Autophagy is a physiological process of self-digestion of nonfunctional organelles and/or macromolecules, limiting redundant and damaged cellular components. This biochemical pathway can be selective or nonselective and guarantees eukaryotic homeostasis through the turnover and recycling of target cellular structures, which are pivotal events during cell growth and differentiation (Menna-Barreto, 2019; Abdrakhmanov et al., 2020; Klionsky et al., 2021). Despite nonselective characteristics during different conditions, such as starvation, numerous selective autophagic examples, including mitophagy, reticulophagy, and xenophagy (degradation by autophagy of mitochondria, endoplasmic reticulum, and pathogens, respectively), have been reported (Abdrakhmanov et al., 2020).

In pathological situations, including in protozoan infections, autophagy is increased to recover the cellular balance (Kirkegaard et al., 2004; Swanson, 2006). However, continuous induction of this pathway can culminate in autophagic cell death (Levine and Yuan, 2005). In protozoa, the autophagic phenotype is often induced in vitro by a great variety of drugs with different well-known mechanisms of action (Menna-Barreto et al., 2009c; Menna-Barreto, 2019).

Molecularly, autophagy is a highly conserved process regulated by autophagy-related genes (ATGs) previously identified in Saccharomyces cerevisiae (Klionsky et al., 2003), and their orthologs were subsequently described in all eukaryotes (Klionsky et al., 2021). Currently, three autophagic types are presented in the literature: macroautophagy (also called autophagy), microautophagy, and chaperone-mediated autophagy (CMA; Menna-Barreto, 2019; Abdrakhmanov et al., 2020).

Macroautophagy is characterized by the presence of double membrane organelles named autophagosomes that will address cellular material to be degraded in lysosomes. In a well-controlled process dependent on Atg proteins, a membrane structure (phagophore) surrounds damaged organelles or macromolecules, giving rise to autophagosomes (Shintani and Klionsky, 2004; Alvarez et al., 2008). The triggering of the process depends on the serine/threonine protein kinase TOR (target of rapamycin), a nutritional availability sensor, and Atg6 (beclin 1 in mammals), which is a phosphatidylinositol 3-kinase (PI-3K; Duszenko et al., 2011).

Unlike macroautophagy, autophagosomes are absent in microautophagy. The cellular material that will be degraded is engulfed by invagination of the lysosomal membrane. As demonstrated by electron microscopy, lysosomes full of small vesicles in their lumen are called multivesicular bodies. Unfortunately, this autophagic type is poorly studied due to the absence of specific markers (Menna-Barreto, 2019; Klionsky et al., 2021).

The most selective autophagic type is CMA, where signal pentapeptides (KFERQ, VDKFQ or QREFK) present in target proteins bind to cytosolic chaperones. The chaperone substrate binds to the lysosomal receptor LAMP-2A, promoting channel formation derived from receptor dimerization. Target molecules are degraded in the lysosomal lumen after entering through this channel (Duszenko et al., 2011).

Since the first description more than half a century ago, ultrastructural characterization remains a valuable tool for autophagic phenotype detection, allowing autophagosome identification without the use of specific markers (Menna-Barreto et al., 2009c; Menna-Barreto, 2019). More recently, knockdown or knockout of autophagic components strategies have also been commonly used. The gold-standard method for monitoring autophagic flux is the detection of Atg8 (LC3 in mammals) by morphological investigation after immunostaining microscopy (presence of LC3 puncta) and/or by immunoblotting (detection of LC3-I and LC3-II; Klionsky et al., 2021). In the present work, we reviewed different aspects of the protozoan autophagy exacerbation, discussing the possible role for drug resistance/susceptibility of these pathogens.

