Diatom Endosymbionts have Shrinking but Stable Genomes Despite Low Coding Density

This study reveals that diatom endosymbionts (spheroid bodies) are in the early stages of genome reduction with stable, shrinking genomes and unique features like the early loss of mobile elements, while exhibiting positive selection in genes related to host interaction and core nitrogen fixation.

The Gist

Imagine a tiny, microscopic world where a single-celled algae (a diatom) and a bacterium have decided to move in together and never leave. This isn't just a casual roommate situation; it's a permanent, life-or-death partnership called endosymbiosis.

For decades, scientists have studied these partnerships in insects (like aphids and their bacterial roommates). They noticed a pattern: over millions of years, the bacterial roommate gets lazy. It throws away its tools, shrinks its house, and becomes completely dependent on the host. Eventually, it becomes so small and specialized that it can't survive on its own.

But here's the mystery: How does this happen? Does the bacteria throw things away first, or does the host start controlling the house first? And does this happen the same way in all types of partnerships?

This paper investigates a brand-new set of roommates: Diatoms (a type of algae) and their bacterial roommates called Spheroid Bodies (SBs). The researchers wanted to catch this process in the "early stages" to see how the integration begins.

Here is the story of what they found, explained simply:

1. The "Shrinking but Messy" House

Usually, when a house gets smaller, you expect it to get tidier. But these bacterial roommates are living in a strange state.

• The Analogy: Imagine a house where the owner has thrown away the furniture (genes) to save space, but they haven't cleaned up the empty boxes and broken furniture pieces (pseudogenes) left behind.
• The Finding: The bacteria's genome (their instruction manual) is shrinking, but it's still very "messy." It has a lot of empty space between the instructions. This is different from older partnerships where the house is tiny and perfectly organized. These bacteria are still in the messy "moving out" phase.

2. The Missing Keys (Host Control)

In a normal bacteria, there are specific "keys" needed to start the engine of cell division (making new bacteria).

• The Analogy: Think of a car. Usually, the driver (the bacteria) has the keys to start the engine and drive.
• The Finding: The researchers found that these bacteria are missing two specific keys (dnaA and mltA) in every single one of the 22 new genomes they studied.
• What it means: It looks like the host (the diatom) has taken the keys away. The bacteria can't start their own engine anymore. The host likely provides the "starter" protein to tell the bacteria when to divide. This is a huge step toward total control by the host.

3. The "No-Tools" Rule (No Mobile Elements)

There is a popular theory that says bacteria shrink because they have too many "jumping genes" (transposases)—like little viruses inside their DNA that copy and paste themselves, causing chaos and forcing the bacteria to delete other things to survive.

• The Analogy: Imagine a library where books are constantly jumping off the shelves and rearranging themselves, causing the librarian to throw away whole sections of the library to make room.
• The Finding: These bacteria have zero jumping genes. They got rid of them very early on.
• Why it matters: This breaks the old rule. It suggests that in this specific partnership, the bacteria didn't shrink because of chaos; they shrank because they simply stopped needing those tools. They are stable, not chaotic, even though they are still losing genes.

4. The "Specialist" Team (Nitrogen Fixing)

These bacteria are hired for one specific job: Nitrogen Fixation. They take nitrogen from the air and turn it into food for the algae.

• The Analogy: Imagine a construction crew where everyone is a general contractor, but one guy is hired specifically to be the master electrician.
• The Finding: Even though the bacteria are losing genes left and right, they kept the "electrician's toolkit" (the genes for nitrogen fixation) perfectly intact. In fact, some of these specific tools are actually improving (evolving) to work better in the new environment.
• The Twist: The bacteria are working while the host is photosynthesizing (making energy from the sun). This creates a lot of oxygen, which is poison to the nitrogen-making machine. The bacteria are evolving their "electrician tools" to be more oxygen-resistant so they don't get poisoned while working.

5. The "Wall" and the "Door"

The researchers looked for genes that are changing rapidly (evolving fast). Usually, genes that change fast are the ones interacting with the outside world.

