Lipid as a potential posttranslational modification of the H volcanii S-layer

The covalent attachment of lipids is a common modification experienced by both eukaryal and bacterial proteins and can involve myristoyl or palmitoyl acyl groups and isoprenyl polymers of various lengths, or aminoglycan-linked phospholipids. These can be added at the amino terminus, the carboxy terminus or at internal residues via ester, thioester, thioether or amide bonds, or through mediating functionalities such as phosphodiesters. Lipid modification of proteins is largely a posttranslational event and serves a variety of roles, such as enhancing membrane affinity of the modified protein, signal transduction, embryogenesis and pattern formation, protein trafficking through the secretory pathway and evasion of the immune response by infectious parasites (Eichler and Adams, 2005).

One of the defining traits of archaea distinguishing them from eukaryotes and bacteria is the chemical composition of their membrane phospholipids (Kates, 1993). Unlike eukaryal and bacterial phospholipids, which are built on a glycerol-3-phosphate backbone, archaeal phospholipids are based on the opposite steroisomer, glycerol-1-phosphate. The archaeal phospholipids contain polyisoprenyl side-chains rather than

the acyl groups employed by eukaryotes and bacteria. The archaeal phospholipids further rely on ether bonds to link the isoprenyl side chains to the glycerol-1-phosphate backbone (Eichler and Adams, 2005). Archaeal phospholipids are generally organised into the bilayer structure that is also present in eukaryal and bacterial cells, although tetra-ether lipid-based monolayers can also be found in thermophilic and hyperthermophilic archaea (Koga et al., 1993). Whereas phospholipids and other polar lipids (phosphoglycolipids, glycolipids and sulfolipids) account for the vast majority of archaeal membrane lipids, archaeal membranes also contain acetone-soluble nonpolar lipid species, primarily neutral squalenes and other isoprenoid-based polymers (Kates, 1978, Kates, 1992).

Lipid-modified proteins have been reported from a wide range of species of archaea. In the haloalkaliphile Natronobacterium pharaonis, a small blue copper protein was proposed to undergo amino-terminal lipid modification based on the presence of the so-called lipopbox sequence motif near the start of predicted amino acid sequence (Mattar et al., 1994). Although widespread in eukaryotes, GPI-anchored proteins have not been observed in bacteria, but have been seen in archaea, as Sulfolobus acidocaldarius was shown to possess three proteins that incorporate radiolabeled precursors of the GPI anchor moiety (Kobayashi et al., 1997).

Growth of Halobacterium cutirubrum and Halobacterium salinarum in the presence of radiolabeled [3H] mevalonate, a precursor for the isoprene building block used to synthesise archaeal lipids (Boucher et al., 2004, Smit and Mushegian, 2000), led to the appearance of several proteins radiolabeled through the covalent attachment of a lipid entity. Sagami and colleagues showed that a derivative of [3_{H] mevalonic acid} is incorporated into a number of specific proteins in H. halobium and H. cutirubrum and that the major radioactive material released by treatment with methyl iodide was an unknown compound, which was analysed by reverse and normal phase high performance liquid chromatographies followed by mass spectrometry (Sagami et al., 1994). The identified parent ion showed the compound to be a diphytanylglyceryl methyl thioether, suggesting that Halobacteria contained specific proteins with a novel type of modification of a cysteine residue of proteins with a diphytanylglyceryl group in thioether linkage (Sagami et al., 1995).

During further analysis of the isoprenoid-modified proteins in H. salinarum using other radiolabeled isoprenyl derivatives to obtain insight into the nature of the isoprenoid binding manner, radioactive PAS-stained peptides recovered from the gel were treated with methyl iodide. No radioactive materials were released from the peptides in the sulfonium salt cleavage reaction or under the acidic conditions at 37°C, however under the acidic conditions at 95°C, radioactive materials were released from the peptides with 90% recovery and revealed that the S-layer glycoprotein is modified by a second novel group, diphytanylglyceryl phosphate, which is attached through an as yet uncharacterised linkage. Amino acid sequencing placed the modification near an O-glycosylated Thr-rich stretch found in the C-terminal region of the protein, upstream of the single transmembrane domain (Kikuchi et al., 1999). Kessel and colleagues had speculated that the unusual structural threonine-rich cluster element in the S-layer of haloarchaea serves as a spacer between the membrane-binding domain and the extracellular domain of the cell-surface glycoprotein, thus creating an interspace that may be regarded as analogous to the periplasmic space of Gram-negative bacteria. The

hydroxyl group of the side chain of serine or threonine next to each membrane-binding domain was therefore suggested as a possible modification site (Kessel et al., 1988).

