1 March 2013
We have read with great interest the study by Rembach and colleagues  on the diagnostic value of serum copper (Cu), ceruloplasmin (CP), and ‘free’ Cu also named non-CP Cu, in the Australian Imaging Biomarkers and Lifestyle Study of Aging (AIBL). When discussing their results, these authors commented that some studies of ours on the relationship between serum non-CP Cu and Alzheimer’s disease (AD) [2-8] were produced in cohorts not adjusted for age and gender. Actually, our studies were all adjusted for age and gender, and finally they involved a large population of more than 860 individuals . In fact, to date we have collected Cu studies from a cohort of about 300 healthy controls, 400 AD, and 169 mild cognitive impairment (MCI) subjects. Specifically, beginning with our studies in 2002, we have used “laboratory reference sera” in every Cu and CP experimental run in addition to the standard controls provided by manufacturers. This procedure allowed us to merge data obtained in different runs during different years. Expressively, we have created a set of five pooled sera from about 20 healthy individuals, well defined in terms of Cu and CP, and we have divided these control pooled sera into thousands of aliquots, stored them at -80°C, and tested them in every run. As reported , the descriptive statistics of this merged dataset indicated: healthy controls (n=303; non-CP Cu mean=0.25, SD=2.32 µmol/L), AD patients (n=399; non-CP Cu mean=2.27, SD=2.29 µmol/L), and MCI subjects (n=169; Non-CP Cu mean=1.36, SD=2.13 µmol/L). After adjusting for age and gender through a general linear model, the estimates were as follows: healthy mean=0.20 (95% CI=-0.09, +0.492), AD mean=2.18 (95% CI=1.91, 2.23), MCI mean=1.38 (95% CI=1.04; 1.73) with each pairwise comparison clearly significant (p<0.001 after Bonferroni correction).
Moreover, when Rembach et al.  considered age and gender as covariates in their statistical models, their relationships observed between non-CP Cu and disease status were all abrogated, so their conclusion (see the Abstract) of decreased levels of non-CP Cu in MCI and AD appears not sustained by their data.
In all previous studies on this matter, the authors have used the equation published by Walshe to calculate non-CP Cu . Rembach et al.  questioned this equation because its utilization in the AIBL cohort produced mostly non-CP Cu values in the negative range. Specifically, they reported that the Walshe equation assumes that the amount of Cu in CP is 0.3% g/g, which they refer to be equal to 7.1 moles of Cu/CP. Actually, Walshe  clearly reported that the number of atoms (moles) of Cu in CP are 6 (in particular, 0.3% g/g of Cu in Cp corresponds exactly to 6.23 atoms which can be approximated to 6) and, thereby, their concern about the Walshe equation and its utilization appear not pertinent. From the above, equation #1 of Rembach et al. (which is actually the Walshe’s equation ) and equation #2, which these authors report to be derived from:
[CP copper]=[CP mole/L]*6 moles of copper/CP mole ,
are the same, as follows:
CP-Cu1(µM)=CP (mg/L) * 1000 *(0.3/100)*[1/63.6 (g/mol)]=
=CP (mg/dL) *10*1000* (0.3/100)*[1/63.6(g/mol)]=
= CP (mol/L)* (6moles Cu/1moles Cp) *10(6)
So, also non-CP Cu (µM)1 and non-CP Cu (µM)2 reported in Table 2 by Rembach et al.  should be the same, differently from what is reported. Moreover, 50% of the control non-CP Cu values calculated with equation #2 are higher than the normal reference range reported for healthy controls .
The diverse results obtained by Rembach et al.  in terms of non-CP Cu in comparison with all the previous studies published [3, 5, 8, 11-17] can more likely be related to the diverse values of CP measured in the AIBL cohort. As a matter of fact, if one considers all the studies published in 1983 through January 2013 on Cu and CP measurements in AD and control individuals [3, 5, 8, 11-17], the mean value of non-CP for healthy controls is in the positive range. This is not true for the Rembach et al. study , which finds a negative mean value (Table 1).
