Abstract

Introduction

Proteolytic enzymes, or proteases, catalyse the cleavage of peptide or isopeptide bonds in proteins. Proteases likely arose early in evolution as merely digestive enzymes necessary for protein degradation and the generation of amino acids in primitive organisms (López-Otín and Bond, 2008). However, apart from being blunt protein distractors, proteases can also act akin to sharp suture scissors by dissecting target polypeptide chains at precise locations into smaller polypeptides, in a process called ‘limited proteolysis’. The result of limited proteolysis is the formation of new post-translationally modified protein species or proteoforms with neo-N- and/or -C-termini (Weng et al., 2019; Liu et al., 2020), which may gain, lose, or show altered biochemical activity compared with their precursor proteins. The actual role and importance of limited proteolysis appears to be greatly underestimated, as it may in fact control myriad biological processes, ranging from seed development and photosynthesis in plants to apoptotic cell death and blood clotting in animals. Acting in a cooperative manner, digestive and limited proteolysis sustain continuous modification of the proteome necessary for normal development and fitness (Minina et al., 2017).

While up to 3% of eukaryotic genes encode proteases that collectively represent the protease degradome of a given organism (Quesada et al., 2008), only a small fraction of proteases are mechanistically understood. This is especially true for plants, where most of the knowledge on proteases is derived from the model organism Arabidopsis thaliana (van der Hoorn, 2008). Still, the majority—approximately 86%—of the protease holotypes in Arabidopsis remain uncharacterized biochemically and are known only as sequences (Rawlings, 2020). The monophyletic clade of green plants, consisting of both land plants and green algae, is divided into two evolutionary lineages, the Chlorophyta (chlorophytes) and Streptophyta (streptophytes), that diverged over a billion years ago (Bremer, 1985; Morris et al., 2018). Chlorophyte proteases are largely unknown territory, which limits our mechanistic comprehension of the role of proteolytic mechanisms in plant biology and hampers integrated understanding of how proteolysis evolved.

With these thoughts in mind, we turned our attention to proteolytic enzymes of Chlamydomonas reinhardtii (hereafter Chlamydomonas), an ancient unicellular model organism belonging to the Chlorophyta that shares ancestral traits not only with higher plants but also with animals. Thus, Chlamydomonas has been an invaluable reference for studies in the areas of light perception, photosynthesis, chloroplast development, ciliary formation, cell motility, and the cell cycle, among others (Harris, 2001). Whereas the haploid genome of Chlamydomonas expedites functional analysis of genes and proteins of interest, this research is often hindered by the low efficiency of nuclear transgene expression (Salomé and Merchant, 2019). Although the establishment of insertional mutant libraries (Li et al., 2016; Cheng et al., 2017) has aided in some cases, mutants for most genes are still unavailable due to insertion in non-coding regions or a lack of insertion. However, more recent developments of cloning technology (Crozet et al., 2018; Emrich-Mills et al., 2021) and genome editing in Chlamydomonas (Picariello et al., 2020), along with the emerging understanding of transgene silencing mechanisms (Neupert et al., 2020) in Chlamydomonas, should facilitate the functional analysis of various genes of interest in this model alga.

In multicellular plants, a search for protein substrates whose cleavage by a specific protease would be central to its biological function is significantly hampered by the difficulties in standardizing protein terminomics experiments, due to the complex cellular heterogeneity within the samples used for proteome isolation. Along with the inherent differences in cell physiology and morphology among diverse cell types and tissues that contribute to the variation in the expression, compartmentalization, and activation of a specific protease, there is also responsiveness to both experimental conditions and proteome isolation treatments to consider (Demir et al., 2018; Dissmeyer et al., 2018). In contrast, Chlamydomonas represents a beneficial model system for protease substrate identification by virtue of its limited number of cell types and the possibility of synchronizing the cell cycle and response to external factors within a homogenous cell population.

We argue that the simple unicellular life cycle of Chlamydomonas, coupled with the ease of cell phenotyping and the above-mentioned genetic and technical advantages, provide a powerful paradigm for systematic studies of proteases and proteolytic pathways. Here, we share a genome-wide survey and classification of Chlamydomonas proteases, compare proteases in Chlamydomonas and in Arabidopsis by catalytic type, and classify Chlamydomonas proteases based on their relatedness to major taxonomic groups. Furthermore, we summarize and discuss the available evidence for biochemical regulation and/or physiological roles of proteases in Chlamydomonas.

