THE THIRD DOMAIN
Almost every microorganism used in industrial biomanufacturing is a bacterium. For bioplastic production, this reflects historical path dependence — E. coli, Corynebacterium, Bacillus, Cupriavidus are the organisms that modern molecular biology was built around — rather than a first-principles choice. Archaea, the third domain of life, possess structural properties that address the specific cost drivers that have kept PHA biopolymers above price parity with petrochemical plastics. One organism in particular, Haloferax mediterranei, has been studied for over four decades and is positioned to move from academic model to industrial chassis.
Why archaea — and why industrial biotechnology has been looking in the wrong domain of life
Carl Woese and George Fox established the tripartite structure of the tree of life in 1977, using 16S rRNA catalogues to show that the microorganisms then lumped as "prokaryotes" actually comprise two lineages as distant from each other as each is from eukaryotes. The formal three-domain taxonomy — Bacteria, Archaea, Eukarya — was codified by Woese, Kandler, and Wheelis in 1990. Archaea were recognised as a separate domain only after decades during which they had been studied primarily as curiosities of extreme environments: hot springs, salt lakes, anoxic sediments. Industrial biotechnology followed the organisms it already knew how to engineer, and archaea remained outside the default toolbox.
For PHA production specifically, four archaeal properties are structurally — not incrementally — advantageous. They reduce feedstock cost, eliminate sterility capex, simplify downstream processing, and remove the endotoxin burden that bacterial chassis carry into biomedical applications. The organism that concentrates all four of these properties is Haloferax mediterranei, an extreme halophile isolated from the Santa Pola solar salterns near Alicante and first described by Rodríguez-Valera, Juez, and Kushner in 1983. Its 3.9 Mb genome (plus three megaplasmids) was fully sequenced by Han and colleagues in 2012, and it is the second-best-characterised haloarchaeon after its sister species Haloferax volcanii.
The overlooked domain
The historical link runs as follows. In 1987, Ishino and colleagues at Osaka University noticed a curious cluster of repeated sequences downstream of the iap gene in E. coli K-12 — the first observation of what would later be called CRISPR arrays, though they made little of it. In 1989, Francisco Mojica began doctoral work at the University of Alicante under Francisco Rodríguez-Valera, studying how salinity affected restriction enzyme activity on the H. mediterranei genome. Mojica, Juez, and Rodríguez-Valera published the first archaeal observation of the same class of clustered palindromic repeats in Molecular Microbiology in 1993. Over the following decade, Mojica's group catalogued these loci across dozens of bacterial and archaeal genomes and, in 2005, independently with Pourcel and Bolotin, proposed that the spacer sequences derived from phage DNA and that the arrays functioned as an adaptive immune memory. The acronym CRISPR was coined by Mojica in correspondence with Ruud Jansen and first used in print in 2002.
The direct genome-editing tools that revolutionised molecular biology come from Streptococcus pyogenes Cas9, not from the H. mediterranei system (Jinek, Charpentier, and Doudna 2012; Cong and Zhang 2013). The relevance of H. mediterranei for engineering is different and more practical: its endogenous Type I-B CRISPR-Cas system has been repurposed in situ for metabolic engineering, without importing heterologous Cas machinery.
Two recent demonstrations are central. In 2021, Lin and colleagues (Chinese Academy of Sciences, published in Communications Biology) used the endogenous Type I-B system to implement CRISPR interference (CRISPRi), achieving tunable gene repression from 25% to 98% and demonstrating that plasmid-borne CRISPRi knockdown of the citrate synthase genes citZ and gltA increased PHBV titer by 76% (from 1.78 to 3.14 g/L). Integrating the crRNA cassette into the chromosome shortened the fermentation cycle and raised PHA productivity by 165% over the wild-type baseline. In 2025, published in FEMS Microbiology Letters, the first CRISPR-mediated gene knockout system in H. mediterranei was established, achieving 27% knockout efficiency on the hlyR4 target — roughly ninefold higher than the ~3% efficiency of conventional homologous recombination in this organism. The absolute knockout efficiency remains modest, and we treat advancing this toolbox as one of the core engineering priorities of the platform.
From the archaeon where CRISPR was first observed in archaea
The industrial default chassis — E. coli, Cupriavidus necator, Halomonas bluephagenesis — are Gram-negative bacteria. Several of their architectural features are directly adverse to low-cost PHA production.
