Crystal structure of Plasmodium vivax macrophage migration inhibitory factor

This article is part of a special issue celebrating early career researchers in structural science.

Plasmodium vivax is a major cause of malaria globally and has recently been transmitted locally in the USA. P. vivax produces homologs of host proteins, including cytokines such as macrophage migration inhibitory factor (MIF). MIF regulates both adaptive and innate immune responses and contributes to the pathogenesis of parasitic infections, including malaria. Plasma concentrations of P. vivax MIF (PvMIF) correlate with the severity of P. vivax malaria. Plasmodium spp. MIFs have been recognized as candidate malaria vaccines. PvMIF, like other protozoan MIFs, binds to host CD74 and can suppress host MIF–CD74 signaling. The production, crystallization and 1.8 Å resolution structure of PvMIF (PDB entry 9b0m, pdb_00009b0m) are reported. PvMIF crystallized in space group P6₃ with a single molecule in the asymmetric unit. The biological unit of PvMIF is the prototypical MIF trimer.

1. Introduction

Over one-third of the world's population are at risk from Plasmodium vivax, the most geographically expansive of the five human-infecting Plasmodium spp. (Battle et al., 2019). The Southeast Asian and Western Pacific regions have a highest prevalence of P. vivax cases (Howes et al., 2016; Weiss et al., 2025). P. vivax was formerly endemic in North America and Europe, and was eradicated by the 1970s; however, in 2023 the first locally transmitted P. vivax malaria cases in two decades were reported (DeVita et al., 2025). Plasmodium spp. are transmitted by bites of infected female Anopheles mosquitoes, and once in the human host, P. vivax infects hepatocytes (liver stage or pre-erythrocytic stage), eventually leading to blood-stage infection or the erythrocytic stage associated with typical malaria symptoms (Khan & Daily, 2022; Chu & White, 2021). P. vivax can persist in human hosts as hypnozoites in the liver, which can cause relapses that can extend over several months or years (Flannery et al., 2022). The cyclical fever and weakness episodes of P. vivax malaria lead to high treatment costs and productivity loss in endemic countries (Baird et al., 2016). Additionally, the dormant liver stage can cause relapse weeks, months or even decades after the first infection, complicating treatment of P. vivax malaria and requiring antimalarial drugs that are effective against both the blood and liver stages. Additionally, the main antimalarial with P. vivax activity, primaquine, is contraindicated in people with glucose-6-phosphate dehydrogenase (G6PD) deficiency (Douglas et al., 2023). This is because primaquine induces hemolytic anemia in people with glucose-6-phosphate dehydrogenase (G6PD) deficiency, which corresponds to approximately 15% of the population living in P. vivax endemic regions (Douglas et al., 2023).

The Seattle Structural Genomics Center for Infectious Diseases and collaborators are investigating P. vivax proteins for therapeutics discovery as well as to obtain mechanistic insights (Vijayan et al., 2021; Mendez et al., 2025; Bolling et al., 2024). One of the proteins of interest is the P. vivax orthologue of human macrophage migration inhibitory factor (MIF). MIF is a cytokine that regulates adaptive and innate responses and has roles in the pathogenesis of parasitic infections, including malaria (Calandra & Roger, 2003). Protozoan parasite MIF homologs mimic their human MIF counterparts (hMIF1, NCBI Accession No. CAG30406.1, and hMIF2, NCBI Accession No. CAG30317.1), facilitating the modulation of host immune responses and suppressing apoptosis-induced cell death (Twu et al., 2014; Ghosh et al., 2019). Multiple structures of hMIF1 and hMIF2 have been determined and reveal biological trimers that bind to CD74 (Rajasekaran et al., 2014; Meza-Romero et al., 2016; Sun et al., 1996). Plasma concentrations of P. vivax MIF (PvMIF) correlated with parasitaemia and severity of the disease in P. vivax malaria patients (Han et al., 2010). Furthermore, Plasmodia spp. MIFs have been recognized as candidate malaria vaccines that protect against severe malaria infection (Baeza Garcia et al., 2018). Additionally, the suppression of MIF–CD74 signaling by Plasmodia spp. MIF was protective against severe malaria infection (Baeza Garcia et al., 2021). As part of ongoing studies to clarify the structures and functions of P. vivax proteins that are therapeutic targets, we present here the cloning, purification, crystallization and structure of PvMIF.

