The CRK2-CYC13 complex functions as an S-phase cyclin-dependent kinase to promote DNA replication in Trypanosoma brucei

Background Faithful DNA replication is essential to maintain genomic stability in all living organisms, and the regulatory pathway for DNA replication initiation is conserved from yeast to humans. The evolutionarily ancient human parasite Trypanosoma brucei, however, lacks many of the conserved DNA replication factors and may employ unusual mechanisms for DNA replication. Neither the S-phase cyclin-dependent kinase (CDK) nor the regulatory pathway governing DNA replication has been previously identified in T. brucei. Results Here we report that CRK2 (Cdc2-related kinase 2) complexes with CYC13 (Cyclin13) and functions as an S-phase CDK to promote DNA replication in T. brucei. We further show that CRK2 phosphorylates Mcm3, a subunit of the Mcm2–7 sub-complex of the Cdc45-Mcm2–7-GINS complex, and demonstrate that Mcm3 phosphorylation by CRK2 facilitates interaction with Sld5, a subunit of the GINS sub-complex of the Cdc45-Mcm2–7-GINS complex. Conclusions These results identify the CRK2-CYC13 complex as an S-phase regulator in T. brucei and reveal its role in regulating DNA replication through promoting the assembly of the Cdc45-Mcm2–7-GINS complex. Supplementary Information The online version contains supplementary material available at 10.1186/s12915-021-00961-1.

Trypanosoma brucei is an early branching protozoan parasite [18] causing human sleeping sickness in sub-Saharan Africa and possesses numerous unusual biological features that might be exploited as drug targets. The regulation of DNA replication in T. brucei, although poorly understood, appears to possess both evolutionarily conserved and trypanosome-specific features [19][20][21]. T. brucei expresses a highly divergent ORC complex, of which only five subunits have been identified [22][23][24], and a conserved CMG complex [22], but its genome does not encode close homologs of Cdc6, Cdt1, and DDK/Cdk7 [25]. It appears that T. brucei may employ a control mechanism for DNA replication that is distinct from that in yeast and animals.
The S-phase CDK, a crucial DNA replication regulator that helps recruitment of the pre-loading complex to the replication origins [3], and its associated cyclin partner have not been discovered in T. brucei. The genome of T. brucei encodes eleven Cdc2-related kinases (CRK1-CRK4 and CRK6-CRK12) and eleven cyclins (CYC2-CYC12) [26], but it remains unknown which CRK-CYC pair functions as an S-phase CDK-cyclin complex to promote DNA replication. Here, we discover CRK2 as an S-phase CDK to complex with a new cyclin, CYC13, to promote DNA replication and S-phase progression in T. brucei. We also uncover a mechanistic role of CRK2 in DNA replication by demonstrating that CRK2 phosphorylates the Mcm2-7 complex subunit Mcm3 to facilitate the interaction of Mcm3 with the GINS complex subunit Sld5. These findings reveal a novel mechanism for CRK2-mediated DNA replication and highlight the unusual biology of cell cycle regulation in this early divergent eukaryote.

CYC13 is the cyclin partner of CRK2 in T. brucei
Hidden Markov Model (HMM) analysis identified 12 cyclin box-containing proteins in T. brucei shown in the superfamily HMM library and genome assignments server (http://supfam.org/SUPERFAMILY/cgi-bin/genome. cgi?sf=47954&listtype=sf&cgi_Tb=yes). Two of these proteins, encoded by Tb927. 11.780 (or Tb11.03.0350) and Tb927. 11.11500 (or Tb11.01.3350), have not been characterized previously. Tb927.11.11500 encodes a CDK-binding protein homologous to human CABLES protein [27], whereas the protein encoded by Tb927. 11.780 contains a cyclin box-like domain (a.a. 272-420; E-value: 2.61e−16) and thus was named CYC13. The cyclin-box domain of CYC13 has an overall 19% sequence identity and a similar structure to the cyclin-box domain of human cyclin A2 (Fig. 1a, b), as predicted by structural modeling using SWISS-MODEL [28]. Cyclin A2 in humans activates both Cdk2 and Cdk1 to regulate S-phase and mitosis, respectively [29]. Sequence analyses of all trypanosome cyclins, CYC2-CYC13, showed that these cyclins do not share an overall sequence similarity (Additional file 1: Figure S1), because they belong to distinct types of cyclins. Phylogenetic analysis of these cyclins placed CYC13 to a distinct clade, to which the closest clade is the mitotic B-type cyclin clade (Fig. 1c).
To identify the CRK partner(s) of CYC13, we first carried out yeast two-hybrid assays to test its interaction with any of the eleven CRKs, and the results showed that CYC13 interacts with CRK2, CRK3, and CRK12 (Fig.  1d). Further, we carried out co-immunoprecipitation to test the in vivo interaction between CYC13 and the three CRKs. To this end, CYC13 was endogenously tagged with a PTP epitope in the trypanosome cells that express CRK2, CRK3, or CRK12 fused with a C-terminal triple HA epitope from their respective endogenous locus. Immunoprecipitation of CYC13-PTP was able to pull down CRK2-3HA, and immunoprecipitation of CRK2-3HA was able to pull down CYC13-PTP (Fig. 1e). However, immunoprecipitation of CYC13-PTP did not pull down CRK3-3HA and CRK12-3HA (Fig. 1f), and immunoprecipitation of CRK3-3HA or CRK12-3HA did not pull down CYC13-PTP (Additional file 1: Figure S2). These results identified CRK2 as the CDK partner of CYC13 in T. brucei.