Cell envelope: Gram-negative bacteria have a two-membrane architecture: an inner cytoplasmic membrane and an outer membrane containing lipopolysaccharide (LPS). LPS is the endotoxin that co-purifies with intracellular PHA and triggers immune responses in mammals; it is the reason every bacterial PHA platform requires dedicated endotoxin-reduction steps for biomedical-grade polymer. Haloarchaea have a single cytoplasmic membrane — a phospholipid bilayer of archaeol-based diether lipids (isoprenoid phytanyl chains linked via ether bonds to sn-glycerol-1-phosphate), dominated by phosphatidylglycerol methyl phosphate (PGP-Me) — covered externally by a two-dimensional glycoprotein S-layer. No outer membrane, no LPS, no peptidoglycan, no pseudomurein. This architectural difference is the basis of the inherent endotoxin-free polymer output that opens the biomedical application segment without the bacterial purification penalty.
Osmoadaptation strategy: Two strategies exist across the tree of life for maintaining cellular function against hypersaline environments. The "compatible solute" strategy, used by moderate halophiles including Halomonas and most halotolerant bacteria, excludes salt from the cytoplasm and synthesises or imports organic osmolytes (ectoine, glycine betaine, trehalose) to balance external osmotic pressure. The "salt-in" strategy, used by haloarchaea and by a few independently evolved halophilic bacteria (Salinibacter ruber), accumulates molar concentrations of K⁺ (up to ~4 M in the cytoplasm) against the external Na⁺-dominated medium, with Cl⁻ imported to maintain charge balance — in Halobacterium, via the light-driven pump halorhodopsin. The salt-in strategy requires proteome-wide adaptation: haloarchaeal proteins carry a large excess of acidic amino acids (aspartate, glutamate) on their surface and lose function at low salt. The practical consequence for industrial engineering is that haloarchaeal enzymes are evolved to function at cytoplasmic ionic strength approaching saturated KCl — the opposite of what non-halophilic bacterial hosts can tolerate.
Contamination barrier: Sterility cost in conventional fermentation is dominated by the need to exclude fast-growing environmental bacteria and fungi. Most Halomonas platforms (H. bluephagenesis, H. campaniensis) operate at 10–60 g/L NaCl — sufficient to exclude common environmental contaminants, but still accessible to salt-tolerant organisms including some Bacillus, Staphylococcus, and Debaryomyces species. H. mediterranei operates at 200–250 g/L NaCl (~3.4–4.3 M), which excludes essentially all non-halophilic organisms on energetic grounds: the ATP cost of maintaining osmotic homeostasis against this medium is prohibitive for any organism lacking pre-adapted salt-in or high-efficiency compatible-solute machinery. A narrow group of organisms grows at this salinity — other haloarchaea, Salinibacter ruber, Halanaerobium, and the halophilic fungi Wallemia ichthyophaga and Hortaea werneckii — but none are plausible fermenter invaders outside solar-saltern environments, and none grow fast enough to compete with H. mediterranei at industrial cell densities. The barrier is not absolute, but it is an order of magnitude stronger than that of moderate halophiles and operationally sufficient for open, non-sterile continuous fermentation.
Genome organisation: Haloarchaea are polyploid. H. volcanii carries approximately 20 genome copies per cell under optimal conditions (range 2–40 depending on growth phase and phosphate availability); H. mediterranei is similarly polyploid, with independent copy number regulation on its main chromosome and megaplasmids (Liu et al. 2013). This has two consequences relevant to industrial engineering. First, polyploidy provides genetic robustness — resistance to desiccation, radiation, and metabolic stress — which is operationally useful at scale. Second, it complicates clean gene knockouts: all genome copies must be edited to fix a phenotype, which is one reason conventional homologous recombination achieves only ~3% knockout efficiency in this organism. A 2023 study by Dattani and colleagues found that unlike H. volcanii, H. mediterranei maintains heteroploidy (different alleles on different chromosome copies) for extended periods in non-selective conditions. This is a double-edged property: it can be exploited for allele-dosage tuning and strain-library construction, but it requires explicit verification of homozygosity before any engineered phenotype can be declared stable.
Archaea versus bacteria: the structural differences that matter
Halophilic Archaea (H. mediterranei)
Cell envelope: No outer membrane. No lipopolysaccharide (LPS). Ether-linked isoprenoid lipids form a monolayer membrane resistant to extreme osmotic stress.