2. Materials and methods

2.1. Macromolecule production

Cloning, expression and purification followed standard SSGCID protocols, as described previously (Serbzhinskiy et al., 2015; Kimble et al., 2024; Srivastava et al., 2024). Briefly, the full-length gene for putative macrophage migration inhibitory factor from P. vivax Salvador I (UniProt A5K093), encoding amino acids 1–116 (PvMIF), was PCR-amplified from cDNA using the primers in Table 1. The gene was cloned into pET-28a, which encodes an N-terminal His-tag and adds 21 additional residues to the expressed protein. Chemical transformation of competent Escherichia coli BL21(DE3) Rosetta cells with the resulting plasmid was followed by small-scale expression tests. 2 l of culture was then grown using established SSGCID protocols.

Table 1
Macromolecule-production information

Source organism	Plasmodium vivax Salvador I
DNA source	Wes Van Voorhis, UW PPG
Forward primer	5′-ATGCCCTGCTGTCAGGTTAGCA-3′
Reverse primer	5′-CCCAAACAGAGAGCCATTAAACG-3′
Expression vector	pET-28a, AVA N-terminal tag
Expression host	E. coli BL21(DE3) Rosetta
Complete amino-acid sequence of the construct produced†	MAHHHHHHMGTLEAQTQGPGSMPCCQVSTNINASDDDAKKALSQIENAISQVLGKPLGYIMSNLDYQKHMRFGGSHDGFCFVRVTSLGGINKSNNSSLADKITKILASTLNVKSERVFIEFKDCSAQNFAFNGSLFG

†Additional vector-derived N-terminal residues are underlined and in bold.

His-PvMIF was purified in a two-step protocol consisting of an immobilized metal (Ni²⁺) affinity chromatography (IMAC) step and size-exclusion chromatography (SEC) at 4°C. All chromatography runs were performed on an ÄKTApurifier 10 (GE Healthcare) using automated IMAC and SEC programs (Bryan et al., 2011). Thawed bacterial pellets (∼25 g) were lysed by sonication in 200 ml lysis buffer [25 mM HEPES pH 7.0, 500 mM NaCl, 5%(v/v) glycerol, 0.5%(w/v) CHAPS, 30 mM imidazole, 10 mM MgCl₂, 1 mM TCEP, 250 µg ml⁻¹ AEBSF, 0.025%(w/v) sodium azide]. After sonication, the crude lysate was treated with 20 µl Benzonase (25 U µl⁻¹) and incubated with mixing at room temperature for 45 min. The lysate was clarified by centrifugation at 11 850g for 1 h using a Sorvall centrifuge (Thermo Scientific). The clarified supernatant was then passed over an Ni–NTA HisTrap FF 5 ml column (GE Healthcare) which had been pre-equilibrated with loading buffer [25 mM HEPES pH 7.0, 500 mM NaCl, 5%(v/v) glycerol, 30 mM imidazole, 1 mM TCEP, 0.025%(w/v) sodium azide]. The column was washed with 20 column volumes (CV) of loading buffer and was eluted with loading buffer plus 250 mM imidazole in a linear gradient over 7 CV. Peak fractions were pooled and concentrated to 5 ml for SEC. A SEC column (Superdex 75, GE) was equilibrated with SEC running buffer [20 mM HEPES pH 7.0, 300 mM NaCl, 5%(v/v) glycerol, 1 mM TCEP]. The peak fractions were collected and analyzed for PvMIF using SDS–PAGE. The protein eluted as a single, monodisperse peak of ∼19 kDa in SEC running buffer. Based on molecular standards, the expected monomer molecular weight is ∼15 kDa. The peak fraction was pooled and concentrated to 12 mg ml⁻¹ using an Amicon purification system (Millipore). Aliquots of 110 µl were flash-frozen in liquid nitrogen and stored at −80°C until use. The expression clone (HepyC.00487.a.B1.GE40934) and recombinant protein (PlviB.00834.a) can be requested from SSGCID (https://www.ssgcid.org/available-materials).

2.2. Crystallization

His-PvMIF crystals were grown by vapor diffusion directly from the Index (Hampton Research) crystallization screen condition as described in Table 2.