The CRK2-CYC13 complex is required for S-phase progression
We carried out RNAi to investigate the function of the CRK2-CYC13 complex in the procyclic form of T. brucei. To assess the efficiency of RNAi, we epitope-tagged CYC13 and CRK2 from their respective endogenous locus in their respective RNAi cell line for western blotting. Induction of CYC13 RNAi and CRK2 RNAi each resulted in the reduction of their protein levels by > 90% after 1 day (Fig. 2a), indicating efficient depletion of the target protein by RNAi. Depletion of CYC13 and CRK2 each caused severe growth defects (Fig. 2b), suggesting that both proteins are essential for cell proliferation. We further analyzed the potential effects of CYC13 RNAi and CRK2 RNAi on cell cycle progression by counting the cells with different numbers of kinetoplasts and nuclei. T. brucei cells at different cell cycle stages can be readily distinguished by the numbers of kinetoplasts and nuclei. Cells at the G1 phase and early S phase of the cell cycle contain one nucleus and one kinetoplast (1N1K), cells at late S phase to early anaphase contain one nucleus and two kinetoplasts (1N2K), and cells at late anaphase to cytokinesis contain two nuclei and two kinetoplasts (2N2K). Knockdown of CYC13 caused an initial increase of 1N1K cells at 24 h and subsequently a decrease of 1N1K and an emergence of two abnormal cell types, the 2N1K cells (1 8%) and 0N1K cells (~11%), at 48 h (Fig. 2c). Unlike knockdown of CYC13, however, knockdown of CRK2 caused a gradual increase of 1N1K cells from~76 tõ 91% after RNAi induction for 48 h, without a significant accumulation of abnormal cell types, the 2N1K cells and the 0N1K cells (Fig. 2c).
To further characterize the cell cycle defects of the two RNAi cell lines, flow cytometry was carried out. Knockdown of CYC13 caused a decrease of G1 cells by 35% and G2/M cells by 13%, and a corresponding increase of S-phase cells by 9%, cells with DNA content greater than 4C by 27%, and anucleate (sub-G1) cells by 12% at 48 h of RNAi (Fig. 2d). The emergence of cells with DNA content between 4C and 8C suggests that the bi-nucleated cells were undergoing the next round of DNA replication before cell division, and the accumulation of anucleate cells indicates aberrant cytokinesis of some of the bi-nucleated cells, which produced 2N1K and 0N1K cells. This profound effect of CYC13 knockdown on the cell cycle suggests that CYC13 may play multiple roles at different cell cycle stages, including S phase. Knockdown of CRK2 caused a decrease of G1 cells by 41% and G2/M cells by 12%, and a corresponding increase of S-phase cells by 44% and cells with DNA content greater than 4C by 8%, but no significant increase of anucleate cells (Fig. 2d). These results suggest that CRK2 plays a role in S-phase progression. We next carried out EdU incorporation assay to examine the effect of CYC13 RNAi and CRK2 RNAi on DNA replication. Non-induced control and RNAi-induced cells were incubated with EdU, and the incorporation of EdU into DNA was monitored by fluorescence microscopy and by counting the number of EdU-positive 1N1K cells (Fig.  2e, f). In the non-induced control cells,~43% of the 1N1K cells were EdU-positive (Fig. 2e, f), indicating that these cells were undergoing DNA replication. In the RNAi-induced cells,~13% and~6% of the 1N1K cells were EdU-positive for CRK2 RNAi and CYC13 RNAi, respectively (Fig. 2e, f), suggesting that the majority of the RNAi-induced cells failed to incorporate EdU into DNA. These results further demonstrated that knockdown of CYC13 and CRK2 impaired DNA replication.
Finally, we carried out immunofluorescence microscopy using KKT13 as an S-phase marker to examine the effect of CYC13 RNAi and CRK2 RNAi on S-phase progression. KKT13 is a kinetochore protein and is only expressed during the S phase of the cell cycle in T. brucei [30], and thus it can serve as an S-phase marker to distinguish the 1N1K cells that are at either G1 phase or S phase. Immunostaining of KKT13 tagged with a Cterminal triple HA epitope from its endogenous locus was performed (Fig. 3), and the intensity of KKT13 fluorescence intensity was quantitated (Additional file 1: Figure S3), enabling us to distinguish between KKT13 fluorescence signal and the background signal for unambiguous counting of KKT13-positive cells. The results showed that the percentage of S-phase cells among the 1N1K cell population was increased by~66% and~70% after CYC13 knockdown and CRK2 knockdown, respectively ( Fig. 3a-c). Since flow cytometry showed that the bi-nucleated cells of the CYC13 RNAi-induced cell population had entered the S phase of the next cell cycle (Fig. 2d), we examined KKT13 expression in the binucleated cells (2N2K cells from the control cell population and 2N2K and 2N1K cells from the CYC13 RNAiinduced cell population). We found that the KKT13positive bi-nucleated cells increased to~85% after CYC13 RNAi induction for 48 h (Fig. 3d,   that the bi-nucleated cells were indeed arrested at the next round of S phase. All together, these results suggest that the CRK2-CYC13 complex plays an essential role in promoting S-phase progression. DNA replication in eukaryotes requires a cohort of DNA replication factors, including the CMG complex that functions as a replicative helicase to unwind the DNA duplex during DNA replication in the S phase of the cell cycle [17]. Since the CRK2-CYC13 complex is required for DNA replication (Figs. 2 and 3), we asked whether CRK2 regulates the CMG complex by interacting and phosphorylating any of the CMG subunits. To test this possibility, we first carried out yeast two-hybrid assays to screen for the CRK2-interacting CMG subunit(s). CRK2 was fused with either the Gal4 DNAbinding domain (BD) or Gal4 activation domain (AD) and paired with each of the eleven CMG complex subunits that were fused with the AD or the BD ( To test whether Mcm3 and Sld5 are substrates of CRK2, in vitro kinase assay was performed with recombinant GST-fused proteins, GST-Mcm3, GST-Sld5, GST-CRK2, and GST-CYC13, all of which were expressed and purified from bacteria, using the non-radioactive thiophosphorylation method [31]. This method involves the use of the anti-ThioP antibody to detect thiophosphorylated proteins [31]. The in vitro kinase assay showed that CRK2 alone was able to phosphorylate Mcm3, and the addition of CYC13 increased the levels of phosphorylated Mcm3 (Fig. 4c), demonstrating that CYC13 promotes the kinase activity of CRK2. Further, to identify the phosphosites, mass spectrometry was carried out after in vitro kinase assay. A total of 1097 peptides, which were represented by 16 unique peptides and covered~76% of the entire protein, were identified for Sld5, and a total of 3997 peptides, which were represented by 91 unique peptides and covered~88% of the entire protein, were identified for Mcm3 (Additional file 1: Figure S4). Two phospho-peptides, which contained phosphorylated Ser-213 and Thr-310, respectively (Fig. 4d     moderate growth defect, resulting in a~11-fold growth reduction after 5 days (Fig. 5d). EdU incorporation assay showed that induction of Mcm3-3′UTR RNAi for 48 h caused a reduction of the EdU-positive 1N1K cells from 48 to~26% (Fig. 5e). When one allele of Mcm3 was replaced with the RNAi-resistant wild-type Mcm3 and then RNAi was induced, PTP-Mcm3 was gradually knocked down and Mcm3-3HA was not affected (Fig.  5c), but cell growth was only partly restored, resulting in a~6-fold growth reduction after 5 days (Fig. 5d). EdU incorporation assay showed that EdU-positive 1N1K cells were reduced from~47 to~41% after RNAi induction for 48 h (Fig. 5e). The lack of full rescue of the growth defect is likely due to the fact that only one allele of the Mcm3 gene expressed Mcm3 protein after RNAi induction and that Mcm3 protein level is so tightly controlled that a slight or moderate modulation of its level impacts cell growth. As a support of the notion that Mcm3 protein level is stringently regulated, ectopic overexpression of Mcm3 in the 29-13 cell line caused a severe growth defect, resulting in a~23-fold growth reduction after 5 days (Additional file 1: Figure S6).