Salt tolerance: 150–300 g/L NaCl. "Compatible solute" and "salt-in" strategy for osmotic balance. Growth medium is uninhabitable for virtually all non-halophilic life.
Cell lysis: Catastrophic osmotic lysis in fresh water. The 200+ g/L NaCl differential drives complete membrane rupture. No mechanical disruption needed.
Ploidy: Polyploid — multiple genome copies per cell (up to ~20 in stationary phase). Increases genetic robustness and recombination potential, but complicates clean gene knockouts.
Endotoxins: None. Archaea do not synthesise LPS. PHA is inherently endotoxin-free.
CRISPR: Endogenous Type I-B CRISPR-Cas system. No foreign Cas proteins required.
Rodríguez-Valera and colleagues isolate a novel extreme halophile from Santa Pola saltern ponds, originally classified as Halobacterium mediterranei and later reclassified to Haloferax mediterranei by Torreblanca in 1986.
Halophilic Bacteria (H. bluephagenesis, H. boliviensis)
Cell envelope: Gram-negative outer membrane with LPS. Ester-linked fatty acid phospholipid bilayer. Standard bacterial architecture.
Salt tolerance: 10–60 g/L NaCl (moderate halophile). Sufficient for non-sterile operation but contamination barrier is weaker than extreme halophiles.
Cell lysis: Moderate osmotic shock possible. Typically requires SDS washing, enzymatic digestion, or engineered outer membrane defects for effective PHA release.
Ploidy: Typically haploid or diploid. Clean gene knockouts are straightforward. Standard bacterial genetics.
Endotoxins: LPS present in outer membrane. Co-purifies with PHA. Requires additional purification for biomedical applications.
CRISPR: Heterologous CRISPR/Cas9 imported. Well-developed toolbox (especially in H. bluephagenesis).
Published non-sterile continuous fermentation in Halomonas is a real and validated achievement — PhaBuilder's 10,000 tpa H. bluephagenesis plant in China is built on it, and continuous runs of 14 to 65 days have been reported without contamination in related Halomonas strains. But the engineered low-salt mutants of H. bluephagenesis developed to reduce NaCl consumption push the operating point down toward salinities at which salt-tolerant contaminants can persist. The moderate-halophile contamination barrier is a cost-of-failure-mode optimisation, not a structural exclusion.
The extreme-halophile regime is structural. At 200–250 g/L NaCl, maintaining intracellular homeostasis requires machinery that takes evolutionary timescales to assemble — either a proteome-wide acidic substitution pattern to tolerate molar K⁺ (salt-in) or compatible-solute biosynthesis running at metabolic cost exceeding 20% of ATP budget for non-adapted organisms. Common environmental bacteria and fungi possess neither. The contamination barrier for H. mediterranei fermentation is not "better than Halomonas" in the sense of being 4× higher salt; it is a different category of exclusion mechanism.
WHY THE CONTAMINATION BARRIER IS QUALITATIVELY, NOT QUANTITATIVELY, DIFFERENT
WHY OSMOTIC LYSIS ECONOMICS ARE NON-LINEAR
Downstream processing cost difference between extreme halophiles and moderate halophiles is not proportional to the salt-concentration difference. It is threshold-dependent.
At 10–60 g/L NaCl (moderate halophile regime), resuspending harvested biomass in fresh water produces a modest osmotic gradient (~0.3–2 MPa). Some cells lyse partially; many remain intact. Effective PHA recovery in H. bluephagenesis requires SDS detergent washing, enzymatic digestion, or engineered outer-membrane porosity, each adding a process step, reagent cost, and yield penalty.
At 200–250 g/L NaCl (extreme halophile regime), the osmotic differential exceeds 17 MPa — approximately 7× seawater osmotic pressure. The single archaeal membrane has no capacity to resist this gradient, and lysis is rapid and substantially complete. Polymer granules release directly into the aqueous phase and are recovered by centrifugation or microfiltration, with no detergent, enzyme, or mechanical disruption. This is not a DSP step that has been simplified — it is a unit operation that has been eliminated. When DSP accounts for 30–50% of total PHA production cost in techno-economic analyses, eliminating the primary DSP step is a structural change to the cost model, not a marginal optimisation. The effect also preserves polymer molecular weight, since the high-shear homogenisation that limits MW in conventional platforms is no longer in the process train.
