Heterologous expression of the human cohesin complex in Saccharomyces cerevisiae results in a dominant-negative phenotype

Heterologous expression of the human cohesin complex in *Saccharomyces cerevisiae* fails to rescue yeast cohesin mutants and instead induces a dominant-negative phenotype by forming dysfunctional hybrid complexes with endogenous yeast cohesin rings, leading to cohesion dysregulation and DNA damage sensitivity.

The Gist

The Big Idea: Trying to Swap a Car Engine

Imagine you have a very reliable, old-fashioned bicycle (the Yeast). It works perfectly. Now, imagine you want to swap out its tiny, simple chain and gears for a massive, high-tech engine from a modern sports car (the Human Cohesin).

The scientists in this paper asked a simple question: "If we take the human version of the 'glue' that holds our DNA together and put it inside a yeast cell, will it work?"

They hoped the human "glue" would be so good that it could replace the yeast's glue entirely. Instead, they found something much more interesting: The human glue didn't just fail to work; it actively broke the bicycle.

---

The Experiment: The "Humanized" Yeast

The Goal:
In biology, scientists often use yeast to study human diseases because yeast cells are simple and easy to grow. If a human gene can fix a broken yeast gene, we call that "complementation." The team tried to build a "humanized" yeast by replacing the yeast's DNA-glue components with human ones.

The Setup:

• The Yeast Glue: Yeast has a complex machine made of four main parts (let's call them the "Big Four") that hold its DNA together during cell division.
• The Human Glue: Humans have a very similar machine, but the parts are slightly different.
• The Plan: The scientists built "neochromosomes" (essentially extra, artificial chromosomes) loaded with human genes. They tried swapping in just one human part, then two, then all four, and even added extra helper parts (up to 15 human genes total!).

The Result: A "Poison Pill" Effect

1. It Didn't Fix the Broken Bike
When the scientists removed the yeast's glue and replaced it with the human glue, the yeast didn't just stop working; it died. The human parts simply couldn't do the job in the yeast environment. It was like trying to start a sports car engine in a bicycle frame; the parts didn't fit right, and the whole thing fell apart.

2. The "Dominant-Negative" Phenomenon (The Poison)
Here is the twist: Even when the yeast still had its own glue working perfectly, adding the human glue made the yeast sick.

• The Analogy: Imagine a team of four people (the yeast) building a wall. If you bring in a team of four strangers (the human) who look similar but don't speak the same language, they don't just stand aside. They jump into the mix, grab the bricks, and try to build. But because they don't know the yeast's blueprint, they build a wobbly, broken wall.
• The Science: The human "glue" parts mixed with the yeast "glue" parts to create hybrid monsters. These mixed-up machines were dysfunctional. They clogged up the system, causing the yeast to struggle with DNA damage and cell division. This is called a dominant-negative effect: the bad human parts "poisoned" the good yeast parts.

The Investigation: How Did They Figure It Out?

The scientists wanted to know why this was happening.

• The "Sticky" Test: They checked if the human glue was actually sticking to the yeast DNA. It was! The human parts were latching onto the yeast DNA, but they were stuck there doing nothing useful.
• The "Handshake" Test: They used a technique called Co-Immunoprecipitation (basically, fishing out proteins to see what they are holding hands with). They found that the human parts were physically grabbing onto the yeast parts.
  • Surprise: Even when they removed the human "Big Four" parts, the remaining human helper parts still tried to shake hands with the yeast parts, though they couldn't stick to the DNA without the main human parts.
• The "Engine" Test: They wondered if the human glue was broken because it couldn't use energy (ATP). They tried breaking the human glue's "engine" on purpose. Surprisingly, it didn't matter. Whether the human glue had a working engine or a broken one, it still poisoned the yeast. The problem wasn't the engine; it was that the human parts just didn't fit the yeast blueprint.

The Conclusion: Why This Matters

1. You Can't Just Swap Complex Systems
This study shows that while yeast and humans are related, their complex machines (like the cohesin glue) have evolved too differently to be swapped out easily. You can't just drop a human engine into a yeast car and expect it to run.

2. A New Tool for Cancer Research
Here is the silver lining. The scientists realized that this "poison" effect might be useful.

• The Analogy: Imagine a cancer cell is a house with a slightly cracked foundation (a mutation in its glue). A normal house (healthy cell) has a strong foundation. If you introduce a "poison" that only breaks houses with cracked foundations, you could kill the cancer without hurting the healthy house.
• The Application: Because the human glue is toxic to yeast that already have glue problems, this suggests that human cohesin mutations might be a target for cancer drugs. If we can find a way to make human cancer cells act like the "broken yeast," we might be able to kill them selectively.

Summary

The scientists tried to replace yeast glue with human glue. It didn't work as a replacement. Instead, the human glue acted like a spoiler, mixing with the yeast glue to create broken machines that stopped the cell from working. While this failed to "humanize" the yeast, it revealed a dangerous weakness in cells with DNA glue problems, offering a new potential strategy for fighting cancer.

Technical Summary

1. Problem Statement

The cohesin complex is essential for sister chromatid cohesion, DNA replication, genome organization, and DNA damage response. While the complex is highly conserved between humans and Saccharomyces cerevisiae (yeast), human cohesin genes are often essential, making in vivo functional studies difficult. Researchers aim to create "humanized" yeast models where human orthologs replace yeast genes to study disease variants and drug responses. However, it is unknown if the human cohesin complex can functionally replace the yeast complex, or if the cross-species interaction of these large, multi-subunit complexes leads to functional incompatibility or toxicity.