Table 2
Crystallization

Method	Vapor diffusion, sitting drop
Plate type	96-well compact Rigaku
Temperature (K)	291
Protein concentration (mg ml^−1)	12
Buffer composition of protein solution	20 mM HEPES pH 7.0, 300 mM NaCl, 5%(v/v) glycerol, 1 mM TCEP
Composition of reservoir solution	25%(w/v) polyethylene glycol 3350, 0.1 M sodium acetate trihydrate pH 4.5
Volume and ratio of drop	0.2 µl:0.2 µl
Volume of reservoir (µl)	40
Cryosolution	20%(w/v) polyethylene glycol 3350, 0.08 M sodium acetate trihydrate pH 4.5, 20%(v/v) glycerol

2.3. Data collection and processing

Data were collected at 100 K using a Dectris EIGER2 XE 9M detector on NSLS-II beamline 19-ID at Brookhaven National Laboratory (Table 3). Data were integrated with XDS (Kabsch, 2010) via autoPROC (Vonrhein et al., 2024) and scaled with AIMLESS (Evans, 2011). Raw X-ray diffraction images have been stored at the Integrated Resource for Reproducibility in Macromolecular Crystallo­graphy at https://www.proteindiffraction.org.

Table 3
Data collection and processing

Values in parentheses are for the outer shell.

Diffraction source	Beamline 19-ID, NSLS-II
Temperature (K)	100
Detector	Dectris EIGER2 XE 9M
Space group	P63
a, b, c (Å)	75.46, 75.46, 36.95
α, β, γ (°)	90, 90, 120
Resolution range (Å)	37.73–1.80 (1.84–1.80)
Total No. of reflections	154623 (9558)
Completeness (%)	100.0 (100.0)
Multiplicity	13.7 (14.1)
〈I/σ(I)〉	11.2 (1.7)
Rr.i.m.	0.124 (1.948)
Rp.i.m.	0.034 (0.514)
CC1/2	0.994 (0.664)

2.4. Structure solution and refinement

The structure was determined by molecular replacement with Phaser (McCoy et al., 2007) from the CCP4 suite of programs (Collaborative Computational Project, Number 4, 1994; Krissinel et al., 2004; Winn et al., 2011; Agirre et al., 2023). The molecular-replacement search model was PDB entry 2wkf (Dobson et al., 2009). As with other SSGCID structures, refinement involved iterative cycles in Phenix (Adams et al., 2011; Liebschner et al., 2019) followed by manual rebuilding of the structure using Coot (Emsley & Cowtan, 2004; Emsley et al., 2010). The quality of the structure was checked with MolProbity (Williams et al., 2018). Data-reduction and refinement statistics are shown in Table 4. Coordinates and structure factors have been deposited in the Worldwide PDB (wwPDB) as entry 9b0m (pdb_00009b0m).

Table 4
Structure refinement

Value in parentheses are for the outer shell.

Resolution range (Å)	37.73–1.80 (1.84–1.80)
No. of reflections, working set	11293 (2664)
No. of reflections, test set	534 (123)
Final Rcryst	0.187 (0.283)
Final Rfree	0.218 (0.287)
No. of non-H atoms
Protein	734
Ligand	13
Water	19
Total	766
R.m.s. deviations
Bond lengths (Å)	0.009
Angles (°)	0.944
Average B factors (Å^2)
Protein	54.9
Ligand	78.4
Water	50.3
MolProbity statistics
Protein geometry
Poor rotamers (%)	0
Favored rotamers (%)	91
Ramachandran outliers (%)	0
Ramachandran favored (%)	100
Ramachandran distribution Z-score	1.96 ± 0.82
C^β deviations > 0.25 Å (%)	0
Bad bonds (%)	0
Bad angles (%)	0
Peptide omegas cis-prolines (%)	0
Additional validations	0/117

3. Results and discussion

The recombinant protein screened for crystallization included 21 additional amino-acid residues at the N-terminus corresponding to the purification tag and cleavage site, resulting in a 137-amino-acid polypeptide (Table 1). The protein crystallized directly from the screen conditions, and the structure of His-PvMIF was determined in space group P6₃, with a single monomer in the asymmetric unit (Fig. 1a). Only 96 of the 137 amino acids had ordered electron density. The final refined model is missing 41 amino-acid residues corresponding to the 21 N-terminal vector-derived residues, a five-residue loop 66-LGGIN-70 and the C-terminal loop 101-DCSAQNFAFNGSLFG-115. This apo structure contains no biologically relevant ligands, as it was neither soaked nor co-crystallized with any such ligands. The secondary structure of PvMIF was determined by PDBSum (https://www.ebi.ac.uk/thornton-srv/databases/pdbsum/) as 24% strand, 36.6% α-helix and 5.2% 3₁₀-helix.

