Yeast diversity in open agave fermentations across Mexico.

Yeasts are a diverse group of fungal microorganisms that are widely used to produce fermented foods and beverages. In Mexico, open fermentations are used to obtain spirits from agave plants. Despite the prevalence of this traditional practice throughout the country, yeasts have only been isolated and studied from a limited number of distilleries. To systematically describe the diversity of yeast species from open agave fermentations, here we generate the YMX-1.0 culture collection by isolating 4524 strains from 68 sites with diverse climatic, geographical, and biological contexts. We used MALDI-TOF mass spectrometry for taxonomic classification and validated a subset of the strains by ITS and D1/D2 sequencing, which also revealed two potential novel species of Saccharomycetales. Overall, the composition of yeast communities was weakly associated with local variables and types of climate, yet a core set of six species was consistently isolated from most producing regions. To explore the intraspecific variation of the yeasts from agave fermentations, we sequenced the genomes of four isolates of the nonconventional yeast Kazachstania humilis. The genomes of these four strains were substantially distinct from a European isolate of the same species, suggesting that they may belong to different populations. Our work contributes to the understanding and conservation of an open fermentation system of great cultural and economic importance, providing a valuable resource to study the biology and genetic diversity of microorganisms living at the interface of natural and human-associated environments.

Comparing the evolutionary dynamics of predominant SARS-CoV-2 virus lineages co-circulating in Mexico.

Over 200 different SARS-CoV-2 lineages have been observed in Mexico by November 2021. To investigate lineage replacement dynamics, we applied a phylodynamic approach and explored the evolutionary trajectories of five dominant lineages that circulated during the first year of local transmission. For most lineages, peaks in sampling frequencies coincided with different epidemiological waves of infection in Mexico. Lineages B.1.1.222 and B.1.1.519 exhibited similar dynamics, constituting clades that likely originated in Mexico and persisted for >12 months. Lineages B.1.1.7, P.1 and B.1.617.2 also displayed similar dynamics, characterized by multiple introduction events leading to a few successful extended local transmission chains that persisted for several months. For the largest B.1.617.2 clades, we further explored viral lineage movements across Mexico. Many clades were located within the south region of the country, suggesting that this area played a key role in the spread of SARS-CoV-2 in Mexico.

CurSa: scripts to curate metadata and sample genomes from GISAID for analysis and display in nextstrain and microreact.

The coronavirus SARS-CoV-2 is the most sequenced pathogen ever, with several million genome copies deposited in the GISAID database. This large amount of genomic information poses non-trivial bioinformatic challenges for those interested in studying the evolution of SARS-CoV-2. One common problem when studying the phylogeny of the coronavirus in its geographical context is to count with accurate information of the location of the samples. However, this information is filled by hand by research groups all over the world and sometimes typos and inconsistencies are introduced in the metadata when submitting the sequences to GISAID. Correcting these errors is laborious and time-consuming. Here, we provide a suite of Perl scripts designated to facilitate the curation of this vital information and perform a random sampling of genome sequences if necessary. The scripts provided here can be used to curate geographic information in the metadata and sample the sequences from any country of interest to ease the preparation of files for Nextstrain and Microreact, thus accelerating evolutionary studies of this important pathogen. CurSa scripts are accessible via: https://github.com/luisdelaye/CurSa/.

Twisting development, the birth of a potential new gene.

Evolution has long been considered to be a conservative process in which new genes arise from pre-existing genes through gene duplication, domain shuffling, horizontal transfer, overprinting, retrotransposition, etc. However, this view is changing as new genes originating from non-genic sequences are discovered in different organisms. Still, rather limited functional information is available. Here, we have identified TWISTED1 (TWT1), a possible de novo-originated protein-coding gene that modifies microtubule arrangement and causes helicoidal growth in Arabidopsis thaliana when its expression is increased. Interestingly, even though TWT1 is a likely recent gene, the lack of TWT1 function affects A. thaliana development. TWT1 seems to have originated from a non-genic sequence. If so, it would be one of the few examples to date of how during evolution de novo genes are integrated into developmental cellular and organismal processes.