G T G T A C G C A A A G T G T G C G C A A A G T G T A C G C A A A
When one allele of Mcm3 was replaced with the RNAi-resistant Mcm3 T310A mutant, the cell line was not growing as well as other cell lines even without RNAi induction (Fig. 5d), suggesting that expression of Mcm3 T310A might cause dominant-negative effects but allowed slower cell proliferation. This notion was supported by the observed severe growth defects caused by ectopic overexpression of Mcm3 T310A in the 29-13 cell line, which resulted in a~400-fold growth reduction after 5 days ( Figure S6). When Mcm3-3′UTR RNAi was induced, PTP-Mcm3 was gradually depleted and Mcm3 T310A -3HA was unaffected (Fig. 5c), but cell growth was inhibited, resulting in a~48-fold growth reduction after 5 days (Fig. 5d), indicating that the Mcm3 T310A mutant failed to restore cell proliferation. EdU incorporation assay showed that without RNAi induction the EdU-positive 1N1K cells only constituted1 2% of the 1N1K population, and after RNAi induction for 48 h EdU-positive 1N1K cells were reduced to~1% of the total 1N1K population (Fig. 5e).
When one allele of Mcm3 was replaced with the RNAi-resistant Mcm3 T310D mutant and then RNAi was induced, PTP-Mcm3 was knocked down and Mcm3 T310D -3HA was unaffected (Fig. 5c), and cell growth was partially restored, resulting in a~5-fold growth reduction after 5 days (Fig. 5d), similar to the wild-type Mcm3 complementation cell line (Fig. 5d). EdU incorporation assay showed that the EdU-positive 1N1K cells were reduced from~51 to~45% after RNAi induction for 48 h (Fig. 5e). When Mcm3 T310D was ectopically overexpressed in the 29-13 cell line, it caused a moderate growth defect, resulting in a~14-fold growth reduction after 5 days ( Figure S6). Altogether, these results suggest that Thr-310 phosphorylation is essential for Mcm3 function in DNA replication.

