Exploring the Impact of Mitonuclear Discordance on Disease in Latin American Admixed Populations

Abstract

Background. The coevolution of nuclear and mitochondrial genomes has guaranteed mitochondrial function for millions of years. The introduction of European (EUR) and African (AFR) genomes into the Ameridian continent during the Columbus exchange in Latin America created an opportunity to naturally test different combinations of nuclear and mitochondrial genomes. However, the impact of potential "mitonuclear discordance" (MND, differences in ancestries) has not been evaluated in Latin American admixed individuals (AMR) affected with developmental disorders, even though MND alters mitochondrial function and reduces viability in other organisms. Methods. To characterize MND in healthy and affected AMR individuals, we used AMR genotype data from the 1000 Genomes Project (n = 385), two cohorts of 22q.11 deletion syndrome patients 22qDS-ARG (n = 26) and 22qDS-CHL (n = 58), and a cohort of patients with multiple congenital anomalies and/or neurodevelopmental disorders (DECIPHERD, n = 170). Based on their importance to mitochondrial function, genes were divided into all mitonuclear genes (n = 1035), high-mt (n = 167), low-mt (n = 793), or OXPHOS (n = 169). We calculated local ancestry using FLARE and estimated MND as the fraction of nuclear mitochondrial genes ancestry not matching the mtDNA ancestry and ∆MND as (MNDoffspring-MNDmother)/MNDmother. Results. Generally, MND showed distinctive population and haplogroup distributions (ANOVA p < 0.05), with haplogroup D showing the lowest MND of 0.49 ± 0.17 (mean ± s.d.). MND was significantly lower in 22qDS-ARG patients at 0.43 ± 0.24 and DECIPHERD patients at 0.56 ± 0.12 compared to healthy individuals at 0.60 ± 0.09 (ANOVA p < 0.05). OXPHOS and high-mt showed the same trend, but with greater differences between healthy and affected individuals. Conclusions. MND seems to inform population history and constraint among affected individuals, especially for OXPHOS and high-mt genes.

Keywords: admixed Latinos; mitochondrial genomics; mitonuclear discordance.

Figure 1

Population- and haplogroup-specific MDN distribution. (A) Distribution of mean MND among 1kGP and Chilean populations. (B) Distribution of mean MND among Latin American macrohaplogroups. ****: p < 0.0001: ***: p < 0.001; *: p < 0.05; ns: not significant.

Figure 2

Mean MND in cohorts of patients with different genetic disorders. We compared healthy individuals to multiple cohorts of patients. The first three distributions correspond to Chilean individuals, whereas the last distribution corresponds to patients from Argentina. (A) Comparison of mean MDN using all mitonuclear genes. (B) Comparison of mean MDN using OXPHOS genes. DRD: DECIPHERD. CHL: Chile. ARG: Argentina. ***: p < 0.001; **: p < 0.01; *: p < 0.05; ns: not significant.

Figure 3

Comparison of ∆MND in affected cohorts. 1kGP-pairs represent healthy individuals, while DRD-pairs and 22q-ARG-pairs represent cohorts affected by rare genetic disorders. 1kGP: 1000 Genomes Project. DRD: DECIPHERD. ARG: Argentina.