Figure 1 The PvMIF monomer. (a) Cartoon representation of a PvMIF monomer colored in rainbow from blue at the N-terminus to red at the C-terminus. (b) Superposed monomers of PDB entry 9b0m (gray) with PDB entries 9mf3 (lemon green), 2wkf (magenta), 4p7m (tan), 3gac (wheat), 2wkb (cyan), 4dh4 (slate blue), 8ap3 (orange), 8ur2 (green), 8vj2 (smudge green) and 2xcz (yellow), the orthologues identified by ENDScript that were used to generate the sausage plot (Dobson et al., 2009; Pantouris et al., 2014; Shao et al., 2010; Sommerville et al., 2013; Spiller et al., 2023; Srivastava et al., 2024; Kimble et al., 2024; Wasiel et al., 2010). (c) PvMIF sausage plot generated by ENDScript. The ribbon (sausage) shows the relative conservation of secondary structure compared with the closest MIF structures. The ribbon thickness reflects the degree of secondary-structure similarity, with thinner ribbons indicating more conserved regions and thicker ribbons indicating lower conservation, as determined by r.m.s.d. alignment of the structures in (c). The ribbon is colored based on sequence conservation, with red indicating identical residues (a), (b) and (c) show the same view and orientation of PvMIF. The C-terminus and N-terminus are indicated by C and N, respectively.

The most similar structures to PvMIF were identified by PDBeFold (https://www.ebi.ac.uk/msd-srv/ssm/) analysis (Krissinel & Henrick, 2004) using default protein secondary-structure element thresholds of 70% (https://www.ebi.ac.uk/pdbe/sites/default/files/documents/service_tutorials/PDBeFold.pdf) as MIF structures (PDB entries 2wkf, 4p7m and 4p7s) from P. falciparum (PfMIF) with an r.m.s.d. of ∼0.7 Å (Dobson et al., 2009; Pantouris et al., 2014). This is unsurprising since PfMIF shares ∼66% sequence identity with PvMIF. The next most similar structure is that of a plant MIF-like protein (PDB entry 8dq6), with an r.m.s.d of ∼1.1 Å but only ∼23% identity to PvMIF (Spiller et al., 2023). Apart from the plant MIF-like protein, the most similar structures include those of a MIF from another Plasmodium spp., P. berghei, which shares ∼64% sequence identity with PvMIF. Structures of MIFs from the rodent parasite P. yoelii (PyMIF; PDB entries 3gac and 3gad) with an r.m.s.d. of ∼0.7 Å and sharing ∼63% identity with PvMIF (Shao et al., 2010) ranked next in similarity (a summary of the outcome of PDBeFold is listed in Supplementary Table S1).

Similar results were obtained with ENDScript using a BLAST+ search performed against the PDBAA database and BLAST+ hits aligned using ClustalO (Gouet et al., 2003; Robert & Gouet, 2014). ENDScript revealed that the structure of PvMIF is most similar to those from other Plasmodium spp., including P. knowlesi (PDB entry 9mf3; Seattle Structural Genomics Center for Infectious Diseases, unpublished work), P. falciparum (PDB entries 2wkf and 4p7m; Dobson et al., 2009; Pantouris et al., 2014), P. yoelii (PDB entry 3gac; Shao et al., 2010) and P. berghei (PDB entry 2wkb; Dobson et al., 2009). The next most similar structures are the MIFs from the protozoan parasite Toxoplasma gondii (PDB entry 4dh4; Sommerville et al., 2013) and a plant MIF-like protein (PDB entry 8ap3; Spiller et al., 2023). These are followed by the structures of Trichomonas vaginalis MIF (PDB entry 8ur2; Srivastava et al., 2024), Onchocerca volvulus MIF (PDB entry 8vj2; Kimble et al., 2024) and Prochlorococcus marinus MIF (PDB entry 2xcz; Wasiel et al., 2010). Comparison of the superposed MIF structures reveals that the missing C-terminal residues in PvMIF correspond to a loop (Fig. 1b). An ENDScript-generated sausage plot reveals that PvMIF is a prototypical MIF, with greatest structural variation in the carboxyl-termini, as indicated by the ribbon thickness in the sausage plot (Fig. 1c). The regions with identical amino-acid residues are shown in red; notably, the N-terminal proline at the tautomerase site is conserved across all structures. Sequence-based secondary alignment also reveals the alignment of secondary-structure elements across all MIF structures, highlighting variations in strand, helix and loop lengths, especially at the carboxyl-termini (Fig. 2).

