(A) Schematic diagram of cDNA cloning for b5R.1, b5R.2, and b5+b5R in human. Amino acid sequences of the classical human b5R (GenBank accession no. P00387) were used as the template for tblastn searches (28) against dbest available at http://www.ncbi.nlm.nih.gov. Matched EST entries with homologous (but not identical) encoded amino acid sequences were used for a further alignment to form a consensus polypeptide sequence. The well-conserved FAD- and NAD(P)H-binding sites (in yellow and green boxes, respectively) were used to anchor the alignment. Three b5R-like cDNA consensus sequences (b5R.1, b5R.2, and b5R.3) were found; b5R.3 seemed to have an open 5′ end. With the 5′ EST entry (GenBank accession no. AA166937) of human b5R.3 as the template for blastn search, one mouse EST entry (AA762399) was found to match with the human b5R.3 at its COOH half and with known cytochrome b5 sequences at its NH₂ half, including the well-conserved two histidines (in red boxes). This mouse EST was further used as the template for another blastn search to find two human EST entries AA400638 and AA360693 corresponding to the b5* domain. The two clusters of human EST sequences for b5* and b5R* domains belong to the same cDNA (renamed as b5+b5R), which has an ORF of 487 aa residues containing both b5* and b5R.3 domains, linked by a hinge of about 100 residues long (colored in blue). The membrane anchors of b5R and b5R.1 are indicated as brown boxes. (B) Sequence alignment for b5+b5R and classical b5 and b5R proteins in human, mouse, and rat. One letter code (standard) is used for each of the 20 aa residues. Efforts were made to minimize the number of gaps by anchoring the well-conserved binding motifs for heme, FAD, and NAD(P)H (colored in red, yellow, and green, respectively). Identical residues are boxed in black, and chemically similar ones in gray. Gaps are indicated by -, unidentified residues in rat b5+b5R by ., and stop codons by *. GenBank accession numbers are: AF169803 (human b5+b5R), P10067 (human b5), P56395 (mouse b5), P00173 (rat b5), P00387 (human b5R), P20070 (rat b5R), H33691 (rat b5+b5R NH₂ end), and AI071770 (rat b5+b5R COOH end). Amino acid sequences for mouse b5+b5R are translated from compository nucleotide sequences of all available EST entries in GenBank, including AA791608, AA762399, AA117503, and AI464992. Sequences for mouse b5R are also from EST entries, including AA870211, W77006, AA168906, and AA762398. All EST-derived sequences may contain error(s) from DNA sequencing. (C) Tissue and cell-line distribution of the human b5+b5R transcript. Both multiple tissue and multiple cell-line Northern blots, purchased from CLONTECH, were hybridized at high stringency to gene-specific cDNA probe for human b5+b5R (reverse transcription–PCR fragment for b5* or b5R*) and for human actin (CLONTECH). Approximately 2 μg poly(A)⁺ RNA was loaded in each lane. HL-60, promyelocytic leukemia; HeLa S3, epitheloid carcinoma (cervix); K-562, chronic myelogenous leukemia; MOLT-4, lymphoblastic leukemia; Raji, Burkitt's lymphoma; SW480, colorectal adenocarcinoma; A549, lung carcinoma; G361, melanoma.

(A) Expression constructs for human b5+b5R protein and its derivatives. The constructs for b5+b5R fusion as well as b5* and b5R* single-domain proteins were made by inserting PCR fragments into the pET-19b vector (Novagen), using NdeI site at the NH₂ end and BamHI site at the COOH end (both sites were included in PCR primers). The b5* domain extends from the NH₂ terminus through the NH₂-PSYPSF-COOH peptide sequences, and the b5R* domain from the NH₂-GQHVYLK-COOH peptide sequences through the COOH terminus of b5+b5R. The b5+b5R.ΔH was constructed by ligating the following DNA pieces: NdeI–NheI fragment for b5*, NheI–BamHI fragment for b5R*, and NdeI–BamHI fragment for pET19b vector. The presence of His₁₀ tag and enterokinase cleavage site at the NH₂ terminus of recombinant protein is indicated. (B) Light absorption spectrum for the human b5+b5R protein. Spectrum was measured at room temperature under air in low-salt buffer, either with 2 μM b5+b5R alone (oxidized) or mixed with additional 0.1 mM NADH (NADH-reduced). Excess NAD(P)H was necessary to maintain the b5+b5R protein in reduced state under aerobic conditions.

(A) Transient expression of the human b5+b5R protein in COS-7 cells. COS-7 cells (from monkey kidney) were transfected with expression construct (in pCR3.1 vector, Invitrogen) for classical b5R with c-myc tagged at the COOH end or b5+b5R tagged at the NH₂ end. Total cell lysate was collected after 2 days of transient expression and subjected to SDS/PAGE analysis. Approximately 10 μg of total protein was loaded onto each lane and the Western blot was hybridized with mouse antibody 9E10 against c-myc tag. (B) Confocal microscopy for subcellular localization of the human b5+b5R protein. COS-7 cells were grown on glass slide for 2 days after the transfection. Mouse antibody 9E10 and a rhodamine-conjugated donkey antimouse antibody were used to hybridize with c-myc-tagged proteins. b5+b5R (Left) and classical b5R (Right) are shown.

(A) NADH-reductase activity with cytochrome c as substrate. Cytochrome c reduction was monitored at 417 nm (Soret peak for reduced cytochrome c) at room temperature under air. Light absorption was followed for 15 min, beginning immediately after mixing the sample. Substrate = 6 μM oxidized cytochrome c; reductant = 1 mM NADH; sample = 0.03 μM b5+b5R (with or without 30 μM DPI), or 0.125 μM cytochrome c reductase. Nearly identical data were obtained three times. A representative time course is shown here. (B) Superoxide production. The production of ROS was measured by the Luminol + HRP method. The b5+b5R protein (0.1 μM) and NADH (0.5 mM) were used as the components for ROS production, whereas Luminol (3 mM) and HRP (0.05 mg/ml) were used for ROS detection. Inhibitors tested: DPI (1 μM or 30 μM) and SOD (1,000 units).