Phosphorylation of Mcm3 at Thr-310 is required for its interaction with Sld5
Having established the functional essentiality of Mcm3 Thr-310 phosphorylation (Fig. 5), we explored the underlying mechanisms by investigating the potential effect of Mcm3 phosphorylation on its interaction with other subunits of the CMG complex. We chose the subunit protein Mcm2 from the Mcm2-7 sub-complex, the subunit protein Sld5 from the GINS sub-complex, and Cdc45 to test their interaction with Mcm3 in CRK2 RNAi cells. We coexpressed Mcm3-3HA with Mcm2-PTP, PTP-Sld5, or Cdc45-PTP in cells harboring the CRK2 RNAi construct, and then performed co-immunoprecipitation to test the interactions in control and CRK2 RNAi cells. We found that immunoprecipitation of Mcm2-PTP was able to pull down Mcm3-3HA and the interaction between them was not affected by CRK2 knockdown (Fig. 6a). Similarly, coimmunoprecipitation of Cdc45-PTP and Mcm3-3HA was also unaffected by CRK2 knockdown (Fig. 6b). However, co-immunoprecipitaiton of PTP-Sld5 and Mcm3-3HA was impaired by CRK2 knockdown (Fig.  6c). CYC13 knockdown also affected the interaction between Mcm3 and Sld5 (Fig. 6d), albeit the effect was weaker than that caused by CRK2 knockdown, likely because CRK2 kinase activity was only partially reduced in the absence of CYC13 (Fig. 4c). Further, to test whether phosphorylation of Thr-310 affects the interaction between Mcm3 and Sld5, we coexpressed PTP-Sld5 and 3HA-tagged wild-type Mcm3, phospho-deficient Mcm3 (Mcm3 T310A ), and phosphomimic Mcm3 (Mcm3 T310D ), and then performed coimmunoprecipitation. The results showed that immunoprecipitation of Sld5 pulled down significantly less amounts of Mcm3 T310A protein, but significantly more amounts of Mcm3 T310D protein than wild-type Mcm3 (Fig. 6e). Altogether, these results suggest that phosphorylation of Thr-310 in Mcm3 by CRK2 is required for Mcm3 interaction with the GINS subunit protein Sld5, but not Cdc45 and the Mcm2-7 complex subunit protein Mcm2.

