Acanthamoeba castellanii exhibits intron retention during encystment

Intron retention (IR) refers to the mechanism of alternative splicing in which an intron is not excised from the mature transcript. IR in the cosmopolitan free-living amoeba Acanthamoeba castellanii has not been studied. We performed an analysis of RNA sequencing data during encystment to identify genes that presented differentially retained introns during this process. We show that IR increases during cyst formation, indicating a potential mechanism of gene regulation that could help downregulate metabolism. We identify 69 introns from 67 genes that are differentially retained comparing the trophozoite stage and encystment after 24 and 48 h. These genes include several hypothetical proteins. We show different patterns of IR during encystment taking as examples a lipase, a peroxin-3 protein, an Fbox domain containing protein, a proteasome subunit, a polynucleotide adenylyltransferase, and a tetratricopeptide domain containing protein. A better understanding of IR in Acanthamoeba, and even other protists, could help elucidate changes in life cycle and combat disease such as Acanthamoeba keratitis in which the cyst is key for its persistence.


Introduction
Alternative splicing (AS) is the process through which primary transcripts can be modified in different arrangements to produce functionally different mature mRNA. AS is responsible for most of the complexity of the eukaryotic proteome (Blencowe 2006). Intron retention (IR) refers to the form of AS where an intron remains in the mature mRNA instead of being spliced out (Grabski et al. 2021). IR is a complex and evolutionarily conserved mechanism of gene regulation, and genes that go through this process are highly regulated (Jacob and Smith 2017;Schmitz et al. 2017). In many cases, the study of IR has been neglected or misinterpreted as noise. However, next-generation sequencing techniques have highlighted the importance of such mechanisms in physiological and pathological processes of eukaryotic organisms (Jacob and Smith 2017).
IR is a mechanism of controlling and enhancing gene expression in eukaryotes (Vanichkina et al. 2018). It is the most prevalent mode of alternative splicing in non-animal eukaryotes (McGuire et al. 2008;Tapial et al. 2017;Grau-Bové et al. 2018). IR is also linked to the down-regulation of gene expression via the nonsense-mediated decay (NMD) pathway (Lykke- Andersen and Jensen 2015;Brogna et al. 2016;Wong et al. 2016).
IR has not been extensively studied in the cosmopolitan free-living amoeba Acanthamoeba castellanii or other organisms of the genus. However, one study compared exon skipping and intron retention in 65 eukaryotic transcriptomes including Acanthamoeba. The goal was to observe the relationship between genome architecture and alternative splicing events. This study showed that IR is more prevalent in 64 of the species than exon skipping, including Acanthamoeba (Grau-Bové et al. 2018). It also included other amoebae such as Naegleria gruberi and Dictyostelium discoideum (Grau-Bové et al. 2018 (Clarke et al. 2013;Chelkha et al. 2018). The genome has at least 56,920 annotated ORF and 15,655 genes reported in AmoebaDB (Amos et al. 2022). Acanthamoeba genes have an average of 6.2 introns (Roy 2006;Clarke et al. 2013).
Other forms of AS in Acanthamoeba have been reported. The SBDS gene (for the Shwachman-Bodian-Diamond syndrome protein) is upregulated during encystation and phagocytosis where it presented two diverse patterns of expression (Wang et al. 2021). Additionally, the cytochrome P450 monooxygenase provides resistance to polyhexamethylene biguanide and other drugs, thanks in part to AS processes (Huang et al. 2021). Acanthamoeba is capable of forming an extremely resilient cyst when conditions are not favorable. In this article, we present the first study related to IR in Acanthamoeba encystment. We show that IR is an important component in gene regulation during encystment. We compared retained introns of Acanthamoeba trophozoites and encysting organisms after 24 and 48 h. We demonstrate that the number of retained introns increases during encystment related to a decrease in metabolism related to the dormant cyst.