• The Analogy: If you live in a house, the walls and the front door are the parts that interact with the neighbors. The kitchen inside doesn't change much.
• The Finding: The genes that are evolving the fastest are the ones that build the cell wall and the membrane (the walls and doors of the bacteria).
• What it means: This is where the host and the bacteria are shaking hands. They are constantly tweaking how they touch and talk to each other to make the partnership work smoothly.

The Big Picture

This paper tells us that the story of how bacteria and hosts become best friends is more complex than we thought.

1. It's not always chaotic: These bacteria didn't shrink because of genetic chaos; they shrank because they lost their "mobile elements" (jumping genes) very early.
2. Control is key: The host is already taking control of the bacteria's reproduction by stealing the "keys" to start cell division.
3. It's a work in progress: These bacteria are in the "early stages" of becoming a permanent organelle (like a mitochondria). They are still messy, but they are on the right track.

In short: The diatom and its bacterial roommate are in the early days of a marriage. The bacteria is still unpacking boxes (losing genes), the host has already taken the keys to the car (controlling division), and they are both learning how to talk to each other through the front door (evolving cell walls). It's a fascinating glimpse into how life creates new, complex forms from simple partnerships.

Technical Summary

1. Problem Statement

The evolution of obligate endosymbiosis involves the integration of hosts and endosymbionts across genetic, metabolic, and cellular levels. While highly derived endosymbionts (e.g., in insects) typically exhibit extremely reduced genomes, the specific mechanisms driving early integration and the order of these changes remain unclear. Most existing models of endosymbiont evolution are based on invertebrate hosts, potentially obscuring universal evolutionary patterns.

The study focuses on Spheroid Bodies (SBs), nitrogen-fixing cyanobacterial endosymbionts of diatoms in the family Rhopalodiaceae. SBs represent an intermediate stage in endosymbiotic evolution: they are younger than insect endosymbionts (~34 Mya vs. >100 Mya) but show signs of metabolic integration (e.g., loss of carbon fixation genes). However, critical questions remain unresolved:

• What is the trajectory of SB genome reduction compared to current models?
• What is the closest free-living relative of SBs?
• How do hosts control SB growth and division?
• Which genes mediate direct host-symbiont interactions?

2. Methodology

The authors significantly expanded the genomic dataset for SBs to address these questions through the following steps:

• Sampling and Sequencing: Cultured diverse Rhopalodiaceae diatom strains from 15 locations across 6 US states (2020–2023). Assembled 22 new SB genomes from short-read metagenomic data, tripling the available SB genome count.
• Genome Assembly & Annotation: Used SPAdes for de novo assembly. Assemblies ranged from closed chromosomes to drafts (up to 477 contigs). Annotations were performed using PROKKA, with specific refinement for pseudogenes using pseudofinder against a custom database of free-living cyanobacteria.
• Phylogenomics: Constructed a genome-wide maximum likelihood phylogeny using 398 single-copy orthologs (30 SB taxa + 5 outgroups) via IQTree2. Gene concordance factors (gCF) were calculated to assess topological support and discordance.
• Pangenome Analysis: Estimated core and pan genomes using Roary and rarefaction curves (micropan package) to determine if the pangenome is open or closed and to estimate the "true" core genome size.
• Metabolic Reconstruction: Assigned KEGG Orthology (KO) IDs to core genes to map metabolic pathways, comparing SBs to free-living relatives (Rippkaea orientalis and "SU2").
• Selection Analysis: Used HyPhy BUSTED (Branch-site Unrestricted Statistical Test for Episodic Diversification) to detect episodic positive selection across 1,278 core genes. Tests included tree-wide analyses and specific branch/clade tests to identify genes under adaptive evolution.

3. Key Results

A. Genome Architecture and Stability

• Low Coding Density: SB genomes range from 2.4 to 3.2 Mbp with a coding density of 0.58–0.68 genes/kb (mean 0.63), significantly lower than free-living bacteria. This is attributed to intergenic regions containing degraded gene fragments (pseudogenes).
• Structural Stability: Contrary to models predicting that mobile elements drive early genome instability, SB genomes lack intact transposases and exhibit high synteny. This suggests an early, complete loss of mobile elements, leading to structural stability despite ongoing gene loss.
• Gene Content: The core genome contains 1,285 genes (estimated "true" core: 1,230 genes). The pan genome is estimated at ~4,989 genes, suggesting it still reflects ancestral content rather than a subset of lost genes.