H. mediterranei is not a newly discovered organism being rushed into application. It has a deep and expanding scientific foundation.
Four decades of science, now entering industry
1980 - 1993
Rodríguez-Valera and colleagues isolate a novel extreme halophile from Santa Pola saltern ponds, originally classified as Halobacterium mediterranei and later reclassified to Haloferax mediterranei by Torreblanca in 1986. Mojica, Juez, and Rodríguez-Valera publish the first observation of clustered palindromic repeats in an archaeal genome (Mol. Microbiol., 1993). The repeats had been seen earlier in E. coli (Ishino 1987) but without recognition of structural commonality.
2002–2005
The acronym CRISPR coined in Mojica–Jansen correspondence (first in print 2002). Mojica, Pourcel, and Bolotin independently propose the adaptive-immunity hypothesis (2005).
2008
Han and colleagues characterise the H. mediterranei PHA synthase as a Class III PhaC/PhaE heterodimer (Appl. Environ. Microbiol.), establishing the enzymatic basis of native PHBV production.
2012
Complete H. mediterranei genome published (Han et al., J. Bacteriol.): 3.9 Mb chromosome plus three megaplasmids (pHM500, pHM300, pHM100), with PHA biosynthesis gene cluster localised to pHM300.
2013
Zhao and colleagues demonstrate that knocking out the exopolysaccharide biosynthesis genes redirects carbon flux to PHA, increasing PHBV accumulation by ~20% and reducing fermentation broth viscosity. Liu and colleagues characterise the minimal replicon of pHM300 and independent copy number regulation of the main chromosome and megaplasmids, formally establishing the polyploid architecture.
2017–2024
Multiple groups (Koller, Rodríguez-Valera, Chen, Hezayen, Alsafadi, among others) demonstrate PHBV production from diverse waste feedstocks: VFAs from anaerobic digestion of food waste (55% higher titer than glucose), hydrolysed cheese whey, crude glycerol from biodiesel, olive mill wastewater, rapeseed meal, vinasse, and confectionery waste. Optimised cultivation reaches ~60–65% PHA by cell dry weight.
2021
Lin et al. (Commun. Biol.) establish CRISPRi in H. mediterranei using the endogenous Type I-B system. Tunable repression 25–98%. Chromosomal citrate-synthase knockdown delivers 165% productivity improvement.
2023
Dattani et al. (G3) characterise prolonged heteroploidy in H. mediterranei — a property with both risks and opportunities for strain engineering.
2025
First CRISPR-mediated gene knockout in H. mediterranei (FEMS Microbiol. Lett.); 27% efficiency, ninefold above homologous recombination.
Now
Exoform is building the first commercial PHA production platform on H. mediterranei, integrating the platform's inherent process advantages with computational enzyme engineering for monomer-specificity control.
The term "halophilic" spans two orders of magnitude in NaCl tolerance and the differences are industrially consequential.
Cupriavidus necator, the classical PHA platform, is a freshwater organism with no salt tolerance. Full sterilisation is mandatory. DSP requires mechanical disruption and solvent extraction. This is the legacy paradigm.
Halomonas bluephagenesis, the Chinese PhaBuilder / NGIB platform, is a moderate halophile operating at 10–60 g/L NaCl. Non-sterile fermentation is feasible and has been industrially validated. However, engineered low-salt mutants (optimised for reduced salt cost) operate at the lower end of this range, narrowing the contamination margin. Cell lysis requires SDS washing.
Haloferax mediterranei is an extreme halophile operating at 200–250 g/L NaCl. The growth medium is a qualitatively different contamination barrier. Lysis is spontaneous in fresh water. The organism belongs to a different domain of life, with distinct membrane chemistry and no LPS endotoxin.
The jump from 60 g/L to 200+ g/L is not a 4× improvement. It is a phase transition in process robustness and downstream economics.
Growth medium for H. mediterranei can be prepared by dissolving solar salt in water — a commodity input at roughly $40–60/tonne — or by evaporative concentration of seawater at coastal facilities. Spent medium recycle has been demonstrated in the published literature (Hermann-Krauss et al., Koller lab) at substitution rates replacing a substantial fraction of fresh mineral medium without yield loss, reducing both salt input cost and saline effluent disposal.
