On 2016 May 11, Daniel Haft commented:

This crystallography paper, from 1998, reports a structure for an aminoglycoside-(3)-N-acetyltransferase whose sequence is shown in GenBank record AAB20441.1. It’s interesting to compare AAB20441.1 to a very different aminoglycoside-modifying enzyme, CAG34229.1, the nucleotidyltransferase ANT(2'')-Ia. These two translations are identical over the first twenty amino acids, MLRSSNDVTQQGSRPKTKLG, but are otherwise unrelated. As is readily apparent from examining AJ746361, the source nucleotide record for the latter protein, both these antibiotic resistance genes occur integrated into integrons of the same family, called class 1.

The N-terminal sequence shared by otherwise unrelated antibiotic resistance genes, starting with a plausible-looking ATG-encoded Met in the appropriate reading frame, raises the question of whether this region might actually be translated, and what its contribution to protein structure might be. This crystallography paper is interesting because the extended region was included (along with an additional engineered N-terminal prefix that aided in protein purification) when the enzyme was expressed for the crystallography study, and therefore was studied experimentally. The authors found the N-terminal region to contain sites “that are exquisitely sensitive to trypsin (Arg-3, Arg-14, Lys-16, and Lys-18), suggesting that the N terminus of the enzyme is … disordered.” The N-terminal extension clearly did not participate in forming an ordered crystal structure, and seemed to neither help nor hinder enzymatic activity.

A number of additional unrelated antibiotic resistance genes occur in class 1 integrons and appear in public sequence databases with (probably faulty) translations that start from the same integron-derived candidate start site, resulting in similar N-terminal sequence extensions. Examples include a class A beta-lactamase (BAE71359.1), a trimethoprim-resistant dihydrofolate reductase (BAD07295.1), and a rifampin ADP-ribosyltransferase (CAR63501.1). Readers of this paper may enjoy knowing that the N-terminal sequence extensions shared by these translations reflect integration of unrelated genes at equivalent sites, not conservation of some structural element that would be visible in solved crystal structures.

On 2016 May 11, Daniel Haft commented:

This crystallography paper, from 1998, reports a structure for an aminoglycoside-(3)-N-acetyltransferase whose sequence is shown in GenBank record AAB20441.1. It’s interesting to compare AAB20441.1 to a very different aminoglycoside-modifying enzyme, CAG34229.1, the nucleotidyltransferase ANT(2'')-Ia. These two translations are identical over the first twenty amino acids, MLRSSNDVTQQGSRPKTKLG, but are otherwise unrelated. As is readily apparent from examining AJ746361, the source nucleotide record for the latter protein, both these antibiotic resistance genes occur integrated into integrons of the same family, called class 1.

The N-terminal sequence shared by otherwise unrelated antibiotic resistance genes, starting with a plausible-looking ATG-encoded Met in the appropriate reading frame, raises the question of whether this region might actually be translated, and what its contribution to protein structure might be. This crystallography paper is interesting because the extended region was included (along with an additional engineered N-terminal prefix that aided in protein purification) when the enzyme was expressed for the crystallography study, and therefore was studied experimentally. The authors found the N-terminal region to contain sites “that are exquisitely sensitive to trypsin (Arg-3, Arg-14, Lys-16, and Lys-18), suggesting that the N terminus of the enzyme is … disordered.” The N-terminal extension clearly did not participate in forming an ordered crystal structure, and seemed to neither help nor hinder enzymatic activity.

A number of additional unrelated antibiotic resistance genes occur in class 1 integrons and appear in public sequence databases with (probably faulty) translations that start from the same integron-derived candidate start site, resulting in similar N-terminal sequence extensions. Examples include a class A beta-lactamase (BAE71359.1), a trimethoprim-resistant dihydrofolate reductase (BAD07295.1), and a rifampin ADP-ribosyltransferase (CAR63501.1). Readers of this paper may enjoy knowing that the N-terminal sequence extensions shared by these translations reflect integration of unrelated genes at equivalent sites, not conservation of some structural element that would be visible in solved crystal structures.