RNA sequencing
Encystment was induced in confluent cultures, changing axenic media to NEM. Samples were taken at 0 (also referred to as trophozoite stage), 24, and 48 hours after inducing encystment. RNA extraction was performed using RNeasy Mini Kit (QIAGEN) following the manufacturer's instructions. RNA quality and purity were assessed using the QUBIT RNA BR Assay Kit (Thermo Fischer Scientific). cDNA libraries were prepared for an automated TruSeq mRNAseq (next-generation shotgun sequencing) from total RNA. Sequencing was performed with a HiSeq-4000 75PE by Edinburgh Genomics.
A. castellanii Neff strain reference genome (FASTA and GTF) was obtained from ENSEMBL Protists (Clarke et al. 2013;Kersey et al. 2017). Quality control was assessed using FASTQC (Andrews 2010). STAR was used to index the genome and align the reads (Dobin et al. 2013). Differential expression analysis was performed using EdgeR (Robinson et al. 2009).

Intron retention analysis
Six samples in total were analyzed, two belonging for each timepoint: 0, 24, and 48 h. iREAD and IRFinder software were used to identify IR events (Middleton et al. 2017;Li et al. 2020). The intersection of the results from both software tools was used to identify retained introns. The previously mapped RNAseq analysis was compared with the intron annotation file created using GTFtools (Li 2018). In both programs, default settings were used to identify and classify an IR event by a minimum retention of the intron in 10% of the transcripts (IR ratio ≥ 0.1). Differential expression of introns was calculated using EdgeR (Robinson et al. 2009). A direct comparison of differential expression of genes and retained introns was performed to assess if the increase of IR events was not caused by the increase of the gene. The comparison is expressed as a Log2IRratio (encystment vs trophozoite) obtained by dividing the IR ratio mean during encystment (either at 24 h or 48 h) and the IR ratio mean at 0 h. AmoebaDB was used when required to identify specific genes (Amos et al. 2022).

Results
The expression of three genes related to the encystment process was compared at the different timepoints (Hirukawa et al. 1998;Dudley et al. 2008;Moon et al. 2009). This was used to prove that encystment was under way at a molecular level and can be observed in (de Obeso Fernandez del Valle 2018). Differential expression of the cyst specific protein 21, encystation mediating serine proteinase, and mitogenactivated protein kinase can be observed in Fig. 1.
IR increased during encystment of Acanthamoeba as shown after 24 and 48 h in Table 1. The total number of retained introns identified across the three treatments was 987 and included 891 genes or slightly over 5% of the genes in the genome. The number of retained introns had an increase of over 30% after 48 h of encystment (633 to 825). However, some of the retained introns are shared in the three samples. There are 226 unique IR events after 24 h and 282 events after 48 h in contrast to the original timepoint of 0 h. There are also 169 introns being differentially retained from 24 to 48 h. We found 110 IR events that are lost during the encystment process after 24 h.
After identifying the retained introns, the expression of genes and retained introns during encystment was compared to verify that the overrepresentation of IR events was not caused by an increase of gene overexpression since overexpression of a gene would lead to overrepresentation of a retained intron. We considered a Log2IRratio (encystment vs trophozoite) above 1 or below − 1 as the threshold for differentially retained introns. After 24 h, we found 33 instances in which the IR event had a Log2IRratio (encystment vs trophozoite) above 1, and 9 below − 1. After 48 h, but not at 24 h, there were 23 IR events that were overrepresented in relation to the gene and only 4 that were underrepresented. A heatmap of the differentially retained introns is shown in Fig. 2. These results are also shown in Fig. 2 as scatterplots comparing proportionality of genes and retained introns. Note that genes showing the lowest change in expression were associated with higher changes in IR during encystment, compared to genes showing greater log2FC, most of which showed minimum changes in IR levels. Of the overrepresented IR events after 24 h, two of them (ACA1_070640 and ACA1_315050) had two introns retained.  Several introns are retained across the organism. In Fig. 3, we show the retention of introns in six selected genes with some of the highest overrepresentation of retained introns. To select these genes, hypothetical proteins were ignored. These examples include a lipase, a peroxin-3 protein, an Fbox domain containing protein, a proteasome subunit, a polynucleotide adenylyltransferase and a tetratricopeptide domain containing protein.
A full list of the 67 genes with differentially retained introns alongside the corresponding differential expression data can be found in the supplementary materials.