Translation Comes First: Ancient and Convergent Selection of Codon Usage Bias Across Prokaryotic Genomes.

Codon usage is the outcome of different evolutionary processes and can inform us about the conditions in which organisms live and evolve. Here, we present R_ENC', which is an improvement to the original S index developed by dos Reis et al. (2004). Our index is less sensitive to G+C content, which greatly affects synonymous codon usage in prokaryotes, making it better suited to detect selection acting on codon usage. We used R_ENC' to estimate the extent of selected codon usage bias in 1800 genomes representing 26 prokaryotic phyla. We found that Gammaproteobacteria, Betaproteobacteria, Actinobacteria, and Firmicutes are the phyla/subphyla showing more genomes with selected codon usage bias. In particular, we found that several lineages within Gammaproteobacteria and Firmicutes show a similar set of functional terms enriched in genes under selected codon usage bias, indicating convergent evolution. We also show that selected codon usage bias tends to evolve in genes coding for the translation machinery before other functional GO terms. Finally, we discuss the possibility to use R_ENC' to predict whether lineages evolved in copiotrophic or oligotrophic environments.

Emergence and widespread circulation of a recombinant SARS-CoV-2 lineage in North America.

Although recombination is a feature of coronavirus evolution, previously detected recombinant lineages of SARS-CoV-2 have shown limited circulation thus far. Here, we present a detailed phylogenetic analysis of four SARS-CoV-2 lineages to investigate the possibility of virus recombination among them. Our analyses reveal well-supported phylogenetic differences between the Orf1ab region encoding viral non-structural proteins and the rest of the genome, including Spike (S) protein and remaining reading frames. By accounting for several deletions in NSP6, Orf3a, and S, we conclude that the B.1.628 major cluster, now designated as lineage XB, originated from a recombination event between viruses of B.1.631 and B.1.634 lineages. This scenario is supported by the spatiotemporal distribution of these lineages across the USA and Mexico during 2021, suggesting that the recombination event originated in this geographical region. This event raises important questions regarding the role and potential effects of recombination on SARS-CoV-2 evolution.

The Alpha Variant (B.1.1.7) of SARS-CoV-2 Failed to Become Dominant in Mexico.

During the coronavirus disease 2019 (COVID-19) pandemic, the emergence and rapid increase of the B.1.1.7 (Alpha) lineage of severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2), first identified in the United Kingdom in September 2020, was well documented in different areas of the world and became a global public health concern because of its increased transmissibility. The B.1.1.7 lineage was first detected in Mexico during December 2020, showing a slow progressive increase in its circulation frequency, which reached its maximum in May 2021 but never became predominant. In this work, we analyzed the patterns of diversity and distribution of this lineage in Mexico using phylogenetic and haplotype network analyses. Despite the reported increase in transmissibility of the B.1.1.7 lineage, in most Mexican states, it did not displace cocirculating lineages, such as B.1.1.519, which dominated the country from February to May 2021. Our results show that the states with the highest prevalence of B.1.1.7 were those at the Mexico-U.S. border. An apparent pattern of dispersion of this lineage from the northern states of Mexico toward the center or the southeast was observed in the largest transmission chains, indicating possible independent introduction events from the United States. However, other entry points cannot be excluded, as shown by multiple introduction events. Local transmission led to a few successful haplotypes with a localized distribution and specific mutations indicating sustained community transmission. IMPORTANCE The emergence and rapid increase of the B.1.1.7 (Alpha) lineage of severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) throughout the world were due to its increased transmissibility. However, it did not displace cocirculating lineages in most of Mexico, particularly B.1.1.519, which dominated the country from February to May 2021. In this work, we analyzed the distribution of B.1.1.7 in Mexico using phylogenetic and haplotype network analyses. Our results show that the states with the highest prevalence of B.1.1.7 (around 30%) were those at the Mexico-U.S. border, which also exhibited the highest lineage diversity, indicating possible introduction events from the United States. Also, several haplotypes were identified with a localized distribution and specific mutations, indicating that sustained community transmission occurred in the country.

