Mass Spectrometry Imaging of In Vitro Cryptosporidium parvum-Infected Cells and Host Tissue

Abstract

Cryptosporidium parvum is a zoonotic-relevant parasite belonging to the phylum Alveolata (subphylum Apicomplexa). One of the most zoonotic-relevant etiologies of cryptosporidiosis is the species C. parvum, infecting humans, cattle and wildlife. C. parvum-infected intestinal mucosa as well as host cells infected in vitro have not yet been the subject of extensive biochemical investigation. Efficient treatment options or vaccines against cryptosporidiosis are currently not available. Human cryptosporidiosis is currently known as a neglected poverty-related disease (PRD), being potentially fatal in young children or immunocompromised patients. In this study, we used a combination of atmospheric pressure scanning microprobe matrix-assisted laser desorption/ionization (AP-SMALDI) mass spectrometry imaging (MSI) and liquid chromatography-tandem mass spectrometry (HPLC-MS/MS) to determine and locate molecular biomarkers in in vitro C. parvum-infected host cells as well as parasitized neonatal calf intestines. Sections of C. parvum-infected and non-infected host cell pellets and infected intestines were examined to determine potential biomarkers. Human ileocecal adenocarcinoma cells (HCT-8) were used as a suitable in vitro host cell system. More than a thousand different molecular signals were found in both positive- and negative-ion mode, which were significantly increased in C. parvum-infected material. A database search in combination with HPLC-MS/MS experiments was employed for the structural verification of markers. Our results demonstrate some overlap between the identified markers and data obtained from earlier studies on other apicomplexan parasites. Statistically relevant biomarkers were imaged in cell layers of C. parvum-infected and non-infected host cells with 5 µm pixel size and in bovine intestinal tissue with 10 µm pixel size. This allowed us to substantiate their relevance once again. Taken together, the present approach delivers novel metabolic insights on neglected cryptosporidiosis affecting mainly children in developing countries.

Keywords: AP-SMALDI; Cryptosporidium parvum; mass spectrometry imaging.

Figure 1

General workflow. Cryptosporidium parvum oocysts, infected intestine and non-infected intestine were obtained from neonatal calves (A). These thick-walled oocysts were then used to generate cell pellets, monolayers and artificially infected intestinal tissues (B). Sections from cell pellets were used to obtain statistically significant markers (C). Markers were visualized with the help of MALDI-MSI in monolayers and host intestinal tissues (D).

Figure 2

Workflow for determination of biomarkers: (A) The left part illustrates an infected section, and the right part the corresponding control sample. Upregulated infection marker LPE (22:4) as [M+H]⁺, m/z 530.3212 ± 5 ppm, in red, each tile is 50 × 50 pixels with a step size of 10 µm. (B) Three technical replicates; LPE (22:4) as [M+H]⁺, m/z 530.3212 ± 5 ppm, each tile is 50 × 50 pixels with a step size of 10 µm. (C) Segment of a heat map generated with Perseus. The color code of the column indicates whether it is a corresponding marker. Red means the signal is significantly increased compared to the other sample group. The red arrow indicates that the visualized marker is one of many markers within the heatmap.

Figure 3

Overview of the annotated lipids, showing abundances of categories of detected HCT-8 markers (fractions of signal numbers (in [%])): (A,B) positive-ion mode; (C,D) negative-ion mode; (A,C) upregulated lipids; (B,D) downregulated lipids—(1) lipid categories, (2) detected lipid classes within the glycerophospholipids category.

Figure 4

Visualization of biomarkers in cell monolayers: MALDI MSI measurements of cell monolayers in positive-ion mode, measured with 5 µm laser focus diameter and step size. The green channel in both images represents the TIC, used for visualization purposes only. The red channel shows the distribution of two different infection markers (±5 ppm mass tolerance) identified earlier by LC-MS/MS measurements of cell pellets or host tissue. (A) Infection marker signal at m/z 766.5720, identified as plasmanyl-PC(O-16:1/20:4) as [M+H]⁺, (B) infection marker signal at m/z 504.3036, identified as LPE(20:3) as [M+H]⁺.

Figure 5

Visualization of infection biomarkers in host tissue: MALDI MSI measurements of C. parvum-infected neonatal bovine intestine in positive-ion mode, measured with 10 µm laser focus diameter and step size in Full Pixel mode. The red and blue channels show the distribution of two infection markers (±5 ppm mass tolerance) identified earlier by LC-MS/MS measurements of cell pellets or host tissue. (A) The green channel represents an ion signal at m/z 756.5513, used for visualization purposes only. Infection marker signal at m/z 504.3036, identified as LPE (20:3) as [M+H]⁺ in red, and m/z 530.3058, identified as LPE (22:4) as [M+H]⁺ in blue. (B) Infection marker signal at m/z 504.3036, identified as LPE (20:3) as [M+H]⁺ in red and m/z 530.3058, identified as LPE (22:4) as [M+H]⁺ in blue. (C) Optical image of the whole intestine section. The area of infection (wrinkles and villi) is outlined in red.

Figure 6

Visualization of biomarkers in experimentally C. parvum-infected neonatal bovine intestinal tissue: MALDI MSI measurements of bovine intestine in positive-ion mode, infected with C. parvum and measured with 12 µm laser focus diameter and step size. The green channel represents an ion signal at m/z 726.5603, used for visualization purposes only. The red and the blue channel show the distribution of two infection markers (±5 ppm mass tolerance) identified earlier by LC-MS/MS measurements of cell pellets. (A) Infection marker signal at m/z 504.3057 (red), identified as LPE (20:3) as [M+H]⁺, (B) optical image of the whole intestine section, (C) zoomed optical image of the measured area. (A2–A6) confirm the interpretation of A for another ten biomarkers. These ten biomarkers were always found in the same area and their distribution was congruent with A. (A2) Infection marker signals at m/z 532.3371 (red) and 558.2955 (blue), annotated by Lipid Maps as LPC (17:0) as [M+Na]⁺ (further annotations possible) and LPC (18:2) as [M+K]⁺ (further annotations possible); (A3) infection marker signals at m/z 540.3059 (red), annotated by Lipid Maps as LPC (18:3) as [M+Na]⁺ (further annotations possible) and 576.3282 annotated by Lipid Maps as LPS (20:0) as [M+Na]⁺ in blue (further annotations possible); (A4) infection marker signals at m/z 640.3220 (red) annotated by Lipid Maps as PS (24:3) as [M+Na]⁺ (further annotations possible) and m/z 618.3399 (blue) annotated by Lipid Maps as PS (24:3) as [M+H]⁺; (A5) infection marker signals at m/z 656.3557 (red) annotated by Lipid Maps as PS (27:5) as [M+H]⁺ and m/z 657.3591 (blue) annotated by Lipid Maps as PI (21:0) as [M+H]⁺; (A6) infection marker signals at m/z 521.3428 (red) annotated by Lipid Maps as DG (25:2;O2) as [M+Na]⁺ and m/z 520.3394 (blue) annotated by Lipid Maps as LPC (18:2) as [M+H]⁺.