Protease degradome of Chlamydomonas versus Arabidopsis

According to the nature of the nucleophile acting during catalysis, proteases are divided into seven major catalytic types: (i) aspartic, (ii) cysteine, (iii) glutamate, (iv) serine, and (v) threonine proteases, as well (vi) metalloproteases and (vii) asparagine peptide lyases. In addition, there are proteases classified as ‘unknown’ and of a ‘mixed catalytic type’ (Rawlings, 2020). While glutamate proteases and asparagine peptide lyases are found only in pathogenic fungi, bacteria, and archaea, the other types of proteases are spread through all domains of life (Jensen et al., 2010; Rawlings et al., 2011).

We used the major database of proteolytic enzymes, MEROPS, release 12.2 (https://www.ebi.ac.uk/merops/index.shtml) and the Chlamydomonas genome annotation, version 5.5 (https://phytozome.jgi.doe.gov/pz/portal.html#!info?alias=Org_Creinhardtii), to retrieve all protease genes and classify the encoded proteins in comparison with Arabidopsis proteases from both MEROPS and TAIR (https://www.arabidopsis.org/). Altogether, 352 and 764 protease-encoding genes constitute the Chlamydomonas and Arabidopsis degradomes, respectively (Tables S1 and S2 at Zenodo; https://zenodo.org/record/5347045#.YS46It-xWUk, Zou and Bozhkov, 2021).Table 1 shows the distribution of Chlamydomonas and Arabidopsis proteases among catalytic types. One striking difference between the two species is the predominant occurrence of metalloproteases in Chlamydomonas, accounting for ~35% of its degradome (versus only ~15% in Arabidopsis). Notably, more than one-third of the Chlamydomonas metalloproteases (45 out of 124, or ~13% of the whole degradome) belong to the gametolysin family M11, which is absent in Arabidopsis (Fig. 1) and other higher plants. The main role of gametolysin is algal cell-wall degradation to release gametes of both mating types as a necessary prelude to gamete fusion (Matsuda, 2013).

Open in a separate window

Fig. 1.

Hierarchical distribution of protease-encoding genes among different clans and families in Chlamydomonas and in Arabidopsis. U, uncharacterized.

By contrast, there are relatively more aspartic and serine proteases in Arabidopsis than in Chlamydomonas (Table 1), that is, ~12% and ~41% versus ~2.8% and ~30% of the respective degradomes. This is mainly due to the extensive A1 (pepsin-like; 69 homologues) and S33 (prolyl aminopeptidases; 76 homologues) families in Arabidopsis, which are represented by only a few homologues in Chlamydomonas (Fig. 1). In higher plants, the considerable expansion of A1 and S33 proteases may correlate with the specific developmental stages or a plethora of biotic and abiotic stresses that a plant must withstand given its sessile lifestyle (in comparison with motile aquicolous unicellular algae such as Chlamydomonas). For example, a typical pepsin-like A1 protease, constitutive disease resistance 1 (CDR1), elevates the plant resistance to Pseudomonas syringae (Xia et al., 2004), whereas a member of the same family, promotion of cell survival 1 (PCS1), is essential for embryogenesis (Ge et al., 2005). Another example is the prolyl aminopeptidase AtPAP1, which underlies Arabidopsis tolerance to salt stress and drought (Sun et al., 2013).

Another interesting difference between the degradomes of the two species is the complete absence of certain protease families in one of them. The Chlamydomonas degradome is devoid of aspartic protease family A11 (copia transposon endopeptidase), threonine protease family T2 (glycosylasparaginase precursor), and metalloprotease families M10 (matrixin), M102 (DA1 peptidase), M49 (dipeptidyl-peptidase III, also known as nudix hydrolase 3), and M38 (isoaspartyl dipeptidase), which are represented by one (M49) or more homologues in Arabidopsis (Fig. 1). The presence of the M10 family in Arabidopsis might be associated with defence against biotrophic and necrotrophic pathogens (Zhao et al., 2017), whereas proteases among the M102 family members are known to be involved in organ size control (Vanhaeren et al., 2016; Wang et al., 2017b). In contrast to land plants, the mobile unicellular green algae are capable of evading potential pathogens.

