against a range of RNA viruses. And again, no resistant viruses emerged in the experiments. 

   This work went a little further than that I mentioned earlier and Bonavia et al were able to pull down the drug target using a type of affinity purification with immobilized samples of the compound (iTRAQ quantitative chemical proteomics). The protein binding most strongly to this may mostly likely be the target. In this case, the two compounds tested both targeted seperated enzymes in the same pathway, the de novo pyrimidine biosynthesis pathway, that is the nucleotides A, T and U (A and U being used extensively by RNA viruses). To assess the safety of targeting this important cellular pathway, the group looked at both in vitro and in vivo rodent models of disease. Rapidly-dividing cells showed high sensitivity to the compound while significant histopathological changes were observed in lymphoid and gut tissues in the cotton rat. This is not good for a potential clinical drug that may be administered to very young babies. 

Pyramidine biosynthesis - a good antiviral target? http://upload.wikimedia.org/

   Sadly, these compounds exhibited no antiviral activity in vivo and rather one increased RSV virus replication. This is possibly down to the potential immunosupressive nature of the target (inhibits rapidly dividing cells - T and B cells) or the ability of circulating pyramidines in the animal's body to rescue the inhibition of their synthesis inside the cell.

   I wonder how the first compound, JMN3-003 would fare in the many animal models of myxovirus infection. And, what would the target be when the chemical pull-down was applied? would it be the same pathway? Both these papers highlight the ability of this approach to identify readily targeted host cell components to reduce virus replication in vitro and potentially overcome the problems of narrow virus targeting and resistance while the second paper goes further and shows us how the results may not be pleasing. The targeting of host factors has obvious implications for patient health and this cotton rat model displays what can go wrong. We should learn from this and expand any potential drug candidates into in vivo modeling quickly and at the start of any experiments, not like what was done with JMN3-003, although that's not to say the same will happen with that compound.



Bonavia, A., Franti, M., Pusateri Keaney, E., Kuhen, K., Seepersaud, M., Radetich, B., Shao, J., Honda, A., Dewhurst, J., Balabanis, K., Monroe, J., Wolff, K., Osborne, C., Lanieri, L., Hoffmaster, K., Amin, J., Markovits, J., Broome, M., Skuba, E., Cornella-Taracido, I., Joberty, G., Bouwmeester, T., Hamann, L., Tallarico, J., Tommasi, R., Compton, T., & Bushell, S. (2011). Organic Synthesis Toward Small-Molecule Probes and Drugs Special Feature: Identification of broad-spectrum antiviral compounds and assessment of the druggability of their target for efficacy against respiratory syncytial virus (RSV) Proceedings of the National Academy of Sciences, 108 (17), 6739-6744 DOI: 10.1073/pnas.1017142108

Krumm, S., Ndungu, J., Yoon, J., Dochow, M., Sun, A., Natchus, M., Snyder, J., & Plemper, R. (2011). Potent Host-Directed Small-Molecule Inhibitors of Myxovirus RNA-Dependent RNA-Polymerases PLoS ONE, 6 (5) DOI: 10.1371/journal.pone.0020069

Yoon, J., Chawla, D., Paal, T., Ndungu, M., Du, Y., Kurtkaya, S., Sun, A., Snyder, J., & Plemper, R. (2008). High-Throughput Screening--Based Identification of Paramyxovirus Inhibitors Journal of Biomolecular Screening, 13 (7), 591-608 DOI: 10.1177/1087057108321089