In protozoa, autophagy was first reported in T. brucei in 1977 by Vickerman and Tetley (1977) based on ultrastructural evidence. To date, parasites under starvation and/or subjected to other stress conditions commonly present autophagic features, including an increase in autophagosome number, multivesicular bodies, and myelin-like structures (Figures 1A–D, 2A–F, 3A,B; Benchimol, 1999; Maia et al., 2007; Corrêa et al., 2009; Menna-Barreto et al., 2009c; Besteiro et al., 2011; Ghosh et al., 2012; Koh et al., 2015; Picazarri et al., 2015; Souto et al., 2016; Nguyen et al., 2017b; Hernández-García et al., 2019; Araujo-Silva et al., 2021; Wu et al., 2021; Zhang et al., 2021). Interestingly, concentric membrane and myelin-like structures share morphological similarities to the phagophore described in yeast and mammals. The endoplasmic reticulum (ER), the main source of the phagophoric membrane, is frequently found surrounding degraded subcellular structures, especially in stressed parasites (Figures 3A,B; Maia et al., 2007; Corrêa et al., 2009; Martins-Duarte et al., 2009; Menna-Barreto et al., 2009a; Duszenko et al., 2011; Busatti et al., 2013; Hernández-García et al., 2019).

Regarding the molecular machinery, the autophagic pathway is well conserved among eukaryotes. Many ATG homologs have been identified in pathogenic protozoa, but some components are lacking or differ from those found in yeast. In trypanosomatids, genes involved in phagophore elongation and degradation of autophagosome cargo were detected by in silico approaches, including the complete Atg8 conjugation system (Atg3, Atg4, Atg7, and Atg8; Figure 4; Herman et al., 2006). Similar data were obtained in Entamoeba and T. vaginalis, where the Atg8 conjugation system was described, and the Atg12 complex is lacking (Picazarri et al., 2015; Hernandéz-García et al., 2019; Huang et al., 2019). In Giardia, bioinformatic analysis revealed the TOR, S6K1, PI3K, Atg1, Atg16, Atg7, Atg8, and Atg18 genes (Cernikova et al., 2020; Wu et al., 2021). Toxoplasma gondii seems to have well-conserved autophagic machinery, presenting several putative orthologs of yeast Atgs, including proteins of the Atg1, TOR, and PI3K complexes and Atg9 and Atg8/Atg12 systems (Lévêque and Besteiro, 2016; Besteiro, 2017). On the other hand, the Atg repertoire in Plasmodium did not reveal the presence of Atg24, TOR kinase, Atg9, Atg6, and Atg16, but the other Atgs found in T. gondii are also present (Lévêque and Besteiro, 2016; Besteiro, 2017).

The first functional analysis of Atgs in protozoa was performed in L. major in 2006 (Besteiro et al., 2006). The role of Atg8 and Atg4 was assessed in T. cruzi in 2008, and its activity was found to be directly related to the differentiation process of the parasite. The isoforms of Atg8 (TcAtg8.1 and TcAtg8.2) and Atg4 (Atg4.1 and Atg4.2) were investigated, showing the localization of Atg8.1 in autophagosomes of parasites under nutritional deprivation (Alvarez et al., 2008). Other Atg8 isoforms were identified in Leishmania spp. and T. brucei (Atg8, Atg8A, Atg8B, and Atg8C) and are regulated by the same two Atg4 isoforms (Rigden et al., 2005; Koopmann et al., 2009; Williams et al., 2009). On the other hand, in trypanosomatids, the Atg12 conjugation system is incomplete, and ATG5, ATG10, and ATG12 are lacking (Herman et al., 2006; Kiel, 2010). In E. histolytica, no Atg regulation was described during starvation, but EhAtg8 has a function in driving phagosome acidification (Picazarri et al., 2015). Atg8 is upregulated during the G. intestinalis differentiation process, and its expression induces parasite encystation (Wu et al., 2021). Moreover, T. vaginalis expresses two Atg8 genes, TvAtg8a and TvAtg8b, both of which retain a functional domain of Atg8. Moreover, TvAtg8a is more highly expressed (Huang et al., 2019).

