Cu was recognised as an essential nutrient after careful laboratory studies in rats and field observations in sheep and cattle in the 1920 s and 1930s(Reference Smith, Dreosti and Smith1). As for other essential metals, Cu is required during perinatal development of the central nervous system(Reference Prohaska2). Cu's importance for brain development was further exemplified by research done on Cu-deficient guinea-pigs where underdeveloped cerebella and hypomyelination were evident(Reference Everson, Tsai and Wang3). In humans, the brain requirement for Cu is best illustrated by Menkes' disease, an X-linked inherited neurodegenerative disorder which was first connected with Cu deficiency in 1972(Reference Danks, Campbell and Stevens4). The molecular reasons for the central nervous system phenotype accompanying Cu deficiency remain unknown.
Humans require Cu as an essential catalytic cofactor for approximately twelve mammalian cuproenzymes(Reference Prohaska, Bowman and Russell5). One is dopamine β-mono-oxygenase (EC 1.14.17.1; DBM) that catalyses the hydroxylation of dopamine (DA) to noradrenaline (NA) using Cu, oxygen and ascorbic acid as additional substrates(Reference Friedman and Kaufman6). DA and NA are important neurotransmitters required for the proper function of the brain. DBM is located in granulated vesicles of both sympathetic nerve terminals, adrenal medulla chromaffin cells, and noradrenergic and adrenergic neurons of brain(Reference Axelrod7). The importance of DBM was demonstrated by the embryonic lethality of DBM knockout mice(Reference Thomas, Matsumoto and Palmiter8). In vivo limitation of DBM following Cu deficiency was first demonstrated in heart tissue in 1967(Reference Missala, Lloyd and Gregoriads9). Normally NA synthesis is limited by tyrosine mono-oxygenase activity; however, under certain circumstances DBM can be rate limiting(Reference Wise, Belluzzi and Stein10).
The Cu–DBM brain connection was concurrently made 2 years after the first report that Menkes' disease was associated with aberrant Cu metabolism. These studies in brindled mice, a genetic homologue of Menkes' disease, and in Cu-deficient rats both reported lower brain NA and brain Cu(Reference Hunt11, Reference Prohaska and Wells12). More recently, it was observed that several brain regions had lower NA levels but higher DA levels in Cu-deficient rats and mice compared with Cu-adequate rodents(Reference Prohaska and Bailey13, Reference Prohaska and Bailey14). Supported by these catecholamine results DBM activity was suggested to be lower in Cu-deficient rodent brain.
Conversely, direct in vitro DBM activity assay of mutant mouse brain showed an increase in DBM activity(Reference Hunt15, Reference Prohaska and Smith16). Additionally, following dietary Cu deficiency in both rats and mice, higher brain DBM activity was reported(Reference Prohaska and Bailey13, Reference Prohaska and Smith16). Mutant blotchy mice, another mottled mutant, also exhibited higher DBM activity in adrenal glands(Reference Hunt15). Young male rats following perinatal Cu deficiency also demonstrated higher adrenal DBM activity(Reference Prohaska and Bailey13). DBM activity should be lower based on catecholamine levels; however, using enzyme assay, DBM activity was higher after Cu deficiency. Thus, a paradox exists.
Increased DBM activity following Cu deficiency is probably due to increased DBM protein abundance rather than augmented levels of inhibitors or lower levels of activators, but this has not been rigorously investigated. Adrenal DBM protein levels in older Cu-deficient rats were higher than in Cu-adequate rats(Reference Prohaska and Brokate17). Additionally, it was reported that adrenal mRNA for DBM was also increased following Cu deficiency(Reference Prohaska and Brokate18). Medulla/pons DBM mRNA was higher in young Cu-deficient females but not males based on Northern blot data, adding further to the confusion(Reference Prohaska and Brokate18). Thus it is still unknown whether an increase in DBM protein in brain of Cu-deficient rats accompanies higher DBM activity.