An Upgrade on the Surveillance System of SARS-CoV-2: Deployment of New Methods for Genetic Inspection.

SARS-CoV-2 variants surveillance is a worldwide task that has been approached with techniques such as Next Generation Sequencing (NGS); however, this technology is not widely available in developing countries because of the lack of equipment and limited funding in science. An option is to deploy a RT-qPCR screening test which aids in the analysis of a higher number of samples, in a shorter time and at a lower cost. In this study, variants present in samples positive for SARS-CoV-2 were identified with a RT-qPCR mutation screening kit and were later confirmed by NGS. A sample with an abnormal result was found with the screening test, suggesting the simultaneous presence of two viral populations with different mutations. The DRAGEN Lineage analysis identified the Delta variant, but there was no information about the other three mutations previously detected. When the sequenced data was deeply analyzed, there were reads with differential mutation patterns, that could be identified and classified in terms of relative abundance, whereas only the dominant population was reported by DRAGEN software. Since most of the software developed to analyze SARS-CoV-2 sequences was aimed at obtaining the consensus sequence quickly, the information about viral populations within a sample is scarce. Here, we present a faster and deeper SARS-CoV-2 surveillance method, from RT-qPCR screening to NGS analysis.

Draft Genome Sequence of a Kazachstania humilis Strain Isolated from Agave Fermentation.

The ascomycetous yeast Kazachstania humilis is an active species in backslopped sourdough and in the spontaneous fermentation of several traditional foods and beverages. Here, we report the draft genome sequence of a K. humilis strain isolated from agave must from a traditional distillery in Mexico.

Phylogenomics and population genomics of SARS-CoV-2 in Mexico during the pre-vaccination stage reveals variants of interest B.1.1.28.4 and B.1.1.222 or B.1.1.519 and the nucleocapsid mutation S194L associated with symptoms.

Understanding the evolution of the SARS-CoV-2 virus in various regions of the world during the Covid-19 pandemic is essential to help mitigate the effects of this devastating disease. We describe the phylogenomic and population genetic patterns of the virus in Mexico during the pre-vaccination stage, including asymptomatic carriers. A real-time quantitative PCR screening and phylogenomic reconstructions directed at sequence/structure analysis of the spike glycoprotein revealed mutation of concern E484K in genomes from central Mexico, in addition to the nationwide prevalence of the imported variant 20C/S:452R (B.1.427/9). Overall, the detected variants in Mexico show spike protein mutations in the N-terminal domain (i.e. R190M), in the receptor-binding motif (i.e. T478K, E484K), within the S1-S2 subdomains (i.e. P681R/H, T732A), and at the basis of the protein, V1176F, raising concerns about the lack of phenotypic and clinical data available for the variants of interest we postulate: 20B/478K.V1 (B.1.1.222 or B.1.1.519) and 20B/P.4 (B.1.1.28.4). Moreover, the population patterns of single nucleotide variants from symptomatic and asymptomatic carriers obtained with a self-sampling scheme confirmed the presence of several fixed variants, and differences in allelic frequencies among localities. We identified the mutation N:S194L of the nucleocapsid protein associated with symptomatic patients. Phylogenetically, this mutation is frequent in Mexican sub-clades. Our results highlight the dual and complementary role of spike and nucleocapsid proteins in adaptive evolution of SARS-CoV-2 to their hosts and provide a baseline for specific follow-up of mutations of concern during the vaccination stage.

RNA polymerases in strict endosymbiont bacteria with extreme genome reduction show distinct erosions that might result in limited and differential promoter recognition.