Together with the above-mentioned family M11 (gametolysin), Arabidopsis also lacks the cysteine protease families C11 (clostripain) and C45 (acyl-coenzyme A:6-aminopenicillanic acid acyl-transferase precursor), metalloprotease family M32 (carboxypeptidase), and serine protease family S77 (prohead endopeptidase), which are all present in Chlamydomonas (Fig. 1). However, information about the role of these proteases in the algal life cycle is still lacking.

In contrast to frequently observed gene duplication events, and hence gene redundancy, in higher plants, including Arabidopsis, Chlamydomonas has a simpler genome with much less frequent gene duplications (Merchant et al., 2007). This is also true for protease-encoding genes. Indeed, among 58 protease families present in both Chlamydomonas and Arabidopsis, 41 (~71%) families have more genes per family in Arabidopsis than in Chlamydomonas (Fig. 1). One representative example is family C14, the metacaspases. There are nine metacaspase genes, AtMC1–AtMC9, in Arabidopsis, which are further split into two structurally distinct types, I and II, represented by three (AtMC1, 2, and 3, or AtMCA-Ia, b, and c according to new nomenclature) and six (AtMC4, 5, 6, 7, 8, and 9, or AtMC-IIa, b, c, d, e, and f) homologues, respectively (Tsiatsiani et al., 2011; Minina et al., 2020). Notably, four type II metacaspase genes, AtMCA-IIa, b, c, and d (formerly named AtMC4, 5, 6, and 7) are products of tandem duplication and are located next to each other within a region of 10.6 kb on Arabidopsis chromosome 1. Additionally, the AtMCA-IIe (AtMC8) gene situated on the same chromosome is related to the tandem repeat of those four genes by a second internal duplication event (Vercammen et al., 2004). This duplication of type II metacaspase genes in Arabidopsis makes it difficult to decipher the whole spectrum of their mechanistic roles and to distinguish between redundant and individual gene-specific functions. In contrast, the Chlamydomonas genome encodes only one member of each type of metacaspase (CrMCA-I and CrMCA-II; Fig. 1), offering an ideal system for studying the ancestral roles of these proteolytic enzymes.

Interestingly, some protease genes are also found duplicated in Chlamydomonas, suggesting that those gene duplication events likely occurred after divergence from its ancestor shared with higher plants. For example, while Arabidopsis has a single-copy gene encoding Deg1 (for Degradation of periplasmic proteins) protease belonging to the Deg/HtrA (for High temperature requirement A) family (S1), its Chlamydomonas orthologue is represented by three copies (Schuhmann et al., 2012). Based on these facts, we conclude that most of the proteases in Chlamydomonas are encoded by single-copy genes, offering a valuable model for genetic studies.

Phylogenetic relatedness of Chlamydomonas proteases

As green algae are believed to share a common ancestor with higher plants, and Chlamydomonas shares some features (e.g. cilia) with animals (Merchant et al., 2007), we wondered how such evolutionary versatility affected the phylogenetic relatedness of the Chlamydomonas degradome as a whole. In this analysis, Chlamydomonas protease sequences were used as queries to search against a non-redundant protein database of the National Centre for Biotechnology Information (NCBI). The protein sequences sharing high similarity (E value <10^–10) were collected for calculating the relative distance through phylogenetic analysis. Based on the closest homologues to algal proteins, all Chlamydomonas proteases were classified into six types: animal, animal and plant, plant, bacterial, green algal, and unclassified (Table S1 at Zenodo). Thus, plant-type proteases are the ones that are most similar to homologues from higher plants, and the same principle determines the ontology of other types, except that animal-and-plant-type proteases are close to the cluster consisting of both animal and plant homologues.

The most abundant (~60%) of Chlamydomonas proteases are of a plant type, followed by green algal type (~16%), and bacterial type (~12%), whereas animal-type along with animal-and-plant-type proteases jointly constitute 4.3% of the Chlamydomonas degradome (Fig. 2A). Interestingly, 7.4% of the Chlamydomonas proteases (26 protein sequences) cannot be ascribed to any phylogenetic type, because they exhibit too high divergence with homologous protein sequences. However, the distribution of all Chlamydomonas proteases into various phylogenetic types presented above does not hold for individual catalytic types. Indeed, all threonine proteases (mostly proteasome subunits) and the vast majority (90%) of aspartic proteases of Chlamydomonas are of the plant type (Fig. 2B). The plant type also predominates among cysteine proteases (~71%), whereas surprisingly large proportions of the metallo- and serine proteases belong to the green algal (~38%) and bacterial (~22%) types, respectively. Finally, the animal type is most frequently found among cysteine proteases (~5.3%; Fig. 2B).

