Pre-clinical Imaging of Invasive Candidiasis Using ImmunoPET/MR

Abstract

The human commensal yeast Candida is the fourth most common cause of hospital-acquired bloodstream infections, with Candida albicans accounting for the majority of the >400,000 life-threatening infections annually. Diagnosis of invasive candidiasis (IC), a disease encompassing candidemia (blood-borne yeast infection) and deep-seated organ infections, is a major challenge since clinical manifestations of the disease are indistinguishable from viral, bacterial and other fungal diseases, and diagnostic tests for biomarkers in the bloodstream such as PCR, ELISA, and pan-fungal β-D-glucan lack either standardization, sensitivity, or specificity. Blood culture remains the gold standard for diagnosis, but test sensitivity is poor and turn-around time slow. Furthermore, cultures can only be obtained when the yeast resides in the bloodstream, with samples recovered from hematogenous infections often yielding negative results. Consequently, there is a pressing need for a diagnostic test that allows the identification of metastatic foci in deep-seated Candida infections, without the need for invasive biopsy. Here, we report the development of a highly specific mouse IgG3 monoclonal antibody (MC3) that binds to a putative β-1,2-mannan epitope present in high molecular weight mannoproteins and phospholipomannans on the surface of yeast and hyphal morphotypes of C. albicans, and its use as a [⁶⁴Cu]NODAGA-labeled tracer for whole-body pre-clinical imaging of deep-seated C. albicans infections using antibody-guided positron emission tomography and magnetic resonance imaging (immunoPET/MRI). When used in a mouse intravenous (i.v.) challenge model that faithfully mimics disseminated C. albicans infections in humans, the [⁶⁴Cu]NODAGA-MC3 tracer accurately detects infections of the kidney, the principal site of blood-borne candidiasis in this model. Using a strain of the emerging human pathogen Candida auris that reacts with MC3 in vitro, but which is non-infective in i.v. challenged mice, we demonstrate the accuracy of the tracer in diagnosing invasive infections in vivo. This pre-clinical study demonstrates the principle of using antibody-guided molecular imaging for detection of deep organ infections in IC, without the need for invasive tissue biopsy.

Keywords: MRI imaging; computed tomography scanning; invasive candidiasis; invasive fungal disease; monoclonal antibody; positron emission tomography.

FIGURE 1

Specificity of mAb MC3 determined by Enzyme-Linked Immunosorbent Assay tests of surface washings containing water-soluble antigens from Candida species and related and unrelated yeasts and molds. ELISA absorbance values at 450 nm for soluble antigens from Candida species in the CTG clade and Saccharomycetaceae, the emerging human pathogen Candida auris (inset), and other related and unrelated yeasts, yeast-like fungi and molds of clinical significance. Wells were coated with 60 μg protein/mL buffer. Bars are the means of three biological replicates ± standard errors and the threshold absorbance value for the detection of antigen is ≥0.1 (indicated by line on graph).

