Proteomic and transcriptomic analyses of early and late-chronic Toxoplasma gondii infection shows novel and stage specific transcripts
The protozoan pathogen Toxoplasma gondii has the unique ability to develop a chronic infection in the brain of its host by transitioning from the fast growing tachyzoite morphology to latent bradyzoite morphology. A hallmark of the bradyzoite is the development of neuronal cysts that are resilient against host immune response and current therapeutics. The bradyzoite parasites within the cyst have a carbohydrate and protein-rich wall and a slow-replication cycle, allowing them to remain hidden from the host. The intracellular, encysted lifestyle of T. gondii has made them recalcitrant to molecular analysis in vivo.
Here, we detail the results from transcriptional and proteomic analyses of bradyzoite-enriched fractions isolated from mouse brains infected with T. gondii over a time course of 21 to 150 days. The enrichment procedure afforded consistent identification of over 2000 parasitic peptides from the mixed-organism sample, representing 366 T. gondii proteins at 28, 90, and 120 day timepoints. Deep sequencing of transcripts expressed during these three timepoints revealed that a subpopulation of genes that are transcriptionally expressed at a high level. Approximately one-third of these transcripts are more enriched during bradyzoite conditions compared to tachyzoites and approximately half are expressed at similar levels during each phase. The T. gondii transcript which increased the most over the course of chronic infection, sporoAMA1, shows stage specific isoform expression of the gene.
We have expanded the transcriptional profile of in vivo bradyzoites to 120 days post-infection and provided the first in vivo proteomic profile of T. gondii bradyzoites. The RNA sequencing depth of in vivo bradyzoite T. gondii was over 250-fold greater than previous reports and allowed us to identify low level transcripts and a novel bradyzoite-specific isoform of sporoAMA1.
KeywordsToxoplasma gondii Cysts Bradyzoites RNA-seq Proteomics
Human foreskin fibroblast
Liquid chromatography-tandem mass spectrometry
Phosphate buffered saline
Principal component analysis
High-throughput RNA sequencing
Spliced Transcripts Alignment to a Reference program
Transcripts per million
The protozoan parasite Toxoplasma gondii is one of the most successful eukaryotic pathogens, infecting approximately a quarter of the world’s population . One of the drivers of its success as a pathogen is the ability to develop a chronic infection in the brain of any warm-blooded host. Within the brain, the parasite undergoes a transformation from the fast-growing tachyzoite form to the slow-growing bradyzoite form . Bradyzoites remain shielded from the host immune system by changing surface protein expression and developing a cyst wall [2, 3]. T. gondii cysts cannot be cleared from the host and current therapeutics are ineffective against chronic infection. This persistence becomes particularly problematic if the host becomes immunocompromised. As the immune pressure wanes, bradyzoites transition back into tachyzoites and start to rapidly replicate, which can lead to encephalitis if left untreated . Despite its clinical importance, the biology of bradyzoites and the molecular components of the cyst structure are not well understood.
T. gondii undergoes a major transformational change during the switch from a tachyzoite to a bradyzoite. The transition has been characterized by both microarray analysis of cysts 21 days post-infection (DPI)  and deep sequencing of cysts 28 DPI , which identified ~ 500–800 genes changing between the different stages in vivo. This data has provided a good snapshot of cysts during early chronic infection; however, bradyzoites have a dynamic growth pattern during chronic infection. Parasite growth measured by ICM3 protein abundance showed a cyclic growth pattern of bradyzoites out to 8 weeks post-infection . These results suggest that cysts are not static but are continually changing and responding to the environment.
The documented changes in cysts throughout chronic infection has led us to question whether the global transcriptional profile of bradyzoites changes over the course of chronic infection. For this study, we isolated T. gondii cysts from the infected brain tissue of mice 21, 28, 90, and 120 DPI using dextran centrifugation. This purification method resulted in enough parasite material to analyze both the global RNA and protein profile of in vivo bradyzoites. While the proteins present on the cyst wall have been recently analyzed , this current study is the first proteomic analysis of T. gondii bradyzoite parasites and includes the transcription profile of T. gondii bradyzoites throughout the time course of chronic infection.
