Liesbeth Vanherp1, Amy Hillen1, Jennifer Poelmans1, Katrien Lagrou2, Greetje Vande Velde1, and Uwe Himmelreich1
1Imaging and Pathology, University of Leuven, Leuven, Belgium, 2Laboratory of Clinical Bacteriology and Mycology, University of Leuven, Leuven, Belgium
Synopsis
Animal models of
cerebral infection by the pathogenic fungi Cryptococcus neoformans and C. gattii were developed and assessed
longitudinally by using anatomical and diffusion MRI as well as MR
spectroscopy. MR spectroscopy identified in vivo biomarkers for potential etiological diagnosis and more
importantly for quantification of the fungal load in living animals. Our results have great potential to assist in the
differential diagnosis of brain lesions in patients, whereby MR spectroscopy is
a safer, non-invasive and rapid method in comparison to traditional invasive
diagnostic methods such as CSF sampling or biopsies.
Introduction
Cryptococcus neoformans and C. gattii are encapsulated yeasts that can
cause life-threatening fungal infections primarily affecting immunocompromised
individuals (1). The main route of infection is by inhalation, but cryptococci can also spread from the
lung to the brain. Thereby, they can cause lesions (cryptococcomas) that are difficult
to distinguish from other pathologies (like cystic brain tumors or bacterial
abscesses) by means of
conventional radiological examinations such as computed tomography (CT) or
magnetic resonance imaging (MRI). Although both cryptococcus species are related,
potential differences in their clinical presentation and disease pathogenesis have
been reported (2). Current techniques used in preclinical research, such as
colony-forming unit (CFU) counting to determine fungal load are limited to snapshots
of the disease and do not allow longitudinal studies of the same animal because
of their invasive character. The aim of this study was to use MRI and MR spectroscopy (MRS) to identify in vivo parameters that can give a quantitative read-out of the
fungal load in living animals at an anatomical and metabolic level over the
course of disease progression. Furthermore, this approach was used to identify
potential differences in pathogenesis between C. neoformans and C. gattii
infections.Materials and Methods
Animal models
Two animal models were used, (1) Cryptococcus strains (C. neoformans H99 and C. gattii R265; 104 fungal
cells) were stereotactically injected into the right striatum of female BALB/c
mice to induce localized lesions and (2) Animals were injected intravenously
with 5 x 104 cryptococcal cells to reflect disseminated disease. For
local injections, mice were scanned at day 5, 8, 10 and 13 post infection. Mice that were infected by i.v. injection were
scanned repeatedly for a period of 1 to 4 weeks.
Imaging
All imaging
experiments were performed using a small animal 9.4 T MRI scanner (Biospec
94/20, Bruker Biospin, Ettlingern, Germany) equipped with 600mT/m gradients and
using a 7cm linearly polarized resonator for transmission and an
actively-decoupled mouse brain surface coil for receiving (Bruker Biospin). We
acquired anatomical 3D MRI (RARE, TR/TE: 1000/36ms, 0.1mm resolution), 2D DWI
(TR/TE: 3s/ 27ms, b-values: 30 to 1500ms,) and single voxel MRS (PRESS, 2 x 2 x
2mm3 voxel, TR/ TE: 1.8s/ 20ms). MR images were analyzed using Paravision 5.1. MRS
results were quantified by using jMRUI to determine absolute metabolite
concentrations (3).
Validation
After the
final time point, brains were removed for confirmation by histology, fungal
load quantification by CFU counts and NMR spectroscopy. Brain metabolites were
assigned and quantified using ex vivo NMR
spectroscopy (400MHz Bruker) as described (4, 5).
Results
Disease
progression was monitored longitudinally with MRI whereby increases in
lesion size correlated with increased fungal load (Figure 1).
Diffusion-weighted MRI showed increased diffusion inside the lesion compared to
normal brain tissue (Figure 2). MRS detected the presence of typical fungal
metabolites like trehalose and mannitol but also lipids inside the cryptococcal
lesions, which was confirmed by ex
vivo NMR spectroscopy (Figure 3). Trehalose concentrations correlated with the fungal
load and a higher in vivo trehalose
production per cell by C. gattii compared to C.
neoformans was revealed (Figure 4). Based on MRS based detectability
limits 104 and 105 cfu per gram tissue were detectable
for C. gattii and C. neoformans, respectively. MRI
was also applied to study infection in the disseminated disease model. Hereby,
a significantly higher amount of brain lesions was detected for C. neoformans infected
animals compared to C. gattii infected animals (Figure 5).Discussion
For the first
time, this study presents a method for non-invasive follow-up and
characterization of cerebral cryptococcomas in a mouse model. While anatomical
MRI and DWI was able to monitor disease progression, MRS identified trehalose as
a characteristic metabolite that can be used for diagnostic purposes. Comparison with MRS-based metabolite profiles from previously
published models of cystic glioblastoma or bacterial abscesses allows
non-invasive distinction between those pathologies from cryptococcoma and
hereby identifies the lesion-causing pathogen (5, 6). In addition, trehalose may act as a
quantitative biomarker for in vivo
assessment of the viable fungal load. Currently,
experiments are in progress to further assess this hypothesis on more cryptococcal
strains and in treatment models.
Our current
findings have great potential to assist in the differential diagnosis of brain
lesions in patients, whereby MRS is a safe, non-invasive and rapid method in
comparison to traditional methods such as CSF sampling or biopsies (7).Acknowledgements
We are thankful for financial support from the European Commission for EU MC ITN TRANSACT 2012 (no. 316679), for the Infect-ERA project CryptoVIEW and from the Flemish Government
FWO for the project G.0869.12N.
References
1. Del Poeta M, Casadevall A. Ten challenges on Cryptococcus and
cryptococcosis. Mycopathologia. 2012, 173:303–10
2. Chen SCA, Meyer W, Sorrell
TC. Cryptococcus gattii infections. Clin.
Microbiol. Rev. 2014, 27:980-1024
3. Osorio-Garcia MI, Sima
DM, Nielsen FU, Himmelreich U, Van Huffel S. Quantification of magnetic resonance
spectroscopy signals with lineshape estimation. J.
Chemometrics 2011; 25: 183–192
4. Croitor-Sava
A, Beck V, Sandaite I, Van Huffel S, Dresselaers T, Claus F, Himmelreich U,
Deprest J. High-resolution 1H NMR spectroscopy discriminates amniotic fluid of fetuses
with congenital diaphragmatic hernia from healthy controls. J.
Proteome Res. 2015, 14: 4502−4510
5. Himmelreich U, Dzendrowskyj
T, Allen C, Dowd S, Malik R, Shehan P, Russell P, Mountford CE, Sorrell TC.:
Cryptococcomas distinguished from gliomas with magnetic resonance spectroscopy
: an experimental rat and cell culture study; Radiology 2001, 220:
122-128
6. Himmelreich U,
Accursso R, Malik R, Dolenko B, Somorjai RL, Gupta RK, Gomes L, Mountford CE,
Sorrell TC. Identification of Staphylococcus
aureus brain abscesses: rat and human studies using H-1 MR Spectroscopy.
Radiology 2005, 236: 261-270
7. Sorrell
TC, Wright LC, Malik R, Himmelreich U.
Application of proton nuclear magnetic resonance spectroscopy to the study of Cryptococcus and cryptococcosis. FEMS Yeast Research 2006, 6: 558-566