In H. volcanii, the S-layer glycoprotein was found to be first synthesised as an immature precursor, possessing a lower apparent molecular mass than the final version of the protein. Pulse-chase radio- labelling and cell-fractionation studies were employed by Eichler and colleagues to reveal that newly synthesised S-layer glycoprotein seemed to undergo a maturation step following the translocation of the protein across the plasma membrane. Post-translational conversion into the mature form of the protein is largely completed within the first 10 min following the appearance of the full-length polypeptide. The S-layer glycoprotein modification was suggested to require Mg2+-dependent membrane association of the protein, as no maturation was detected in the absence of Mg2+, which is known to be involved in attachment of the S- layer glycoprotein to the membrane (Sumper et al., 1990). This processing step was detected as an increase in the apparent molecular mass of the S-layer glycoprotein and was suggested to be unaffected by inhibition of protein synthesis and unrelated to glycosylation of the protein. Maturation resulted in an observed increase in hydrophobicity of the protein as revealed by enhanced detergent binding and modification of the H. volcanii S-layer glycoprotein was suggested to offer an explanation for these observations (Eichler, 2001). Eichler and Konrad suggest that the H. volcanii S-layer glycoprotein was also modified by a mevalonic acid- based lipid moiety and that such modification was responsible for the maturation of the H. volcanii S-layer glycoprotein. H. volcanii cells were therefore incubated with a [3_{H]-radiolabelled version of the isoprene} precursor, mevalonic acid. The [3_{H] mevalonic acid label appeared to remain associated with the S-layer} glycoprotein peptide fragment after de-lipidation, proteolytic digestion and SDS-PAGE. Konrad and Eichler proposed that neither phosphate ester nor pyrophosphate ester lipids were released under the mild acidic conditions, which led to the suggestion that the mevalonic acid-derived moiety could be covalently linked to to the protein, rather than simply being loosely associated with it (Konrad and Eichler, 2002).

It was hypothesised from the studies described above that the H.volcanii S-layer was likely to be substituted with a lipid moiety in the mucin-like domain. However, the S-layer has a C-terminal hydrophobic sequence which has all the properties required for it to be a robust membrane-spanning domain (Lechner and Sumper, 1987, Sumper et al., 1990, Wakai et al., 1997). It was therefore puzzling that an additional membrane anchor in the form of a covalent lipid would be required. To cast more light on this issue we felt that it was important to rigorously establish whether the S-layer is, indeed, covalently modified with lipid. The chemical nature of the lipid association with the S-layer glycoprotein is therefore also explored in this thesis (Section 3.6). 1.2.3.1 Characterisation of polar lipids of the halophilic archaeon H. volcanii

As part of a study to identify novel lipids with immune adjuvant activity, a structural comparison was made by Sprott and colleagues between the polar lipids from two halophiles, Haloferax volcanii and Planococcus H8 (Sprott et al., 2003). The total polar lipids of H. volcanii accounted for 83.5% by weight and fast atom bombardment mass spectrometry (FAB-MS) and thin layer chromatography were used to identify and quantify the major signals corresponding to the phospholipids. H. volcanii polar lipid extracts consisted of

44% archaetidylglycerol methylphosphate, 35% archaetidylglycerol, 4.7% of archaeal cardiolipin, 2.5% archaetidic acid, and 14% sulphated glycolipids 1 and 2 as are summarised in Table 1.1.

Lipid moiety Name Structure % Abundance m/z +Na

PA archaetidic acid 2.5 731.5 PG archaetidylglycerol 35.2 805.6 PGP-CH3 archaetidylglycerol methylphosphate 43.8 899.5 921.5 S-GL-1 sulphate-diglycosyl archaeol-1 5.2 1055.6 S-GL-2 sulphate-diglycosyl archaeol-2 8.7 1770.2 1792.2

Table 1.1 – Summary of H. volcanii lipid bilayer population.

The table describes the total lipid population of H. volcanii, as characterised described by Sprott and colleagues (Sprott et al., 2003).