OUP user menu

Acylated cholesteryl galactosides are ubiquitous glycolipid antigens among Borrelia burgdorferi sensu lato

Gunthard Stübs, Volker Fingerle, Ulrich Zähringer, Ralf R. Schumann, Jörg Rademann, Nicolas W.J. Schröder
DOI: http://dx.doi.org/10.1111/j.1574-695X.2011.00827.x 140-143 First published online: 1 October 2011


Lyme disease (LD) is the most common tick-borne disease in the Northern hemisphere. It is caused by Borrelia burgdorferi sensu lato, in particular, B. burgdorferi sensu stricto, Borrelia garinii, and Borrelia afzelii. However, other genospecies have been implicated as causative factors of LD as well. Borrelia burgdorferi exhibits numerous immunogenic lipoproteins, but due to strong heterogeneity, the use of these proteins for serodiagnosis and vaccination is hampered. We and others have identified acylated cholesteryl galactosides (ACGal) as a novel glycolipid present in B. burgdorferi sensu stricto, B. afzelii, and B. garinii. ACGal is a strong antigen and the majority of patients display anti-ACGal antibodies in the chronic stages of LD. However, it is unknown whether ACGal is present in other presumably pathogenic B. burgdorferi genospecies. Therefore, we performed an analysis of the total lipid extracts of a wide spectrum of genospecies of B. burgdorferi sensu lato using thin-layer chromatography as well as Western blot and dot-blot assays. We show that ACGal is present in substantial quantities in all B. burgdorferi genospecies tested. Therefore, this molecule might improve the serological detection of rarely pathogenic genospecies, and may be used as a protective vaccine regardless of the prevailing genospecies.

  • Lyme disease
  • Borrelia burgdorferi
  • glycolipid
  • antigen
  • vaccine
  • serology

Lyme disease (LD) is a multisystemic, often chronic infectious disease prevalent in Europe, North America, and Asia. In endemic areas, LD reaches an incidence of up to 160 cases per 100 000 (Berglund et al., 1995; Strle, 1999). The clinical manifestations are divided into early and late manifestations: early localized disease is characterized by erythema migrans. Disseminated early disease primarily encompasses neuroborreliosis, lymphocytoma, or myocarditis. Late manifestations predominantly comprise Lyme arthritis, acrodermatitis chronica atrophicans, and rarely late neuroborreliosis (Huppertz et al., 1999). LD diagnosis is based on the clinical signs and serodiagnosis using ELISA and confirmative immunoblots (Wilske et al., 2007).

A number of Borrelia burgdorferi sensu lato genospecies are etiologic agents of LD causing early localized as well as early and late stages of disseminated disease: B. burgdorferi sensu stricto, Borrelia afzelii, and Borrelia garinii. Furthermore, the OspA serotype 4 of B. garinii, which has been associated with LD affecting the skin and CNS (Wilske et al., 1993), was recently delineated as a novel genospecies, Borrelia bavariensis (Margos et al., 2009). To the contrary, Borrelia spielmanii causes the localized stage while disseminated disease caused by this agent has not been reported as yet (Fingerle et al., 2008). Three further genospecies are rarely found in skin biopsies and only in single cases in CSF or cardiac tissue. The pathogenic potential of these genospecies, namely Borrelia bissettii (Fingerle et al., 2008; Rudenko et al., 2008), Borrelia valaisiana (Diza et al., 2004), and Borrelia lusitaniae (Collares-Pereira et al., 2004), remains unclear. Furthermore, eight genospecies of the group have been found only in ticks or in animals, for example Borrelia japonica (Kawabata et al., 1993), and are considered nonpathogenic for humans.