Effect of IR in specific proteins
Gene expression patterns have been studied in life cycle events of different protozoa and in Acanthamoeba for under certain conditions (Moon et al. 2011;Maciver et al. 2019). IR events during Acanthamoeba encystment were overrepresented in 67 genes. Of these, we focused on six genes with some of highest changes in IR ratio that were not annotated as hypothetical proteins. We found that three Fig. 3 Six genes with retained introns during encystment. In the bottom part of each RNA-seq density plot, in blue, you can observe normal architecture of the gene representing exons with thicker lines. The differentially retained introns are highlighted in gray. A plus sign ( +) next to the gene_ID represent 5′ to 3′ direction from left to right, while a minus sign ( −) represents the opposite direction. To the left, you can see the name of the gene as obtained from AmoebaDB. RNA seq density plots were generated using Spark (Kurtenbach and William Harbour 2019) of those genes code for proteins related to degradation of proteins and lipids. The encystment process consists of a first stage characterized by degradation and autophagy (Moon et al. 2008a, b;Leitsch et al. 2010). A second stage follows with the production of cyst specific proteins and structures (Leitsch et al. 2010). The transcript for an FBox domain containing protein (ACA1_106350) retained the tenth intron of the gene. FBox machinery is important in defense and protection, such as cyst formation. Legionella pneumophila is capable of surviving inside Acanthamoeba with the assistance of the high jacking of the ubiquination machinery (Price et al. 2009). The Fbox machinery participates in the mediation of ubiquination of proteins to be degraded by the 26S proteasome. The proteasome subunit alpha type 6 (ACA1_260270) which is part of the 26S proteasome, also presented IR events, although the inclusion of the intron in the mRNA did not generate a premature stop codon. This shows a potential source for potential variability as it generated two different amino acid sequences between mRNA with and without the intron.
A lipase (ACA1_326860) presented retained introns during the trophozoite stage interestingly presented a particular expression of transcripts (Fig. 3). The lipase has 17 exons, of which 14 were underrepresented in transcripts during trophozoite stage before the retained intron. This shows a possible upregulation of the mature transcript during encystment even with a small overrepresentation of the retained intron, and a clear example of IR as gene regulation mechanism during encystment. In contrast, the peroxin-3 protein (ACA1_151280) has a retained intron during encystment and potentially affects peroxisome biogenesis and therefore might alter oxidative process such as the ones related to lipids. Peroxisomes, and therefore peroxin-3, have been identified in several species of Acanthamoeba and other free-living amoebae and are also involved in protein import, which could be affected during encystment (González-Robles et al. 2020).
A polynucleotide adenylyltransferase (ACA1_096220) presented one retained intron which could probably affect the polyadenylation of other mRNAs. In this case, the downregulation of the intron occurred after 48 h. Additionally, the tetratricopeptide repeat domain containing protein (ACA1_150720) helps modulate interactions with different proteins, which in this case is near impossible to establish. However, this can be related to the downregulation and decrease in metabolic function during the latent stage. Moreover, an ankyrin repeat protein (ACA1_315050; seen in the supplementary materials) presented two differentially retained introns and reinforces the hypothesis as this protein family is the most abundant protein-protein interaction motif (Al-Khodor et al. 2010). These discoveries highlight the lengths we have to cover to understand the AS events, transcriptome variation and their effect on the proteome during Acanthamoeba encystment, and almost certainly similar organisms with life cycles that include drastic life cycle events such as encystment.

