As the first author on K. honarmand Ebrahimi et al, PLoS ONE (2012) who did all the experiments, I should mention that we were puzzled by all the mistakes and inconsistencies that we observed in the work of Duce et al, Cell (2010). Some of these are listed below and have been mentioned in our work.
We have recently obtained the construct for production of E2 domain of APP. We have plans to study the ferroxidase activity of E2 domain and we will publish our results on this topic in future.

The bottom line of our work is the suggestion that other groups who are working on the Alzheimer disease and who work with APP, should re-evaluate the work that is done by Duce et al, under appropriate experimental conditions and with appropriate experimental tools to prove or disprove the ferroxidase activity of APP as claimed by Duce et al.

More comments on the work published by Duce et al:

1. the ferroxidase activity reported by Duce et al for the E2 domain is circa 10 µM/min in Figure 2A, but it is 16 µM/min in Figure 2E. The experimental conditions appear to be similar, and this means an uncertainty of circa 50%.

2. The activity of E2 domain (Figure 2A and Figure 2E) is half that of the activity of the FD1 peptide (Figure 2D). This is inconsistent with the statement in the text that both FD1 and the E2 domain had 40% of the activity of APP.

3. the maximum activity in Figure 2C for the FD1 peptide (from the data points) is 70 µM/min/µM (70*0.2=14 µM) for Fe(II) concentrations above 70 µM, but in Figure 2D it is circa 27 µM/min for an Fe(II) concentration of 110 µM. This is again an uncertainty of 50%.

4. the pH used by Bakers & Boyer in 1986 was 6.0, however Duce et al have used a pH of 7.2 for the ferroxidase activity of APP, CP, E2 domain, and the FD1 peptide. It is possible that Duce et al have followed the procedure of Baker & Boyer (pH 6) to measure the background oxidation of Fe(II) by molecular oxygen and incorporation of Fe(III) into transferrin, and they have used pH 7.2 to measure the ferroxidase activity of other proteins and the FD1 peptide. At pH 6.0 we have observed the non-enzymatic oxidation of Fe(II) in the presence of transferrin to be negligible. But this is certainly not true at pH 7.2

5. The final concentration of Hepes buffer was not 50 mM. Duce et al have added 200 µl of HBS buffer (150 mM NaCl and 50 mM Hepes) to a final volume of 1000 µl. This means a final buffer concentration of 10 mM, i.e. a very limited buffer capacity. pH change affects the measurements significantly.