B. Phylogeny and Evolutionary Origins

• Closest Relative: SBs form a monophyletic clade. The closest free-living relative is identified as "unicellular cyanobacterium SU2" (marine, Zanzibar), followed by the freshwater Rippkaea orientalis. This supports an independent origin of SBs from the UCYN-A lineage.
• Transmission: The SB phylogeny largely mirrors host phylogeny, suggesting predominantly vertical transmission.
• Habitat Transition: The phylogeny indicates a single transition from marine to freshwater habitats, with SBs and hosts co-diversifying.

C. Metabolic Capacity

• Nitrogen Fixation: All essential nif genes are retained, except nifJ (lost, likely due to stable intracellular iron).
• Carbon Metabolism: Pathways are incomplete. Glycolysis lacks phosphofructokinase (pfk), and the TCA cycle is truncated. The oxidative pentose phosphate pathway remains complete, likely to regenerate NADPH for nitrogenase.
• Vitamin/Cofactor Synthesis: Chlorophyll and phylloquinone synthesis are absent. Mo cofactor synthesis is lost (likely to prioritize Mo for Fe-Mo cofactor in nitrogenase). Vitamin B12 synthesis is present in most SBs but lost in Epithemia s. str. strains, indicating functional divergence.

D. Host Control Mechanisms

• Missing "Essential" Genes: Two genes typically essential for bacterial cell division are universally absent in all SB genomes:
  1. dnaA: Required for initiating DNA replication.
  2. mltA: A lytic transglycosylase involved in peptidoglycan (PG) degradation and cell wall integrity.
• Implication: The loss of these genes suggests the host has taken control of SB replication and division, potentially using host-encoded proteins to initiate these processes, thereby regulating symbiont population density and metabolic cost.

E. Positive Selection and Host Interaction

• Selection Pressure: Core genes are under moderately strong purifying selection (mean dN/dS = 0.173), indicating functional constraints despite genetic drift.
• Positively Selected Genes: 54 genes showed evidence of episodic positive selection.
  • Functional Categories: 21 genes are membrane-associated or involved in cell wall/membrane metabolism (transporters, signal transduction, PG synthesis).
  • Nitrogenase Adaptation: Surprisingly, three core nitrogen fixation genes (nifH, nifK, nifS) are under positive selection. This may indicate adaptation to the host's photosynthetic environment, specifically evolving an oxygen-insensitive nitrogenase to cope with concurrent host photosynthesis and N-fixation.

4. Key Contributions

1. Expanded Genomic Dataset: Provided a three-fold increase in SB genomes, enabling robust core/pan genome estimation and high-resolution phylogenetics.
2. Challenging Evolutionary Models: Demonstrated that SBs differ from the standard model of endosymbiont evolution. Instead of mobile-element-driven instability, SBs exhibit early loss of all transposases followed by stable, high-synteny genomes with low coding density.
3. Identification of Control Loci: Identified dnaA and mltA as universal absences, providing specific molecular targets for how hosts control endosymbiont division.
4. Insights into Host-Symbiont Interface: Highlighted that genes under positive selection are enriched for membrane/cell wall proteins and core metabolic enzymes (nif), suggesting active co-evolution at the physical interface and within the core metabolic machinery.
5. Evolutionary Context: Positioned SBs as an early-stage model for the evolution of the "nitroplast" (UCYN-A), showing that while they share functional similarities, SBs are less reduced and retain a larger pan-genome.

5. Significance

This study provides a critical window into the early stages of endosymbiont genome reduction. It reveals that genome stability can precede extreme reduction if mobile elements are lost early, challenging the view that transposase proliferation is a prerequisite for initial genome shrinkage. The findings suggest that host control over symbiont division is established early via the loss of specific replication and cell wall genes. Furthermore, the identification of positive selection on core nitrogenase genes implies that metabolic integration involves rapid adaptation to the host's physiological environment (e.g., oxygen levels). These insights refine our understanding of how obligate symbioses establish and are maintained in unicellular eukaryotic hosts, offering a comparative framework for studying the evolution of organelles.