Discussion
Tremendous efforts have been undertaken in the past to understand the functions of the expanded repertoire of cyclins and CRKs in T. brucei and to dissect the CRKregulated cellular pathways in T. brucei [32][33][34][35][36][37][38][39][40][41][42][43][44][45][46], but the S-phase cyclin-CRK pair had been overlooked, making the understanding of the CDK-mediated regulation of DNA replication in T. brucei impossible. T. brucei possesses a unique molecular composition of the DNA replication initiation machinery [22][23][24]47] and lacks several key regulators of DNA replication [19,48], suggesting distinct mechanisms and regulatory pathways for DNA replication. With the discovery of the CYC13-CRK2 pair as a regulator of S-phase progression in T. brucei, we are now able to delineate the signaling cascade controlling DNA replication in this early branching eukaryote, which might provide novel insights into the understanding of the evolution of the DNA replication machinery in eukaryotes.
Flow cytometry analysis of CRK2 RNAi cells detected a major S-phase peak (Fig. 2d), indicating the arrest of cells at S phase. Since RNAi induction was initiated in asynchronous cells containing~28% of the G2/M-phase cells (Fig. 2d), it suggests that these G2/M cells were able to divide, but the daughter cells were then arrested in the S phase of the next cell cycle. Hence, knockdown of CRK2 did not inhibit G2/M progression and cytokinesis. Flow cytometry analysis of CYC13 knockdown cells, however, detected an S-phase peak, a post-G2/M peak, and a smaller sub-G1 peak (Fig. 2d). We interpreted these observations as follows. Upon CYC13 RNAi induction, G1 cells progressed and then were arrested in S phase, producing uni-nucleated cells with DNA content greater than 2C but less than 4C (S-phase peak). However, G2/M cells re-initiated DNA replication without undergoing cytokinesis and then were arrested in the S phase of the next cell cycle, thus generating bi-nucleated cells with DNA content greater than 4C (post-G2/M peak). The anucleated cells (sub-G1 peak), together with the 2N1K cells (Fig. 2c), were generated through aberrant division of some of the bi-nucleated cells. Thus, CYC13 plays one role through activating CRK2 to promote S-phase progression and another role through activation of an unidentified CRK to promote cytokinesis. This unidentified CRK might be among the remaining eight CRKs that showed no interaction with CYC13 in yeast two-hybrid assays, but it may interact with CYC13 in vivo in trypanosomes. This subject is out of the scope of the current work and was not pursued further. The identification of Mcm3 as a substrate of CRK2 (Fig. 4) and the finding that Mcm3 phosphorylation promotes Mcm3-Sld5 interaction (Fig. 6) provided the evidence to support the role of CRK2 in regulating S-phase progression. Mcm3 is an essential subunit of the CMG complex, the replicative helicase required for DNA replication initiation in eukaryotes [17]. T. brucei has a conserved CMG complex [22], but lacks the homolog of the Dbf4-dependent kinase (DDK), the major protein kinase responsible for Mcm2-7 phosphorylation [3]. The CRK2 phosphosite on Mcm3, Thr-310, resides in the N-terminal intrinsically disordered region that is1 00 a.a. away from the Walker A motif of the AAA + ATPase domain, indicating that this phosphorylation is unlikely to impact its ATPase activity. The Mcm4-Mcm6-Mcm7 sub-complex has helicase activity, and the Mcm2-Mcm3-Mcm5 sub-complex may play regulatory roles [57]. In T. brucei, only Mcm4, but not other Mcm proteins, possesses in vitro helicase activity [22]. Within the CMG complex, Cdc45 interacts with Mcm2, and the GINS sub-complex interacts with Mcm3 and Mcm5 [58]. Therefore, our finding that Mcm3 interacts with the Sld5 subunit of the GINS sub-complex in T. brucei (Fig. 6) suggests that T. brucei CMG complex has similar subunit arrangements as in other eukaryotes. Based on the well-established order of loading of Mcm2-7, Cdc45, and GINS to the replication origins and the known protein-protein interactions among the CMG subunits [22,59], we propose the following model to depict the order of assembly of the CMG complex and its regulation by CRK2 in T. brucei. Mcm2-7 is among the first set of DNA replication regulators to be loaded onto the replication origins. This is followed by the loading of Cdc45 onto the vicinity of Mcm2 and Mcm5 through a DDK-independent mechanism and, finally, after phosphorylation of Thr-310 in Mcm3 by CRK2, the GINS sub-complex is loaded onto the vicinity of Mcm3 and Mcm5 to form the CMG complex, thus leading to the initiation of DNA replication (Fig. 7). When CRK2 is knocked down, Mcm3 is not phosphorylated and the GINS sub-complex is not loaded onto the vicinity of Cdc45 and Mcm2-7; therefore, assembly of the CMG Subsequently, CRK2 phosphorylates Mcm3 at Thr-310 and Ser-213, and phosphorylation of Thr-310 promotes the recruitment of the GINS complex to the Mcm2-7 complex, leading to the formation of the CMG complex and then the initiation of DNA replication. When CRK2 is deficient, the GINS complex is not recruited, and the CMG complex is not formed, thereby preventing the initiation of DNA replication complex is impaired and DNA replication is inhibited (Fig. 7). In this regard, CRK2 plays an essential role in DNA replication by facilitating the assembly of the CMG complex through promoting the interaction between Mcm3 and Sld5.

Conclusions
We have identified CYC13-CRK2 pair as an S-phase cyclin-CDK in the early branching trypanosomes, and we provided evidence to show that CYC13-CRK2 promotes DNA replication initiation and S-phase progression. CYC13-CRK2 phosphorylates Mcm3, a subunit of the Mcm2-7 complex, and this phosphorylation promotes the interaction of Mcm3 with Sld5, a subunit of the GINS complex, thereby facilitating the assembly of the Cdc45-Mcm2-7-GINS complex.

