Bacterial and eukaryotic membranes contain terpenoids mainly as regulatory components. While hopanoids and steroids are the two major membrane terpenoids, several specific lineages, including anaerobic protists and ciliates, alternatively produce a unique structural homolog called tetrahymanol (Takishita et al., 2012; Banta et al., 2015). In some species, the abundance of regulatory terpenoids is comparable to that of fatty acyl phospholipids (van Meer et al., 2008; Belin et al., 2018).

Steroids in eukaryotes are the most extensively studied terpenoids for their roles in membrane organization. Eukaryotic-specific membrane dynamics is based on the triad of distinct lipid components – fatty acyl phospholipids, sphingolipids and steroids (van Meer et al., 2008; Sezgin et al., 2017). The planar polycyclic structure of steroids, particularly cholesterol, has an ability to decrease the fluidity and the permeability of host membranes and laterally compress membranes, forming a liquid-ordered microdomains (Cheng and Smith, 2019). Further, it has increasingly been suggested that eukaryotic membranes are kept near the critical state (Shaw et al., 2020), which enables eukaryotic cells to respond to external stimuli in a highly dynamic way (Hammond et al., 2005; Levental et al., 2010). Individual steroids have varying effects, due to their subtle structural differences. For instance, C-24 alkylated steroids have elevated interactions with other lipid components, relative to cholesterol, enabling host membranes to have a wider resilience against temperature fluctuations and thus providing a selective advantage for host algae and plants (Beck et al., 2007). Hopanoids have a similar membrane packing ability (Sáenz et al., 2012; Sáenz et al., 2015), but to a lesser degree than cholesterol, because of the presence of additional methyl groups in the rings.

Bacterial membranes have also been suggested to have a eukaryotic-like heterogeneous membranes (functional membrane microdomains; FMMs) (Bramkamp and Lopez, 2015; Lopez and Koch, 2017), although the critical status of FMMs is currently unclear. It is suggested that FMMs are induced by terpenoids, including hopanoids, carotenoids and farnesol (Feng et al., 2014; Bramkamp and Lopez, 2015; García-Fernández et al., 2017; Cheng and Smith, 2019). Also, sphingolipids are distributed in some bacteria (Stankeviciute et al., 2022) and hence those bacteria might have eukaryotic-like membrane regulations based on the combination of sphingolipids and hopanoids. The presence of membrane heterogeneity has also been suggested in the archaeal membrane that is entirely composed of linear terpenoids, but its dynamics remains underexplored (Bagatolli et al., 2000; Salvador-Castell et al., 2020; Tourte et al., 2020).

Other terpenoids may also engage in membrane organization and protein-lipid interactions locally and/or temporarily. Quinones, polyprenoids and C_<25 small terpenoids can alter the membrane fluidity and/or permeability, although their physiological significance is not fully understood (Hartley and Imperiali, 2012; Camargos et al., 2014; Sévin and Sauer, 2014; Pham et al., 2015; Mishra et al., 2021). Also, C₁₅ or C₂₀ prenylation of eukaryotic Ras proteins not only anchors the proteins in membranes, but also mediates protein-protein and protein-lipid interactions (Sinensky, 2000).

Endomembrane systems are observed in all three domains of life and often have divergent lipid compositions from outer cellular membranes. In archaea, the hyperthermophilic anaerobe Ignicoccus hospitalis contains two cytoplasmic regions enclosed by outer and inner membranes. These two membranes have different terpenoid profiles and are utilized for different metabolic activities (Flechsler et al., 2021; van Wolferen et al., 2022). The presence of such a double membrane is suggested for several additional archaea, but the overall distribution remains punctate in the domain (Klingl, 2014). In contrast, bacteria are well known to have several unique endomembrane systems, including thylakoids and chromatophores in photosynthetic bacteria (Jürgens et al., 1992; Simonin et al., 1996; Orf and Blankenship, 2013; Mullineaux and Liu, 2020), anammoxosomes in anammox bacteria (van Niftrik et al., 2004) and magnetosomes in magnetotactic bacteria (Barber-Zucker and Zarivach, 2017). Also, eukaryotes are well known to universally possess a variety of membrane-bound organelles, including mitochondria, endoplasmic reticulum, Golgi apparatus and plastids.

However, terpenoids are not necessarily present in those endomembranes, even if terpenoids are present in cellular membranes. This heterogeneity presumably reflects the physiological requirements of individual endomembranes and/or their evolutionary origins. For instance, hopanoids are found in thylakoids in cyanobacteria, but are mostly absent in other bacterial endomembranes (Rattray et al., 2008; Schüler, 2008). Also, steroids are only minor components in mitochondria and endoplasmic reticulum in animals and fungi (van Meer et al., 2008). Mitochondria also lack sphingolipids and this might reflect the membrane composition of ancestral symbiotic alphaproteobacteria. Steroids are similarly absent in eukaryotic thylakoids, possibly reflecting their cyanobacterial origin (Hölzl and Dörmann, 2019).