Twomey et al.  proposed 6.6 as the theoretical optimal mean value of the Cu:CP ratio in healthy subjects. This ratio is an index of the accordance of the Cu and CP pair for each serum measurement. As reported in Table 1, the Cu:CP ratio for the Rembach et al. study  is lower than 6.6, which indicates an underestimation of Cu with respect to CP values, pointing out that in the AIBL study, the CP values are not in line with the other studies. Genetic factors may account for this difference which seems to move the AIBL CP values toward higher values, especially in relation to total serum Cu. In particular, ethnicity may explain this evidence. As described by the AIBL group, Australia is a geographically isolated country with a different ethnic mix and AIBL participants reflected this admixture . Serum CP levels are affected by genetics . Several studies have demonstrated that genetic variation among human populations is associated with significant functional differences in health-related aspects . Among human populations, Australians showed a very complex genetic structure , which, thereby, may explain the differences between AIBL CP values and those from the other studies. Otherwise, existing CP concentration methods are not standardized . On this basis, we can conclude that the data showed by the AIBL study are not sufficient to sustain the authors’ statement that non-CP Cu is not a robust discriminator for AD. More precisely, we agree with them that the calculation is composite; it is susceptible to Cu, CP, and their reciprocal variations which should move in concert to fit their physiological interaction. In other words, as depicted in this Letter to the Editor, the main problem concerns the calculations and not the occurrence of a Cu dysfunction in AD. This is why we have developed a direct fluorescent assay to measure serum non-CP Cu . This assay uses a coumarin fluorescent probe, for which a decay in fluorescence emission is recorded when it binds [Cu++]. This probe allows one to overcome the weakness of the non-CP Cu calculation and the uncertainty of the results.
Rosanna Squittia,b, Patrizio Pasqualettic, Mariacristina Siottod, Renato Polimantie, Mariacarla Ventrigliaa,b, Nicola Colabufof
aDepartment of Neuroscience, AFaR – Fatebenefratelli Hospital, Isola Tiberina, Rome, Italy
bDepartment of Neurology, “Campus Biomedico” University, Rome, Italy
cMedical Statistics & Information Technology, AFaR – Fatebenefratelli Hospital, Isola Tiberina,
dDon Carlo Gnocchi Foundation ONLUS, Milan, Italy
eDepartment of Biology, University of Rome “Tor Vergata”, Rome, Italy
fDipartimento Farmacia, Università degli Studi di Bari ALDO MORO, Bari, Italy
Copper (Cu), ceruloplasmin (CP), CP-bound copper (CP-Cu), non-CP-bound copper (Non-CP Cu), and Cu:CP ratio in healthy subjects as reported in all the studies published from 1983 through January 2013 on Cu and CP measurements in Alzheimer’s disease and control individuals.
*Data concerning CP-Cu have been calculated following the Walshe equation . Non-CP Cu was calculated by subtracting CP-Cu from Cu as reported by Walshe  and expressed in terms of negative or positive values. The Rembach et al. study  is the only study reporting mean value <0.
**Cu:CP ratio is calculated as reported in Twomey et al. . These authors provide the equation as follows:
[copper µM]* [132000 g/mol]/[ CPmg/dL*10(4)]
Twomey et al.  propose that 6.6 is the theoretical optimal mean value of this ratio in healthy subjects. In Table 1, the values consistently higher than 6.6 are reported in pale pink, and they represent an overestimation of copper with respect to CP values. Values consistently lower than 6.6 are represented in pale blue and represent an underestimation of copper with respect to CP values.
 Rembach A, Doecke JD, Roberts BR, Watt AD, Faux NG, Volitakis I, Pertile KK, Rumble RL, Trounson BO, Fowler CJ, Wilson W, Ellis KA, Martins RN, Rowe CC, Villemagne VL, Ames D, Masters CL, Bush AI (2013) Longitudinal analysis of serum copper and ceruloplasmin in Alzheimer's disease. J Alzheimers Dis34, 171-182.
 Salustri C, Squitti R, Zappasodi F, Ventriglia M, Bevacqua MG, Fontana M, Tecchio F (2010) Oxidative stress and brain glutamate-mediated excitability in depressed patients. J Affect Disord 127, 321-325.
 Squitti R, Barbati G, Rossi L, Ventriglia M, Dal Forno G, Cesaretti S, Moffa F, Caridi I, Cassetta E, Pasqualetti P, Calabrese L, Lupoi D, Rossini PM (2006) Excess of nonceruloplasmin serum copper in AD correlates with MMSE, CSF [beta]-amyloid, and h-tau. Neurology 67, 76-82.
 Squitti R, Bressi F, Pasqualetti P, Bonomini C, Ghidoni R, Binetti G, Cassetta E, Moffa F, Ventriglia M, Vernieri F, Rossini PM (2009) Longitudinal prognostic value of serum "free" copper in patients with Alzheimer disease. Neurology 72, 50-55.
 Squitti R, Ghidoni R, Scrascia F, Benussi L, Panetta V, Pasqualetti P, Moffa F, Bernardini S, Ventriglia M, Binetti G, Rossini PM (2011) Free copper distinguishes mild cognitive impairment subjects from healthy elderly individuals. J Alzheimers Dis 23, 239-248.
 Squitti R, Lupoi D, Pasqualetti P, Dal Forno G, Vernieri F, Chiovenda P, Rossi L, Cortesi M, Cassetta E, Rossini PM (2002) Elevation of serum copper levels in Alzheimer's disease. Neurology 59, 1153-1161.