The purpose of these experiments was to extend earlier observations on brain and adrenal DBM, measuring enzyme activity, mRNA, protein and catecholamine levels in a single perinatal Cu deficiency rat experiment. Additionally, a tissue innervated with sympathetic nerves, the vasa deferentia, was studied. Rat and mice adrenal NA levels are different when Cu-deficient animals are compared with Cu-adequate animals. It was not known if DBM protein levels changed in a similar manner; thus some analyses on Cu-deficient mice were conducted. A second purpose of these studies was to test the hypothesis that depletion of NA was associated with increased transcription of DBM mRNA. Prior studies by several groups using reserpine to deplete catecholamines had made this prediction(Reference McMahon, Geertman and Sabban19, Reference Reis, Joh and Ross20).
Experimental methods
Animal care and diets
Holtzman rats and Hsd:ICR (CD-1) outbred albino mice were purchased commercially (Harlan Sprague–Dawley, Indianapolis, IN, USA). Rodents were maintained on Cu-adequate (Cu+) or Cu-deficient (Cu−) dietary treatment consisting of a Cu-deficient modified AIN-76A diet (Teklad Laboratories, Madison, WI, USA) that contained 0·36 mg Cu/kg by analysis. All dams, minimum of five litters per treatment group, and offspring were fed the Cu− diet, but Cu+ groups drank water supplemented with cupric sulfate (20 mg Cu/l), and Cu− groups drank deionised water. Cu deprivation was started at embryonic day 7 for the rats and at birth (postnatal day 0) for mice. Litter size was culled to ten pups on postnatal day 2(Reference Pyatskowit and Prohaska21). All animals were maintained on a 12 h light cycle (07.00 to 19.00 hours) at 24°C with 55 % relative humidity and had free access to diet and water. The University of Minnesota Animal Care Committee approved all protocols.
Both male rats (postnatal day 24–postnatal day 26) and male mice (postnatal day 27–postnatal day 28) were weighed before anaesthetised with diethyl ether. The rodents were killed by decapitation. Medulla oblongata/pons, remainder of brain, liver samples, hearts, adrenal glands and vasa deferentia were harvested, weighed and processed for biochemical analysis or frozen in liquid N2 and stored at − 75°C until used.
Biochemical analyses
Selected tissues and diet were wet-digested with concentrated HNO3 (Trace Metal grade; Fisher Scientific, Pittsburgh, PA, USA). Digested material was suspended in 0·1 m-HNO3 and analysed for total Cu content by flame atomic absorption spectroscopy (model 1100B; Perkin-Elmer, Norwalk, CT, USA). Protein levels of tissue samples were measured using a modified Lowry method(Reference Markwell, Haas and Bieber22).
DBM activity was determined by a modified spectrophotometric method previously described(Reference Prohaska and Smith16). Adrenal, medulla/pons and vasa deferentia tissues were homogenised in 49, 9 or 24 volumes of 5 mm-potassium phosphate (pH 7·0) containing 0·2 % Triton X-100. Conversion of tyramine to octopamine was determined as described elsewhere(Reference Kato, Kuzuya and Nagatsu23).
Analysis of dopamine β-mono-oxygenase mRNA expression
Total RNA was extracted from frozen samples using a TRI reagent® kit (Ambion, Austin, TX, USA), following manufacturer recommendations, including all optional steps. The purity of RNA was established spectrophotometrically and by RNA gels(Reference Prohaska and Brokate18). Contaminating DNA was removed using a DNA-free™ kit (Ambion), following manufacturer recommendations and cDNA was synthesised using Omniscript® Reverse Transcriptase (Qiagen®, Valencia, CA, USA) and amplified with a SYBR Green I kit (Roche, Indianapolis, IN, USA).
Copy number of rat tissue mRNA of DBM, and glyceraldehyde 3-phosphate dehydrogenase (GAPDH) and/or ribosomal 18S, both used as controls, was determined by quantitative real-time (qRT)-PCR. Primers for DBM and GAPDH are listed in Table 1. Rat 18S primer pairs were 5′ CTG TGA TGC CCT TAG ATG TCC 3′ (forward) and 5′ GCT TAT GAC CCG CAC TTA CTG 3′ (reverse). Mass of the transcripts was determined by comparison with gel purified PCR product standards. Molecular weight was determined by entering full DNA sequence including the forward and reverse primers into a molecular-weight calculator to determine the complementary strand, in addition to the genes' molecular weight(Reference Kibbe24). Dividing the mass of the gene transcript by its molecular weight and multiplying it by Avogadro's number calculated copy number.