Enrichment of bradyzoite cysts decreases host protein interference
StringTie program predicts novel T. gondii bradyzoite transcripts
RNA-seq reads per sample. The number of total reads sequenced from each sample and the number of which mapped to the T. gondii genome
Most identified proteins had high gene expression
Transcriptomic analysis of purified bradyzoites results in a higher number of differentially expressed genes
T. Gondii transcription remains static after 3 months post-infection
Comparing purified bradyzoites throughout chronic infection, 59 transcripts (from 50 genes) change at 90 DPI and 109 transcripts (from 93 genes) change by 120 DPI relative to 28 days (Fig. 6a, Additional files 7 and 8). No transcripts change significantly between 90 and 120 DPI. Of the differentially expressed transcripts, 51 transcripts were differentially expressed at both 90 and 120 DPI: 15 increasing in abundance and 36 decreasing (Fig. 6b). Only 8 transcripts change specifically at 90 days, whereas 58 transcripts are specific for 120 days: 18 increasing and 40 decreasing in expression. 33 of the differentially expressed transcripts have an average TPM value < 10 among all purified bradyzoite timepoints. This group of 33 included the transcript for SAG1 which had an average TPM value of 16 at 28 days, decreasing 3.5-fold and 4.8-fold at 90 and 120 DPI, respectively. This low level of bradyzoite expression is compared to the average TPM values of 8000 and 13,000 for tachyzoites and acute whole brain tissue respectively. Enriched GO terms among the genes that are reduced in expression during late-chronic infection relate to protein phosphorylation and protein modification. Specifically at 120 DPI, GO terms related to polysaccharide processes were enriched, suggesting much of the developmental changes are complete by 28 DPI.
Although the host is largely asymptomatic during chronic T. gondii infection, bradyzoites are actively replicating within cysts . Transcriptional activity of parasites during early chronic infection is high; however, the abundance of host material has hindered global analyses of T. gondii proteins and late-chronic transcripts. Our ability to collect tissue cysts and separate them from host material has allowed us the resolution to analyze low abundant T. gondii transcripts and to reproducibly identify many T. gondii peptides.
Dextran enrichment greatly increased the relative T. gondii peptide abundance. Unpurified infected brain found 3.4% of total unique peptides mapping to T. gondii, which increased to 30% for many of the enriched samples (Fig. 2a). Although some samples yielded fewer T. gondii peptides, comparing the most consistent samples between timepoints identified 2040 unique peptides in all samples, mapping to 366 different proteins (Fig. 2). While pepsin digestion removed host material and release bradyzoites from the cysts, many cyst wall proteins were likely lost due to the procedure. Despite the pepsin treatment, some known cyst wall markers, such as CST1 and GRA2, were seen [18, 19]. Also due to the enrichment procedure, we were unlikely to identify host proteins that were associated with the parasites or the cyst wall. It is interesting to note that two host transporter proteins associated with antigen processing (TAP1 and TAP2) were identified in every sample by MS/MS. Transcripts for these proteins are highly expressed in the brain during chronic infection , suggesting the TAP proteins may not be interacting directly with the cyst. In addition, 31 host proteins identified in our datasets were also identified from in vitro T. gondii cyst wall fractions (Additional file 2, right column) , suggesting there may be an association with the cyst wall among these proteins.
For many of the identified T. gondii proteins, the RNA expression levels were high during chronic infection (Fig. 5). These genes include the bradyzoite specific markers BAG1 and LDH2 [20, 21], as well as constitutively expressed genes such as Act1 and TubA1. Interestingly, 100 of the 366 proteins had low level gene expression during chronic infection. These results likely indicate either short-lived transcripts or long-lived proteins. Alternatively, as bradyzoites within tissue cysts are heterogeneously replicating in vivo , some proteins may be produced by a subset of metabolically active parasites or those undergoing reactivation into the tachyzoite state.
Cyst isolation has allowed us to identify transcripts from in vivo parasites at a depth similar to parasites in vitro. This depth allowed examination of alternative transcripts produced by bradyzoites from early to late chronic infection. Comparing timepoints with two biological replicates from each, over 100 transcripts were differentially expressed from early to late chronic infection. No transcripts change between 90 and 120 DPI, suggesting that, at the global level, the parasites are at a steady state by 90 DPI. This result is consistent with many host transcripts decreasing expression from 28 to 90 DPI and remaining similar from 90 to 180 DPI .