The presence of genes encoding TOR kinases (TOR1 and TOR2) and their respective complexes (TORC1 and TORC2) was also detected in pathogenic trypanosomatids; however, their functions were not fully investigated. Following treatment with rapamycin, a classical autophagic inducer, an increased number of autophagosomes derived from TORC2 inhibition in T. brucei were observed, blocking parasite replication (Barquilla et al., 2008; Denninger et al., 2008). In Giardia and Trichomonas, rapamycin increased the number of Atg8 puncta, which may regulate parasite growth and differentiation into cysts (Hernandéz-García et al., 2019; Wu et al., 2021). A putative ortholog of mammalian TOR kinase is present in the T. gondii genome (TgTOR, TGME49_116440). To assess whether TgTOR is a component of the amino acid sensing mechanism, the parasites were treated with rapamycin, mimicking amino acid deprivation. Increasing drug concentrations induced a dose-dependent accumulation of Atg8 puncta, suggesting an increase in autophagic activity in the parasites (Besteiro et al., 2011) and a dose-dependent fragmentation of the parasite’s mitochondria (Ghosh et al., 2012).

As previously mentioned, autophagy seems to be particularly important for the success of the protozoan life cycle, especially during the differentiation steps (Alvarez et al., 2008; Besteiro et al., 2011; Lévêque et al., 2015; Picazarri et al., 2015; Smith et al., 2021; Wu et al., 2021). Trypanosoma cruzi epimastigotes are submitted to limited nutrients after their migration to the triatominae rectum, a crucial step for the occurrence of metacyclogenesis (Alvarez et al., 2008; Duszenko et al., 2011). Recently, Losinno et al. (2021) described the involvement of acidocalcisomes during this differentiation process. Autophagic vesicles released from this organelle fuse to reservosomes, contributing to reservosomal acidification and consequently increasing the hydrolytic activity of the cysteine protease cruzipain, resulting in parasite self-proteolysis (Losinno et al., 2021). Further experiments on the role of this self-processing and cruzipain activation in parasite differentiation and infection in both invertebrate and vertebrate hosts must be performed. In Leishmania spp., metacyclogenesis is also regulated by autophagy (Besteiro et al., 2006; Williams et al., 2006). Atg4.2 deletion blocked autophagic flux, leading to the accumulation of lipidated Atg8 and a decrease in the percentage of promastigotes under differentiation (Schoijet et al., 2017) as well as during L. mexicana amastigogenesis (Williams et al., 2006). In this parasite, megasomes are lysosome-like organelles that play a crucial role during the differentiation process. The deletion of two megasomal cysteine peptidases (CPA and CPB) also led to the impairment of amastigogenesis, and a high number of autophagosomes were found in mutant parasites (Williams et al., 2006; Kiel, 2010). At least for trypanosomatids, the involvement of autophagy in the regulation of virulence and infectivity in vertebrate hosts is clear (Besteiro et al., 2006), but the related molecular processes still need to be elucidated.

Due to the remarkable differences among vertebrate and invertebrate hosts, it is common sense that nutrient availability, temperature, and pH, among other environmental conditions influence protozoan metabolism, including ATP production (Kiel, 2010). Interestingly, the regulation of some of these metabolic adaptations occurs by autophagy. Li and He (2014, 2017) showed that autophagic flux participates in acidocalcisome acidification and that the blockage of organelle biogenesis also impairs the autophagic pathway in T. brucei. In G. intestinalis, Atg8 interacts with other proteins, such as myeloid leukemia factor (MLF) and FYVE domains, participating in the protein metabolism pathway and processing mitosomal and encystation proteins (Wu et al., 2021). In Entamoeba, EhAtg8 is involved in the incorporation, scavenging and intracellular trafficking of nutrients (Picazarri et al., 2015). On the other hand, glucose restriction in T. vaginalis induces autophagy associated with TvAtg8 expression and autophagosome-like formation (Huang et al., 2019).