Strict endosymbiont bacteria present high degree genome reduction, retain smaller proteins, and in some instances, lack complete functional domains compared to free-living counterparts. Until now, the mechanisms underlying these genetic reductions are not well understood. In this study, the conservation of RNA polymerases, the essential machinery for gene expression, is analyzed in endosymbiont bacteria with extreme genome reductions. We analyzed the RNA polymerase subunits to identify and define domains, subdomains, and specific amino acids involved in precise biological functions known in Escherichia coli. We also perform phylogenetic analysis and three-dimensional models over four lineages of endosymbiotic proteobacteria with the smallest genomes known to date: Candidatus Hodgkinia cicadicola, Candidatus Tremblaya phenacola, Candidatus Tremblaya Princeps, Candidatus Nasuia deltocephalinicola, and Candidatus Carsonella ruddii. We found that some Hodgkinia strains do not encode for the RNA polymerase α subunit. The rest encode genes for α, ß, ß', and σ subunits to form the RNA polymerase. However, 16% shorter, on average, respect their orthologous in E. coli. In the α subunit, the amino-terminal domain is the most conserved. Regarding the ß and ß' subunits, both the catalytic core and the assembly domains are the most conserved. However, they showed compensatory amino acid substitutions to adapt to changes in the σ subunit. Precisely, the most erosive diversity occurs within the σ subunit. We identified broad amino acid substitution even in those recognizing and binding to the -10-box promoter element. In an overall conceptual image, the RNA polymerase from Candidatus Nasuia conserved the highest similarity with Escherichia coli RNA polymerase and their σ70. It might be recognizing the two main promoter elements (-10 and -35) and the two promoter accessory elements (-10 extended and UP-element). In Candidatus Carsonella, the RNA polymerase could recognize all the promoter elements except the -10-box extended. In Candidatus Tremblaya and Hodgkinia, due to the α carboxyl-terminal domain absence, they might not recognize the UP-promoter element. We also identified the lack of the ß flap-tip helix domain in most Hodgkinia's that suggests the inability to bind the -35-box promoter element.

Chromosome-Level Genome Assembly of the American Cranberry (Vaccinium macrocarpon Ait.) and Its Wild Relative Vaccinium microcarpum.

The American cranberry (Vaccinium macrocarpon Ait.) is an iconic North American fruit crop of great cultural and economic importance. Cranberry can be considered a fruit crop model due to its unique fruit nutrient composition, overlapping generations, recent domestication, both sexual and asexual reproduction modes, and the existence of cross-compatible wild species. Development of cranberry molecular resources started very recently; however, further genetic studies are now being limited by the lack of a high-quality genome assembly. Here, we report the first chromosome-scale genome assembly of cranberry, cultivar Stevens, and a draft genome of its close wild relative species Vaccinium microcarpum. More than 92% of the estimated cranberry genome size (492 Mb) was assembled into 12 chromosomes, which enabled gene model prediction and chromosome-level comparative genomics. Our analysis revealed two polyploidization events, the ancient γ-triplication, and a more recent whole genome duplication shared with other members of the Ericaeae, Theaceae and Actinidiaceae families approximately 61 Mya. Furthermore, comparative genomics within the Vaccinium genus suggested cranberry-V. microcarpum divergence occurred 4.5 Mya, following their divergence from blueberry 10.4 Mya, which agrees with morphological differences between these species and previously identified duplication events. Finally, we identified a cluster of subgroup-6 R2R3 MYB transcription factors within a genomic region spanning a large QTL for anthocyanin variation in cranberry fruit. Phylogenetic analysis suggested these genes likely act as anthocyanin biosynthesis regulators in cranberry. Undoubtedly, these new cranberry genomic resources will facilitate the dissection of the genetic mechanisms governing agronomic traits and further breeding efforts at the molecular level.

Driven progressive evolution of genome sequence complexity in Cyanobacteria.