Yeast two-hybrid assay
Yeast two-hybrid assays were performed according to our published procedures [22]. Briefly, the full-length coding sequences of CYC13, the eleven CRKs, and the eleven subunits of the CMG complex were each cloned into the pGAD vector for expression of Gal4 activation domain fusion proteins (prey) or into the pGBK vector for expression of Gal4 DNA-binding domain fusion proteins (bait). Yeast strain AH109 (mating type a) and Y187 (mating type α) were transformed with the prey and the bait plasmids, respectively. Yeast mating was carried out by mixing the AH109 and Y187 strains in YPDA medium and incubating on the SD-Leu-Trp plates at 30°C for 24 h. To test the interaction between prey and bait proteins, each combination stain was spotted in three 10-fold serial dilutions onto SD-Leu-Trp (2 drop-outs or 2DO), SD-Leu-Trp-His-Ade (3 drop-outs or 3DO), or SD-Leu-Trp-His-Ade (4 drop-outs or 4DO) plates, and then incubated at 30°C for up to 3 days. Growth of yeast on the 3DO and 4DO plates indicates interaction between the prey and the bait proteins. The empty vectors of pGAD and pGBK were used as the negative control, and the pair of SV40-p53 was used as the positive control.
To generate the RNAi cell line for CYC13 and CRK2, the full-length coding sequence of CYC13 and a 417-bp DNA fragment (nucleotides 609-1025) from the coding region of CRK2 were each amplified by PCR and then cloned into the stem-loop RNAi vector, pSL [61], to generate the pSL-CYC13 and pSL-CRK2 constructs, respectively. The RNAi plasmid was linearized by NotI digestion and transfected into the 29-13 strain by electroporation. Transfectants were selected with 2.5 μg/ml phleomycin and cloned by limited dilution in a 96-well plate containing SDM-79 medium supplemented with 20% heat-inactivated FBS and appropriate antibiotics. RNAi was induced by incubating the cells with 1.0 μg/ml tetracycline.

Ectopic overexpression of Mcm3 and its mutants
The full-length coding sequence of Mcm3 and its Thr-310 phosphodeficient and phosphomimic mutants (Mcm3 T310A and Mcm3 T310D ) was cloned into the pLew100-3HA vector. The resulting plasmids were each transfected in the 29-13 strain by electroporation. Transfectants were selected 2.5 μg/ml phleomycin and cloned by limited dilution as described above. Cells were incubated with 1.0 μg/ml tetracycline to induce the overexpression of 3HA-tagged Mcm3 and its mutants.
In situ epitope tagging of proteins Endogenous epitope tagging of proteins was performed using the one-step PCR-based epitope tagging method [62]. CRK2 and KKT13 were each tagged with a Cterminal triple HA epitope in the pSL-CRK2 cell line (both proteins) and the pSL-CYC13 cell line (KKT13 only), and transfectants were selected with 1 μg/ml puromycin and then cloned by limiting dilution. CYC13 was tagged with a C-terminal PTP epitope in the pSL-CYC13 cell line, and transfectants were selected with 10 μg/ml blasticidin and then cloned by limiting dilution.
For co-immunoprecipitation between CYC13 and CRK2, CRK3, or CRK12, CYC13 was tagged with a Cterminal PTP epitope in the Lister427 stain, and transfectants were selected with 40 μg/ml G418 and cloned by limiting dilution. Subsequently, the clonal cell line was used for tagging of CRK2, CRK3, or CRK12 with a Cterminal triple HA epitope, and transfectants were selected with 1 μg/ml puromycin and then cloned by limiting dilution.
For co-immunoprecipitation between Mcm3 and Cdc45 and between Mcm3 and Sld5, Mcm3 was tagged with a C-terminal triple HA epitope in the pSL-CRK2 RNAi cell line and the pSL-CYC13 cell line, and transfectants were selected with 1 μg/ml puromycin and cloned by limiting dilution. Subsequently, the clonal cell lines were used for tagging of Mcm2 or Cdc45 with a Cterminal PTP epitope or Sld5 with an N-terminal PTP epitope, and transfectants were selected with 10 μg/ml blasticidin and then cloned by limiting dilution.

Flow cytometry
Flow cytometry was performed according to published procedures [43]. Cells were collected by centrifugation, washed with PBS, and then fixed with ethanol for 30 min at room temperature. Fixed cells were collected by centrifugation, washed once with PBS, and then suspended in PBS containing 10 μg/ml DNase-free RNase and 20 μg/ml propidium iodide. Cells were analyzed with a FACScan analytical flow cytometer (BD Biosciences) to quantitate the DNA content. The percentage of cells in different stages of the cell cycle was calculated by using the Kaluza Analysis software.