Photosynthetic endomembranes (thylakoids and chromatophores) contain carotenoids either as part of the photosynthetic machinery, or as non-bound components. Carotenoids span both leaflets of a bilayer and have an ability to vertically compress membranes (Cheng and Smith, 2019). The fluidity of thylakoid membranes is in fact controlled by carotenoids, rather than hopanoids. The imbalance between carotenoids, other lipids and proteins, hampers the proper conformation of thylakoid structures (Bykowski et al., 2021). Carotenoids are also directly involved in photosystem assembly in cyanobacteria (Tóth et al., 2015). Hence, thylakoid membranes seem to be regulated differently from other (endo)membranes of non-photosynthetic organisms.

The presence of lipid membranes is not limited to cellular organisms. Many viruses have lipid membranes either as outer envelopes or as capsid-enclosed inner membranes. Membrane-containing viruses are unevenly distributed among eukaryotic, bacterial and archaeal viruses and viral membranes likely do not share a common ancestry (Poranen et al., 2015; Mäntynen et al., 2019). Viral membranes are thought to have evolved to overcome the surface barriers of host organisms (Poranen et al., 2015; Omasta and Tomaskova, 2022) and/or cope with environmental stress (Baquero et al., 2021). The presence of terpenoids in membranes is known for archaeal and eukaryotic viruses thus far. Archaeal viruses have archaeal-type membrane terpenoids, while eukaryotic viruses have steroids. Viral membranes are commonly acquired by budding of host cellular membranes and this causes a similar lipid composition of viral membranes to that of host membranes (Waheed and Freed, 2010). However, viral membrane formation within host cells is also known (Baquero et al., 2021). In this case, the lipid composition of the viral membrane is highly divergent from that of the host membrane. For instance, positive-strand RNA viruses manipulate the host lipid biosynthesis to optimize the lipid microenvironment for the viral replication (Nagy, 2022). It is currently unknown if viral membrane terpenoids are directly involved in viral entry and replication processes. In contrast, some structural homologs of steroids (e.g. oleanoic acid and betulinic acid) in host membranes are known to interact with viral membrane proteins as antiviral agents and prevent viral-host membrane fusion (Si et al., 2018).

Various forms of membrane-bound extracellular vesicles (EVs) are secreted by cellular organisms, utilizing host membrane lipids (Deatherage and Cookson, 2012; Gill et al., 2019). EVs include plasmid vesicles to virus-like particles and eukaryotic exosomes. EVs facilitate the transfer of genetic materials and other metabolites between cells. EVs share some similarity with enveloped viruses and in fact a shared origin for certain types of archaeal EVs and viruses has been suggested (Nolte-’t Hoen et al., 2016). Further, infectious virus-like plasmids that are surrounded by membrane vesicles were recently discovered from haloarchaea (Erdmann et al., 2017). Terpenoids are present in EVs from all three domains, reflecting host cellular membrane compositions (Berry et al., 1993; Skotland et al., 2020; Liu et al., 2021). However, the role of terpenoids in EVs remains unexplored.

Resting cells differentiate from vegetative cells to preserve genetic materials and other important metabolites under adverse environmental conditions. Resting cells are observed in multiple lineages of bacteria, including endospores in the phylum Bacillota, exospores in actinobacteria and akinetes in cyanobacteria. Endospores in Bacillota are coated with lipid membranes that contain a unique terpenoid called baciterpenol A that is a structural homolog of hopanoids and steroids (Sato, 2013; Willdigg and Helmann, 2021) ( Figure 2 ). In fact, baciterpenol A is produced by class II TC in aerobic members of the phylum. Baciterpenol A increases the rigidity of the host membrane and thus the resistance against oxidative stress (Bosak et al., 2008). In contrast, polycyclic terpenoids have not been found in exospores, even though they are widespread in hopanoid-producing actinobacteria (e.g. Streptomyces). In turn, hopanoids are widespread in akinetes and even a correlation between a certain type of hopanoids and akinete formation is suggested (Doughty et al., 2009). Exospore formation was recently reported even for haloarchaea (Tang et al., 2023). Haloarchaea are unique because they horizontally acquired a large number of bacterial genes (Gophna and Altman-Price, 2022), including those for carotenoids and even fatty acids that are otherwise bacterial signatures (Dibrova et al., 2014). Hence, the membrane dynamics of haloarchaea may be divergent from that of other archaea.