In apicomplexan parasites, Atg8 exerts unique functions. In addition to its cytosolic or, in the case of stressful conditions, vesicular location, it also localizes to the apicoplasts of Toxoplasma and Plasmodium (Kitamura et al., 2012; Kong-Hap et al., 2013; Tomlins et al., 2013; Lévêque et al., 2015). In T. gondii, Atg8 is responsible for the proper segregation of the organelle by tethering it to the centrosomes during the replication of the parasite (Lévêque et al., 2015). In Plasmodium, the protein contributes to apicoplast formation and maintenance (Tomlins et al., 2013). Given that the apicoplast is important for the synthesis of isoprenoid precursors and fatty acids, which are essential for parasite survival (Sheiner et al., 2013), Atg8 has an indirect but important role in apicomplexan metabolism.

Autophagy is also involved in the control of mitochondrial functionality and phospholipid homeostasis in protozoa (Besteiro et al., 2006; Williams et al., 2012; Vanrell et al., 2017). Trypanosoma cruzi epimastigotes submitted to an acidic environment or nutritional deprivation showed intense autophagic activity, ROS generation, and mitochondrial remodeling. These conditions reproduce the triatomine rectum environment, which is crucial for differentiation to the metacyclic form, suggesting a direct correlation between autophagy and mitochondrial remodeling during the process. On the other hand, insect blood digestion promotes transitory alkaline conditions. Our group also demonstrated that alkaline medium led to early exacerbation of autophagy and mitochondrial impairment. These features recovered over time, indicating a survival mechanism to increase the autophagic flux for the removal of damaged structures (Pedra-Rezende et al., 2021). Leishmania major promastigotes deficient in Atg5, which are not capable of forming autophagosomes, also presented a remarkable decrease in their virulence in vitro and in vivo. Strong mitochondrial dysfunction was also observed in these mutants together with the increased phosphatidylethanolamine (PE) content and ROS production, suggesting a conjugation of mitochondrial PE to Atg8 for autophagosome biogenesis (Williams et al., 2012). In T. gondii, starvation leads to mitochondrial dysfunction with consequent impairment of host cell invasion capacity and accumulation of Atg8 puncta (Besteiro et al., 2011; Ghosh et al., 2012). The use of mutated and conditional knockout parasites revealed that Atg8 and Atg3 are essential for Atg8 lipidation and autophagosome formation (Besteiro et al., 2011).

Regarding the selectivity of the autophagic pathway, selective degradation of T. brucei glycosomes (peroxisome-like) was proposed during the differentiation from bloodstream trypomastigotes into procyclic forms. This target organelle is crucial for parasite survival and is involved in bioenergetic metabolism and antioxidant defenses, corroborating the pivotal role of pexophagy in this trypanosomatid (Herman et al., 2008; Brennand et al., 2015).

In general, autophagic phenotypes are very conserved, at least in pathogenic protozoans. Ultrastructural evidence (autophagosome formation, myelin-like structures and ER surrounding organelles) and Atg expression are frequently assessed in parasites under autophagic stimuli (Figures 5A–D, 6).

Following the recent guidelines for monitoring autophagy, autophagic phenotypes in protozoa after treatment with drugs were assessed by different techniques, including Western blotting and electron and fluorescence microscopy (Klionsky et al., 2021). Innumerous compounds from distinct classes induced different autophagic phenotypes (to different degrees) in protozoan parasites.

One of the most frequent morphological features of autophagy identified in these parasites is the presence of concentric membrane structures or myelin-like structures. Ultrastructural studies noted this structure in protozoa treated with all classes of compounds (Vannier-Santos et al., 1995; Benchimol, 1999; Braga et al., 2005; Granthon et al., 2006; Uzcátegui et al., 2007; Menna-Barreto et al., 2009a,b; Carvalho et al., 2010; Bombaça et al., 2018, 2021; Martínez-García et al., 2018; Araujo-Silva et al., 2021).