FIGURE 2

Characterisation of the MC3 antigen, its epitope and spatial distribution in yeast and pseudohyphal cells. (A) ELISA absorbance values at 450 nm for purified C. albicans cell wall mannan. Each point is the mean of three replicates ± standard errors. (B) Western immunoblot with MC3 using purified C. albicans cell wall mannan (Mannan) and soluble antigens extracted from immunogen of C. albicans SC5314 (SC5314). Wells were loaded with 1.6 μg of protein. M_r denotes molecular weight in kDa. (C) Western immunoblot with MC3 using soluble antigens in surface washings of the C. albicans wild-type strain NGY152 and the mannan mutant strains. Wells were loaded with 1.6 μg of protein. The asterisk (^∗) indicates the low molecular weight band of ∼12 kDa, assumed to be PLM, which is absent in extracts from the Δpmr1 mutant. (D) ELISA absorbance values at 450 nm for culture fluids from the C. albicans wild-type strains SC5314 and NGY152 and the mannan mutant strains. Wells were coated with 60 μg protein/mL buffer. Each bar is the mean of three biological replicates ± standard errors and bars with the same letter are not significantly different at p < 0.001. (E) Corresponding western immunoblot with MC3 using culture fluids from the C. albicans wild-type strains SC5134 and NGY152 and mannan mutant strains. Wells were loaded with 1.6 μg of protein. The asterisk (^∗) indicates the low molecular weight band of ∼12 kDa which is only present in culture fluids of the mutant Δmnt1,2. (F) ELISA absorbance values at 450 nm for heat-treated antigens from C. albicans SC5314. Wells were coated with 60 μg protein/mL buffer. Each bar is the mean of three biological replicates ± standard errors and bars with the same letter are not significantly different at p < 0.001. There was no significant effect of heating on MC3 recognition, indicating that its antigen is heat-stable. (G) ELISA absorbance values at 450 nm for periodate-treated antigens (white bars) and control-treated antigens (gray bars) extracted from lyophilised cells of C. albicans SC5134. Wells were coated with 60 μg protein/mL buffer. Each bar is the mean of three biological replicates ± standard errors and bars with different letters are significantly different from one another at p < 0.001. The significant reduction over time of MC3 recognition of periodate-treated antigens shows that the antibody’s epitope is carbohydrate and contains vicinyl hydroxyl groups. (H–M) Photomicrographs of C. albicans SC5314 blastospores and hyphae immunostained with MC3 (I,K) or TCM control only (M) followed by anti-mouse polyvalent Ig fluorescein isothiocyanate conjugate. Bright-field images of C. albicans SC5134 blastospores (H) and blastospores and hyphae (J,L). (I–M) Same fields of view as (H–L), but examined under epifluorescence. Scale bars = 4 μm. Note the intense staining of cell walls of blastospores and hyphae with MC3 (I,K), but lack of staining with TCM only (M), demonstrating specific recognition of C. albicans yeast and hyphal cells by the antibody.

FIGURE 3

(A) In vivo biodistribution of [⁶⁴Cu]NODAGA-MC3 in PET/MR imaging at 48 h p.i. Coronal Maximum Intensity Projection (MIP), MR and fused PET/MR images of PBS-treated (control) mice, and C. albicans-infected and C. auris-infected mice injected with the tracer. The acquired images reveal specific uptake of the tracer in the left and right kidneys of C. albicans-infected mice, but not in the kidneys of PBS-treated (control) mice. The specificity of the tracer was further demonstrated using a strain of C. auris which, while reactive with MC3 in vitro using ELISA (Figure 1), was non-infective in the i.v. challenge model. Here, uptake of the tracer in the kidneys was similar to the uptake found in the PBS control mice. (B) Quantification of the PET images for the left and right kidneys of the three different groups of mice at 3, 24, and 48 h p.i. Significantly higher uptake of the [⁶⁴Cu]NODAGA-MC3 tracer is seen in the left and right kidneys of C. albicans-infected mice compared to PBS-treated (control) and C. auris-infected animals at 24 and 48 h p.i. Data are expressed as mean ± SD %ID/cc. Group differences were examined using one-way ANOVA, followed by post hoc Tukey–Kramer, with significant differences at ^∗p < 0.05. (C) Ex vivo biodistribution at 48 h p.i. Significantly higher uptake of the [⁶⁴Cu]NODAGA-MC3 tracer in the left and right kidneys of C. albicans-infected mice, compared to the kidneys of corresponding PBS-treated (control) and C. auris-infected animals, confirms the PET/MR imaging and in vivo biodistribution data at 48 h p.i. In addition to increased uptake of the tracer in the kidneys, significantly higher uptake was also shown in the colons, muscles, and brains of C. albicans-infected mice compared to C. auris-infected and PBS-treated mice, significantly higher uptake in the spleens of C. albicans-infected mice compared to C. auris-infected mice, but significantly reduced uptake in the blood of C. albicans-infected mice compared to PBS-treated (control) animals. Data are the results of N = 4–5 animals per group and expressed as the mean ± SD %ID/g.