Lightrun

The gene encoding the mature form of human FGF21 was optimized for expression in Escherichia coli, fused with N-terminal His-tag, commercially synthetized (GeneArt), and subcloned into the pET3c expression vector. Mutant recombinant genes were constructed using the NEBaseChanger v1.2.9 method and Q5 high-fidelity polymerase (New England Biolabs) according to the web program's instructions and the polymerase manufacturer's recommendations. Mutagenesis of FGF21 was done by whole plasmid PCR using the plasmid pET3c::His-fgf21, as a template and two reverse-complement oligonucleotides carrying desired mutations. The PCR products were transformed into NEB5α(DH5α) chemically competent E. coli cells. The resulting plasmids were isolated from overnight cultivation of DH5α cells using a GeneJet plasmid miniprep kit (Thermo Fischer Scientific). Multiple-point mutants (three-point mutant, FGF21–3PM; four-point mutant, FGF21–4PM; five-point mutant, FGF21–5PM) were constructed using the same protocol as single-point mutants. The following plasmids pET3c::His-fgf21-Q104M, pET3c::His-fgf21–3PM, and pET3c::His-fgf21–4PM were used as templates for the construction of FGF21–3PM, FGF21–4PM, and FGF21–5PM, respectively. The nucleotide sequences of the constructed FGF21 variants were verified commercially by the Sanger sequencing (LightRun, Eurofins Genomics).

Viral genes UL27 and US7, encoding the glycoproteins B (gB) and I (gI), respectively, were sequenced for areas with maximal variability between 17+ (JN555585.1) and the recently full-genome sequenced Finnish clinical isolates [3 (link)]. The primers designed for UL27, 5′-CGGTGGTCTCCAGGTTGTTG-3′ (reverse) and 5′-TGGTCTACGACCGAGACGTT-3′ (forward), were used to sequence nucleotides 55,132–155,897 according to the numbering of the 17+ strain sequence, JN555585.1. The primers designed for US7, 5′-ACGTGTTACGCGTATGGGTC-3′ (forward) and 5′-TATACCAACAGGGGAGGCGT-3′ (reverse), were used to sequence nucleotides 140,222–140,960 according to the numbering of the 17+ strain sequence JN555585.1. The target sequences were amplified from viral stocks using Phusion polymerase (Thermo Scientific, Waltham, MA, USA). The product from the PCR was confirmed with an analytical agarose gel run. The resulting DNA from the PCR run was purified using a GeneJET PCR Purification Kit (ThermoFisher, Waltham, MA, USA), according to the manufacturer’s protocol. The purified DNA was sequenced by LightRun, Eurofins Genomics, Ebersberg, Bayern, Germany.

Variants of OmpX from E. coli containing loop inserts were generated by blunt-end ligation of PCR products. A pET3b plasmid containing ompX without the signal peptide, a T7 promotor and ampicillin resistance (Pautsch et al., 1999 (link)) or an identical plasmid containing the complete OmpX sequence (with signal peptide (Arnold et al., 2007 (link))) was used as a template for the PCR. Primers were designed to amplify the plasmid with overhangs containing the intended insert. Primers are listed in Supplementary Material S1 (Supplementary Table S1). Ligation products were transformed into calcium-competent E. coli TOP10 cells (Invitrogen) and the DNA sequence of the inserts was confirmed with Sanger sequencing (LightRun, Eurofins).
For in vivo studies of OmpX folding, a variant of E. coli BL21 had to be generated that lacked the wildtype OmpX gene. This was done using P1 transduction as described (Saragliadis et al., 2018 (link)), with E. coli BL21 as the acceptor strain and an OmpX knockout from the Keio collection (Baba et al., 2006 (link)) as the donor strain. This gave rise to the kanamycin-resistant strain E. coli BL21ΔOmpX. Absence of the ompX gene was verified by Western blotting of whole-cell lysates using an OmpX rabbit antiserum (Arnold et al., 2007 (link)) and E. coli BL21 as a positive control (data not shown).