Effect of IR at the species level
Several early developed protists or unicellular eukaryotes have very few introns and include the genera Giardia, Entamoeba, and Trichomonas (McGuire et al. 2008). Apicomplexans, such as Toxoplasma gondii and Plasmodium falciparum, have shown that IR might be a widespread mechanism for differentiation and could partly explain the increase in IR in Acanthamoeba (Lunghi et al. 2016). Some eukaryotic species have a high retention of introns (Ner-Gaon et al. 2004). For example, IR can affect around 80% of protein coding genes in humans (Middleton et al. 2017). Other mammals have shown the presence of such events between 50 and 75% of their genes (Braunschweig et al. 2014). Although IR is not the main route of alternative splicing in most animals, microscopic animal parasites such as Echinococcus present high volumes of IR retention (33-36% of genes presented some sort of alternative splicing), highlighting the diversity that exists in IR and alternative splicing across organisms (Liu et al. 2017). Additionally, the potato cyst nematode Globodera rostochiensis uses IR to regulate chorismate mutase which plays an important role in pathogenesis (Lu et al. 2008). IR has even been linked to multicellularity through Creolimax fragantissima (de Mendoza et al. 2015). The phenomenon has been shown to happen in all kingdoms and viruses that present introns in their genome (McGuire et al. 2008). Acanthamoeba showed small IR at 5% in comparison to other organisms. Nevertheless, several of these genes might play a crucial role in encystment, survival and differentiation. Encystment in Acanthamoeba leads to a vegetative cyst stage that helps protect against stressors, leading to a decrease in metabolic pathways and functions. In Saccharomyces cerevisiae, IR has been linked to starvation and stress response, where spliced introns "clutter" the spliceosome inhibiting its function and playing a role in transcripts regulation and yeast growth (Morgan et al. 2019;Parenteau et al. 2019).
Studies in other protists or potentially infectious organisms have been recorded. For example, IR has been studied in Trypanosoma evansi, the agent of "surra." IR was the main form of alternative splicing in T. evansii, where 819 of 820 genes analyzed presented it. These genes mainly have functions related to RNA processing and cellular components. This occurs mainly during the parasitic stage (Zheng et al. 2019). Entamoeba histolytica, the causative agent of amoebiasis, has shown that its main route of alternative splicing is IR (Valdés et al. 2014). IR has also been linked to virulence of E. histolytica through the splicing factor U2AF84 (Gonzalez Blanco et al. 2021).
IR studies are a novel way of analyzing RNAseq data that could provide new insights in Acanthamoeba and other microorganisms of interest (Zheng et al. 2020). The study of IR provides important insight in regulating gene expression that could increase the understanding of the biology of different organisms and the development and persistence of infection (Lunghi et al. 2016). For example, IR events are regulated by methylation processes creating a link between the comprehension of epigenetic regulation and IR (Wong et al. 2017). IR has been demonstrated to play an important role in several diseases; therefore, it should not be overlooked (Wong et al. 2016). IR studies and the insight provided in gene expression might open the possibility for new therapies against Acanthamoeba diseases, since encystment has been identified as the key to persistence (Lorenzo-Morales et al. 2015). Understanding IR and other ways of alternative splicing offer insight into the transcriptional landscape of these organisms (Monteuuis et al. 2019). As new data is being generated through the increased use of next generation sequencing and other high-throughput techniques, processes like IR and alternative splicing are valuable resources for understanding the biology of organisms such as Acanthamoeba.
Further research is required to better understand this process and the impact in general of alternative splicing in Acanthamoeba and other pathogenic protists. This should help combat different diseases caused by amoebae such as Acanthamoeba keratitis. Additionally, a more in-depth study of IR events in different conditions and microorganisms might help comprehend the process.
Supplementary information.
Funding The study was supported by Mexico's National Council of Science and Technology (CONACYT) and a University of Edinburgh College of Medicine and Veterinary Medicine PhD Studentship (to A. d. O. F. d. V.). Edinburgh Genomics is partly supported through core grants from NERC (R8/H10/56), MRC (MR/K001744/1), and BBSRC (BB/J004243/1).

Data Availability
The information of the discussed genes can be found in the supplementary materials. Everything else can be provided upon request.

Declarations
Ethics approval Not applicable.

Consent to participate Not applicable.
Consent for publication Not applicable.