LPS ligand and culture additives improve production of monomeric MD-1 and 2 in Pichia pastoris by decreasing aggregation and intermolecular disulfide bonding

Abstract

Myeloid differentiation proteins MD-1 and MD-2 have both been shown to form a heterogeneous collection of oligomers when expressed in absence of their respective receptor, RP105 and TLR4. The biological relevance of these oligomers is not clear. Only monomeric proteins have been found to be active and able to trigger an immune response to endotoxin by modulating the TLR4 pathway. In this study, we produced variants of MD-1 and MD-2 in Pichia pastoris. To minimize the time and expense of initial expression tests, small-scale cultures have been set up to allow the rapid identification of the highest expressing clone and the optimal expression conditions. The expression vectors used, the site of linearization and the locus of integration affected the yield of transformation. Next we screened culture additives and found that they significantly increased the fraction of monomeric proteins secreted in the culture medium (up to 15% of the total MD protein produced). We confirmed their presence by size-exclusion chromatography. Optimal anti-aggregation agents were protein-dependent except for LPS that presented stabilizing effects for all MD proteins. Contrary to previous reports, this study suggests that MD-1 can bind to LPS.

Fig. 1

Optimal plasmid fragments for Pichia transformation. Several combinations of linearized plasmids were transformed into Pichia strain GS115. pPIC9K linearized at its 5′AOX1 (with SacI and PmeI) or HIS4 locus (with SalI) were transformed, along with pPICZα linearized at its 5′AOX1 locus (with SacI). Transformation reaction 1: pPIC9K blunt ends at 5′AOX1 with PmeI; reaction 2: pPIC9K sticky ends at 5′AOX1 with SacI; reaction 3: pPICZα sticky ends at 5′AOX1 with SacI; reaction 4: double selection with pPIC9K sticky ends at HIS4 with SalI and pPICZα sticky ends at 5′AOX1 with SacI; reaction 5: triple selection with pPIC9K sticky ends at HIS4 with SalI and both pPIC9K and pPICZα with sticky ends at their respective 5′AOX1 loci with SacI. The number of antibiotic-resistant colonies isolated after transformation of each combination of linearized plasmids was recorded.

Fig. 2

Effect of His phenotype on protein expression. (A) 48 colonies derived from transformation reaction 1 (His⁺ phenotype) were tested for expression levels in small-scale trials. (B) 48 colonies derived from transformation reaction 3 (His⁻ phenotype) were tested for expression levels in small-scale trials. The supernatant from each colony was analyzed by dot blot, and each signal was quantified as an integrated density percent of the total signal using ImageJ®.

Fig. 3

Time course of expression. A typical time course of expression was performed on a single colony for each recombinant Pichia strain. Samples were taken at various time points throughout expression over a week and analyzed by anti-His Western blot in reducing conditions. Supernatants (A) and cell lysates (B) from GS115/hMD-2 cultures are represented here.

Fig. 4

Single colony screening for the detection of the highest-expressing colony. 48 recombinant colonies of strain GS115 and KM71H expressing each variant were induced in small-scale expression screens, and the relative expression level of each colony was compared by anti-His dot blotting. Colonies 1–24 (A) and 25–48 (B) expressing an MD variant in GS115 were analyzed, along with colonies 1–24 (C) and 25–48 (D) expressing the same MD variant in KM71H. The highest-expressing colony of each strain, which was selected for all future expression studies, is circled. The MD variant shown here is HE82–89.

Fig. 5

Optimization of induction. The highest-expressing colony of each protein in each strain was expressed in 24 different culture conditions (see Supplementary Table 1 for culture medium composition), on a small scale, and supernatants were analyzed by dot blotting followed by signal quantification with ImageJ®. The relative expression in each condition is indicated by the integrated density percent of the total signal present for proteins expressed in Pichia strain GS115 (left panels) and KM71H (right panels). (A) Effect of pH. Values between 5.0 and 8.0 have been tested in BMMY in one unit increments. (B) Effect of methanol. Daily addition of increasing concentrations (0.5%; 1.0%; 1.5%; 2.0%) of methanol have been tested. (C) Effect of yeast extract and peptone. Complex BMMY and minimal buffered media in absence and in the presence of 1% glycerol, BMG and BMM, respectively, have been compared here. (D) Effect of glucose. Increasing concentrations (0%; 0.1%; 0.2%; 0.5%) of glucose have been tested in either BMMY or MM for each construct as indicated. (E) Effect of glycerol. Increasing concentrations (0%; 0.1%; 0.2%; 0.5%) of glycerol have been tested in either BMMY or MM on each construct.

Fig. 6

Effect of culture media additives on the prevention of intermolecular disulfide formation. A recombinant colony of each construct was induced in the absence of culture additives (−) and in the presence of a given anti-aggregation agent at the following concentrations: 0.2 M MgSO₄; 0.15 M (NH₄)₂SO₄; 0.5 M NDSB-201; 1.5% glycine; 0.1 M NaSCN; 0.5 M KCl; 0.01 mg/ml LPS from E. coli O127:B8 (Sigma L3129). Samples of supernatants harvested from each construct were analyzed by Western blotting in non-reducing conditions. (A) GS115 cultures and (B) KM71H cultures.

Fig. 7

Inclusion of additives in culture medium and purification buffers allows isolation of monomeric protein. Typically, expression cultures were prepared, and the harvested supernatant was concentrated and buffer-exchanged by tangential flow filtration prior to protein purification. After Nickel affinity chromatography, the sample was concentrated, and purified by gel filtration on a HiLoad 16/60 Superdex 200 column. (A) In absence of additive proteins from an 8L-culture of GS115–hMD-2 in BMMY medium eluted in the void volume only (blue chromatogram; aggregates eluted at 43.3 ml). Optimization of the culture medium (MM + 0.2% glycerol instead of BMMY) and inclusion of 0.2 M MgSO₄ significantly improved the yield of monomeric hMD-2 protein production (red chromatogram; aggregates eluted at 42.9 ml and monomers at 96.8 ml). The nature of the second peak was confirmed by anti-His dot blotting. (B) Gel-filtered proteins taken from both peak fractions with elution profiles corresponding to aggregates and monomeric proteins were analyzed on a non-reducing Coomassie-stained SDS–PAGE gel.