Contribution of Complex I NADH Dehydrogenase to Respiratory Energy Coupling in Glucose-Grown Cultures of Ogataea parapolymorpha

Abstract

The thermotolerant yeast Ogataea parapolymorpha (formerly Hansenula polymorpha) is an industrially relevant production host that exhibits a fully respiratory sugar metabolism in aerobic batch cultures. NADH-derived electrons can enter its mitochondrial respiratory chain either via a proton-translocating complex I NADH-dehydrogenase or via three putative alternative NADH dehydrogenases. This respiratory entry point affects the amount of ATP produced per NADH/O₂ consumed and therefore impacts the maximum yield of biomass and/or cellular products from a given amount of substrate. To investigate the physiological importance of complex I, a wild-type O. parapolymorpha strain and a congenic complex I-deficient mutant were grown on glucose in aerobic batch, chemostat, and retentostat cultures in bioreactors. In batch cultures, the two strains exhibited a fully respiratory metabolism and showed the same growth rates and biomass yields, indicating that, under these conditions, the contribution of NADH oxidation via complex I was negligible. Both strains also exhibited a respiratory metabolism in glucose-limited chemostat cultures, but the complex I-deficient mutant showed considerably reduced biomass yields on substrate and oxygen, consistent with a lower efficiency of respiratory energy coupling. In glucose-limited retentostat cultures at specific growth rates down to ∼0.001 h^-1, both O. parapolymorpha strains showed high viability. Maintenance energy requirements at these extremely low growth rates were approximately 3-fold lower than estimated from faster-growing chemostat cultures, indicating a stringent-response-like behavior. Quantitative transcriptome and proteome analyses indicated condition-dependent expression patterns of complex I subunits and of alternative NADH dehydrogenases that were consistent with physiological observations.IMPORTANCE Since popular microbial cell factories have typically not been selected for efficient respiratory energy coupling, their ATP yields from sugar catabolism are often suboptimal. In aerobic industrial processes, suboptimal energy coupling results in reduced product yields on sugar, increased process costs for oxygen transfer, and volumetric productivity limitations due to limitations in gas transfer and cooling. This study provides insights into the contribution of mechanisms of respiratory energy coupling in the yeast cell factory Ogataea parapolymorpha under different growth conditions and provides a basis for rational improvement of energy coupling in yeast cell factories. Analysis of energy metabolism of O. parapolymorpha at extremely low specific growth rates indicated that this yeast reduces its energy requirements for cellular maintenance under extreme energy limitation. Exploration of the mechanisms for this increased energetic efficiency may contribute to an optimization of the performance of industrial processes with slow-growing eukaryotic cell factories.

Keywords: Hansenula polymorpha; NADH; P/O ratio; bioenergetics; bioreactor; chemostat; proteomics; respiration; retentostat; transcriptomics.

FIG 1

Putative respiratory chain structure of Ogataea (para)polymorpha illustrating routes to couple direct NADH oxidation to ATP formation. Respiratory complex I (C I) and possibly an internal alternative NADH dehydrogenase(s) (Ndh2-i) oxidize NADH in the mitochondrial matrix (MM). NADH generated in the cytosol can be directly oxidized by an external alternative NADH dehydrogenase(s) (Ndh2-e). Shuttles, consisting of a corresponding pair of cytosolic and mitochondrial dehydrogenases, might exist which can indirectly translocate NADH over the inner mitochondrial membrane (IMM). All NADH-oxidizing respiratory enzymes donate electrons (red arrows) to the quinone pool (Q), from which they are funneled linearly through the rest of the respiratory chain, consisting of complex III (C III), cytochrome c (C), and complex IV (C IV), before reduction of oxygen to water. Contrary to many other complex I-harboring yeasts, O. (para)polymorpha does not possess an alternative oxidase (15). Respiratory complexes I, III, and IV, but not Ndh2, contribute to the proton gradient across the inner mitochondrial membrane which is utilized by mitochondrial F_oF₁ ATPase, complex V (C V), for formation of ATP. The dashed line represents the outer mitochondrial membrane. IMS, intermembrane space.

FIG 2

Biomass accumulation profile and viability of aerobic, glucose-limited retentostat cultures of wild-type O. parapolymorpha CBS11895 and the congenic complex I-deficient strain IMD010. The retentostat phase was initiated from steady-state chemostat cultures (D = 0.025 h⁻¹) at time zero. Depicted are the measured biomass dry weight concentrations (circles) and culture viability based on propidium iodide staining (diamonds) of two independent cultures each of strains CBS11895 (closed symbols) and IMD010 (open symbols), as well as the predicted biomass accumulation profiles of CBS11895 (dashed line) and IMD010 (dotted line) based on m_S and Y_X/S^max values estimated from chemostat cultures grown at 0.1 and 0.025 h⁻¹. The mean biomass concentration of CBS11895 was significantly higher than of that of IMD010 at each equivalent sampling point (Student's t test, P < 0.05).

