What They Saw
They sequenced the DNA upstream of hupDE and found four open reading frames, which they called hupABYC. The AB and C were similar to E. coli genes, but the Y wasn't, so they called it Y (for Ynknown, I guess). These were all homologous to A. vinelandii genes though, and in the same order.
Then they tried knocking out each of these. Each knockout was unable to oxidize hydrogen, even in the presence of methylene blue as an electron acceptor.
They also made a fusion of HupL (the structural subunit) and beta-galactosidase, then knocked out hupY or hupB to see if this changed the expression of hupL. Beta-galactosidase activity rose a little bit, like 25% in each, but it didn't seem either was an important regulator.