The −451 to −309 region of the BACE1 5′UTR contains no inhibitory elements. (A) Plasmid pB1P-234m was constructed to contain the fourth uATG mutated to ATT, and pB1P-234δm contains an additional deletion of the third uATG-containing region from −451 to −309 of pB1P-234m. (B) Cells were cotransfected with pB1P-234, pB1P-234m, or pB1P-234δm and pCMV-Rluc. Firefly luciferase activity was measured at 24 h, and Renilla luciferase activity was used to normalize for transfection efficiency. The values represent means ± standard error of the mean (n = 3). There is no difference in luciferase activity between the mutants and the wild-type control, P > 0.05 by ANOVA.

Cleavage of eIF4G inhibits BACE1 uATG- and physiological ATG-initiated translation. (A) Plasmid pB1P+4 or pB1P-403 was cotransfected into N2a cells with the empty vector or pCMV2A and pCMV-Rluc. Cells were lysed in 1× reporter buffer, and the luciferase assay was performed. Luciferase activity was expressed as a percentage of that of the vector control. The values represent means ± standard error of the mean (n = 3), P < 0.001 by t test. (B) HEK293 cells were transfected with 500 ng of plasmid pB1P+4 and 1 ng of pCMV-Rluc for 24 h and then treated with TRAIL. Cells were harvested in 1× passive lysis buffer for the luciferase assay. The values represent means ± standard error of the mean (n = 3), P < 0.001 ANOVA with the posthoc Newmann-Keuls test. TRAIL treatment significantly reduced luciferase activity.

Sequence features of the human BACE1 gene minimal promoter and 5′UTR. (A) Nucleotide sequence of the minimal promoter and 5′UTR of the human BACE1 gene. The minimal promoter region corresponds to bp −1310 to −400 upstream of the physiological ATG of the human BACE1 gene, with the A designated as position +1 (2). The adenine in bold at −691 represents the transcription start site. The uATGs are indicated with both underlining and bold, and their corresponding stop codons are in italics. The fourth and fifth uATGs are in the same reading frame as the physiological ATG and their uORFs are shown in a box. The uAUGs of the corresponding upstream ORFs are also indicated. (B) Schematic diagram of the BACE1 gene uORFs. The arrow represents the transcription initiation site at −691, and +1 is the position of the adenine of the physiological start codon of the BACE1 gene ORF. Six uORFs are indicated by rectangular boxes, with the uAUGs represented by the black bars. The first, second, third, and sixth uORFs do not contain the uAUG in the same reading frame as the physiological start codon of BACE1 and are indicated by the hatched box. The uAUGs of the fourth and fifth uORFs are in the same reading frame as the physiological start codon of BACE1, and the clear box represents these uORFs.

Inhibition of BACE1 translation by the fourth uAUG. (A) Plasmids pBIP+4m and pBIP+4dm were constructed by site-directed mutagenesis using wild-type full-length 5′UTR plasmid pB1P+4 as a template. pBIP+4m contains the fourth uATG mutated to ATT, and pBIP+4dm has mutations of the third and fourth uATGs to ATT. (B) The new constructs were transfected into cells for luciferase assays. Firefly luciferase activity was measured at 24 h, and Renilla luciferase activity was used to normalize for transfection efficiency. The values represent means ± standard error of the mean (n = 3). *, P < 0.001 by ANOVA with the posthoc Newmann-Keuls test. (C) Western blotting was used to detect luciferase protein expression in cells transfected with the wild-type and mutant plasmids. BACE1-Myc protein in cells cotransfected with pBACE1-mycHis was detected with anti-Myc antibody 9E10 and used as the control. (D) Quantitative analysis of luciferase protein levels. Values are means ± standard error of the mean (n = 3). The protein levels are expressed as a percentage of the level in the wild-type pB1P+4 control. *, P < 0.001 by ANOVA with the posthoc Newmann-Keuls test. The mutation of the fourth uAUG in the full-length 5′UTR constructs increased downstream gene expression.

Fourth uAUG functions efficiently as a translation initiation codon. (A) Plasmid pB1P-237 was constructed with a deletion of the fifth uATG codon and the stop codon for both the fourth and fifth uATGs of BACE1 from plasmid pB1P-234. Plasmid pB1P-237 contains 63 bp of extra nucleotides in the BACE1 5′UTR relative to pB1P-330. (B) HEK293 cells were transfected with either pB1P-330, pB1P-237, or a combination of pB1P-330 and pB1P-237. Luciferase fusion proteins were detected with antiluciferase antibody. Plasmid pB1P-237 expressed a fusion luciferase with a higher molecular weight than that of pB1P-330. The middle lane clearly shows that there are two distinct luciferase protein bands from lysates of cells transfected with plasmid pBI-330 or pBI-237. (C) Plasmid pB1P-235 contains an extra nucleotide downstream of position −234, resulting in a frameshift mutation in the luciferase ORF. Plasmid pB1P-235m was generated to have the fourth uATG codon of the BACE1 gene in plasmid pB1P-235 mutated to ATT. (D) Cells were transfected with pB1P-234, pB1P-235, or pB1P-235m and pCMV-Rluc for 24 h. Luciferase activity was measured 24 h after transfection, and Renilla luciferase activity was used to normalize for transfection efficiency. The values are means ± standard error of the mean (n = 3) and are expressed as normalized luciferase activity relative to that of the wild-type pB1P-234 control. *, P < 0.001 by ANOVA with the posthoc Newmann-Keuls test.