We and others have identified 6-O-acylated cholesteryl β-d-galactopyranosides (ACGal) as the major glycolipids in B. burgdorferi sensu stricto, B. afzelii, and B. garinii (Ben-Menachem et al., 2003; Schröder et al., 2003; Stübs et al., 2009). On the other hand, Borrelia hermsii — the causative agent of relapsing fever — contains 6-O-acylated cholesteryl β-d-glucopyranosides (ACGlc) (Stübs et al., 2009). It was demonstrated that ACGal is a strong and highly specific antigen in the late stage of LD, with anti-ACGal immunoglobulin G (IgG) antibodies detected in 83–96% of the patients (Jones et al., 2009; Stübs et al., 2009), and the antigenic nature of ACGal has been confirmed by chemical synthesis (Stübs et al., 2010). These data imply that ACGal could improve serodiagnostics, and may act as a basis for vaccine development. However, to date, it is unclear whether detection of or vaccination with ACGal would encompass LD-causing genospecies other than B. burgdorferi sensu stricto, B. afzelii, and B. garinii. On the other hand, the function of ACGal in B. burgdorferi is not elucidated, and the report that acylated cholesteryl α-d-glucosides in Helicobacter pylori are associated with immune evasion (Wunder et al., 2006) raises the question of whether ACGal are involved in the pathogenesis of LD. Therefore, in this study, we wanted to determine whether ACGal is a feature of other genospecies of B. burgdorferi sensu lato, including those associated with all stages of LD as well as B. spielmanii as an agent of localized LD.

The following Borrelia strains were grown under microaerophilic conditions in 9 mL of BSK-H medium at 33 °C as described previously (Preac-Mursic et al., 1986): B. burgdorferi s.s. strain B31, B. afzelii PKo, B. bavariensis PBi, B. garinii A and TN, B. spielmanii PSig II, B. bissettii DN 127, B. lusitaniae Poti B2 and Poti B3, B. valaisiana VS 116 and UK, B. japonica HO 14, B. hermsii HS 1. The methods and materials for harvesting and extraction of bacteria have been described in detail earlier. In brief, the cells were harvested, lyophilized, and disintegrated using an ultrasonic rod and the lipids were extracted by a Folch extraction (Folch et al., 1957). The total lipids were dissolved and spotted in about equal amounts on a thin-layer chromatogram (TLC). Synthetic ACGal was applied as a reference (Stübs et al., 2010). The chromatography was performed in chloroform/methanol 85 : 15 v/v. The lipids were visualized on the TLC by molybdenum stain. The dried TLC was immersed in buffer and blotted onto a polyvinylidene difluoride (PVDF) membrane using a hot iron. The membrane was blocked with a skim milk/phosphate-buffered saline solution and incubated for 13 h at 4 °C with a 1 : 750 diluted serum of LD patients in the late stage. The membrane was incubated for 1.5 h at room temperature with a 1 : 50 000 dilution of a secondary, horseradish peroxidase-conjugated anti-human IgG antibody. The serum antibody binding was detected using enzymatic chemoluminescence to expose and subsequently develop X-ray films. Dot blots and Borrelia lysates were generated as described previously (Stübs et al., 2010): ACGal, Borrelia lysate and total lipids were spotted on PVDF membranes and incubated with pooled sera (n=4) from patients diagnosed with LD, syphilis as well as leptospirosis at 4 °C for 15 h. Detection with secondary antibodies was performed via chemoluminescence.

