NEW YORK (GenomeWeb) – In a study published online today in mBio, a University of California at San Francisco-led team described white blood cell gene expression signatures that may prove useful for diagnosing acute and sub-acute Lyme disease.
The team, led by Charles Chiu, a laboratory medicine researcher at UCSF, followed 29 individuals with Borrelia burgdorferi-caused Lyme disease, collecting blood samples at their time of diagnosis, at the end of a three week course of antibiotic treatment, and again six months later. At each of these time points, their RNA sequencing experiments uncovered sets of genes with elevated or diminished white blood cell expression in the Lyme disease patients compared with 13 unaffected control individuals.
The team is now in the process of trying to identify the minimum set of genes that could be tested using more targeted panel sequencing, in the hopes of establishing a clinically useful diagnostic test in the future, Chiu told GenomeWeb. It also plans to continue testing the same individuals up to two years after infection to better understand the course of Lyme disease infection.
In the current study, for example, the group saw similar gene expression patterns in seemingly resolved Lyme disease cases as they did in a handful of individuals who reported persistent Lyme disease symptoms after antibiotic treatment.
Even so, neither the resolved nor persistent cases had returned to the baseline white blood cell expression patterns found in control individuals, suggesting the infection may lead to lingering immune responses, if not lasting physical symptoms.
There is ongoing controversy about whether Lyme disease can persist as a chronic condition, Chiu explained.
Past research hints that some 10 to 20 percent of individuals diagnosed with Lyme disease develop lasting symptoms such as fatigue, joint pain, confusion, meningitis, facial nerve problems, and, in some cases, dangerous heart muscle inflammation and damage. Post-treatment Lyme disease syndrome (PTLDS) is considered for cases involving functional decline coupled with these long-term symptoms.
The mBio study was not designed to address the existence of chronic Lyme disease, or PTLDS, Chiu noted. Rather, the work was part of a longitudinal Lyme disease study that set out, in part, to search for a host response signature coinciding with acute Lyme disease infection, which is notoriously tricky to diagnose.
"The gold standard test for Lyme disease is an antibody test," he said. "But … you typically have to wait for a period of time — it can be from several days to weeks — before the person mounts an antibody response that's detectable. So in the setting of an acute infection, the antibody test may not be positive."
Antibody tests also tend to be non-specific, he added, while efforts to diagnose Lyme disease through direct molecular detection of B. burgdorferi DNA are often complicated by the bacteria's transient appearance in the blood.
In an effort to find specific host responses to Lyme disease, the team collected blood samples from 29 individuals diagnosed with Lyme disease at a clinic in Maryland and from 13 unaffected individuals matched for age, sex, and ethnic background.
After doing Illumina RNA sequencing on peripheral blood mononuclear cells from case and control samples, the researchers narrowed in on more than 1,200 genes with higher- or lower-than-usual levels in white blood samples from the Lyme disease patients at diagnosis compared to samples from control individuals.
Expression differences persisted after three weeks of treatment with the antibiotic doxycycline, when the team identified some 670 genes with elevated expression in Lyme disease compared to control samples and another 390 genes with diminished expression in the patient group.
The differentially expressed genes found at this second time point overlapped to a large extent with those uncovered in diagnostic samples, Chiu noted, suggesting such early stages of disease may reveal the most promising expression targets for future diagnostic panel development.
"When a patient has recovered and no longer has fever, no longer has a rash, you can still detect this very strong inflammatory host response to the infection," Chiu said.
"That's very attractive, because what it suggests is that a diagnostic … based on host response might fit into what we call the 'window period' — between when you might be able to detect the bacteria and when you might be able to detect an antibody response to it," he explained, noting that existing diagnostic approaches may be effective during this stretch of time.
The diagnostic and post-treatment expression patterns were largely distinct from those described previously in white blood cells from individuals with other types of infection, the team noted, though the acute Lyme disease signature shared some features with that found in individuals with acute influenza infections.
Symptoms in 15 of the 29 individuals were successfully resolved after antibiotic treatment, the researchers reported, and just four individuals reported PTLDS with functional decline after six months.
The team did not see expression differences between individuals with PTLDS and resolved infection after six months, Chiu noted, though he cautioned that larger studies or studies of more specific blood cell types may be needed to detect subtle biological changes in the PTLDS group.
And while similar gene expression profiles turned up in samples from fully resolved cases, those with some persistent symptoms, and PTLDS cases, these individuals still showed differential expression at 686 genes compared to the control group, six months after treatment.
Again, the late Lyme disease signature overlapped to some extent with expression patterns present at diagnosis, the team noted, though the genes and pathways detected at this later stage increasingly overlapped with those implicated in host response to autoimmune conditions such as rheumatoid arthritis or systemic lupus erythematosus.
Chiu and his team are continuing to follow the same patient group and plan to assess samples collected at intermediate stages of infection — around three or four months after treatment — as well as samples collected up to two years after Lyme disease diagnosis.
For potential diagnostic development, "we're looking at the first and the second time point, because we want to develop an acute diagnostic within the zero to three-week time period," Chiu said.
And while the current analysis focused on host responses to the Lyme disease-causing bacteria B. burgdorferi, he and his colleagues are interested in scrutinizing host responses to other tick-borne infections as well — from the pathogens behind human granulocytic anaplasmosis, babesiosis, and spotted fevers.
They would also like to test host response to other Borrelia species, including the new Lyme disease-causing species introduced this week in Lancet Infectious Diseases, if they can access appropriate patient samples.
"It's our hope that we can develop distinctive signatures for a wide variety of tick-borne infections, not just Borrelia burgdorferi," Chiu said. "Ultimately, the goal would be to develop a panel that could diagnose tick-borne infections solely on the basis of host response — and that panel would be complementary to existing diagnostics."