One of the transcripts which had the largest change from early to late infection was for sporoAMA1. The StringTie program identified two transcripts for sporoAMA1 (MSTRG.4390.2 and TGME49_315730-t26_1). Both transcripts were significantly upregulated at 90 and 120 DPI from 28 DPI, and MSTRG.4390.2 had the largest increase at near 7-fold increased at both late-chronic timepoints compared to 28 DPI. SporoAMA1 is the sporozoite specific paralog of the AMA1 protein and a major component of the moving junction during entry into the host cell [14, 15]. The N-terminal region of sporoAMA1 includes a binding domain to the T. gondii protein RON2 [14, 22] to help form the moving junction complex. The annotated gene for sporoAMA1 includes 9 exons (transcript TGME49_315730-t26_1); however, the bradyzoite reads do not align until exon 7 (Fig. 7b). In comparison, RNA-seq data from T. gondii infected cat intestines showed that the reads to sporoAMA1 do not align until after exon 4 (Fig. 7). Microarray evidence of this same region from sporulated oocysts , show reads aligning to all nine exons of sporoAMA1 (Additional file 9). This result suggests that sporoAMA1 has stage specific isoforms, with the expression of the shortest isoform starting in late chronic infection and longer isoforms expressed during different stages of the cat cycle.
The functions of the shortened sporoAMA1 isoforms are unknown. For each of these transcripts, an in frame start codon is present 150–300 base pairs after the reads start aligning to the genome, highlighting that these shortened transcripts may be translated. It is unlikely that either of these two short forms plays a role in invasion as the RON2 binding region is not present in either isoform (Additional file 9, Figure S1B). In support of a non-invasion function for sporoAMA1, there are almost no transcripts sporoRON2 (TgME49_265120) transcripts expressed either in bradyzoite cysts or sexual stages in the cat intestine. The shortest isoform would produce an 8 kDa protein during late chronic infection that contains the cytoplasmic region and the conserved serines/threonines. The sequential phosphorylation of these residues is essential for Plasmodium falciparum to efficiently invade red blood cells . Thus the 8 kDa bradyzoite isoform may be phosphorylated and playing a regulatory role. However, no sporoAMA1 peptides were found in any of the six chronic infection cyst samples generated in this study. While this result could indicate that the short 8 kDa isoform is unstable or degraded in the sample preparation process, it could also indicate that the approximately 400 bp RNA molecule is a long non-coding RNA. Perhaps this RNA functions similar to the Herpes encoded Latency Associated Transcripts (LATs), which are essential to maintain viral latency and reactivation . The generation of an antibody to this shared cytoplasmic domain will help determine if the short isoforms are translated as well as their potential function.
Further future studies are also needed to determine the host interactions with the bradyzoite cyst during long-term chronic infection. Upwards of 90% of the RNA-seq reads obtained in this study aligned to the T. gondii genome, whereas the 75% of the peptides aligned to the mouse. This difference is likely due to the instability of the host mRNA and the stability of the host proteins during sample processing. The enrichment of certain host proteins may indicate their stable interaction with the cysts. We have not yet examined if any identified host proteins are directly associated with the parasites or the cyst wall. The membrane associated host transporters TAP1 and TAP2 were identified in each sample by MS/MS. This result may be a consequence of their high expression in the brain during infection , or it may indicate a direct interaction with the cyst. T. gondii interaction with TAP1 and TAP2 could block antigen processing to help hide the cyst during chronic infection or it might assist in nutrient acquisition across the cyst wall. Our dataset had 31 host proteins (Additional file 2, right column) that were also identified in in vitro T. gondii cyst wall fractions , suggesting that these host proteins may be an association with the cyst wall. Examination of these host proteins potentially interacting with cysts will be an important area of future research as it may indicate how cysts can reside in their warm-blooded host for so long.
The isolation of in vivo bradyzoites has allowed us to look deeply at both the transcript and protein content of the parasites. This depth has allowed us to identify novel bradyzoite specific isoforms, one of which is sporoAMA1. Over the course of chronic infection, we have found that the global transcriptional profile remains constant after the first month of chronic infection. This result is similar to the protein profile, in which most of the proteins identified are found in each timepoint. This dataset is the first global protein profile of in vivo bradyzoites. Together with the RNA-seq data, they provide a better understanding of the transcriptional and translational activity of T. gondii throughout infection.
