Figure 5. Direct and indirect disruption of complex IV increases APOE expression.

(A) Direct (COX17-20, HIGD1A) and indirect (SLC25A3, SLC31A1, ATP7A) mechanisms required for complex IV assembly. (B) MesoScale ELISA determinations of human APOE in wild-type and HEK293 cell clones null for the genes indicated in blue font. Shown are APOE content in lysates and conditioned media. (C) MesoScale ELISA determinations of human APOE in wild-type and two independent COX17Δ/Δ mutant SH-SY5Y cell clones. (D) MesoScale ELISA determinations of human APOE in wild-type and two independent SLC31A1Δ/Δ mutant SH-SY5Y cell clones were studied. (E) MesoScale ELISA determinations of human APOE in wild-type and two independent ATP7AΔ/Δ mutant SH-SY5Y cell clones transfected with the endocytosis-deficient ATP7A-LL construct. (F) MesoScale ELISA determinations of human APOE in wild-type and two independent ATP7AΔ/Δ mutant SH-SY5Y cell clones. (G) MesoScale ELISA determinations of human APOE in wild-type and SLC25A3Δ mutant HAP1 cells were studied. N=8 for B and n=4 for C-G. (H) NanoString mRNA quantification of human APOE and TBP transcripts in wild-type, and two independent mutant clones of either COX17Δ/Δ or SLC31A1Δ/Δ mutant SH-SY5Y cells. Wild type and SLC31A1Δ/Δ cells were treated with vehicle or 200 micromolar of the copper chelator bathocuproinedisulfonic acid (BCS). TBP was used as a housekeeping control transcript. n=3. (I) Seahorse basal cellular respiration across different genotypes normalized to the corresponding wild-type cell. (J) Correlation between APOE in conditioned media with either basal cellular respiration (OCR) or the extracellular acidification rate determined by Seahorse (ECAR, n=3–9). Simple linear regression fit and 95% confidence interval is shown. All data are presented as average ± SEM. For B to F, and H One-way ANOVA followed by Benjamini-Krieger-Yekuiteli multiple comparison corrections (FDR = 5%). (G) unpaired t-test with Welch’s correction.

Figure 5—figure supplement 1. Total cellular copper in wild type and diverse mutant cells.

Figure 5—figure supplement 2. Upregulation of APOE expression and secretion in SLC25A20 mutant cells.

(A) qRT-PCR quantification of SLC25A20 and housekeeping control genes (PCBP1 and RER1) in wild-type and SLC25A20Δ mutant cells. Average ± SEM, One-way ANOVA followed by Dunnett’s multiple comparisons test. (B) Seahorse stress test in wild-type and SLC25A20Δ HAP1 cells. Arrows show sequential addition of oligomycin, FCCP, and rotenone-antimycin during the Seahorse stress test. (C) Depicts basal, ATP-dependent and maximal oxygen consumption rates. p value, Mann-Whitney U test, n=4. (D) Immunoblots with OxPhos antibody mix in mitochondrial fractions from wild-type and SLC25A20Δ cells. Left panels represent two exposures and right panel shows blot with COX4 antibodies. (E) MesoScale ELISA determinations of human APOE in wild-type and SLC25A20Δ mutant cells. Shown are APOE content in lysates and conditioned media. p value, Mann-Whitney U test, n=4. All graphs depict average ± SEM.

Figure 5—figure supplement 3. Extracellular acidification rate (ECAR) in wild-type and diverse mutant cells.

(A) ECAR was measured by Seahorse stress test in wild-type and mutant HAP1, SH-SY5Y, and HEK293 cells. Arrows mark the sequential addition of oligomycin, FCCP, and rotenone-antimycin during the Seahorse stress test. Average ± SEM, Two-way ANOVA followed by Bonferroni multiple comparisons test, p value represents the effect of genotype, n of experiments is in parentheses. (B) MesoScale ELISA determinations of human APOE in wild-type and PARK2Δ/Δ mutant SH-SY5Y cells. PARK2 mutants were selected as controls where there is strong acidification of media as measured by ECAR. Unpaired t-test with Welch’s correction.

Figure 5—figure supplement 4. Citrate effects on APOE expression in HAP1 cells.

(A) Diagram of citrate pathways and genes studied. (B) Fireworks analysis of co-essentiality genes to identify genes connected to ACLY (Amici et al., 2021). Green denote hubs and purple marks MitoCarta annotated nodes. ACLY is a component of the SLC25A1 interactome, see Gokhale et al., 2021. (C) qRT-PCR quantification of citrate transporter SLC13A5 and a housekeeping control gene (PCBP1) in wild-type and SLC25A1Δ HAP1 cells. (C) MesoScale ELISA determinations of human APOE in wild-type and SLC25A1Δ HAP1 cells incubated with vehicle, 2.5 mM citrate for 48 hr. Average ± SEM, One-way ANOVA followed by Dunnett’s multiple comparisons test. (D) qRT-PCR quantification of ACLY and a housekeeping control gene (PCBP1) in wild-type and ACLYΔ mutant cells. Average ± SEM, One-way ANOVA followed by Dunnett’s multiple comparisons test. (E) MesoScale ELISA determinations of human APOE in wild-type and ACLYΔ HAP1 cells. Unpaired t-test with Welch’s correction.

Figure 5—figure supplement 5. AMPK, mitochondrial stress, and redox responses in wild type and SLC25A1Δ HAP1 cells.

(A) Activity of AMPK in wild-type (lanes 1–2) and SLC25A1Δ (lanes 3–4) HAP1 cells was assessed with antibodies against AMPK, phospho-AMPK, and SLC25A1. Cells were treated with AICAR at 0.4 mM concentration for 72 hr to induce the activation of AMPK (lanes 2 and 4). (B) RNAseq determinations of FGF21 and GDF21 as well ATF transcription factors in wild-type and mutant cells. (C) Cis-regulatory sequence analysis with iRegulon to infer transcriptional responses activated in the SLC25A1Δ and SLC25A4Δ upregulated transcriptome (1429 transcripts p<0.001 and log2 fold of change = 1) (Janky et al., 2014). Significance threshold is set at 3. There are no ATF3, 4, or 5 target genes in the SLC25A1Δ upregulated transcriptome. (D) Activation of the mitochondrial stress response with doxycycline modestly increases APOE in conditioned media from SLC25A1Δ cells but not in wild-type cells. MesoScale ELISA determinations of human APOE in wild-type and SLC25A1Δ cells treated in the absence or presence of doxycycline at a concentration of 9.75 micromolar for 48 hr. APOE was measured in cell lysates and media. (E–F) Activation of the mitochondrial stress response with doxycycline in wild-type (D) and SLC25A1Δ (E) HAP1 cells. Doxycycline treatment was performed as in C. Gene expression was determined by qRT-PCR measuring the transcription factors ATF3-5 and their selected target genes ASNS, CHAC1, CHOP, PCK2, and WARS. RER1 and PCBP1 were used as housekeeping control transcripts. n=3. (G). Cells were treated with 2 mM N-acetyl cysteine (NAC) for 48 hr, n=3. RER1 was used as a control gene and glutathione-disulfide reductase (GSR) as a NAC-sensitive gene reporter. (E to G) show data as a ratio between drug/vehicle. All figures depict average ± SEM. For F and G, One-way ANOVA followed by Šydák’s multiple correction. (D and E) One-way ANOVA followed by Benjamini-Krieger-Yekuiteli multiple comparison corrections (FDR = 5%).