Gene expression profiling of monocytes in the HIV/tuberculosis-associated immune reconstitution inflammatory syndrome.

Scriptie/masterproef: Doctoral Thesis


Tuberculosis (TB), caused by Mycobacterium tuberculosis (Mtb), is one of the most common
opportunistic infections (OIs) in HIV patients, in particular in developing countries. Restoration and
preservation of immune function by combination antiretroviral therapy (cART) have led to a
significant reduction in prevalence and associated mortality of OIs such as TB, and improved survival
of HIV-TB coinfected patients. However, in some instances cART is associated with complications
and paradoxical clinical deterioration. One of the more common but least understood complications
associated with cART in HIV-TB coinfected patients is the TB-associated Immune Reconstitution
Inflammatory Syndrome (TB-IRIS), which typically manifests as a paradoxical clinical deterioration
in patients already stabilized on anti-TB therapy. This syndrome significantly complicates the
management of both HIV and TB treatment and represents a great burden for the resource-limited
health systems in developing countries.
Despite satisfactory control of viral replication and reconstitution of CD4+ T cell counts, and despite
absence of active TB infection, patients with TB-IRIS show worsening symptoms of TB (such as
fever, return of cough) and/or detrimentally exuberant inflammatory reactions such as abscess
formation within lymph nodes, soft tissues or solid organs. TB-IRIS can occur as paradoxical or as
unmasking condition. Paradoxical TB-IRIS occurs in patients on successful TB treatment before
cART. Unmasking TB-IRIS occurs in patients without apparent TB when they start cART and may
represent either reactivation of latent infection or worsened symptoms of TB that was not previously
diagnosed as active disease or was subclinical.
The precise mechanisms underlying the development of TB-IRIS remain incompletely understood and
no prognostic/diagnostic test for TB-IRIS has been developed. Most studies to date on the
immunopathology of TB-IRIS have focused on components of the adaptive arm of the immune
system, such as mycobacterial-specific T cells, Th1/Th2 cytokine profiling, and T-regulatory cells.
However, findings have frequently been confusing and contradictory between studies and/or study
populations. Other components of the immune system have only been explored cursorily, even though
the predominantly early occurrence of TB-IRIS (during the initial months and even weeks of cART)
suggests a possible role for the innate arm of the immune system in severely immunosupressed
patients. This notion has been increasingly supported by recent evidence, including the overwhelming
presence of macrophages in TB-IRIS lung biopsy, Natural Killer cell activation during TBIRIS,
and dysregulation of pro-inflammatory cytokines associated with innate immune cell function.
Combined with the documented wide range of monocyte/macrophage dysfunctions during
HIV infection, HIV-Mtb coinfection, and cART, we considered it likely that MOs/macrophages are of
importance in the development of TB-IRIS.The current work therefore aimed to investigate the following aspects of potential involvement of
myeloid cells in the development of TB-IRIS, using monocytes (MOs) as a study model: (1) to
confirm that MOs are involved in TB-IRIS development; (2) to search for MO-associated factors that
represent (prognostic and diagnostic) biomarkers for the development of TB-IRIS; and (3) to
investigate how MOs functionally contribute to cART and combined cART/anti-TB treatment
responses in patients who do and do not develop TB-IRIS. Accordingly, our work consisted of three
main parts: a study on the different MO subsets present in PBMCs of TB-IRIS patients, a human
genome-wide (HGW) microarray analysis, and a focused functional investigation. Our work was
conducted as a nested case-control study, integrated in a larger cohort study on TB-IRIS, involving an
HIV-positive patient population attending Mulago hospital in Kampala, Uganda. Due to the specifics
of our study population, this work focused on the development of paradoxical TB-IRIS only.
In order to obtain an overview of the extant MO sub-populations during TB-IRIS, we performed flow
cytometric analysis on PBMC samples from TB-IRIS patients and matched HIV+/TB+ non-IRIS
controls. We observed proportional differences of MO subpopulations between patients and controls,
specifically for the CD14hi/CD16- MO subset, providing the first indication of dysregulation of
myeloid cell functionality during TB-IRIS. As this MO subset is also readily available in peripheral
blood samples, it was selected as the focus of our study. Protocols for isolation of this MO population
from cryopreserved PBMC samples and purification of high-quality RNA from minute cell samples
were optimised.
To compare gene expression between patients who developed TB-IRIS and control patients, a HGW
microarray profiling was conducted, in which the transcriptome of MOs of TB-IRIS patients and
controls was analysed at two time-points: at cART initiation (baseline) and at week two post-cART
initiation. This study was among the first to show the possibility of using gene expression to classify
TB-IRIS patients and controls, both at the approximate time of the TB-IRIS-defining event and before
initiation of cART (i.e. as a predictive classification). This finding is an important proof-of-concept
that gene expression can be used as a classifier for predisposition to TB-IRIS, and moreover that this
predisposition manifests at least partially at the level of the circulating MO population. This
observation should allow a much more refined analysis of the underlying mechanisms of
predisposition to TB-IRIS, such as putative genetic determinants, than the relatively general precART
predictors identified so far, such as a low nadir CD4 count. At the practical level, such a
predisposition could be exploited for the development of a prognostic test or testing algorithm, which
could allow identification of patients at high risk of developing TB-IRIS, thus greatly facilitating
clinical management of co-infected patients. In addition to the relevance of MO for the purpose of
classifying TB-IRIS patients, the fact that MO-associated genes are expressed differentially,
combined with the functional mapping of such differentially expressed MO genes, indicates a
potential functional contribution of MO/myeloid cells to the development of TB-IRIS. A number of
MO-associated biological processes or functions were found to be perturbed in TB-IRIS patients - in
particular, an apparent dysregulation in both anti- and pro-inflammatory processes in MOs was
Amongst these modulated pathways, the complement system was selected as focus for further
investigation, due to the extent of its dysregulation and its potential relevance. An independent
technological setup was used for confirmation of complement cascade dysregulation at the mRNA
level, and key protein components of the complement system were examined further where the tools
and samples were available. We showed that even before initiation of cART, the complement system
was dysregulated in HIV-TB co-infected patients who were predisposed to TB-IRIS. Detailed analysis
revealed differences between TB-IRIS patients and matched non-TB-IRIS cases, at the level of the
balance between the effector C1Q and the inhibitor C1-INH, both before and two weeks after cART
initiation. Our results suggest that inappropriate control of complement activation could be associated
with the "flaring up" of inflammation observed during TB-IRIS.
Even though a great number of research gaps remain to be filled, the data from our microarray
analysis has opened several potential directions for further functional investigation of the disease
mechanisms. Integrating our findings and other reports on TB-IRIS suggests the underlying
immunopathogenesis of the syndrome is likely a multifactorial process resulting from, amongst
others, availability of antigen, sensitivity towards antigen, degree of reduction of microbicidal activity
and aberrant activation of inflammatory effector cells leading to an imbalance between pro-and antiinflammation
Datum Prijs2 jul 2013
BegeleiderPatrick De Baetselier (Promotor), Rafael Van Den Bergh (Advisor), Geert Raes (Advisor) & Luc Kestens (Co-promotor)

Citeer dit