The increased number of autophagosomes, a determinant of macroautophagy occurrence, is another very recurrent phenotype detected in treated protozoa (Braga et al., 2005; Granthon et al., 2006; Uzcátegui et al., 2007; Dos Santos et al., 2010; Schurigt et al., 2010; Besteiro et al., 2011; Sengupta et al., 2011; Fernandes et al., 2012; Ghosh et al., 2012; Veiga-Santos et al., 2013; Souto et al., 2016; Nguyen et al., 2017b; Scariot et al., 2017; Lim et al., 2018; Martínez-García et al., 2018; Hernandéz-García et al., 2019; Huang et al., 2019; de Paula et al., 2020; Silva et al., 2020; Nishi et al., 2021; Zhang et al., 2021).

As previously mentioned, Atg8 is considered the gold standard method for autophagic evaluation. Treatment of L. donovani with cryptolepine induced an increase in the number of Atg8 puncta (Sengupta et al., 2011). Similar findings were observed in T. brucei treated with L-leucine methyl ester and bacteriocin AS-48 (Koh et al., 2015; Martínez-García et al., 2018). In T. gondii, starvation or treatment with the drug monensin or some antimalarial compounds induces a time-dependent accumulation of Atg8 puncta (Besteiro et al., 2011; Lavine and Arrizabalaga, 2012; Varberg et al., 2018). Similar results were obtained when parasites were treated with the ER stress-inducing agents dithiothreitol, brefeldin A, or tunicamycin (Nguyen et al., 2017a). Alternatively, the autofluorescent compound monodansyl cadaverine (MDC) is typically employed to assess autophagy, despite the nonspecificity of this marker. Several groups observed an increase in MDC labeling after the treatment of parasites with different compounds (Sengupta et al., 2011; Fernandes et al., 2012; Lazarin-Bidóia et al., 2013; Veiga-Santos et al., 2013; Desoti et al., 2014; Scariot et al., 2017, 2019; de Paula et al., 2020; Bortoleti et al., 2021; Nishi et al., 2021; Zhang et al., 2021).

ER profiles surrounding cytoplasmic structures and organelles are also commonly detected in treated protozoa. Fernandes et al. (2012) demonstrated that treatment with triazolic naphthoquinone led to the appearance of ER in close contact with reservosomes of T. cruzi epimastigotes. Especially in this case, the authors described the Golgi as an alternative source of the phagophoric membrane (Fernandes et al., 2012). In L. amazonensis, elatol and amiodarona induced a similar phenotype and pronounced swelling of the mitochondrion and destabilization of the plasma membrane (Dos Santos et al., 2010; Macedo-Silva et al., 2011). In T. gondii, treatment with thiolactomycin analogs induced dramatic morphological changes in parasite shape and intracellular organization, including abnormal amounts of concentric membranes expanded throughout the parasite cytoplasm, possibly representing ER profiles (Martins-Duarte et al., 2009).

To evaluate the specific role of autophagy in a drug mechanism of action, one of the most common experimental protocols is the treatment of parasites with autophagic inhibitors, such as wortmannin and 3-MA (Pasquier, 2016; Klionsky et al., 2021). In T. cruzi, the use of these inhibitors completely abolished the trypanocidal effect of naphthoimidazoles, reinforcing autophagy as part of the mechanism of cell death induced by these compounds (Menna-Barreto et al., 2009a). Similar results were obtained with T. gondii, where preincubation with 3-MA efficiently abrogated the effects of monensin on mitochondrial fragmentation (Lavine and Arrizabalaga, 2012). Interestingly, 3-MA was not able to prevent antimalarial compounds from the medicines for malaria venture malaria box MMV2- or MMV3-induced mitochondrial disruption, indicating that autophagy may occur downstream of mitochondrial fragmentation via PI3K-independent mechanisms (Varberg et al., 2018).

