Basically, they engineered a vector which, when transfected to E. coli, produces transcripts of -and thus proteins to- a fusion gene which produces these transmembrane α helices conjugated to the staphylococcal nuclease. In other words, the E. coli are just factories for producing proteins:
Expression, Extraction, and Purification of SN/GpA - For high
levels of SN/GpA production, pT7SN/GpA was transformed into E.
coli MGT7 (kindly provided by D. LeMaster), containing the plasmid
And the reasoning therein was proposed in the discussion,
We report here the use of a chimeric protein to show that
the presence of just the transmembrane domain of GpA, fused
to a normally monomeric soluble protein, is sufficient to
mediate the dimerization of this artificial membrane protein
What they mean by readily purified is that the protein can be extracted relatively easily, and in the experimental method they explain some commonplace methods of purification:
One-step purification of milligram quantities of SN/GpA from the
extract could be achieved by reversed-phase HPLC utilizing an acetonitrile/isopropyl
alcohol/water gradient on a semipreparative Vydac
C4 column. Alternatively, purification of larger quantities was
achieved using two rounds of cation-exchange chromatography...
...The chimera SN/GpA131, but not truncated forms, could alterNaCl,
50 mM Tris-HC1, 5 mM EDTA, 1 mM PMSF, 0.025% NaNa,
natively be extracted by sonication of the pellet after cell lysis in 1 M
pH 7.9, containing no detergent. This extraction was of similar
efficiency to that with Lubrol, and was utilized in the preparation of
material for generation of transmembrane peptide by trypsin treatment.