DBM, dopamine β-mono-oxygenase; CCS, Cu chaperone for superoxide dismutase; COX I, cytochrome c oxidase subunit 1; COX IV, cytochrome c oxidase subunit 4; CTR1, Cu transporter (SLC31A1); PAM, peptidylglycine α-amidating mono-oxygenase; SOD1, Cu,Zn superoxide dismutase.
* Mean value was significantly different from that of the Cu-adequate group (P < 0·05; Student's t test).
† Genes and their respective forward and reverse primers were designed for real-time PCR. Copy numbers of specific mRNA were determined by quantitative real-time PCR and compared in Cu-adequate and Cu-deficient rats.
Dopamine β-mono-oxygenase antibody characterisation
DBM antisera were developed in rabbits (Sigma Genosys, The Woodlands, TX, USA) against KLH-peptide, N-SEPPESPFPYHIPLD-C, corresponding to amino acid residues 43–57 and 44–58 in rats and mice, respectively. Anti-DBM antibody was affinity purified from final bleed serum using a SulfoLink® kit (Pierce, Rockford, IL, USA) according to manufacturer recommendations. The affinity purified DBM antibody was diluted 1:500 for Western blot analysis.
Antibody specificity was determined using two approaches, antigen blocking and mobility shift upon removal of carbohydrate. Blocking of the antibody was accomplished using Rat N-term DBM peptide (SEPPESPFPYHIPLD) with DBM primary antibody diluted 1:500 with either 12 μg of peptide or 12 μg of bovine serum albumin, as a control. Carbohydrate removal from DBM protein was accomplished using peptide:N-glycosidase (PNGase) F (New England BioLabs, Ipswich, MA, USA) according to manufacturer recommendations.
Western blot analysis
Samples for Western blot analysis of DBM, actin and Cu chaperone for superoxide dismutase (CCS) were prepared by homogenising frozen rat adrenal gland, medulla/pons and vasa deferentia, and mouse adrenals and vasa deferentia in 99, 9, 24, 39 or 9 volumes, respectively, of 50 mm-potassium phosphate buffer (pH 7·0) containing 0·2 % Triton X-100 and protease inhibitors (Protease Inhibitor Cocktail; Sigma Chemical, St Louis, MO, USA). Homogenates were centrifuged at 10 000 g for 10 min at 4°C and supernatant fractions were saved for analysis at − 75°C. Fractionation was completed on 10 % SDS-PAGE gels. Transfer of protein to 0·2 μm nitrocellulose membranes and processing for immunoblotting was described previously(Reference Prohaska and Brokate17). Membranes were stained with Ponceau S (Sigma Chemical) to verify equal protein loading.
Protein levels of CCS were evaluated using affinity purified rabbit anti-hCCS characterised previously, at a 1:500 dilution(Reference West and Prohaska25). Membranes were also probed for actin to verify equal loading of protein. Chemiluminescence was captured using high-speed blue X-ray film (Lake Superior X Ray Inc., Duluth, MN, USA) and densitometry was carried out using a Kodak Image Station 2000M and Molecular Imaging Software (version 4.0.4; Kodak, New Haven, CT, USA).
Catecholamine analysis
Catecholamines were extracted from tissues by homogenising with nine volumes of 0·05 m-HClO4 containing 0·3 μm-3,4-dihydroxybenzylamine(Reference Pyatskowit and Prohaska26). Catecholamines were eluted from alumina with 200 μl of a 60:40 mixture of 0·2 m-acetic acid and 0·04 m-H3PO4(Reference Teman and Prohaska27). Separation of NA, DA and 3,4-dihydroxybenzylamine was accomplished using reverse-phase ion pairing HPLC with electrochemical detection using a mobile phase of 0·5 mm-1-octanesulfonic acid, 0·1 mm-EDTA and 0·1 m-KH2PO4 with 6·5–8 % methanol. The pH was adjusted to 3·0 with 0·1 m-phosphoric acid. Samples were separated on a 3·2 × 15 mm cartridge guard column and a 4·6 × 250 mm analytical column (ODS-II 5 μm; Regis, Morton Grove, IL, USA). Output was recorded and peaks were integrated using Peak Simple Software (Chrom Tech, Inc., Apple Valley, MN, USA)(Reference Pyatskowit and Prohaska21).