Progressive evolution, or the tendency towards increasing complexity, is a controversial issue in biology, which resolution entails a proper measurement of complexity. Genomes are the best entities to address this challenge, as they encode the historical information of a species' biotic and environmental interactions. As a case study, we have measured genome sequence complexity in the ancient phylum Cyanobacteria. To arrive at an appropriate measure of genome sequence complexity, we have chosen metrics that do not decipher biological functionality but that show strong phylogenetic signal. Using a ridge regression of those metrics against root-to-tip distance, we detected positive trends towards higher complexity in three of them. Lastly, we applied three standard tests to detect if progressive evolution is passive or driven-the minimum, ancestor-descendant, and sub-clade tests. These results provide evidence for driven progressive evolution at the genome-level in the phylum Cyanobacteria.

Inferring Horizontal Gene Transfer with DarkHorse, Phylomizer, and ETE Toolkits.

In this chapter, we describe how to use DarkHorse2.0 to search for xenologs in the genome of the cyanobacterium Synechococcus elongatus PCC 7942. DarkHorse is an implicit phylogenetic method that uses BLAST searches to identify proteins having close homologs of unexpected taxonomic affiliation. Once a set of putative xenologs are identified, Phylomizer is used to reconstruct phylogenetic trees. Phylomizer reproduces all the necessary steps to perform a basic phylogenetic analysis. The combined use of DarkHorse and Phylomizer allows the identification of genes incorporated into a given genome by HGT.

Transcriptome analysis of bolting in A. tequilana reveals roles for florigen, MADS, fructans and gibberellins.

BACKGROUND: Reliable indicators for the onset of flowering are not available for most perennial monocarpic species, representing a drawback for crops such as bamboo, agave and banana. The ability to predict and control the transition to the reproductive stage in A. tequilana would represent an advantage for field management of agaves for tequila production and for the development of a laboratory model for agave species. RESULTS: Consistent morphological features could not be determined for the vegetative to reproductive transition in A. tequilana. However, changes in carbohydrate metabolism where sucrose decreased and fructans of higher degree of polymerization increased in leaves before and after the vegetative to reproductive transition were observed. At the molecular level, transcriptome analysis from leaf and shoot apical meristem tissue of A. tequilana plants from different developmental stages identified OASES as the most effective assembly program and revealed evidence for incomplete transcript processing in the highly redundant assembly obtained. Gene ontology analysis uncovered enrichment for terms associated with carbohydrate and hormone metabolism and detailed analysis of expression patterns for individual genes revealed roles for specific Flowering locus T (florigen), MADS box proteins, gibberellins and fructans in the transition to flowering. CONCLUSIONS: Based on the data obtained, a preliminary model was developed to describe the regulatory mechanisms underlying the initiation of flowering in A. tequilana. Identification of specific promoter and repressor Flowering Locus T and MADS box genes facilitates functional analysis and the development of strategies to modulate the vegetative to reproductive transition in A. tequilana.

Evolution of Base Excision Repair in Entamoeba histolytica is shaped by gene loss, gene duplication, and lateral gene transfer.

During its life cycle, the protist parasite Entamoeba histolytica encounters reactive oxygen and nitrogen species that alter its genome. Base excision repair (BER) is one of the most important pathways for the repair of DNA base lesions. Analysis of the E. histolytica genome revealed the presence of most of the BER components. Surprisingly, this included a gene encoding an apurinic/apyrimidinic (AP) endonuclease that previous studies had assumed was absent. Indeed, our analysis showed that the genome of E. histolytica harbors the necessary genes needed for both short and long-patch BER sub-pathways. These genes include DNA polymerases with predicted 5'-dRP lyase and strand-displacement activities and a sole DNA ligase. A distinct feature of the E. histolytica genome is the lack of several key damage-specific BER glycosylases, such as OGG1/MutM, MDB4, Mag1, MPG, SMUG, and TDG. Our evolutionary analysis indicates that several E. histolytica DNA glycosylases were acquired by lateral gene transfer (LGT). The genes that encode for MutY, AlkD, and UDG (Family VI) are included among these cases. Endonuclease III and UNG (family I) are the only DNA glycosylases with a eukaryotic origin in E. histolytica. A gene encoding a MutT 8-oxodGTPase was also identified that was acquired by LGT. The mixed composition of BER genes as a DNA metabolic pathway shaped by LGT in E. histolytica indicates that LGT plays a major role in the evolution of this eukaryote. Sequence and structural prediction of E. histolytica DNA glycosylases, as well as MutT, suggest that the E. histolytica DNA repair proteins evolved to harbor structural modifications that may confer unique biochemical features needed for the biology of this parasite.

