1. Temperature B-factor analysis used in the paper is based on the premises that (1) knot knots contains no more than 12% of non-proton atoms, and (2) there are three or four such atoms per residue. According to (1) and (2), the percentage of non-proton atoms in knot essentially ranges from 6 to 11% in our analysis.

It is important to note that our analysis of knot substructure considers the

range that covers knot atoms or residues, not the number of knot atoms. Because

of considering the range rather than the number, the question of how many

percentage of non-proton atoms exactly belonging to knot atoms in either a

mesophile or a thermophile is not a major issue as far as our conclusions

regarding knot substructure are concerned.

The above point is further confirmed by our consistent findings with all

four pairs of thermophile-mesophile comparisons. Regardless whether knot atoms

in thermophile as defined above have a moderately percentage of non-proton

atoms than those in mesophile or vice versa, our conclusions regarding distinctive

knot substructuresbetween thermophile and mesophile are the same. No such

a distinction is found in either halophile-mesophile or psychrophile-mesophile

comparison.

  1. As we know, a key issue in X-ray crystallography is to obtain enough proteins for the structural determinations. In the majority of cases, X-ray people cannot purify enough “native”proteins directly from issues or organisms. To overcome this problem, they use recombinant techniques. This makes it possible forthem to get enough amounts of proteins. Suchobtained proteins have been demonstrated to have identical sequences, same folded structures, and exhibitenzyme activity as the “native” proteins. Alarge majority of proteins reported in PDB are prepared using recombinant techniques, and are called as being “engineered” in PDB. They are different from mutant proteins.

Mutant proteins are modified proteins, where one or more amino acids

have been replaced. Site-specific mutagenesis is involved in their modification.

They are called “mutant” proteins to distinguish from the unmodified “wild” type

proteins. Mutant proteins do not have identical amino sequences or folded

structures as native proteins. Most of “mutant” proteins are lack of enzyme

activity. It should be noted that “mutant” and “engineered” are different things.

The source of proteins reported in PDB for mesophilic or thermophilic

proteins are listed as with or without “engineered”. In the four pairs of proteins

used in the paper for mesophile-thermophile comparisons, the proteins have a

different combination of sources: without “engineered” for mesophile and with

“engineered” for thermophile, with “engineered” for meosphile and without

“engineered” for thermophile, and with “engineered” for both mesophile

and thermophiles. However, regardless which of the above sources, consistent

results and conclusions areobtained for all four pairs. This evidenceconfirms that

an inclusion of “engineered” proteins in our temperature B-factor analysisshould

not be an issue of concern.

  1. The issue regarding the heat of mixing of ethanol with water close to zero at 354oK may affect the idea of matrix contraction at higher temperature in the hyperthermophile. Hence, it is an issue of concern and requires the attention.

However, figures 9-11 in the paper compare the changes of B-values from

knot to matrix substructure between mesophile and thermophile at 4oC or room

temperature. The results show that matrix substructure is less expanded in

thermophile than mesophile atthat temperature. Other figures are limited to

address knot substrutures only. Thus, our findings in the paper are unable to

address the question of matrix contraction - expansion at higher temperature in

hyperthermophiles. This issue is beyond the scope of our studies in the paper.

  1. This paper reveals distinctive substructures of knot atoms and residues between thermophilc and mesophilic proteins. Our findings are important in providing a new insight into the stability of thermophiles and the role of knot substructure. Nevertheless, the results and conclusions obtained in the paper can not be analyzed or synthesized to address a larger issue beyond the scope of this investigation.

In knot substructure, actually, we do not really know exactly how many

atoms per residue: four, three or two per residue? However, we know that there

are four atoms participating in the peptide bond. They are C, O, N and C (alpha).

When these four atoms are all knot atoms, they should have lowest B-values.

Therefore,fouratoms per residue should be investigated. To expand the range of

knot atoms, three atoms per residue are also investigatedin the paper.Consistent

results and conclusions are obtained in both cases, suggesting thatour findings are

not a result of arbitrary choice of the number of atoms per residue. Moreover, it

should be noted that our analysis of knot substructure is based the range that

covers knot atoms or residues, rather than the number of knot atoms (see

comment 1 above).