FIG 3

Biomass-specific glucose uptake rates (q_S) during aerobic glucose-limited retentostat cultivation of wild-type O. parapolymorpha CBS11895 and the congenic complex I-deficient strain IMD010. Depicted q_S values are means ± standard errors of the means (error bars smaller than symbol size) of two independent retentostat cultures each of CBS11895 (closed circles) and IMD010 (open circles) and were directly calculated from biomass accumulation. The values plotted at time zero correspond to the q_S value in the steady-state chemostat cultures at 0.025 h⁻¹ that preceded the retentostat cultures. Horizontal lines indicate the maintenance energy requirements (m_S) calculated from chemostat cultures grown at 0.1 and 0.025 h⁻¹ for strains CBS11895 (dashed) and IMD010 (dotted). With the exception of the values calculated at 7 and 9 days of cultivation, the q_S values of CBS11895 were significantly lower than those of IMD010 at each equivalent sampling point (Student's t test, P < 0.05).

FIG 4

Depicted m_S values are means ± standard errors of the means of two independent retentostat cultures each of strains CBS11895 (closed circles) and IMD010 (open circles) and were calculated via linear regression from sets of corresponding μ and q_S values (directly calculated from biomass accumulation) from five adjacent sample points. The values plotted at time zero correspond to m_S values determined by chemostat cultivation at 0.1 and 0.025 h⁻¹, also represented by dashed (CBS11895) and dotted (IMD010) lines. With the exception of time points at 16 (CBS11895) as well as at 7 and 23 (IMD010) days of cultivation, m_S values of both strains were found significantly lower than those determined by chemostat cultivation (Student's t test, P < 0.05).

FIG 5

Transcriptional response of O. parapolymorpha to a lack of functional respiratory complex I. Green (upregulated) and red (downregulated) numbers indicate how many genes were found significantly differentially expressed (absolute log₂ fold change of >2; FDR < 0.001) in strain IMD010 (disrupted complex I Nubm subunit) compared to levels in strain CBS11895 (wild type) in glucose-grown batch cultures (A) and glucose-limited chemostat (0.1 and 0.025 h⁻¹) and late-stage retentostat cultures (0.001 h⁻¹) (B). Boxed in green are the most highly enriched GO terms within the set of upregulated genes in IMD010 under batch conditions (based on 275 out of 409 genes for which an S. cerevisiae ortholog could be identified) (see Table S2 in the supplemental material for an extended list). Numbers inside circles represent the specific growth rate/dilution rate of the different cultures.

FIG 6

Transcriptional adaptation of wild-type O. parapolymorpha CBS11895 and congenic complex I-deficient strain IMD010 to increasingly lower specific growth rates. (A) The bottom graphs show mean-normalized expression levels of genes identified to be positively (left) and negatively (right) correlated with specific growth rates in strains CBS11895 and IMD010, based on samples taken from glucose-limited chemostat (0.1 and 0.025 h⁻¹) and late-stage retentostat (0.001 h⁻¹) cultures (data presented as means ± standard deviation). Venn diagrams at the top indicate the overlap between genes identified to be positively (left) and negatively (right) correlated with specific growth rates identified for CBS11895 and IMD010. Numbers in parentheses indicate genes for which an S. cerevisiae ortholog could be identified. (B) Significantly enriched GO terms identified in the sets of genes with growth rate-correlated expression. Colors and lowercase letters correspond to Venn diagrams in panel A. For each set, the two most highly enriched GO terms of a category (biological process, molecular function, and cellular component) are listed, except for set b, for which all significantly enriched GO terms are shown (see Table S3 in the supplemental material for extended list).

FIG 7

Mean-normalized transcript and protein abundances of essential complex I subunits (A) and alternative NADH dehydrogenases (B) in O. parapolymorpha strains CBS11895 (wild type) and IMD010 (disrupted complex I Nubm subunit). Samples were taken from duplicate independent aerobic, glucose-grown batch (0.37 h⁻¹), chemostat (0.1 and 0.025 h⁻¹), and late-stage retentostat (0.001 h⁻¹) cultures. Transcript and protein abundances were mean normalized separately for each gene and strain. Gray, protein not detected based on criteria described in Materials and Methods. (C) The location and catalyzed reactions of the enzymes. Localization of the three alternative NADH dehydrogenases is unknown, and any of the enzymes could be internally (MM) or externally (IMS) localized. MM, mitochondrial matrix; IMM, inner mitochondrial membrane; IMS, intermembrane space.