Differentiating Infections Based on Gene Signatures

Author: Julie Spitzer

Institution:  University of Wisconsin-Madison


Sniffling, sneezing, coughing and wheezing, the tissue box empties as the infection worsens and a trip to the doctor may only do as much as a simple decongestant, depending on what kind of infection you body is fighting that is. Knowing, however, which type of illness has taken over your body’s insides may now be discovered much more easily, leading to better treatment practices and vaccination make ups. Recently, researchers have found that gene behaviors, or rather gene signatures, are able to show whether an illness is viral or bacterial. Such information is relevant as it will help diagnose and treat the illness. This information is particularly helpful in identifying how individuals will respond to certain vaccinations.

A group of scientists, led by Purvesh Khatri of Stanford University, have recently published a study in the December issue of Immunity outlining their work and findings.

In general, when a person becomes infected with an illness, the person’s genes behave differently. Subsequently, the host response, or rather the way the body works in order to defend against the infection, will differ if the infection is bacterial or viral. Thus, creating a distinction is vital so that the optimal treatment can be carried out.

Bacteria and viruses creep into the body as pathogens. When illness strikes, an immune response is triggered. Bacteria often hide between cells, combatted by antibodies. Viruses must enter the cell and then be fought off by cytotoxic T lymphocytes (http://www.imgt.org/IMGTeducation/Tutorials/ImmuneSystem/UK/the_immune_system.pdf).

Their research techniques included looking at publicly available gene expression datasets.

According to Khatri, public datasets were utilized in order to save time and cost. Additionally, these data were gathered from real infections, not in vitro infections to account for natural human responses.

First, they found a robust 396-gene meta-virus signature (MVS). This host signature is the same throughout various respiratory viral infections. By recognizing such a signature, a distinction between viral infections, bacterial infections, and healthy controls could be made. Additionally, the researchers were also able to distinguish between symptomatic subjects and asymptomatic subjects even before symptoms began. All of such may help act as diagnostic or prognostic biomarkers.

“The idea was we [wanted] to find the common transcriptional response across multiple ... viruses and the rationale for that was two-fold. One was we wanted to see if that allows us to learn any new biology and the second was … for in the clinical setting,” said Khatri about the work conducted in the study.

To find MVS, or the common transcriptional signature for all respiratory viral infections, three gene expression datasets from 205 human blood samples and three viral infections (influenza, human rhinovirus and respiratory syncytial virus) were analyzed.

“There was one subject that was asymptomatic, so never showed any symptoms but when they would do a nasal swab, everyday they did a nasal swab, that subject was always showing viral … Then there was another subject that was the opposite of that: showing all the signs of [an] infection but whenever they would do [a] nasal swab, it would never show presence of virus and we were able to identify both of them accurately,” said Khatri.

“So we were identifying asymptomatic [patients] as infected and symptomatic [patients] as non-infected and that was really really fascinating for me that we could diagnose at an individual level out of 300 samples we looked at so it tells you the precision and the accuracy of what we had found,” said Khatri.

Secondly, the researchers identified an 11-gene influenza meta signature (IMS) capable of further differentiating influenza infected samples from other viral infections. This means the host response was a biomarker in influenza-pneumonia patients. Furthermore, IMS was found to help predict a response to the influenza vaccine.

The IMS was found by testing five influenza gene expression datasets of 292 samples. This was necessary to show that there were “virus-specific signatures encompassing smaller subsets of genes.”

The IMS increases in influenza vaccine recipients, meaning that IMS could act as a biomarker for vaccination. The study found that there were differences in vaccine responses between males and females; innately, humoral and cell-mediated. They concluded that this is likely due to testosterone and its immunosuppressive role. Hence, males respond sooner that females. The data had shown that males IMS scores had changed after just one day while females scores had changed after three.

Khatri explains that although previous studies may be correct in that testosterone does play a part in suppressing the immune system, testosterone may not be the only reason for varied vaccine responses. “Its not that testosterone is inhibitory just that the dynamics are different… we should be sampling more frequently in the first three days…when is it that the males respond and when is it that the females respond,” said Khatri.

MVS data may be useful for developing treatment regimens, particularly anti-viral drugs. Previously, such drugs were created from a “one-bug-one-drug philosophy,” but by utilizing these data, better, or rather stronger, drugs could be created to combat broader, more complex viruses.

In regards to future studies, Khatri and others aims to collect prospective samples and test how good this technique is at differentiating between viral and bacterial infections. Additionally, they aim to work with host responses. If a common host pattern is determined, it may be targeted so that the virus life cycle, or mutations, can be predicted. Hence, new drugs may be developed.

Through the discovery of the MVS and IMS gene signatures, illnesses will be diagnosed quicker, better treatment options will be explored and better vaccines will be developed. So one may now only hope to say goodbye to extended sick days and coughing spells.