In vitro GST pull-down
The full-length coding sequences of Mcm3 and Sld5 were each cloned into the pGEX-4 T-3 vector (Clontech), and the resulting plasmids were transformed into the E. coli BL21 strain. Expression GST-fusion proteins was induced with 0.2 mM IPTG for 4 h at room temperature and purified through binding to the Glutathione Sepharose 4B beads (GE HealthCare) according to the manufacturer's instructions. GST-Mcm3 and GST-Sld5 bound on the Glutathione Sepharose beads were incubated with cleared lysate of T. brucei cells expressing CRK2-3HA at room temperature for 30 min. T. brucei cell lysate was prepared by lysing 10 8 cells with 1 ml cell lysis buffer (25 mM Tris-Cl, pH 7.6, 100 mM NaCl, 1 mM DTT, 1% NP-40, and protease inhibitor cocktail) on ice for 30 min and then removing the insoluble pellets by centrifugation. 900 μl of cell lysate was used for pull-down, and the remaining 100 μl of cell lysate was set aside as the input sample. Beads were washed five times with the cell lysis buffer, and proteins were eluted by boiling the beads in 35 μl of 1x SDS-PAGE sampling buffer for 5 min. Eluted proteins (25 μl) and the input sample (15 μl) were fractionated by SDS-PAGE, transferred onto a PVDF membrane, and immunoblotted with anti-HA antibody. Purified GST was used as the negative control.

In vitro kinase assays
The full-length coding sequences of CRK2 and CYC13 were each cloned into the pGEX-4 T-3 vector, and the resulting plasmids were transformed into E. coli BL21 strain. Recombinant GST-fused CRK2, CYC13, Mcm3, and Sld5 proteins were purified as described above. In vitro kinase assay was carried out using the method developed by the Kevin Shokat group, which uses a semisynthetic epitope for detection of thiophosphorylated kinase substrates [31]. In a 30-μl reaction setup, 10 μg purified GST-Mcm3, GST-Sld5, GST-CRK2, and GST-CYC13 were mixed in the kinase assay buffer (10 mM HEPES, pH 7.6, 50 mM NaCl, and 10 mM MgCl 2 ) in the presence of 1 mM ATP-γ-S for 30 min at room temperature. Subsequently, 1.5 μl of 50 mM alkylating agent p-nitrobenzylmesylate (PNBM) was added into the kinase assay solution and incubated for 60 min at room temperature. Reaction was stopped by adding 1x SDS sampling buffer, and samples were separated by SDS-PAGE, transferred onto a PVDF membrane, and blotted with the monoclonal anti-ThioP antibody, which recognizes the thiophosphate ester (1:5000 dilution, Thermo-Fisher). GST-fusion proteins on the membrane were stained with coomassie brilliant blue.
For mass spectrometry analysis of potentially phosphorylated GST-Mcm3 and GST-Sld5, in vitro kinase assay was performed using 1 mM ATP and no PNBM was added to the reaction. GST-fusion proteins were separated on SDS-PAGE and stained with coomassie brilliant blue. The protein band corresponding to GST-Mcm3 and GST-Sld5 was excised from the SDS-PAGE gel and analyzed by LC-MS/MS to identify phosphopeptides.

In-gel tryptic digestion and LC-MS/MS analysis of protein band
In-gel digestion of proteins was carried out as described previously [61]. Protein band was digested with trypsin (160 ng, Sigma-Aldrich) at 37°C for 4 h, and peptides were extracted with 50 ml of 50% acetonitrile, 5% formic acid. Extracted peptides were dried using SpeedVac, reconstituted in 2% acetonitrile with 0.1% formic acid, and then injected onto Thermo LTQ Orbitrap XL (Thermo-Fisher Scientific, Bremen, Germany). Samples in 2% (v/v) acetonitrile and 0.1% (v/v) formic acid were analyzed on an LTQ Orbitrap XL (Thermo-Fisher Scientific) interfaced with an Eksigent nano-LC 2D plus ChipLC system (Eksigent Technologies, Dublin, CA). Sample was loaded onto a ChromXP C18-CL trap column (200 mm i.d. × 0.5 mm length) at a flow rate of 3 ml/min. Reversed-phase C18 chromatographic separation of peptides was carried on a on a ChromXP C18-CL column (75 mm i.d × 10 cm length) at 300 nl/min. The LTQ Orbitrap was operated in the data-dependent mode to simultaneously measure full-scan MS spectra in the Orbitrap and the five most intense ions in the LTQ by CID, respectively. In each cycle, MS1 was acquired at target value 1E6 with resolution R 1/4 100,000 (m/z 400) followed by top five MS2 scan at target value 3E4. The mass spectrometric setting is as follows: spray voltage was 1.6 KV, charge state screening and rejection of singly charged ion were enabled. Ion selection thresholds were 8000 for MS2, 35% normalized collision energy, activation Q was 0.25 and dynamic exclusion was employed for 30 s. Raw data files were processed and searched against the T. brucei database using the Mascot search engine. The search conditions used peptide tolerance of 10 p.p.m. and MS/MS tolerance of 0.8 Da with the enzyme set as trypsin and two missed cleavages permitted.