2. Methodology

The authors employed a systematic complementation strategy using synthetic biology and genetic manipulation in yeast:

• Single and Multi-Gene Complementation: Human orthologs of cohesin core subunits (SMC1A, SMC3, RAD21, STAG2), loaders, and accessory factors were expressed in yeast temperature-sensitive (Ts) mutants of the corresponding yeast genes (smc1, smc3, mcd1, irr1, eco1, pds1).
• Neochromosome Engineering: To test multi-subunit replacement, the authors utilized synthetic neochromosomes containing varying combinations of human genes (up to 15 genes), including:
  • hC: 4 core subunits.
  • hCL/hCL2A/hCL6A: Core subunits plus loaders and accessory factors (including WAPL, CDCA5, ESPL1, PTTG1).
  • h3P: Human paralogs (STAG1, PDS5A, ESCO2).
• Plasmid Shuffle Assays: To test if human cohesin could replace yeast cohesin entirely, they created strains with genomic deletions of up to 9 yeast cohesin-associated genes, supported by a yeast neochromosome (yCL3A). They then attempted to swap the yeast neochromosome for human neochromosomes using 5-FOA counter-selection.
• Phenotypic Analysis: Growth assays (spot and liquid culture) were performed at permissive and restrictive temperatures, and in the presence of DNA damaging agents (MMS).
• Mechanistic Investigation:
  • Chromatin Fractionation: To determine if human proteins load onto yeast chromatin.
  • Co-Immunoprecipitation (Co-IP): Using FLAG-tagged yeast Smc1 and antibodies against human subunits to detect physical interactions.
  • Single-Molecule Protein Sequencing (SMPS): Used on a Quantum-Si Platinum platform to identify specific yeast subunits in human-cohesin pulldowns, overcoming the lack of commercial antibodies for all yeast subunits.
  • Mutagenesis: Introduction of ATP-binding and hydrolysis-deficient mutations in human SMC proteins to test the dependency of the phenotype on ATPase activity.

3. Key Contributions

• Demonstration of Dominant-Negative Toxicity: The study establishes that human cohesin cannot functionally replace yeast cohesin. Instead, its expression in wild-type yeast creates a severe dominant-negative phenotype, causing growth defects and DNA damage sensitivity.
• Identification of Hybrid Complexes: The authors provide direct evidence that human SMC proteins (SMC1A/SMC3) physically interact with yeast subunits (Smc1/Smc3) to form dysfunctional "hybrid" cohesin rings.
• Mechanism of Toxicity: The dominant-negative effect is driven by the loading of these hybrid complexes onto yeast chromatin. The toxicity is dependent on the presence of human SMC proteins but is independent of ATP binding or hydrolysis.
• Synthetic Lethality: The expression of human cohesin is synthetically lethal in yeast strains with pre-existing mild cohesin defects (e.g., rad61Δ, ctf8Δ) or DNA repair defects (rad52Δ), highlighting the vulnerability of cells with compromised cohesin function to heterologous cohesin expression.

4. Results

• Failure of Complementation: Human cohesin orthologs, whether expressed singly or as complex combinations (up to 15 genes), failed to rescue the viability of yeast Ts mutants or strains with genomic deletions of cohesin genes.
• Dominant-Negative Phenotype: Expressing human core subunits (hC) in wild-type yeast caused significant growth defects and increased sensitivity to MMS. This phenotype was exacerbated by the co-expression of human loaders (hCL), suggesting that human subunits successfully load onto yeast DNA but fail to function correctly.
• Hybrid Complex Formation:
  • Chromatin fractionation showed human core subunits associate with yeast chromatin.
  • Co-IP and SMPS confirmed that human SMC1A and SMC3 interact with yeast Smc1 and Smc3.
  • Crucially, even when human non-SMC subunits (RAD21, STAG2) were expressed without human SMCs, they co-IPed with yeast Smc1, though they did not load onto chromatin or cause toxicity. This indicates that human SMCs are the primary drivers of the dominant-negative effect by facilitating the loading of hybrid rings.
• ATP Independence: Mutations in human SMC proteins that abolish ATP binding (K38I) or hydrolysis (E1144Q) did not suppress the dominant-negative phenotype or the chromatin loading, indicating the toxicity is structural rather than enzymatic.
• Cell Cycle Defects: Expression of human cohesin led to an accumulation of large-budded cells (G2/M arrest) and an increase in Rad52-YFP foci, indicating replication stress and DNA damage.

5. Significance

• Limits of Humanization: The study highlights the complexity of replacing multi-subunit complexes across species. Unlike metabolic pathways (e.g., glycolysis) which can sometimes be humanized, structural complexes like cohesin appear highly sensitive to subunit incompatibility, likely due to the precise stoichiometry and interaction interfaces required for ring closure and DNA entrapment.
• New Research Platform: Despite failing as a complementation tool, the dominant-negative phenotype offers a novel platform. The synthetic lethality observed between human cohesin expression and specific yeast mutations suggests that cohesin dysfunction could be a therapeutic target.
• Cancer Therapy Implications: The finding that human cohesin expression selectively kills cells with pre-existing cohesin defects (synthetic lethality) suggests a potential strategy for targeting cancers harboring somatic cohesin mutations. By introducing a "poison" subunit (human cohesin), one could selectively eliminate cancer cells while sparing normal cells with functional cohesin.
• Methodological Advancement: The successful use of Single-Molecule Protein Sequencing (SMPS) to identify protein interactions in the absence of specific antibodies demonstrates a powerful new tool for studying cross-species protein complexes.