Evidence of the Red-Queen Hypothesis from Accelerated Rates of Evolution of Genes Involved in Biotic Interactions in Pneumocystis.

Pneumocystis species are ascomycete fungi adapted to live inside the lungs of mammals. These ascomycetes show extensive stenoxenism, meaning that each species of Pneumocystis infects a single species of host. Here, we study the effect exerted by natural selection on gene evolution in the genomes of three Pneumocystis species. We show that genes involved in host interaction evolve under positive selection. In the first place, we found strong evidence of episodic diversifying selection in Major surface glycoproteins (Msg). These proteins are located on the surface of Pneumocystis and are used for host attachment and probably for immune system evasion. Consistent with their function as antigens, most sites under diversifying selection in Msg code for residues with large relative surface accessibility areas. We also found evidence of positive selection in part of the cell machinery used to export Msg to the cell surface. Specifically, we found that genes participating in glycosylphosphatidylinositol (GPI) biosynthesis show an increased rate of nonsynonymous substitutions (dN) versus synonymous substitutions (dS). GPI is a molecule synthesized in the endoplasmic reticulum that is used to anchor proteins to membranes. We interpret the aforementioned findings as evidence of selective pressure exerted by the host immune system on Pneumocystis species, shaping the evolution of Msg and several proteins involved in GPI biosynthesis. We suggest that genome evolution in Pneumocystis is well described by the Red-Queen hypothesis whereby genes relevant for biotic interactions show accelerated rates of evolution.

Testing the Domino Theory of Gene Loss in Buchnera aphidicola: The Relevance of Epistatic Interactions.

The domino theory of gene loss states that when some particular gene loses its function and cripples a cellular function, selection will relax in all functionally related genes, which may allow for the non-functionalization and loss of these genes. Here we study the role of epistasis in determining the pattern of gene losses in a set of genes participating in cell envelope biogenesis in the endosymbiotic bacteria Buchnera aphidicola. We provide statistical evidence indicating pairs of genes in B. aphidicola showing correlated gene loss tend to have orthologs in Escherichia coli known to have alleviating epistasis. In contrast, pairs of genes in B. aphidicola not showing correlated gene loss tend to have orthologs in E. coli known to have aggravating epistasis. These results suggest that during the process of genome reduction in B. aphidicola by gene loss, positive or alleviating epistasis facilitates correlated gene losses while negative or aggravating epistasis impairs correlated gene losses. We interpret this as evidence that the reduced proteome of B. aphidicola contains less pathway redundancy and more compensatory interactions, mimicking the situation of E. coli when grown under environmental constrains.

Emergence as an outbreak of the HIV-1 CRF19_cpx variant in treatment-naïve patients in southern Spain.

