The unfolded protein response in the protozoan parasite Toxoplasma gondii features translational and transcriptional control

Abstract

The unfolded protein response (UPR) is an important regulatory network that responds to perturbations in protein homeostasis in the endoplasmic reticulum (ER). In mammalian cells, the UPR features translational and transcriptional mechanisms of gene expression aimed at restoring proteostatic control. A central feature of the UPR is phosphorylation of the α subunit of eukaryotic initiation factor-2 (eIF2) by PERK (EIF2AK3/PEK), which reduces the influx of nascent proteins into the ER by lowering global protein synthesis, coincident with preferential translation of key transcription activators of genes that function to expand the processing capacity of this secretory organelle. Upon ER stress, the apicomplexan parasite Toxoplasma gondii is known to induce phosphorylation of Toxoplasma eIF2α and lower translation initiation. To characterize the nature of the ensuing UPR in this parasite, we carried out microarray analyses to measure the changes in the transcriptome and in translational control during ER stress. We determined that a collection of transcripts linked with the secretory process are induced in response to ER stress, supporting the idea that a transcriptional induction phase of the UPR occurs in Toxoplasma. Furthermore, we determined that about 500 gene transcripts showed enhanced association with translating ribosomes during ER stress. Many of these target genes are suggested to be involved in gene expression, including JmjC5, which continues to be actively translated during ER stress. This study indicates that Toxoplasma triggers a UPR during ER stress that features both translational and transcriptional regulatory mechanisms, which is likely to be important for parasite invasion and development.

Fig 1

Method used to measure changes in mRNA levels and polysome association in response to ER stress. Toxoplasma tachyzoites were purified from host cell monolayers and exposed to 10 μM tunicamycin (+) or DMSO vehicle (−) for 1 h. To measure changes in transcript abundance, total RNA was isolated from the treated parasites, and gene transcript levels were measured by microarray hybridization. Additionally, to measure translational changes as judged by mRNA association with polysomes, lysates were generated from parasite populations treated with tunicamycin or the DMSO vehicle control and subjected to sucrose gradient centrifugation. RNA was isolated from the sucrose fraction, namely, the free ribosomes and monosomes or the polysome fractions, and gene transcript levels were measured by microarray analysis.

Fig 2

UPR genes are induced by multiple ER stress agents. Freshly purified tachyzoites were exposed to 10 μM tunicamycin (TUN), 5 μM A23187, or vehicle (DMSO) for 1 h. Levels for two UPR gene transcripts, a calreticulin family domain member (TGME49_077230) and a putative E3 ubiquitin ligase (TGME49_095670), were measured by qPCR. The histograms represent statistically significant changes (P < 0.01) between ER-stressed samples and the DMSO control, with standard errors (SE) indicated by the error bars.

Fig 3

ER stress represses global translation initiation in Toxoplasma. (A) Equal amounts of lysate from parasites treated with tunicamycin (TUN) or vehicle (DMSO) were separated in a 4 to 12% polyacrylamide gel and transferred for Western blotting with antibodies that specifically recognize TgIF2α phosphorylated at Ser-71 (TgIF2α-P) or total TgIF2α protein. (B) Polysome profiles were generated from parasites exposed to 10 μM tunicamycin (TUN) or no treatment (NT) for 1 h. Fractions containing free RNA, ribosomal subunits, and the 80S monoribosomes were collected and pooled into a single tube (termed free ribosomes and monosomes). A second pool of fractions containing mRNAs engaged with three or more ribosomes (large polysomes) were collected for subsequent microarray analysis.

Fig 4

Polysome association of Toxoplasma gene transcripts is suggested to be reduced in response to ER stress. (A) Histogram representing the number of different gene transcripts (y axis) associated with polysomes (% polysomal RNA) from parasites cultured in the presence of tunicamycin (TUN) or vehicle (black and white bars, respectively). The arrow denotes a subset of mRNAs that are poorly translated in the absence of stress. (B) The percentage of polysomal RNA of the preferentially translated (i.e., enhanced) transcripts (following exposure to tunicamycin) represented in a histogram in the presence of tunicamycin or vehicle (black and white bars, respectively).

Fig 5

Subset of transcriptional regulators is suggested to be preferentially translated in response to ER stress. (A) Pie chart displays GO categories of genes that were suggested to be preferentially translated following treatment with tunicamycin (P < 0.02). (B) A list of preferentially translated AP2 factors and chromatin remodeling factors that are suggested to be preferentially translated during ER stress. The chart indicates enhanced association with polysomes (percent shift to polysome fraction) following tunicamycin treatment and the predicted number of uORFs within the 5′-UTR of each preferentially translated mRNA. (C) Parasites were subjected to ER stress by treatment with 10 μM thapsigargin or control vehicle (DMSO) and incubated with [³⁵S]Met/Cys to radiolabel the synthesized proteins in each sample for 90 min. Two percent of each sample (input) was resolved by SDS-PAGE, followed by autoradiography (second panel from top). Coomassie blue staining indicated equal loading of input protein between the DMSO- and thapsigargin-treated samples (bottom panel). JmjC5-HA was immunoprecipitated from each lysate using anti-HA-coupled agarose beads and subjected to autoradiography after being resolved by SDS-PAGE (top panel). Immunoblotting with antibodies that specifically recognize TgIF2α phosphorylated at Ser-71 (TgIF2α-P) was performed to confirm induction of ER stress by thapsigargin treatment (third panel from the top).

Fig 6

Bioinformatic analyses of Toxoplasma gene transcripts suggested to be preferentially translated in response to ER stress. (A) Comparison of the Kozak sequences for predicted initiation codons from all Toxoplasma genes, as well as those subject to preferential translation or translational repression in response to tunicamycin exposure. (B) Length of the annotated 5′-UTRs (top), 3′-UTRs (middle), and ORF (bottom) for all annotated gene transcripts (gray bar), those suggested to be subject to preferential translation (enhanced) (black bar), and those that are translationally repressed (white bar). An asterisk denotes statistically significant difference (P ≤ 0.05). CDS, coding DNA sequence.