Figure 2 ENDScript with a BLAST+ search performed against the PDBAA database and BLAST+ hits aligned using ClustalO reveals the nearest structural neighbors of PvMIF. Identical and conserved residues are highlighted in red and yellow, respectively. The different secondary-structure elements shown are α-helices (α), 310-helices (η), β-strands (β) and β-turns (TT) (Gouet et al., 1999, 2003). Regions corresponding to the MIF1 residues implicated in CD74 binding are shown in purple boxes.

CD74 binding by parasite MIFs facilitates their suppression of human MIF–CD74 signaling (Baeza Garcia et al., 2021; Leng et al., 2003). ESPript analysis with Clustal Omega alignment of PvMIF with PfMIF and human MIF1 and MIF2 enables the comparison of residues implicated in CD74 binding (Fig. 3). CD74 binding requires the formation of the prototypical MIF trimer (Leng et al., 2003; Meza-Romero et al., 2016). Experimental and docking studies of human MIF1 (hMIF1) implicate chain A residues 50-FGGSEP-55, K76, 90-SPDR-93 and 109-NNS-111 and chain B residues 34-PQ-35, 108-WNN-110 and 111-STFA-114 in CD74 binding (Meza-Romero et al., 2016). Only five of the MIF1 residues implicated in CD74 binding are conserved in both PvMIF and PfMIF. The five residues are Phe50, Gly51, Gly52, Ser53 and Arg94 in MIF1, which correspond to Phe51, Gly52, Gly53, Ser54 and Arg95 in PvMIF and PfMIF (Fig. 3).

Figure 3 ESPript structure-based alignment with Clustal Omega sequence alignment of PvMIF (PDB entry 9b0m), human MIF1 (PDB entry 1mif), human MIF2 (PDB entry 1ddt) and PfMIF (PDB entry 2wkf). Human MIF residues implicated in CD74 binding are indicated with green boxes. The secondary-structure elements are as follows: α-helices are shown as large coils, 310-helices are shown as small coils labeled h, β-strands are shown as arrows labeled β and β-turns are labeled TT. The identical residues are shown in a white font on a red background, conserved residues in a red font and conserved regions in blue boxes. The sequence alignment was performed using ClustalW and was followed by ESPript structure alignment.

The protomer of PvMIF forms the prototypical MIF trimer with two symmetry mates (Fig. 4). The PvMIF trimer aligns well with hMIF1 (r.m.s.d. of ∼1.1 Å for 210 residues), MIF2 (r.m.s.d. of ∼11.3 Å for 213 residues) and PfMIF (r.m.s.d. of ∼10.5 Å for 240 residues) as measured by PyMOL. Additionally, the PvMIF trimer retains a similar interface and packing to hMIF (Sun et al., 1996) and hMIF2 (Rajasekaran et al., 2014), thereby allowing the loop regions and residues implicated in CD74 binding to remain exposed.

Figure 4 (a) Cartoon representation of the biological unit of PvMIF (gray) reveals a prototypical MIF trimer. (b) Superposed trimers of PvMIF (PDB entry 9b0m, gray) with human MIF1 (hMIF1; PDB entry 1mif, purple), human MIF2 (hMIF2; PDB entry 3ker, cyan) and PfMIF (PDB entry 4p7m, yellow). The prototypical MIF trimers of (c) hMIF1 (PDB entry 1mif, purple; Sun et al., 1996), (d) hMIF2 (PDB entry 3ker, cyan; Rajasekaran et al., 2014) and (e) PfMIF (PDB entry 4p7m) are shown. Proteins were aligned with PyMOL (DeLano, 2002).

4. Conclusion

PvMIF shares primary-, secondary-, tertiary- and quaternary-structural features with human MIFs. Five of the human MIF1 amino-acid residues implicated in CD74 binding are conserved in PvMIF. Additional studies are required to determine how PvMIF binds CD74.