Open in a separate window

Fig. 2.

Classification of Chlamydomonas proteases into phylogenetic types. (A) Contribution of different phylogenetic types into the whole Chlamydomonas degradome. (B) Distribution of different phylogenetic types among major catalytic types of Chlamydomonas proteases.

Biochemistry and physiological role of proteases in Chlamydomonas

Green-algal-type and bacterial-type proteases

About 16% of all proteases in Chlamydomonas are found exclusively in the green algal lineage (Fig. 2A) and are probably the products of de novo originated genes. Most of the green-algal-type proteases are gametolysin isoforms with conserved zinc-binding sites (Fig. S3 at Zenodo). The remaining 12% of Chlamydomonas proteases have a close relationship with homologues from bacteria (Fig. 2A), pointing to HGT that occurred after the separation between Chlamydomonas and higher plants. Curiously, one Chlamydomonas serine protease from the S1 family (gene locus Cre06.g267750; V8 proteinase) contains a C-terminal trypsin-like domain and an N-terminal Kazal-type serine protease inhibitor domain, while its closest homologues from bacteria contain only the trypsin-like domain (Fig. 3). Interestingly, the Chlamydomonas genome also encodes Kazal-type serine protease inhibitors, small proteins with one or more Kazal domains, each being 40–60 amino acids in length. However, it remains to be seen whether these inhibitors interact with and block the proteolytic activity of the V8 proteinase.

Open in a separate window

Fig. 3.

Phylogenetic relationship and domain composition of a trypsin-like peptidase (Cre06.g267750) and related proteases from bacteria and green algae. The search was made against the non-redundant (nr) database from NCBI and protein sequences were used for further alignment and phylogeny analysis. The phylogenetic and domain analyses were performed using MEGA X with the neighbour-joining method (Kumar et al., 2018) and Pfam (El-Gebali et al., 2019), respectively. The scale bar indicates the number of amino acid substitutions per site.

Systematic research on copper regulation in Chlamydomonas identified a Regulator of Sigma-E Protease (RSEP1), closely related to RseP, which is a bacterial membrane metalloprotease involved in transmembrane signalling (Castruita et al., 2011). RSEP1 contains a conserved HExxH motif embedded in a transmembrane helix for metal ion binding. Eukaryotic RSEP1 proteases are found exclusively in Chlamydomonadales, suggesting an early HGT from bacteria. In Chlamydomonas, the expression of RSEP1 is induced by copper depletion in a global copper-sensing transcription factor CRR1-dependent manner (Castruita et al., 2011). It has been suggested that chloroplast-localized RESP1 degrades plastocyanin to release the copper for survival in a copper-deficient environment (Castruita et al., 2011; Kropat et al., 2015), but the genetic evidence for such a role of RESP1 is still missing.

Gametolysin and sporangin

Two cell-wall-digesting proteolytic enzymes, bacterial-type sporangin (family S8 of the SB clan) and green-algal-type gametolysin (family M11 of the MA clan), serve important roles in the accomplishment of Chlamydomonas asexual and sexual cycles, respectively. While sporangins from green algae are closely related to bacterial serine proteases (Fig. S4 at Zenodo) and therefore might be a result of an HGT event, gametolysins are found exclusively in Chlamydomonas and its close but multicellular relative Volvox carteri.

Under favourable growth conditions, Chlamydomonas reproduces itself asexually by fission, when a single mother cell undergoes one to three divisions, producing two, four, or eight daughter cells encircled by the mother (sporangial) cell wall (Fig. 4). Sporangin is capable of breaking down the sporangial cell wall during hatching, and its expression is specifically induced during the S/M phase of the asexual cell cycle (Kubo et al., 2009). Under adverse conditions, Chlamydomonas initiates its sexual life cycle, in which haploid gametes of plus and minus mating types fuse to generate a diploid zygote that will divide into four vegetative cells when growth conditions return to normal. Gametolysin is secreted to digest and remove the gamete cell wall, thus allowing mating to commence (Fig. 4). It is noteworthy that gametolysin also exhibits lytic activity towards the cell walls of vegetative and sporangial cells (summarized in Matsuda, 2013).

