Insect cells are superior to Escherichia coli in producing malaria proteins inducing IgG targeting PfEMP1 on infected erythrocytes

Abstract

Background: The PFD1235w Plasmodium falciparum erythrocyte membrane protein 1 (PfEMP1) antigen is associated with severe malaria in children and can be expressed on the surface of infected erythrocytes (IE) adhering to ICAM1. However, the exact three-dimensional structure of this PfEMP1 and its surface-exposed epitopes are unknown. An insect cell and Escherichia coli based system was used to express single and double domains encoded by the pfd1235w var gene. The resulting recombinant proteins have been evaluated for yield and purity and their ability to induce rat antibodies, which react with the native PFD1235w PfEMP1 antigen expressed on 3D7PFD1235w-IE. Their recognition by human anti-malaria antibodies from previously infected Tanzanian donors was also analysed.

Methods: The recombinant proteins were run on SDS-PAGE and Western blots for quantification and size estimation. Insect cell and E. coli-produced recombinant proteins were coupled to a bead-based Luminex assay to measure the plasma antibody reactivity of 180 samples collected from Tanzanian individuals. The recombinant proteins used for immunization of rats and antisera were also tested by flow cytometry for their ability to surface label 3D7PFD1235w-IE.

Results: All seven pAcGP67A constructs were successfully expressed as recombinant protein in baculovirus-infected insect cells and subsequently produced to a purity of 60-97% and a yield of 2-15 mg/L. By comparison, only three of seven pET101/D-TOPO constructs expressed in the E. coli system could be produced at all with purity and yield ranging from 3-95% and 6-11 mg/L. All seven insect cell, but only two of the E. coli produced proteins induced antibodies reactive with native PFD1235w expressed on 3D7PFD1235w-IE. The recombinant proteins were recognized in an age- and transmission intensity-dependent manner by antibodies from 180 Tanzanian individuals in a bead-based Luminex assay.

Conclusions: The baculovirus based insect cell system was distinctly superior to the E. coli expression system in producing a larger number of different recombinant PFD1235w protein domains and these were significantly easier to purify at a useful yield. However, proteins produced in both systems were able to induce antibodies in rats, which can recognize the native PFD1235w on the surface of IE.

Figure 1

The structure and characteristics of PFD1235w. (A) Domain structure of PFD1235w with amino acid domain boundaries indicated. Recombinant PFD1235w protein domains produced in (B) baculovirus-infected insect cells (solid lines) and (C) E. coli (dotted lines). (D) Purity (%), (E) yield (mg/ml), (F) table of theoretical and observed molecular weight in kDa of insect cell and E. coli produced domains. Lines ending with a diamond in (B) and (C) indicates recombinant protein inducing IE surface reactive antibodies in rats and (*) constructs tested by Luminex. NM: not measured, NA: not applicable, NP: not produced.

Figure 2

Purity of recombinant PFD1235w protein domains expressed in insect cells and E. coli. (A and C) Sodium dodecyl sulfate-polyacrylamide (SDS-PAGE) gel electrophoresis of 2 μg/lane and (B and D) Western blotting of 0.2 μg/lane recombinant protein. (A and B) Insect cell produced protein (1-2) NTS-CIDR1α, (3-4) DBL1α-CIDR1α, (5-6) CIDR1α, (7-8) DBL2β, (9-10) DBL3β, (11-12) DBL4γ, (13-14) DBL5δ-CIDR2β, (15-16) DBL5δ, (17-18) CIDR2β and (C and D) E. coli produced (1-2) DBL1α-CIDR1α, (3-4) CIDR1α, (5-6) DBL2β, (7-8) DBL3β, (9-10) DBL4γ, (11-12) DBL5δ-CIDR2β, (13-14) DBL5δ. M: ProSieve Color Protein marker (Lonza). Samples were reduced using DTT (+) or non-reduced (-). Arrows indicate identified protein band.

Figure 3

IE surface reactivity of sera raised against insect cell and E. coli produced domains of PFD1235w. (A) IE surface reactivity of rat antiserum raised against recombinant PFD1235w as indicated in the legend and measured by flow cytometry. (B) Antibodies against C-terminal located PFD1235w domains show higher IE surface reactivity than antibodies against N-terminal domains. Circles and triangles in (B) show antibody reactivity of plasma from rats immunized with insect cell and E. coli produced domains, respectively. Two-three rats were immunized with different PFD1235w domains and plasma was obtained as described in the Methods section. Antibody reactivity to infected erythrocytes surface expressing PFD1235w was measured by flow cytometry and data given as MFI _sample/MFI_{sample prebleed}. The dotted line shows the cut-off defined as described in the Methods section.

Figure 4

IE surface reactivity of sera raised against insect cell produced domains of PFD1235w. Shown is flow cytometry of PFD1235w expressing IE labelled with PFD1235w domains-specific rat antisera (green histograms) or with preimmunization control sera (grey histograms). The corresponding immunofluorescence microscopy surface reactivity is shown by inserts for antisera against CIDR1α, DBL3β, DBL4γ, DBL5δ and DBL5δ-CIDR2β. Preimmunization sera and antisera against VAR2CSA [37] were negative in both assays.

Figure 5

Recognition of native PFD1235w expressed on the surface of 3D7_PFD1235w IE. The surface antibody reactivity of rat antisera raised against (A1 and B1) insect cells and (A2 and B2) E. coli produced DBL1α-CIDR1α and DBL4γ was measured by flow cytometry. (C) The antibody surface reactivity of rat antisera raised against insect cell produced DBL3β. Antisera were pre-incubated with excess homologous immunizing protein, heterologous produced protein, PBS or irrelevant protein as indicated. The percentage of reactivity left following depletion was calculated as described in the Methods section.

Figure 6

Human antibody reactivity to PFD1235w. Percentage antibody responders to recombinant PFD1235w domains produced in baculovirus-infected insect cells (A1-C1) and in E. coli (A2-C2). The antibody reactivity of plasma obtained from individuals living in three Tanzanian villages (A) Magamba, (B) Ubiri, and (C) Mgome of low, moderate, and high malaria transmission intensity was measured by Luminex. Each age group (2-4, 5-9, 10-14, and 15-19 years) included 15 individuals. The percentage responders was categorized into six different intervals: 0-15 (white), 16-30 (lighter gray), 31-45 (light gray), 46-60 (dark grey), 61-75 (darker grey), and >76 (black).