Translation initiation by the physiological BACE1 AUG codon. (A) Schematic diagram of plasmid constructs containing the BACE1 minimal promoter and the full-length 5′UTR. Plasmid pB1P+4 contains the physiological BACE1 ATG in the same reading frame as the downstream luciferase coding sequence, and pB1P+6 contains two more nucleotides at the 3′ end of the 5′UTR, which result in a frameshift in the downstream luciferase coding sequence. (B) Luciferase activity was measured 24 h after cells were cotransfected with the plasmids and pCMV-RLuc. Renilla luciferase activity was used to normalize for transfection efficiency. The values represent means ± standard error of the mean (n = 3). *, P < 0.001 by t test. The frameshift mutation abolished luciferase activity. (C) Western blotting shows that pB1P+4 has robust luciferase protein expression. However, there is no detectable level of luciferase protein in cells transfected with pB1P+6, indicating that the frameshift disrupted luciferase synthesis. pBACE1-mycHis was cotransfected with the luciferase constructs, and the BACE1-Myc proteins were used as the transfection control. (D) Quantitative RT-PCR was performed to determine the level of firefly luciferase mRNA. pBACE1-mycHis was cotransfected, and the exogenous BACE1-Myc transcript was used as the transfection efficiency control. (E) Quantitative analysis of the ratio of luciferase mRNA to BACE1-myc mRNA. The mRNA levels were quantitated by Kodak Image Analysis. Shown are the means ± standard error of the mean. There is no difference in mRNA level between the wild type and the frameshift mutant, P > 0.05 by t test.

Translation initiation by the uAUGs of the human BACE1 5′UTR. (A) Schematic diagram of plasmid constructs containing the BACE1 minimal promoter and the wild-type or frameshift mutant 5′UTR. The BACE1 uATG in plasmids pB1P-403, pB1P-330, pB1P-234, and pB1P-113 is in the same reading frame as the downstream luciferase coding sequence, and pB1P-401, pB1P-331, pB1P-235, and pB1P-114 were generated so that the BACE1 uATG would not be in the same reading frame as the downstream luciferase ORF, which causes a frameshift in the downstream luciferase coding sequence. The uORFs are indicated. (B) The constructs were cotransfected with pCMV-RL into cells for 24 h. Firefly luciferase activity was measured, and Renilla luciferase activity was used to normalize for transfection efficiency. The values represent means ± standard error of the mean (n = 3). *, P < 0.001 by ANOVA with the posthoc Newmann-Keuls test. (C) The luciferase reporter plasmids were also transfected with pBACE1-mycHis for Western blot analysis. Firefly luciferase was detected with antiluciferase antibody, and BACE1-Myc protein was detected with antibody 9E10 as the transfection control. (D) Quantitative RT-PCR assays of the level of firefly luciferase mRNA. pBACE1-mycHis was cotransfected, and exogenous BACE1-Myc transcript was used as the transfection efficiency control. (E) Quantitative analysis of luciferase protein levels. Values are means ± standard error of the mean (n = 3). The protein levels are normalized to the BACE1-Myc control and expressed as a percentage of the level in pB1P-403-transfected cells. *, P < 0.01 relative to the wild-type by t test. (F) Quantitative analysis of luciferase mRNA levels. Values are means ± standard error of the mean (n = 3). The luciferase mRNA levels are normalized to the BACE1-Myc control and expressed as a percentage of the level in pB1P-403-transfected cells. There is no difference in mRNA level between the wild types and the frameshift mutants, P > 0.05 by ANOVA.

Functional deletion analysis of the human BACE1 5′UTR. (A) Schematic diagram of plasmid constructs containing the BACE1 minimal promoter and a series of truncated 5′ leader sequences. The horizontal line indicates the BACE1 minimal promoter with the 5′ leader sequence; the arrow indicates the BACE1 transcriptional start site. The box with the arrow labeled LUC represents the coding sequence of the luciferase reporter gene. The numbers indicate the endpoints of each construct, with +1 as the adenine of the physiological start codon of the BACE1 gene. (B) HEK293 cells were transiently cotransfected with the deletion plasmids and pCMV-RLuc. Firefly luciferase activity was measured 24 h after transfection, and Renilla luciferase activity was used to normalize for transfection efficiency. The values represent means ± standard error of the mean (n = 3). *, P < 0.001 by analysis of variance (ANOVA) with the posthoc Newmann-Keuls test. (C) Western blots were performed to determine the protein level of the downstream firefly luciferase cistron with antiluciferase antibody. pBACE1-mycHis was transfected with the deletion constructs, and the BACE1-Myc proteins were detected with antibody 9E10 against the Myc motif as the transfection control. (D) Quantitative RT-PCR was performed to determine the mRNA level of firefly luciferase. pBACE1-mycHis was transfected with the deletion constructs and exogenous BACE1-myc transcript was used as a transfection efficiency control. The ratio of luciferase protein to BACE1-Myc protein (E) and the ratio of luciferase mRNA to BACE1-myc mRNA (F) were quantitated by Kodak Image Analysis. Shown are the means ± standard error of the mean. *, P < 0.001 versus pB1P-403 by ANOVA with the posthoc Newmann-Keuls test.