Open in a separate window

Fig. 4.

Schematic illustration of the roles of sporangin and gametolysin in Chlamydomonas. Depending on the environmental conditions, Chlamydomonas can undergo either asexual (left) or sexual (right) reproduction cycles. The asexual cycle requires repeated alternating light/dark periods and replete nutrients, and is composed of two broad phases: the light-dependent cell growth (or G1) phase and the dark-dependent cell division (or S phases and mitoses, S/M) phase. Following S/M, daughter vegetative cells hatch out of the mother cell wall digested by sporangin (depicted by the black dotted oval). Under adverse conditions, such as nitrogen deprivation (-N) under light, the vegetative cells transform into gametes of mating type plus or minus. Two gametes of different mating types fuse to form a quadriflagellate cell in a process requiring the proteolytic activity of gametolysin, which removes gametic cell walls (depicted by the black dotted ovals). The quadriflagellate cell loses its cilia and becomes a mature zygote with a thick cell wall. Repletion of nitrogen (+N) induces the zygote to undergo meiosis, which will generate four haploid cells, two of each mating type. Whether proteases are involved in the liberation of haploid cells from the zygote cell wall remains unknown.

While there are only six genes encoding sporangin isoforms (Fig. S4 at Zenodo), the gametolysin family contains more than 40 members (Fig. 1). In Chlamydomonas, the sexual cycle represents not only a mode of reproduction, but also a strategy of overcoming adverse conditions (Suzuki and Johnson, 2002). Therefore, a large number of gametolysins present in Chlamydomonas might provide a robust mechanism facilitating cell-wall lysis not only in the secreting gametic cell, but also in surrounding cells, especially those with the opposite mating type, and in this way sustain algal fitness and survival under adverse conditions.

One representative of each of the gametolysin and sporangin families has been studied biochemically in more detail, and they were found to exhibit distinct P1 substrate specificity, potentially accounting for distinct target cell-wall specificity (Table 2). While gametolysin cleaves peptide bonds preferentially after hydrophobic residues (Matsuda et al., 1990), sporangin, similar to many other serine proteases, requires arginine or lysine at the P1 position (Matsuda et al., 1995). Gametolysin is a zinc-binding metalloprotease and requires metal ions for catalysis; accordingly, metal-binding chelators such as EDTA inhibit gametolysin activity (Matsuda et al., 1990). Although metal ions are presumably not required for the catalytic activity of serine proteases, they might be required for stabilization of the active conformation of the proteases. In line with this notion, it has been shown that EDTA inhibits the proteolytic activity of sporangin (Matsuda et al., 1995).

Concluding remarks

Chlamydomonas has emerged as a unique model representing an early diverged ancestor of higher plants and maintaining some features of the last eukaryotic common ancestor, such as cilia. Thus, Chlamydomonas offers a great evolutionary perspective for research on the mechanisms of cell motility and multicellular pattern formation, besides providing a comparable platform for studying metabolic and signalling pathways operating in higher plants. Since these mechanisms and pathways involve proteolytic regulation, and many proteases of Chlamydomonas are encoded by single-copy genes, the value of this model organism for studying proteolysis is difficult to overestimate.

In this review, we have attempted to offer a comprehensive survey of the protease degradome in Chlamydomonas that may be useful for conceiving future research endeavours. The functional studies performed thus far (Table 2) cover just a tiny fraction of Chlamydomonas proteases whose functions still need to be linked to specific proteolytic events. Overall, protease substrates and proteome modifications caused by proteolysis in Chlamydomonas remain unknown. This lack of knowledge calls for future studies of the Chlamydomonas protease degradome, which should be facilitated by the recent advent of N-terminomics (Luo et al., 2019), positional scanning substrate combinatorial library (PS-SCL) screening (Poręba et al., 2014), and methods for live imaging of protease activity (Fernández-Fernández et al., 2019).

Supplementary Material

Acknowledgements

References