Statistical analysis
Mean values with their standard errors were calculated. Student's unpaired two-tailed t test was used to compare data between the two dietary treatments, α = 0·05 and α = 0·01. Variance equality was evaluated by the F test. All data were processed using Microsoft Excel™ (Redmond, WA, USA). Immunoblot data normalisation was accomplished by assigning a value of 1·0 to the mean pixel density of the Cu+ samples. All individual Cu+ and Cu− density values were then recalculated before graphing for ease of comparison. Correlation analysis for scatter plot was calculated using Excel™ and a Pearson product moment correlation coefficient table of critical values, for two-tailed test and α = 0·05.
Results
Rodent biochemical characteristics
Following perinatal Cu deficiency, Cu status in the rat experiment was analysed by evaluating a number of characteristics: body weight, cardiac hypertrophy (heart:body weight ratio), brain Cu and liver Cu (Table 2). All of the characteristics showed a statistical difference between the Cu− and Cu+ animals. Body weights in the Cu− rats were significantly lower than in the Cu+ rats. Cardiac hypertrophy occurred in Cu+ rats compared with Cu+ rats, as evident by the 2·5 times greater heart:body weight ratio. Cu− rats had a major reduction in brain Cu compared with Cu+ rats (81 %) consistent with neuronal Cu deficiency. Cu deficiency in a peripheral tissue was evident as well. Cu− rats showed a severe reduction in liver Cu compared with Cu+ rats (94 %). These four characteristics demonstrate that the dietary model in the rats studied was successful in achieving rats of two different Cu states. The same four characteristics were measured in the mouse experiment at postnatal day 27-postnatal day 28 (Table 2). Results were similar to those of the rat, except that there was no change in body weight due to Cu deficiency. Compared with Cu+ mice, Cu− mice had a heart:body weight ratio 1·5 times higher, a 77 % reduction in brain Cu and a 71 % reduction in liver Cu.
Cu+, Cu-adequate; Cu − , Cu-deficient.
** Mean value was significantly different from that of the Cu+ animals of the same species (P < 0·01; Student's t test).
Dopamine β-mono-oxygenase antibody specificity
Two experiments were performed to determine DBM antibody specificity. Affinity purified DBM antibody detected a single immunoreactive band of approximately 75 kDa in adrenal glands from rats and mice, similar to the size previously characterised(Reference Wallace, Krantz and Lovenberg28). The antibody detected multiple immunoreactive bands for medulla oblongata/pons, all smaller in size than 75 kDa (Fig. 1). In experiment 1, two Western blots with adrenal glands, medulla oblongata/pons, and liver, as a negative control, were performed: one in which the antibody was treated with bovine serum albumin and one treated with DBM blocking peptide (Fig. 1 (a)). Antibody treated with bovine serum albumin results in a single adrenal gland band (75 kDa) and multiple medulla oblongata/pons bands. The lane loaded with liver had no immunoreactive band. In contrast, the blot treated with DBM peptide clearly shows that the band at 75 kDa in the adrenal gland lane and the slightly smaller band in the medulla oblongata/pons lane are blocked by DBM peptide.
In experiment 2, a Western blot was performed with adrenal gland and medulla oblongata/pons samples. Cu− samples indicated by ( − ) and Cu+ samples (+) above the blot were compared (Fig. 1 (b)). Samples before loading were either treated with PNGase (+) or just buffer ( − ). The adrenal gland sample not treated with PNGase displayed the usual immunoreactive band detected at 75 kDa. In contrast, the adrenal gland sample treated with PNGase migrated further, due to the removal of carbohydrate from DBM. Both the Cu− and Cu+ medulla oblongata/pons samples yielded a pattern approximately the same as before (Fig. 1 (a)), when not treated with PNGase (Fig. 1 (b)). The upper band in both Cu− and Cu+ medulla oblongata/pons samples disappeared when treated with PNGase. The second band was more dense when treated with PNGase. This is consistent with the removal of carbohydrates from DBM and a similar migration shift of DBM seen for adrenal glands.