The stained TLC (Fig. 1a) revealed that all analyzed Borrelia genospecies exhibited a similar lipid pattern. The lipid structures were assigned according to published chemical analyses (Stübs et al., 2009). The most intensely stained glycolipids in the B. burgdorferi s.l. group were ACGal as indicated by the synthetic reference (lanes 1–2) and its nonacylated counterpart cholesteryl β-d-galactopyranoside (CGal). Cholesteryl β-d-glucoside (CGlc) was present with a slightly higher retention factor (Rf) with regard to the latter. In B. burgdorferi s.l. CGlc comprises about one fifth of the amount of CGal whereas in B. hermsii (lacking CGal) it is the only nonacylated cholesteryl glycoside. Mono-α-d-galactosyl diacylglycerol stained weakly, but it was present in the total lipids of all strains including B. hermsii in comparable amounts. The immunostained membrane of the blotted lipids (Fig. 1b) showed only a clear signal in lanes 1–2 with synthetic ACGal and lanes 3–15 covering the 13 B. burgdorferi sensu lato genospecies. No matching immunostaining was observed for B. hermsii, confirming former results that its ACGlc is not cross-reactive with ACGal (Stübs et al., 2009). All other lipids were nonreactive with serum IgG antibodies under these conditions. To assess the specificity of ACGal, it was analyzed with sera derived from patients with serologically confirmed infection with Treponema pallidum or Leptospira spp. The dot blots (Fig. 1c) demonstrate that LD sera recognize synthetic ACGal, the total lipids of B. burgdorferi sensu lato as well as the borrelial lysate. In contrast, antibodies against ACGal could not be detected in pooled sera from patients with T. pallidum or Leptospira infection.

Figure 1

Thin-layer chromatography and dot-blot analysis of borrelial total lipids. (a) Synthetic ACGal, the total lipids from the specified Borrelia burgdorferi sensu lato strains, and from Borrelia hermsii (containing ACGlc instead of ACGal) were separated on a TLC and visualized as described in the text. np, Not pathogenic, indicates total lipids from Borrelia japonica HO, a strain considered to be nonpathogenic to humans. (b) The lipids were transferred from the TLC to a PVDF membrane, incubated with patient sera and secondary anti-IgG antibodies, followed by development of an X-ray film. 1, 2 µg synthetic ACGal; 2, 5 µg synthetic ACGal; 3, B. burgdorferi s.s. B31; 4, Borrelia garinii A; 5, Borrelia afzelii PKo; 6, Borrelia spielmanii PSig II; 7, Borrelia lusitaniae Poti B2; 8, B. lusitaniae Poti B3; 9, Borrelia valaisiana VS 116; 10, B. valaisiana UK; 11, B. burgdorferi s.s. B31; 12, B. garinii TN; 13, Borrelia bavariensis PBi; 14, B. japonica HO 14; 15, Borrelia bissettii DN 127; 16, B. hermsii HS 1. (c) PVDF membranes were spotted with synthetic ACGal (1 µg), Borrelia lysate (1.5 µg) as well as lipid extracts from B. burgdorferi s.s. B31 (Bbu), B. afzelii PKo (Baf), B. garinii A (Bga), B. hermsii HS 1 (Bhe), B. bavariensis PBi (Bba), B. spielmanii Psig II (Bsp), B. valaisiana UK (Bva), and B. bissettii DN 127 (Bbi). Membranes were incubated with pooled sera (n=4 each) from patients diagnosed for leptospirosis (LS, upper panel, diluted 1 : 100), syphilis (S, middle panel, diluted 1 : 100) as well as Lyme disease (LD, lower panel, diluted 1 : 1600), followed by detection with anti-IgG antibodies.

Our data show that ACGal is present in significant quantities in all B. burgdorferi sensu lato genospecies tested, including the common genospecies causing all stages of disease, B. spielmanii causing localized infection only, as well as B. japonica as a nonpathogenic agent. Therefore, using ACGal in serodiagnosis, while potentially enhancing sensitivity, would not bear the risk of missing certain genospecies. It furthermore offers an excellent specificity because it is not recognized by sera from patients suffering from other spirochaetoses. Also, these data support the notion that ACGal may be a promising vaccine target because antibodies recognizing this molecule detect all known B. burgdorferi sensu lato genospecies. In addition, our data do not support a pivotal role of ACGal in LD pathogenesis, but indicate that these glycolipids are important for maintaining the integrity and function of the cell membrane in Borrelia.


We would like to thank Cecilia Hizo-Teufel (Bavarian Health and Food Safety Authority) for cultivating the Borrelia strains as well as Barbara Graf and Janine Zweigner (Institute of Microbiology and Hygiene, Charité — Universitätsmedizin Berlin) for providing patient sera.


  • Editor: Artur Ulmer


View Abstract