Protozoan diseases still represent a significant challenge, demanding specific public health strategies, especially in low-income countries. Many of them are considered neglected diseases that impair physical and cognitive development, limiting individual productivity and resulting in economic issues (World Health Organization, 2015; Centers for Disease Control and Prevention, 2021d). Different parasite forms from a great variety of subpopulations also contribute to the increase in the drug resistance of pathogens, reinforcing the necessity of a continuous search for alternative compounds with anti-protozoan activity. During preclinical tests, cellular, molecular, and biochemical information about the targets of the novel compounds is critical for the characterization of drug safety and specificity.

Indeed, a detailed description of the mechanisms of action is still lacking, even for clinical drugs. As an example, the trypanocidal actions of benznidazole and nifurtimox, which are employed for Chagas disease treatment, are still not completely understood more than 50 years after their discovery (Menna-Barreto and De Castro, 2016). Furthermore, many new compounds with potent anti-protozoan activity do not exhibit predicted biological effects (Chatelain and Ioset, 2011; Gaspar et al., 2015), emphasizing the importance of the identification of effective molecular targets. Despite multidisciplinary efforts involving high-throughput screenings to discover novel candidates for the treatment of protozoan diseases (Annang et al., 2015; Peña et al., 2015), few studies have elucidated their mechanisms of action. One of the main reasons is the lack of efficient and practical tools to assess molecular drug targets in vitro and in vivo.

Electron microscopy was extensively employed in the first identification of primary drug targets in treated parasites, such as organelles and cellular structures (Corrêa et al., 2009; Menna-Barreto and De Castro, 2016). Ultrastructural analysis may allow inferences about the action of the compounds. Many studies also suggest the triggering of cell death processes as part of the drug mechanism (Maia et al., 2007; Corrêa et al., 2009; Menna-Barreto et al., 2009a,b; Menna-Barreto, 2019). Despite all ultrastructural evidence, the occurrence of programmed cell death in protozoan parasites is very controversial due to the absence of precise information about biochemical and molecular events, especially those involved in regulatory processes (Menna-Barreto, 2019). Until the convincing identification of the executioners, these phenotypes should be classified as unregulated processes or incidental necrosis (Proto et al., 2013; Menna-Barreto, 2019).

The treatment of different protozoa with distinct classes of drugs led to a convergent autophagic phenotype. Ultrastructural lesions in organelles, such as mitochondria, reservosomes, or hydrogenosomes, are usually closely associated with the appearance of endoplasmic reticulum profiles and an increase in the number of autophagosomes (Menna-Barreto et al., 2009a,b; Hernandéz-García et al., 2019). Together with the formation of myelin-like structures, these are the most recurrent autophagy-related phenotypes described (Figure 7). Based on morphological findings, the autophagic role in cell death has been postulated, but a description of the regulatory events involved is lacking (Bera et al., 2003; Baehrecke, 2005; Menna-Barreto et al., 2009a). It is well known that autophagy plays a pivotal role in homeostasis maintenance in eukaryotes, including protozoan parasites. Drugs impair different molecular pathways, causing the loss of the balance in the turnover of crucial cellular structures. Such an imbalance promotes an exacerbation of nonselective autophagy, culminating in nonspecific damage (Baehrecke, 2005). This is a reasonable explanation for the appearance of the autophagic phenotype in treated parasites regardless of the drug and/or mechanism of action involved (Menna-Barreto et al., 2009b; Menna-Barreto, 2019). In summary, autophagy represents survival machinery responsible for the removal of cellular structures damaged by compounds.

RM-B conceived the work and drafted the manuscript. YP-R and RM-B wrote trypanosomatids subjects. RM and IM wrote the apicomplexan part. VM wrote about anaerobic parasites. All authors contributed to the article and approved the submitted version.

This work was supported by CNPq, CAPES, FAPERJ, and FIOCRUZ.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.