Adrenal gland analysis
Adrenal gland homogenates from postnatal day 24 male rats were used to determine DBM activity which was greatly influenced by Cu deficiency. Activity, in vitro, of DBM from Cu− adrenal gland samples was 2·4-fold higher than that of Cu+ samples (Fig. 2 (a)).
Total RNA from adrenal glands was isolated from postnatal day 25 male rats of both treatment groups and qRT-PCR was performed to determine the copy numbers of DBM and GAPDH, an expression control. The number of copies of DBM transcript per 1000 copies of GAPDH transcript was significantly higher, 0·7-fold, in Cu− compared with Cu+ samples (Fig. 2 (b)). GAPDH copy number was not affected by diet.
To extend data for DBM activity and transcript levels, adrenal gland DBM protein abundance of postnatal day 24 male rats was measured by Western immunoblot. The DBM immunoreactive band is clearly detected at 75 kDa. It was apparent that DBM protein abundance was markedly higher in Cu− than the Cu+ samples. Cu status was determined by measuring CCS protein abundance. CCS abundance has been shown to increase when Cu is limiting(Reference Prohaska, Broderius and Brokate29). In adrenal glands CCS abundance was markedly higher in Cu− than Cu+ tissue (Fig. 2 (c)). Actin, a loading control, was not changed by treatment (Fig. 2 (c)). Mean DBM density of Cu− adrenal glands was 1·7-fold higher than that of the Cu+ samples (Fig. 2 (d)). Mean CCS density in Cu− adrenal glands was even higher, a 3·6-fold difference (Fig. 2 (d)).
Medulla oblongata/pons analysis
DBM activity in medulla oblongata/pons from postnatal day 24 male rats was greatly influenced by Cu deficiency and was 0·9-fold higher in Cu− medulla oblongata/pons than Cu+ medulla oblongata/pons (Fig. 3 (a)). Specific activity in medulla oblongata/pons was much lower than for adrenal glands.
Total RNA from medulla oblongata/pons was isolated from postnatal day 25 male rats of both treatment groups and qRT-PCR was performed to determine the copy numbers of both DBM and GAPDH. DBM transcript abundance, per 1000 copies of GAPDH, was significantly higher, 0·5-fold, in Cu− compared with Cu+ rats (Fig. 3 (b)). GAPDH copy number was not different between the two groups. There was very low expression of DBM mRNA in medulla oblongata/pons compared with adrenal glands. In Cu+ rats, abundance in medulla oblongata/pons was less than 2 per 1000 GAPDH, whereas in adrenal glands, abundance exceeded 250. This fact made Western immunoblot detection of medulla oblongata/pons DBM very challenging as was shown in Fig. 1.
Medulla oblongata/pons DBM protein abundance of postnatal day 26 male rats was measured by Western immunoblot and the immunoreactive DBM band was detected slightly below the 75 kDa marker. It was apparent that both DBM and CCS protein abundance was higher in Cu− than Cu+ samples (Fig. 3 (c)). Actin was not changed by treatment (Fig. 3 (c)). The augmentation of DBM in Cu− medulla oblongata/pons tissue was not as high as in adrenal glands. The mean DBM density of Cu− samples was 0·7-fold higher than that of the Cu+ samples (Fig. 3 (d)). The mean CCS density in Cu− medulla oblongata/pons samples displayed a similar 0·8-fold increase.
Vasa deferentia analysis
DBM activity measured in homogenates from vas deferens from postnatal day 25 male rats was not statistically influenced by Cu deficiency (Fig. 4 (a)).
However, surprisingly the number of copies of DBM transcript per 1000 copies of GAPDH transcript was significantly higher in Cu− compared with Cu+ vas deferens (Fig. 4 (b)). The Cu− samples were 2·4-fold higher than the Cu+ samples. The GAPDH copy number was not different between the two groups. Relative mRNA expression of DBM in Cu+ samples in vas deferens was similar to medulla oblongata/pons.
DBM activity in the vas deferens suggested that there was no enhancement in DBM protein abundance, in contrast to adrenal glands and medulla oblongata/pons, but DBM transcript data suggested otherwise. Thus, vas deferens DBM protein abundance of postnatal day 24 male rats was measured by Western immunoblot and, like DBM activity, was not statistically higher in Cu− samples. Vas deferens actin levels were not altered by diet (Fig. 4 (c)). Mean CCS density did reveal a modest and statistically significant 0·4-fold higher abundance in Cu− compared with Cu+ vas deferens extracts consistent with marginal Cu deficiency in vas deferens (Fig. 4 (d)).
Mouse analysis
Mouse adrenal gland and vas deferens DBM protein abundance was analysed to determine if mice respond similarly to rats when subjected to Cu deficiency. Adrenal gland tissue was from postnatal day 27 male mice and vas deferens tissue from postnatal day 28 male mice. The two tissues were evaluated separately. It was apparent that both DBM and CCS protein abundance in adrenal glands was higher in the Cu− samples compared with Cu+ samples (Fig. 5 (a)). Mouse vas deferens DBM protein was detected in Cu− samples but not in Cu+ samples (Fig. 5 (b)). CCS protein was clearly more abundant in Cu− vas deferens samples, similar to rat tissue. Ponceau S stain was used as a loading control in mouse tissue because actin data were unreliable in mice. Consistent with immunoblot data, mean adrenal gland DBM activity was higher in Cu− samples (129 (sem 19) nmol/h per mg) compared with Cu+ samples (76·3 (sem 7·3) nmol/h per mg) for mouse adrenal glands (n 4; P < 0·05).
Expression of copper-related genes
Copper-related genes in addition to DBM were studied to determine if the DBM transcription abundance enhancement was unique. Medulla oblongata/pons relative mRNA expression in postnatal day 25 male rats was determined for the following genes: DBM, CCS, cytochrome c oxidase subunit 1 (COX I), cytochrome c oxidase subunit 4 (COX IV), Cu transporter (SLC31A1) (CTR1), peptidylglycine α-amidating mono-oxygenase (PAM) and Cu,Zn superoxide dismutase (EC 1.15.1.1) (SOD1) using primer pairs for qRT-PCR (Table 1). Only DBM showed a statistical difference between Cu− and Cu+ samples. Even though protein abundance for the CCS was elevated in Cu deficiency (Fig. 3 (c)), CCS mRNA was not altered by diet. The abundance of DBM mRNA in medulla oblongata/pons was very low compared with cuproenzymes SOD1 or cytochrome c oxidase, and similar to PAM.
Catecholamine analysis
Catecholamines were extracted from postnatal day 26 male rat adrenal glands, medulla oblongata/pons and vas deferens, and from P24 male rat heart. HPLC analyses demonstrated that diet did not have a significant effect on total NA in adrenal glands. However, total DA and DA:NA ratio (100 × DA/NA) were significantly higher in Cu− compared with Cu+ rats (Fig. 6 (a)), a 2-fold increase for both. Cu deficiency also had a significant effect in medulla oblongata/pons. NA concentration was lower by 25 % in Cu− compared with Cu+ tissue and DA concentration and DA : NA ratio was higher by 1·2-fold and 2·2-fold, respectively, in Cu− compared with Cu+ medulla oblongata/pons (Fig. 6 (b)). Cu deficiency also had a significant effect in vas deferens and heart. Total NA in vas deferens was lower by 29 % in Cu− compared with Cu+ rats and DA concentration and DA:NA ratio were elevated by approximately 0·6-fold and 1·1-fold, respectively, in Cu− compared with Cu+ tissue (Fig. 6 (c)). Total NA in heart was lower by 64 % in Cu− compared with Cu+ rats and DA concentration and DA:NA ratio were markedly higher by 15-fold and 48-fold, respectively, in Cu− compared with Cu+ rats (Fig. 6 (d)). Collectively, these catecholamine data suggest that Cu deficiency limits DBM activity in vivo.
Correlation of dopamine β-mono-oxygenase mRNA expression v. noradrenaline concentration
To compare DBM mRNA elevation with NA depletion in tissues, heart DBM copy number was first determined by qRT-PCR. There was no statistical difference between DBM copy number in Cu+ and Cu+ heart samples when normalised to 1000 copies of GAPDH (Fig. 7 (a)). However, unlike other tissues studied, GAPDH copy number was slightly higher in Cu+ than Cu+ heart. Thus, 18S ribosomal RNA abundance was determined and when using an average of both loading controls, GAPDH and 18S, there was a modest increase (0·4-fold) in DBM copy number (P < 0·05). DBM mRNA abundance of heart was the lowest of all tissues studied, even lower than medulla oblongata/pons (Fig. 3 (b)).
DBM mRNA elevation may be due to NA reduction; therefore, DBM mRNA elevation in four Cu− tissue samples was plotted v. NA reduction (Fig. 7 (b)). The correlation coefficient for these four data pairs was calculated to be 0·2, and not statistically significant.
Discussion
The model for perinatal Cu deficiency in the present experiments produced rats that were severely deficient based on growth retardation, cardiac hypertrophy, and highly reduced brain and liver Cu levels. The results from adrenal, medulla oblongata/pons, vas deferens and heart strongly confirm and extend limited previous data that dietary Cu deficiency leads to an augmentation in DBM mRNA levels. Presently no data are available to evaluate DBM mRNA degradation in Cu− rats. Thus, we suggest that there is increased transcription of DBM following Cu deficiency. Past studies of Cu-deficient rat adrenals showed that there was a significant elevation of DBM mRNA evaluated by Northern blots(Reference Prohaska and Brokate18). The same study showed an elevation in DBM mRNA in brain of Cu− female but not male rat pups. It is possible that multiple factors contribute to the up-regulation of DBM mRNA abundance as message levels are transcriptionally regulated by several factors including cAMP, glucocorticoids, bradykinin, nicotine and immobilisation stress(Reference McMahon and Sabban30).
However, reduction of NA levels in Cu− tissues seemed like the most logical candidate for the increased transcription of DBM. Induction of rat adrenal DBM mRNA can be achieved by a treatment with reserpine, a drug that depletes catecholamines(Reference McMahon, Geertman and Sabban19). DBM activity and protein levels in the brain of rats were elevated when treated with reserpine, while catecholamines were lower(Reference Reis, Joh and Ross20). Consistently lower NA levels in medulla oblongata/pons of Cu-deficient rodents support the hypothesis that depleted NA levels drive the increased DBM mRNA transcription. However, our present results and past work show some inconsistencies in the hypothesis, particularly in adrenal tissue. Studies in older rats subjected to Cu deficiency after weaning found that rat adrenal NA was lower and DBM activity higher following Cu deficiency(Reference Prohaska and Brokate18, Reference Hesketh31). However, the present studies, as well as others, showed no significant reduction in NA content and yet increased DBM mRNA(Reference Prohaska and Brokate18). Additionally, Cu-deficient cattle showed an adrenal NA reduction but with no change in DBM activity(Reference Hesketh32). Rat heart NA levels showed the greatest depletion of all the tissues studied in Cu− samples in the present studies, yet only a modest change in DBM mRNA and only when DBM copy number was normalised to two controls. The reduction in NA content and the DBM mRNA elevation in the four tissues studied in the present experiments were not statistically correlated. Thus, the mechanism for increased DBM mRNA transcription is still not well defined. One possibility is altered glucocorticoids since DBM transcription is augmented by glucocorticoids. This is an ideal candidate because glucocorticoids are elevated in both the adrenal glands and plasma of Cu− rats(Reference Fields, Lewis and Lure33). However, Cu− mice also have higher DBM protein and activity, and perhaps mRNA levels. However, Cu− mice do not have higher glucocorticoids in plasma(Reference Prohaska, Downing and Lukasewycz34). Higher DBM mRNA content in Cu− tissues could be due to increased transcript stability rather than enhanced synthesis. Future research is needed to fully understand the regulation of DBM following Cu deficiency.
However, our data clearly indicate that increased DBM mRNA abundance is associated with an increase in DBM protein. The present data confirm that in young Cu-deficient rats increased adrenal DBM mRNA and increased DBM protein are evident(Reference Prohaska and Brokate17, Reference Prohaska and Brokate18). Novel data report that DBM protein abundance in the medulla oblongata/pons by Western blot is higher in DBM protein in Cu− young male rats. Mouse adrenal and vas deferens samples also demonstrated higher DBM protein, demonstrating that mice responded in a similar manner to rats to Cu deprivation.
Higher DBM protein in Cu− animals is consistent with higher in vitro DBM activity and confirms similar changes in DBM activity and protein reported previously for adrenal glands from young rats(Reference Prohaska and Brokate17). The present results also show that higher DBM activity in vitro and higher DBM protein were evident in both adrenal and medulla/pons tissue of Cu− rats. The magnitude of change in both activity and protein are very similar (adrenal activity 2·4-fold v. protein 1·7-fold change and medulla/pons activity 0·9-fold v. protein 0·7-fold change). Though there was no significant difference in Cu− rats compared with Cu+ rats for vas deferens for either DBM activity or protein levels, a trend of change in both was similar (activity 0·2-fold v. protein 0·14-fold change). Note that a very modest change for CCS in Cu− vas deferens was detected, suggesting a more marginal Cu deficiency in this tissue. There was an excellent correlation in Cu− rat tissues between augmented DBM protein and DBM in vitro activity in Cu− medulla/pons, adrenal glands and vas deferens of 0·98 (P < 0·05). This would suggest that the higher DBM activity observed in tissues from Cu− mammals is due to a higher abundance of DBM protein rather than changes in endogenous inhibitors or subtle kinetic differences observed previously(Reference Prohaska and Smith16, Reference Hesketh31).
Despite higher DBM activity in vitro, catecholamine results suggest that DBM activity in vivo is lower in Cu− animals. As shown in the present study, NA is significantly lower in medulla/pons, vas deferens and heart of Cu− rats. DA levels were significantly higher in all four of the tissues studied in Cu− rats. The DA:NA ratio was significantly higher; approximately 2-fold greater in adrenals, medulla and vas deferens in Cu− rats compared with their controls. The ratio in heart was even higher, showing a 48-fold increase in Cu− rats. This is totally consistent with radiotracer studies done in Cu− rat heart suggesting that DBM activity in vivo was impaired by Cu deficiency(Reference Missala, Lloyd and Gregoriads9). Catecholamine changes in mouse heart and medulla oblongata/pons also respond in a similar manner to Cu deficiency(Reference Pyatskowit and Prohaska21). Present and past data support the hypothesis that there is higher apo-DBM in tissues from Cu-deficient mammals but function is limited by lower Cu availability. Data in the present studies were obtained from a severe restriction in dietary Cu. Thus, the impact of more marginal Cu limitation to DBM function is unknown. Our work and that of others have indicated that DBM limitation can occur in a variety of mammals and that the response is both species and organ specific(Reference O'Dell, Smith and King35–Reference Schoenemann, Failla and Rosebrough37).
The DA:NA ratio in serum can provide valuable information to the clinical community. Clinical and pathological features of Menkes' disease often reflect decreased activities of cuproenzymes, including DBM and others(Reference Kaler38). A significantly raised DA:NA ratio is a promising test for neonatal infants suspected of Menkes' disease(Reference Kaler, Holmes and Goldstein39). Additionally, a higher DA:NA ratio was shown in 100 % of the first six cases of DBM deficiency in humans(Reference Robertson, Haile and Perry40). Thus, the DA:NA ratio is a useful diagnostic tool for DBM limitation.
DBM is an important cuproenzyme. The low copy number of DBM mRNA and immunoblot data suggest that it is not a very abundant protein in the central nervous system. However, despite its low abundance, its function clearly is vital to the livelihood of animals, as demonstrated by the DBM knockout mouse studies(Reference Thomas, Matsumoto and Palmiter8). In humans, patients that have DBM deficiency have cardiovascular disorders and severe orthostatic hypotension. Regardless of the mechanisms for controlling DBM levels our data and others are congruent that DBM function in vivo requires adequate Cu; thus DBM is a key cuproenzyme in mammals.
Acknowledgements
We thank Margaret Boderius, Joshua Pyatskowit and Anya Gybina for their excellent technical assistance. This research was supported by National Institutes of Health (NIH) grant HD-039708.
K. N. carried out most of the analytical work and J. P. designed the experiments. Both authors participated in writing the manuscript.
Neither of the authors has any conflicts of interest.