Mcm3 RNAi by targeting the 3′UTR and functional complementation
To generate the Mcm3-3′UTR RNAi cell line, a 497-bp DNA fragment from the 3′UTR of Mcm3 gene, which does not overlap with the downstream gene, was amplified by PCR and cloned into the pZJM vector [63]. The resulting plasmid, pZJM-Mcm3-3′UTR, was linearized by NotI digestion and transfected into the 29-13 strain by electroporation. Trasfectants were selected with 2.5 μg/ml phleomycin and cloned by limited dilution as described above. To monitor the efficiency of Mcm3-3′UTR RNAi, Mcm3 was tagged with an N-terminal PTP epitope at one of its endogenous loci using PCR-based epitope tagging approach [62]. Transfectants were selected with 1 μg/ml puromycin, and cloned by limiting dilution as described above.
To generate the Mcm3 RNAi complementation cell lines, a 1920-bp DNA fragment (nucleotides 517-2436) of the coding sequence of Mcm3, Mcm3 T310A , and Mcm3 T310D was cloned into the pC-3HA-BSD vector, which contains the 3′UTR from TbRPA1 gene. The resulting plasmids were each linearized by restriction digestion with XcmI and electroporated into the cells containing the pZJM-Mcm3-3′UTR construct and expressing PTP-Mcm3 from one of its endogenous loci (see above). Transfectants were selected with 10 μg/ml blasticidin and then cloned by limited dilution as described above. Clonal cell lines were verified by sequencing of the PCR product amplified using the forward primer targeting Mcm3 sequence and the reverse primer targeting TbRPA1 3′UTR sequence, and by western blotting with anti-HA antibody to detect 3HA-tagged Mcm3 and its mutants expressed from one Mcm3 locus and with anti-protein A antibody to detect PTP-tagged Mcm3 expressed from the other Mcm3 locus (see Fig. 5a, b for details).

Co-immunoprecipitation
Co-immunoprecipitation was performed as described previously [64]. Briefly, 10 8 cells were lysed in 1.0 ml cell lysis buffer (25 mM Tris-Cl, pH 7.6, 100 mM NaCl, 1 mM DTT, 1% NP-40, and protease inhibitor cocktail) for 30 min on ice. Cell lysate was cleared by centrifugation, and cleared cell lysate (0.9 ml) was incubated with 20 μl IgG Sepharose 6 fast flow beads (GE Healthcare) for 1 h at 4°C. The remaining 100 μl of cell lysate was set aside as the input sample. Proteins bound to the IgG beads were eluted by boiling the beads in 35 μl of 1x SDS sampling buffer. Subsequently, 25 μl of the eluted proteins and 15 μl of the input sample were loaded onto a SDS-PAGE gel for protein separation. Proteins were transferred onto a PVDF membrane and immunoblotted with anti-HA antibody and anti-protein A antibody (Sigma-Aldrich). Intensity of the protein band was measured with the ImageJ software.

EdU incorporation assay
The Click-iT™ EdU Alexa Fluor 488 imaging kit (Ther-moFisher Scientific) was used to detect the incorporation of EdU into DNA. Cells were cultured in SDM-79 medium supplemented with 30 mM EdU for 3 h. Cells were collected by centrifugation, washed with PBS, adhered to glass coverslips, and fixed with 4% paraformaldehyde. Cells on the coverslips were used for EdU detection according to the manufacturer's instructions. Cells were visualized under a fluorescence microscope, and EdU-positive cells were counted.

Immunofluorescence microscopy
Cells were washed once with PBS, settled onto glass coverslips for 30 min at room temperature, and fixed with cold methanol for 30 min at − 20°C. Cells were rehydrated with PBS and incubated with the blocking buffer (3% BSA in PBS) for 1 h at room temperature. Cells were incubated with the FITC-conjugated anti-HA monoclonal antibody (1:400 dilution, Sigma Aldrich, Clone HA7) for 1 h at room temperature and washed three times with PBS. Coverslips were washed three times with PBS and mounted with DAPI-containing VectaShield mounting medium (Vector Labs). Immunostained cells were imaged with an inverted fluorescence microscope (Olympus IX71) equipped with a cooled CCD camera (model Orca-ER, Hamamatsu) and a PlanApo N 60 × 1.42-NA lens. Images were acquired using the Slide-book5 software and processed using the Adobe Photoshop software.

Statistical analysis
Statistical analysis was performed using the chi-square test in the Microsoft Excel software. Detailed n values for each panel in the figures were stated in the corresponding figure legends. For immunofluorescence microscopy, images were randomly taken, and all cells in the image were counted.