Mouse infection with T. gondii
The ME49 strain of T. gondii was cultured as tachyzoites in human foreskin fibroblast (HFF) cells. Six to ten week-old CBA/J (JAX) were infected intraperitoneally with 1 × 104 tachyzoites to establish chronic toxoplasmosis.
Collection of tissue bradyzoites
At their respective timepoint, the mice were euthanized by carbon dioxide inhalation in a manner approved by the institutional animal ethics committee, and the whole brains of infected CBA/J mice at 21, 28, 90, 120, and 150 days post-infection were harvested and homogenized in phosphate buffered saline (PBS) using mortar and pestle. Several brain tissues were homogenized together as a single replicate to obtain enough parasite material. Homogenized tissue was serially passed through increasing gauge needles (18, 20, 22 respectively 5x each) and filtered using 150 μm filter. The volume of the homogenate was increased to 40 mL using PBS and centrifuged for 10 min at 2000 x g. Pellet was resuspended in 3 mL of 20% Dextran (150,000 average molecular weight; Alfa Aesar) per brain and separated into tubes so that 3 or fewer brains were in each. Samples were centrifuged for 10 min at 400 x g to pellet the cysts. Host material and dextran were removed, and the cyst pellet was resuspended with 5 μg/mL pepsin in a 1% NaCl solution at pH 2.1 to release the bradyzoites from the cyst (1 mL per brain). Cysts were digested for 5 min at 37 °C, then neutralized with a 1x volume of 1% Na2CO3. Bradyzoites were centrifuged for 10 min at 250 x g and resuspended in PBS then counted using hemocytometer.
Bradyzoites were preserved in 6 M GuHCl prior to processing. Protein was isolated by boiling at 100 °C for 5 min before precipitating in 90% methanol. Protein was pelleted by centrifugation at 10,000 x g for 5 min, the supernatant was discarded, and the pellet was resuspended in lysis buffer (8 M urea, 100 mM tris pH 8, 40 mM 2-chloroacetamide, 10 mM 2-carboxyethyl phosphine). Samples were incubated for 10 min at RT before diluting to < 2 M urea with 50 mM tris pH 8. Trypsin (Promega, Madison, WI) was added at a mass ratio of 1:100::trypsin:protein and the sample was incubated overnight at RT with gentle rocking. A second bolus of trypsin (1:200::trypsin:protein) was added the next morning and incubated for an additional hour before desalting with reversed phase cartridges (StrataX, Phenomenex, Torrence, CA). Peptide yield was calculated by bicinchoninic acid (BCA) assay (Thermo Peirce, Rockford, IL). Two microgram of each sample was analyzed by LC-MS/MS on a Q-IT-OT Fusion Lumos mass spectrometer (Thermo Fisher, San Jose, CA) equipped with a 75/360 μm inner/outer diameter fused silica capillary column with a laser pulled electrospray tip packed with BEH C18 (130 Å pore, 1.7 μm particle size, 35 cm, Waters Corp, Milford, MA). Eluting peptides entered into the mass spectrometer following positive mode electrospray ionization. MS1 survey scans were performed in the orbitrap (240 K resolution, AGC target 1e6, 100 ms maximum injection time). MS2 analysis of HCD generated (25% NCE) product ions were performed in the ion trap (Rapid resolution, AGC target 4e4, 18 ms maximum injection time). Monoisotopic precursor selection and dynamic exclusion (60 s) were enabled. Raw data were searched against a concatenated database of mouse and T. gondii proteins protein sequences through the MaxQuant computational platform (version 126.96.36.199) operating the Andromeda search algorithm . Precursor and fragment mass tolerances were set to 50 ppm and 0.5 Da, respectively. Carbamidomethylation of cysteine was imposed as a fixed modification and oxidation of methionine was allowed as a variable modification. Protein inference was performed from peptide spectrum matches (PSMs) based on the rules of parsimony . Search results were filtered to a 1% false discovery rate at both the PSM and protein levels.