BACKGROUND: CRF19_cpx is a complex circulating recombination form (CRF) of HIV-1. We describe the characteristics of an outbreak of the CRF19_cpx variant among treatment-naïve patients in southern Spain. METHODS: The study was undertaken at the Virgen de la Victoria Hospital, a reference centre for the analysis of HIV-1 genotype in Malaga (Spain). Subtyping was performed through REGA v3.0 and the relationship of our CRF19_cpx sequences, among themselves and regarding other reference sequences from the same variant, was defined by phylogenetic analysis. We used PhyML program to perform a reconstruction of the phylogeny by Maximum Likelihood method as well as further confirmation of the transmission clusters by Bayesian inference. Additionally, we collected demographic, clinical and immunovirological data. RESULTS: Between 2011 and 2016, we detected 57 treatment-naïve patients with the CRF19_cpx variant. Of these, 55 conformed a very well-defined transmission cluster, phylogenetically close to CRF19_cpx sequences from the United Kingdom. The origin of this subtype in Malaga was dated between 2007 and 2010. Over 50% of the patients presented the non-nucleoside reverse transcriptase inhibitor G190A resistance mutation. This variant was mostly represented by young adult Spanish men who had sex with men. Almost half of them were recent seroconverters, though a similar percentage was diagnosed at a late state of HIV infection. Five cases of AIDS and one non-AIDS defined death occurred during follow-up. The majority of patients treated with first-line combination antiretroviral therapy (ART) responded. CONCLUSIONS: We report the largest HIV-1 CRF19_cpx cohort of treatment-naïve patients outside Cuba, almost all emerging as an outbreak in the South of Spain. Half the cases had the G190A resistance mutation. Unlike previous studies, the variant from Malaga seems less pathogenic, with few AIDS events and an excellent response to ART.

Natural selection drove metabolic specialization of the chromatophore in Paulinella chromatophora.

BACKGROUND: Genome degradation of host-restricted mutualistic endosymbionts has been attributed to inactivating mutations and genetic drift while genes coding for host-relevant functions are conserved by purifying selection. Unlike their free-living relatives, the metabolism of mutualistic endosymbionts and endosymbiont-originated organelles is specialized in the production of metabolites which are released to the host. This specialization suggests that natural selection crafted these metabolic adaptations. In this work, we analyzed the evolution of the metabolism of the chromatophore of Paulinella chromatophora by in silico modeling. We asked whether genome reduction is driven by metabolic engineering strategies resulted from the interaction with the host. As its widely known, the loss of enzyme coding genes leads to metabolic network restructuring sometimes improving the production rates. In this case, the production rate of reduced-carbon in the metabolism of the chromatophore. RESULTS: We reconstructed the metabolic networks of the chromatophore of P. chromatophora CCAC 0185 and a close free-living relative, the cyanobacterium Synechococcus sp. WH 5701. We found that the evolution of free-living to host-restricted lifestyle rendered a fragile metabolic network where >80% of genes in the chromatophore are essential for metabolic functionality. Despite the lack of experimental information, the metabolic reconstruction of the chromatophore suggests that the host provides several metabolites to the endosymbiont. By using these metabolites as intracellular conditions, in silico simulations of genome evolution by gene lose recover with 77% accuracy the actual metabolic gene content of the chromatophore. Also, the metabolic model of the chromatophore allowed us to predict by flux balance analysis a maximum rate of reduced-carbon released by the endosymbiont to the host. By inspecting the central metabolism of the chromatophore and the free-living cyanobacteria we found that by improvements in the gluconeogenic pathway the metabolism of the endosymbiont uses more efficiently the carbon source for reduced-carbon production. In addition, our in silico simulations of the evolutionary process leading to the reduced metabolic network of the chromatophore showed that the predicted rate of released reduced-carbon is obtained in less than 5% of the times under a process guided by random gene deletion and genetic drift. We interpret previous findings as evidence that natural selection at holobiont level shaped the rate at which reduced-carbon is exported to the host. Finally, our model also predicts that the ABC phosphate transporter (pstSACB) which is conserved in the genome of the chromatophore of P. chromatophora strain CCAC 0185 is a necessary component to release reduced-carbon molecules to the host. CONCLUSION: Our evolutionary analysis suggests that in the case of Paulinella chromatophora natural selection at the holobiont level played a prominent role in shaping the metabolic specialization of the chromatophore. We propose that natural selection acted as a "metabolic engineer" by favoring metabolic restructurings that led to an increased release of reduced-carbon to the host.