 Squitti R, Pasqualetti P, Cassetta E, Dal Forno G, Cesaretti S, Pedace F, Finazzi-Agro A, Rossini PM (2003) Elevation of serum copper levels discriminates Alzheimer's disease from vascular dementia. Neurology 60, 2013-2014.
 Squitti R, Pasqualetti P, Dal Forno G, Moffa F, Cassetta E, Lupoi D, Vernieri F, Rossi L, Baldassini M, Rossini PM (2005) Excess of serum copper not related to ceruloplasmin in Alzheimer disease. Neurology 64, 1040-1046.
 Colabufo N, Squitti R (2012) Patent: Method and kit for determination of free copper in serum. European Patent Office (EPO) (RO/EP), Patent number: PCT-SAFE Version 3.51.056.232 MT/FOP 20121001/0.20.5.19
 Walshe JM (2003) Wilson's disease: the importance of measuring serum caeruloplasmin non-immunologically. Ann Clin Biochem 40, 115-121.
 Zappasodi F, Salustri C, Babiloni C, Cassetta E, Del Percio C, Ercolani M, Rossini PM, Squitti R (2008) An observational study on the influence of the APOE-epsilon4 allele on the correlation between 'free' copper toxicosis and EEG activity in Alzheimer disease. Brain Res 1215, 183-189.
 Squitti R, Ventriglia M, Barbati G, Cassetta E, Ferreri F, Dal Forno G, Ramires S, Zappasodi F, Rossini PM (2007) 'Free' copper in serum of Alzheimer's disease patients correlates with markers of liver function. J Neural Transm 114, 1589-1594.
 Molaschi M, Ponzetto M, Bertacna B, Berrino E, Ferrario E (1996) Determination of selected trace elements in patients affected by dementia.Arch Gerontol Geriatr 22 Suppl 1, 39-42.
 Snaedal J, Kristinsson J, Gunnarsdottir S, Olafsdottir, Baldvinsson M, Johannesson T (1998) Copper, ceruloplasmin and superoxide dismutase in patients with Alzheimer's disease . a case-control study. Dement Geriatr Cogn Disord 9, 239-242.
 Agarwal R, Kushwaha SS, Tripathi CB, Singh N, Chhillar N (2008) Serum copper in Alzheimer's disease and vascular dementia. Indian J Clin Biochem 23, 369-374.
 Brewer GJ, Kanzer SH, Zimmerman EA, Celmins DF, Heckman SM, Dick R (2010) Copper and ceruloplasmin abnormalities in Alzheimer's disease. Am J Alzheimers Dis Other Demen 25, 490-497.
 Arnal N, Cristalli DO, de Alaniz MJ, Marra CA (2010) Clinical utility of copper, ceruloplasmin, and metallothionein plasma determinations in human neurodegenerative patients and their first-degree relatives. Brain Res 1319, 118-130.
 Twomey PJ, Viljoen A, House IM, Reynolds TM, Wierzbicki AS (2007) Copper:caeruloplasmin ratio. J Clin Pathol 60, 441-442.
 Gardener S, Gu Y, Rainey-Smith SR, Keogh JB, Clifton PM, Mathieson SL, Taddei K, Mondal A, Ward VK, Scarmeas N, Barnes M, Ellis KA, Head R, Masters CL, Ames D, Macaulay SL, Rowe CC, Szoeke C, Martins RN (2012) Adherence to a Mediterranean diet and Alzheimer's disease risk in an Australian population. Transl Psychiatry 2, e164.
 Tang WH, Hartiala J, Fan Y, Wu Y, Stewart AF, Erdmann J, Kathiresan S, Roberts R, McPherson R, Allayee H, Hazen SL (2012) Clinical and genetic association of serum paraoxonase and arylesterase activities with cardiovascular risk. Arterioscler Thromb Vasc Biol 32, 2803-2812.
 Polimanti R, Piacentini S, Manfellotto D, Fuciarelli M (2012) Human genetic variation of CYP450 superfamily: analysis of functional diversity in worldwide populations. Pharmacogenomics 13, 1951-1960.
 Ballantyne KN, van Oven M, Ralf A, Stoneking M, Mitchell RJ, van Oorschot RA, Kayser M (2012) MtDNA SNP multiplexes for efficient inference of matrilineal genetic ancestry within Oceania. Forensic Sci Int Genet 6, 425-436.
 Infusino I, Valente C, Dolci A, Panteghini M (2010) Standardization of ceruloplasmin measurements is still an issue despite the availability of a common reference material. Anal Bioanal Chem 397, 521-525.