The halophilic spectrum: not all salt tolerance is equal
The seawater opportunity: At 200+ g/L NaCl, the growth medium can be prepared by dissolving solar salt (a low-cost commodity) in water, or by concentrating seawater. For coastal production facilities, this further reduces input costs and freshwater consumption. The saline spent medium is recyclable in subsequent fermentation cycles and published work has demonstrated that spent broth substitutes for a substantial fraction of fresh mineral medium.
An organism whose time has arrived
For four decades, H. mediterranei has been the domain of academic research. Characterised, sequenced, engineered in proof-of-concept studies, but never developed into a commercial production platform. Three constraints held it back, and three recent developments have removed them.
First, the genetic toolbox for haloarchaea was primitive relative to E. coli or C. necator. Conventional homologous recombination at ~3% efficiency made rational strain engineering slow. This is no longer true: the 2021 CRISPRi system and the 2025 CRISPR-KO system together now support the kind of iterative metabolic engineering that drove the rise of bacterial chassis in the 1990s and 2000s.
Second, the economics of PHA production had not yet become dominated by feedstock and DSP costs — categories in which H. mediterranei has structural advantages over every bacterial platform. As the sector moves toward scaled, cost-competitive production, these categories are now where the competition is decided.
Third, the concentration of halophilic PHA production in China, built on a single organism and a single university's patent estate, has created a strategic vulnerability for Western markets that rely on biodegradable polymers. Industrialising H. mediterranei on different chemistry, different IP, and different supply chains is not a preference — it is a necessity.
Exoform exists to close the remaining gaps between what H. mediterranei can do and what industrial-scale bioplastic production requires. Its structural advantages — waste-feedstock utilisation, osmotic lysis, non-sterile operation, LPS-free polymer — are inherent. They do not need to be engineered in. What needs to be engineered is higher cell density, broader copolymer specificity, and the process-control infrastructure required for consistent, scalable production.
That is our work.
Frequently asked questions
What is the difference between halophilic archaea and halophilic bacteria for bioplastic production?
Halophilic archaea (such as Haloferax mediterranei) and halophilic bacteria (such as Halomonas bluephagenesis) both enable non-sterile fermentation, but differ in several structurally important ways. Archaea operate at much higher salt concentrations (200+ g/L vs 10–60 g/L), providing a stronger contamination barrier. Archaea lyse more completely in fresh water due to the larger osmotic differential. Archaea do not produce LPS endotoxins, yielding biomedical-grade PHA without extra purification. Archaea have distinct membrane chemistry (ether-linked lipids vs ester-linked), distinct genetic systems, and are polyploid (multiple genome copies), which can increase genetic robustness. Bacteria currently have more mature genetic engineering toolboxes and achieve higher cell densities.
Do archaea produce endotoxins (LPS)?
No. Lipopolysaccharide (LPS) endotoxins are exclusive to Gram-negative bacteria. Archaea have fundamentally different cell envelope structures — their membranes are composed of ether-linked isoprenoid lipids rather than ester-linked fatty acid lipids, and they lack the outer membrane where LPS is located in bacteria. PHA produced by archaea such as Haloferax mediterranei is inherently endotoxin-free, which is critical for biomedical applications where LPS contamination triggers immune responses.
Extreme halophiles (like H. mediterranei, growing at 200+ g/L NaCl) have three structural advantages over moderate halophiles (like Halomonas spp., growing at 10–60 g/L NaCl): (1) A much stronger contamination barrier — at 200+ g/L NaCl, essentially no competing organisms can survive, while moderate halophile conditions can still support growth of some salt-tolerant contaminants. (2) More effective osmotic lysis — the larger osmotic differential produces more complete cell disruption in fresh water, improving PHA recovery without mechanical assistance. (3) Seawater supplementation — extreme halophilic media can be prepared using concentrated seawater or solar salt, potentially reducing freshwater consumption more effectively than moderate-salt systems.
What are the advantages of extreme halophiles over moderate halophiles for industrial fermentation?
Can archaea be used in non-sterile industrial fermentation?
Yes. Halophilic archaea are excellent candidates for non-sterile industrial fermentation because their growth conditions — high salt concentration (150–300 g/L NaCl), often combined with high pH — are inhospitable to virtually all non-halophilic contaminants. This has been demonstrated experimentally with multiple haloarchaeal species. Open, unsterile fermentation eliminates the need for autoclaving feedstocks and bioreactors, sterile-grade air filtration, and clean-room infrastructure, significantly reducing both capital and operating costs.