Generation of RNA and RNA-seq
Trizol was added to bradyzoites for the preservation of RNA, and the RNA extracted using phenol/chloroform separation and isopropanol precipitation. Samples include brain enriched bradyzoites at 28 DPI (n = 2), 90 DPI (n = 2), and 120 DPI (n = 2). Ten nanogram of total RNA was processed using the SMART-Seq v4 Ultra Low Input RNA Kit (Takara), using 7 rounds of cDNA amplification. One nanogram of cDNA was indexed using the Nextera XT Index Kit (Illumina). RNA libraries were multiplexed and run across 2 lanes for paired end, 125 base pair sequencing by Illumina HiSeq 2500 at the University of Wisconsin Biotechnology Center.
Tachyzoite RNA was processed from the ME49 parasites grown in human foreskin fibroblasts (HFF) cells. Just before lysis of the parasite vacuole, infected HFF cells were scraped in trizol and RNA extracted as stated above. RNA libraries were prepared using a TruSeq RNA Library Prep kit v2 set A (Illumina) and sequenced using paired-end sequencing as stated above.
The sequencing reads were processed to remove low quality reads (Trimmomatic, v0.39 ). The tachyzoite, purified bradyzoite, and whole brain acute and chronic  sequencing reads were aligned to the T. gondii ME49 genome (ToxoDB.org) using the Spliced Transcripts Alignment to a Reference program (STAR, v2.7.0f, ). The following STAR parameters were changed from the default: mismatch max = 2, intron min = 70, and intron max = 100,000. The STAR output bam file from each sample was fed into the StringTie program (v1.3.6, ), along with the current T. gondii annotation file (ToxoDB.org) and using a minimum transcript coverage value of 150. The sequencing reads were remapped using the new gtf file output from StringTie. Transcript abundance was estimated using RNA-seq by Expectation-Maximization (RSEM v1.2.31, ). Fold change was calculated using the geometric mean values between samples for transcripts with at least 10 RSEM gene counts in any one sample. The rlog transformed DESeq2 values were used for the PCA plot and PCA calculated using “plotPCA” function of the DESeq2 package. GO term enrichment was perfomed using ToxoDB. Terms were limited to biological processes with a p-value < 0.05 and false discovery rate < 0.05.
We would like to thank Patrick Cervantes for providing RNA-seq data from tachyzoite culture.
All authors contributed to the design of the experiments. ALG and LJK produced and analyzed the RNA-seq data. GMW and JJC performed the mass spectrometry data collection and interpretation of proteomic analyses. All authors read and approved the final manuscript.
This work was funded by the National Institutes of Health (P41 GM108538, JJC; 1R01AI144016–01, LJK; T32AI55397, ALG; T32GM008349, GMW). Funding bodies had no role in the design of the study and collection, analysis, and interpretation of data and in writing the manuscript.
Animals were treated in compliance with the guidelines set by the Institutional Animal Care and Use Committee (IACUC) of the University of Wisconsin School of Medicine and Public Health (protocol #M005217), which adheres to the regulations and guidelines set by the National Research Council. The University of Wisconsin is accredited by the International Association for Assessment and Accreditation of Laboratory Animal Care.
Consent for publication
The authors declare that they have no competing interests.
- 7.Watts E, Zhao Y, Dhara A, Eller B, Patwardhan A, Sinai AP. Novel Approaches Reveal that Toxoplasma gondii bradyzoites within Tissue Cysts Are Dynamic and Replicating Entities In Vivo. mBio. 2015;6. https://doi.org/10.1128/mBio.01155-15.
- 16.Ramakrishnan C, Maier S, Walker RA, Rehrauer H, Joekel DE, Winiger RR, et al. An experimental genetically attenuated live vaccine to prevent transmission of Toxoplasma gondii by cats. Sci Rep. 2019;9. https://doi.org/10.1038/s41598-018-37671-8.
- 17.Fritz HM, Buchholz KR, Chen X, Durbin-Johnson B, Rocke DM, Conrad PA, et al. Transcriptomic analysis of Toxoplasma development reveals many novel functions and structures specific to Sporozoites and Oocysts. PLoS One. 2012;7. https://doi.org/10.1371/journal.pone.0029998.
Open AccessThis article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated.