Tryptic Peptide Analysis of Ceruloplasmin in Dried Blood Spots Using Liquid Chromatography–Tandem Mass Spectrometry: Application to Newborn Screening

Abstract

Background: Newborn screening to identify infants with treatable congenital disorders is carried out worldwide. Recent tandem mass spectrometry (MS/MS) applications have markedly expanded the ability to screen for >50 metabolic diseases with a single dried blood spot (DBS). The feature that makes metabolic disorders particularly amenable to screening is the presence of specific small-molecule metabolites. Many treatable disorders such as Wilson disease, however, are characterized by absent or diminished large proteins in plasma or within circulating blood cells, for which there are currently no cost-effective screening methods.

Methods: We developed an assay for quantifying ceruloplasmin (CP) in DBS for newborn screening of Wilson disease. CP-specific peptides from DBS samples digested by trypsin were quantified using isotopically labeled peptide internal standards and liquid chromatography–triple quadrupole mass spectrometry (LC-MS/MS).

Results: The calibration curve was linear from 20 to 95 mg/dL (200–950 mg/L). Intraassay imprecision (mean CV) for CP concentrations of 25, 35, and 55 mg/dL (250, 350, and 550 mg/L) was 9.2%, 10.7%, and 10.2%, respectively. Interassay imprecision for 19 different batches was 8.9%, 5.8%, and 6.9%. A method comparison study on previously tested patient samples for CP gave comparable results with lower limit of quantification, around 0.7 mg/dL (7 mg/L).

Conclusions: Our study supports that newborn screening for Wilson disease is feasible using LC-MS/MS assay for CP quantification in DBS after tryptic digestion. This approach should be applicable to newborn screening for other treatable genetic conditions, such as primary immunodeficiencies, that have large proteins as biomarkers.

Wilson disease (WD)1 is an autosomal recessive disorder of abnormal copper transport caused by mutations in the gene encoding the copper transporting ATPase, ATP7B (ATPase, Cu²⁺-transporting, β polypeptide)2 (1). ATP7B transports cytosolic copper into the lumen of the trans-Golgi network, where it is incorporated into apoceruloplasmin to form the 6-copper–dependent ferroxidase, ceruloplasmin (CP), a 132-kDa protein. CP is then transported into the blood, where it is the major copper-containing protein, carrying 90% of serum copper (2)(3). Mutations in ATP7B lead to reduced conversion of apoceruloplasmin to CP, which in turn results in markedly reduced CP in the blood (2)(3). Decreased serum CP concentrations can therefore be used as an effective screen for patients at risk of developing WD.

ATP7B also plays a critical role in excretion of excess copper from the body by relocalizing from the trans-Golgi network to the canalicular membrane of hepatocytes, where it transports cytosolic copper into the bile (3)(4). Mutations in ATP7B cause reduced excretion of copper into the biliary canaliculi, resulting in toxic accumulation of copper in hepatocytes and then other organs.

WD has an estimated incidence of 1 in 30 000 individuals but is much more common in certain populations (5). WD is progressive and ultimately fatal if untreated, which is a devastating fact since treatment is readily available (6). Patients treated preemptively with oral chelating agents such as trientine combined with a low-copper diet can live normal lives. Patients who are not diagnosed until after the onset of significant neurological symptoms or liver cirrhosis have a poor prognosis. WD is therefore an appealing candidate for early detection through newborn screening, although no population-based screening tests have been implemented.

Several pilot studies have been conducted measuring CP in the dried blood spots (DBS) of children using ELISAs to identify the patients with presymptomatic WD (7)(8)(9)(10). In 1 study, the mean (SD) CP concentrations in the DBS of healthy individuals, WD carriers, and WD patients were as 30.5 (9.5) mg/dL [305 (95) mg/L], 23.1 (6.8) mg/dL [231 (68) mg/L], and 3.2 (1.8) mg/dL [32 (18) mg/L], respectively (7). It has been shown, however, that the ceruloplasmin concentration is physiologically low in newborns (10), prompting the question whether screening for WD in the newborn period would be possible. Kroll et al. (11) answered this question with a retrospective study showing that WD patients had much lower CP concentrations in their original newborn blood spots than age-matched controls. The concentration of CP in the DBS of the 2 WD patients was 2.6 and 2.8 mg/dL (26 and 28 mg/L), respectively.

The mean value of CP in DBS observed in these pilot studies varied between studies, suggesting that antibody-based immunoassays could have significant challenges, including difficulties with the elution of CP from the blood spot, antibody selection, and calibration. The plasma volume in a punch can vary with hematocrit in newborns. For these reasons, it is important to develop a test with adequate precision to quantify the low concentration of CP in DBS from WD patients.

Newborn screening to identify patients with metabolic disorders was pioneered by Robert Guthrie in the 1960s for the detection of phenylketonuria (PKU) (12). Tandem mass spectrometry (MS/MS) has now been applied to newborn screening, expanding the ability to screen for >50 different metabolic diseases. Most states have now adopted MS/MS-based screening in their own state laboratories.

Most current screening approaches can only identify disorders in which a specific amino acid or small-molecule metabolite accumulates in the serum. However, many treatable disorders are characterized by absent or significantly reduced concentrations of proteins or large molecules in plasma or within circulating blood cells. Here, we used proteolytic digestion of DBS followed by tandem mass spectrometry analysis to identify newborns with absent or decreased concentrations of the large protein biomarker ceruloplasmin.

Materials and Methods

materials

We purchased 3-[3-(1,1-bisalkyloxyethyl)pyridin-1-yl]propane-1-sulfonate (PPS) Silent Surfactant from Protein Discovery, Inc.; Trypsin Gold (mass spectrometry grade) from Promega Corporation; dithiothreitol (DTT) from Fermantas, Inc.; HPLC-grade water, acetonitrile (ACN) (Optima grade), and Certified ACS PLUS hydrochloric acid (HCl) from Fisher Scientific; calcium chloride (CaCl₂), ammonium bicarbonate, and formic acid (FA) from Fluka; custom-made peptides for standards, ALYLQYTDETFR (P2), DIFTGLIGPM[K¹³C₆,¹⁵N₂] (P-1-IS), and ALYLQYTDETF[R¹³C₆¹⁵N₄] (P-2-IS) from Sigma-Genosys; and human ceruloplasmin from Athens Research and Technology.

ceruloplasmin characterization

Purified CP was digested with trypsin and initially analyzed by quadrupole time-of-flight (Q-TOF) mass spectrometer (Ultima; Waters) and ion trap mass spectrometer (LCQ; Thermo). We then subjected the data to searches by Sequest and Mascot software packages to identify peptides that were characteristic for CP. Precursor ions of 3 unique peptides were chosen based on peak intensity, ion score, number of product ions with a higher m/z than the precursor ion, and lack of cysteine residues that form disulfide bonds that could interfere with digestion. Next, we selected 4 product ions for each precursor ion. The collision energy, liquid chromatography (LC) gradient, and length of LC-MS/MS assay were then optimized for the detection of CP in the complex mixture of peptides in digested human serum using ultra-performance liquid chromatography (UPLC)–triple quadrupole mass spectrometer (Acquity UPLC–Quattro Premier; Waters). Finally, we used the optimized assay to detect and quantify CP peptides in DBS.

standard and control samples

We prepared calibrators by spiking venous blood collected in EDTA tubes with appropriate amounts of CP to obtain the final concentrations of 20, 30, 55, 70, and 95 mg/dL (200, 300, 550, 700, and 950 mg/L) plus the basal concentration. We prepared 3 quality controls by spiking venous blood with purified CP to final concentrations of 25, 35, and 55 mg/dL (250, 350, and 550 mg/L). Standards and quality controls were spotted onto filter paper cards (Whatman Protein Saver 903), allowed to dry at room temperature for 4 h, and then stored at −20 °C.

optimization of trypsin digestion

We optimized the denaturing conditions by testing various proportions of ACN and PPS Silent Surfactant in the reaction mixture. We tested trypsin digestions at various ACN concentrations ranging from 0% to 80%. The trypsin concentration was optimized by testing the trypsin to purified CP (wt:wt) ratios of 1:5, 1:10, 1:20, 1:30, 1:50, 1:70, and 1:100.

We tested various lengths of digestion times to determine the most efficient reaction time that resulted in complete digestion of CP. Digestions with a 1:20 trypsin to purified CP (wt:wt) ratio were performed at 37 °C for 1, 2, 4, 6, and 8 h and overnight.

sample preparation

One 3-mm punch was taken from the DBS of each of the standards, quality controls, and WD patients. Patient samples used for method correlation were the same as in a previous study (7)(11). Protein extraction, reduction, and digestion were performed in a single step. Punched DBS were placed in individual microcentrifuge tubes to which we added 100 μL of 0.1% PPS in 50 mmol/L ammonium bicarbonate. Next, we added 5 μL of 105 mmol/L DTT and 43.75 μL of concentrated ACN to make final concentrations of 5 mmol/L and 25%, respectively. We then added 13.6 μL of 2 μg/μL trypsin to make a 1:20 ratio of trypsin to total protein (wt:wt). The mixture was brought to a final volume of 175 μL with water and then incubated in a 37 °C water bath overnight. After digestion, we added 2 μL of 6 ng/μl (10 pmol) internal standard (P-1-IS) dissolved in 0.1% FA to each sample. PPS was cleaved by adding 3.5 μL of concentrated HCl, making a final concentration of 275 mmol/L. The solution was incubated at 37 °C for 30 min. Precipitation in an ice bath was performed by adding 900 μL of concentrated ACN to remove extra impurities, thus reducing the background noise in the mass spectrometry data. Finally, the samples were dried under a stream of nitrogen and reconstituted in 20 μL of 5% ACN in water with 0.1% FA.

lc-ms/ms conditions

We separated 7 μL of each digested sample using a Waters Acquity UPLC BEH C18 column (2.1 × 50 mm, 1.7-μm particles) at 0.1 mL/min flow rate and 45 °C. Peptides were eluted with a linear gradient of 5% to 25% ACN with 1% FA in water with a total cycle time of 7.5 min. The effluent from the LC column was directly coupled to the electrospray source of a Waters Quattro Premier triple-quadrupole mass spectrometer operated in the positive ion mode.

ceruloplasmin quantification

We obtained 5-point calibration curves by peak integration of extracted ion chromatograms of each peptide and corresponding isotopically labeled internal standard. The 3 quality control sera were run with each set of calibrators. We used QuanLynx software to integrate, calibrate, and quantify the concentration of CP in DBS samples.

imprecision

The imprecision of the LC-MS/MS assay was tested using DBS calibration curves. We evaluated interassay imprecision by making 19 different calibration curves with quality controls over the course of 30 days and intraassay imprecision using 19 replicates quantified with the same calibration curve. A correlation study was performed using 7 deidentified patient DBS from a previous ELISA study.

Results

precursor ion and internal standard selection

Purified CP was digested with trypsin to select appropriate peptides and precursor ions that were characteristic to CP. The absence of a band at 132 kDa on an 18% polyacrylamide gel was used to verify the completeness of CP digestion. We then analyzed the digested samples by Q-TOF and ion trap MS to determine which precursor and product ions were most likely to be detected by LC-MS/MS. Precursor ions were chosen only if they were found to be significant (having high peak intensities) by both Q-TOF and ion trap MS. Precursor ions were then selected based on XCorrelation >2.5, “clean” spectra with a low amount of overlapping daughter ion peaks, and number of intense product ion peaks with a higher m/z than the precursor ion. We used Sequest and Mascot data to further limit the number of candidate precursor ions based on ion score (>40) and the statistical significance of that score (expect number <0.05). Based on these criteria, we selected the following 3 peptides: ⁵⁴⁸DIFTGLIGPMK⁵⁵⁸ (P-1), ⁷⁰ALYLQYTDETFR⁸¹ (P-2), and ⁴⁴LISVDTEH SNIYLQNGPDR⁶² (P-3). We selected 4 multiple reaction monitoring (MRM) transitions for each peptide (Table 1 ).

Table 1.

Sequences of selected peptides and isotopically labeled internal standards and their precursor and productions used in MRM transitions.

Digested peptide/internal standard	Peptide sequence	Molecular weight	Precursor ion (m/z)	Product ions (m/z)
P-1	548DIFTGLIGPMK558	1190	596.5	963.5, 715.4, 545.3, 432.2
P-2	70ALYLQYTDETFR81	1518	760.6	1059.5, 931.4, 768.4, 348.2
P-3	44LISVDTEHSNIYLQNGPDR62	2174	1087.3	1209.6, 962.5, 799.4, 558.3
P-1-IS	DIFTGLIGPM[K13C6,15N2]	1198	600.7	971.5, 723.5, 553.4, 440.3
P-2-IS	ALYLQYTDETF[R13C615N4]	1528	765.7	1069.5, 941.4, 778.4, 358.3

Once the CP MRM transitions were identified, we analyzed purified CP by LC-MS/MS to see if these transitions could be detected. All P-1 and P-2 MRM transitions were detected with adequate peak intensity; however, P-3 was not detected. Next, we analyzed digested human serum to determine if P-1 and P-2 could be detected in a complex mixture of digested peptides. Both P-1 and P-2 were observed with adequate peak intensities in digested serum.

Peptides isotopically labeled at the C-terminus amino acid were used as internal standards for P-1 and P-2 (P-1-IS and P-2-IS), DIFTGLIGPM[K¹³C₆,¹⁵N₂] and ALYLQYTDETF[R¹³C₆¹⁵N₄], respectively (Table 1 ). Fig. 1 shows the full-scan MS spectra and precursor-scan MS/MS spectra for P-1-IS and P-2-IS. DBS were digested and analyzed by LC-MS/MS to see if P-1, P-2, and the corresponding internal standards could be detected in the matrix of whole blood. Internal standards were added after the digestion, at the same time as HCl was added to cleave PPS Silent Surfactant. P-1 and P-1-IS were both detected when analyzed by LC-MS/MS. P-2 had substantially lower peak intensity in digested DBS than in digested purified CP. It also had much lower peak intensity in DBS than P-1. Furthermore, P-2-IS was not detected with any measurable peak intensity. Adding P-2-IS later in the sample preparation procedure was tested to determine if P-2-IS was being lost in the early steps of sample processing. P-2-IS was still undetected when added at the end of the sample preparation procedure just before injection onto the UPLC column.

Figure 1.

Precursor ion and product ion MS spectra of P-1 internal standard (A) and P-2 standard (B).

optimization of trypsin digestion

The 25% ACN digestion gave the highest CP peptide peak intensities when analyzed by LC-MS/MS (Fig. 2A ). Similarly, we tested digestions in 0.1% and 0.2% PPS. The 0.1% and 0.2% PPS reactions had similar total ion current (TIC) peak intensities. The ratios of 1:5, 1:10, and 1:20 had high TIC peak intensities. The 1:30, 1:50, 1:70, and 1:100 ratios yielded lower TIC peak intensities, indicating that these digestions had not gone to completion (Fig. 2B ). We then used the 1:10 and 1:20 ratios to digest the total proteins in human blood. As shown in Fig. 2C , the TIC of P-1 increased over time and reached a plateau after the 8-h incubation, indicating that the digestion is complete by 8 h and can be run overnight.

Figure 2.

P-1.

(A), TIC peak intensity vs amount of ACN in purified CP digestion mixture where 25% ACN resulted in the highest peak intensity. (B), TIC peak intensity vs the ratio of trypsin to purified CP in the digestion mixture, where the 1:5, 1:10, and 1:20 ratios resulted in the highest peak intensities. (C), TIC peak intensity increased with the time of digestion. Purified CP was digested with 1:20 trypsin to CP (wt:wt).

ms and lc optimization

MS parameters were optimized using infusion of synthetic peptides. Cone voltage and collision energy for each of the P-1 and P-2 transitions were tuned individually.

To shorten the method to an analysis time that can facilitate a newborn screening sample load, we optimized the column length, mobile phase gradient, and gradient time. We tested 2 Acquity UPLC BEH C18 columns (2.1 × 50 mm and 2.1 × 100 mm) using water with 0.1% formic acid as solvent A and ACN with 0.1% formic acid as solvent B. The method using the longer column started with a 50-min time, which was shortened to 16 min. The shorter column allowed for 9.5- and 7-min method times with 5%–50% and 5%–29% solvent B gradients, respectively, without substantial reduction in P-1 peak intensity. However, the peak intensity of P-2 decreased as the method time was shortened (Fig. 3 ).

Figure 3.

Profile of P-1 and P-2 from digested DBS based on size of LC column, gradient, and length of method.

(A), 2.1 × 100 mm Acquity BEH C18 column with 50-min method time. (B), 2.1 × 100 mm Acquity BEH C18 column with 16-min method time. (C), 2.1 × 50 mm Acquity C18 column with 7-min method time.

validation results

Intraassay and interassay imprecision expressed as % CV are provided in Table 2 . Linearity was established between 20 and 95 mg/dL (200 and 950 mg/L), with correlation coefficient (R²) ≥0.98. Stability of DBS was tested previously, and long-term stability was observed in specimens maintained at −20 °C for years (11).

Table 2.

Intra- and interassay imprecision (% CV) for P-1.

MRM transition (m/z)	CP concentration
		25 mg/dL (250 mg/L)	35 mg/dL (350 mg/L)	55 mg/dL (550 mg/L)
Intraassay, n = 19
596.5 > 432.2	9.4	10.5	12.3
596.5 > 545.3	8.3	9.2	8.3
596.5 > 715.4	10.0	13.9	12.2
596.5 > 963.5	9.1	9.1	7.8
Interassay, n = 19
596.5 > 432.2	8.7	5.4	7.5
596.5 > 545.3	10.2	6.7	7.5
596.5 > 715.4	9.4	4.4	6.2
596.5 > 963.5	7.1	6.6	6.2

We analyzed DBS samples from 7 deidentified WD patients and 1 carrier using the calibration curves. Results with a signal-to-noise ratio (S/N) <10 were considered unreliable data. Of the results with S/N >10, all values obtained by LC-MS/MS were in the range 0–2 mg/dL (0–20 mg/L) of the concentration, comparable with the values measured previously by sandwich ELISA (Table 3 ). The lower limit of quantification (LOQ) was estimated to be approximately 0.7 mg/dL (7 mg/L) based on patient samples 2 and 3, in which the MS/MS results with S/N >10 were 0.63 and 0.86 mg/dL (6.3 and 8.6 mg/L) for transitions 595 > 432 and 596 > 963, respectively.

Table 3.

DBS patient sample correlation between LC-MS/MS and ELISA.

Sample	LC-MS/MS, mg/dL							ELISA, mg/dL
		596 > 432	596 > 545	596 > 715	596 > 963
1	1.07	0.811	2.261	1.13	1
2	0.63	0.991	0.581	0.86	1
3	0.86	1.111	0.921	0.92	2
4	1.50	2.251	1.631	1.41	2
5	1.80	1.841	1.391	1.25	3
6	1.79	5.441	1.74	1.61	3
7	1.45	1.461	1.511	1.37	3
Carrier	24.74	26.76	26.51	24.79	NA2

1

S/N < 10.

2

NA, not available.

Discussion

We developed an LC-MS/MS method to quantify CP using a standard LC flow rate, electrospray ionization source, and LC-MS/MS, all of which are increasingly being used in clinical laboratories for newborn screening. Using the method, we were able to quantify the low concentrations of CP present in the DBS obtained from WD patients.

Previous studies have shown that application of protein cleavage with isotope dilution mass spectrometry can be used to quantify proteins in clinically important matrices such as human serum or plasma (13)(14)(15)(16)(17)(18)(19). The use of MRM mode was also found to help increase the specificity of response for the peptide being analyzed (17). Whereas previous studies have quantified serum proteins after depletion of the most abundant proteins, or have used larger amounts of plasma or serum, the assay developed here is unique in detecting CP in DBS from filter paper (approximately 3.2 μL blood equivalent) without any enrichment and with a shorter run time. The peptide cleaved from CP represents the concentration of the intact protein in the original sample, provided the cleavage was complete. For this reason, the selection of a proper peptide was the key to a successful assay. The peptide selection turned out to be related to the LC conditions, because shortening the assay time increased the chance for ion suppression in the complex matrix of the whole blood.

Peptides P-1, P-2, and P-3 were carefully selected based on the LC-MS/MS results from Q-TOF and LCQ mass spectrometers. Yet these peptides showed different behavior on the triple quadrupole LC-MS/MS system. Intensities of transitions were noticeably affected by gradient, column length, and method time. Peptide 3 was below the detection limit altogether. Additionally, the shorter method decreased the intensity of P-2 to a point that it cannot be used for quantification. This is because P-2 elutes from the LC column early in the method along with many impurities, resulting in ion suppression of the peptide.

Anderson and Hunter (17) tested the utility of MRM for the simultaneous quantification of many high- to medium-abundance plasma proteins. In their study, a tryptic CP peptide ⁴²⁷EYTDASFTNR⁴³⁶ (m/z 602.3) was identified using nano-LC-MS/MS with a cycle time longer than 60 min. This early-eluting peptide showed good signal and peak intensity from pure CP digestions independent of the LC method length in our system. However, the signal from DBS digestions was weaker in the 50-min method and totally obscured in the final 7-min run. This further supports the evidence of suppression by impurities in the sample mixture that elute early from the LC column.

Another important step in designing a screening test by LC-MS/MS is the sample preparation. The protocol in this study was carefully optimized to ensure that the blood spot digestions were reproducible, because leftover undigested CP will result in an underestimated concentration of CP and increase the false positives in the screening of WD. Effective digestion of proteins often requires dissociation of protein complexes, denaturation of proteins, and solubilization of hydrophobic proteins. Traditionally, salts such as urea or guanidinium HCl, or detergents such as sodium dodecylsulfate (SDS), have been used to denature and solubilize proteins before digestion. These compounds are incompatible with LC-MS/MS, however, and must be removed before analysis. Therefore, we used the acid-labile, MS-compatible detergent PPS in combination with ACN for the extraction and digestion of proteins in DBS for this study (20). ACN can inactivate proteases such as trypsin if used at too high concentrations. We found that digestion reactions were most efficient when performed in 25% ACN. However, 25% ACN digestions of DBS resulted in an increased amount of background noise when analyzed by LC-MS/MS, which masked the P-1-IS. Therefore, we added an ACN precipitation step to the sample preparation protocol, effectively removing the extra impurities.

It is crucial to keep the LC-MS/MS method time as short as possible for newborn screening given the sample volumes to be processed daily. We developed a 7-min method without decreasing peak intensity of P-1 and P-1-IS, and further reduction of run time can be achieved by using a multicolumn LC system to carry out the separation of multiple samples simultaneously (16).

The results from 8 DBS samples obtained from 7 WD patients and 1 carrier quantified using the LC-MS/MS method showed similar CP concentrations to those previously obtained using sandwich ELISA. For those samples with extremely low concentrations of CP, the signal of the weaker MRM transitions, 596 > 715 and 596 > 545, were at or below the level of quantification. Nevertheless, the patients with WD can be accurately detected, even at very low levels of CP, by using the 2 stronger MRM transitions, 596 > 432 and 596 > 963.

Another congenital disorder characterized by lowered CP concentrations is Menkes disease, an X-linked disorder of copper transport, providing a similar opportunity for newborn screening. Treatment of individuals with classic Menkes disease with subcutaneous injections of copper histidine or copper chloride before 10 days of age normalized developmental outcomes in some individuals and improved neurologic outcome (21). If untreated, most patients die within 3 years.

In summary, there are a number of disorders for which newborn screening would dramatically improve clinical and functional outcomes, but for which adequate methods are lacking for mass population screening. This study shows that newborn screening for WD disease is feasible using LC-MS/MS to quantify CP in the DBS of newborns. This tryptic peptide screening could be useful for the screening of particular large proteins known to be absent or reduced in specific congenital disorders for newborns.

Author Contributions:All authors confirmed they have contributed to the intellectual content of this paper and have met the following 3 requirements: (a) significant contributions to the conception and design, acquisition of data, or analysis and interpretation of data; (b) drafting or revising the article for intellectual content; and (c) final approval of the published article.

Authors’ Disclosures of Potential Conflicts of Interest:No authors declared any potential conflicts of interest.

Role of Sponsor: The funding organizations played no role in the design of study, choice of enrolled patients, review and interpretation of data, or preparation or approval of manuscript.

1

Nonstandard abbreviations: WD, Wilson disease; CP, ceruloplasmin; DBS, dried blood spots; PKU, phenylketonuria; MS/MS, tandem mass spectrometry; PPS, 3-[3-(1,1-bisalkyloxyethyl)pyridin-1-yl]propane-1-sulfonate; DTT, dithiothreitol; ACN, acetonitrile; FA, formic acid; Q-TOF, quadrupole time-of-flight; LC, liquid chromatography; UPLC, ultra-performance liquid chromatography; MRM, multiple reaction monitoring; TIC, total ion current; S/N, signal-to-noise ratio; LOQ, limit of quantification.

2

Human gene: ATP7B, ATPase, Cu²⁺-transporting, β polypeptide.