Bloodstream- Versus Tick-Associated Variants of a Relapsing Fever Bacterium

See allHide authors and affiliations

Science  19 Jun 1998:
Vol. 280, Issue 5371, pp. 1938-1940
DOI: 10.1126/science.280.5371.1938


The relapsing fever spirochete, Borrelia hermsii, alternates infections between a mammal and a tick vector. Whether the spirochete changes phenotypically in the different hosts was examined by allowing the tick vector Ornithodoros hermsi to feed on mice infected with serotype 7 or serotype 8 of B. hermsii.Upon infection of ticks, the spirochetal serotype-specific variable major proteins (Vmps) 7 and 8 became undetectable and were replaced by Vmp33. This switch from a bloodstream- to tick-associated phenotype could be induced in culture by a decrease in temperature. After tick-bite transmission back to mice, the process was reversed and the spirochetes resumed expression of the same Vmp present in the previous infectious blood meal.

Relapsing fever was recognized as a human disease back in the time of Hippocrates (1). Endemic relapsing fever is caused by numerous species of the spirochete Borrelia and occurs throughout the world in many discrete enzootic foci where the spirochetes are maintained primarily in rodents and ticks of the genus Ornithodoros(2). In western North America, B. hermsii, which is transmitted by the tick O. hermsi, causes human disease that is likely more prevalent than the number of reported cases (3). Human infections result from the bites of these fast-feeding ticks, which usually occur at night and unbeknownst to their victims.

The periodic cycling of acute and afebrile episodes that led to the naming of relapsing fever is associated with dramatic changes in the abundance of spirochetes circulating in the patient's blood (spirochetemia) (1). Accompanying each of these cyclic population changes in the number of spirochetes is a change in serotype and major immunogenic lipoprotein associated with the outer surface of the spirochete. These variable major proteins (Vmps) are composed of two multigene families that have recently been renamed as variable large or small proteins (Vlps and Vsps) (4); however, here we refer to them collectively as Vmps.

In the HS1 strain of B. hermsii, the single vmpgene being expressed at any one time is located in an expression locus near the telomere of a 28-kb linear plasmid (5). A copy of the expressed gene and all other unexpressed vmp genes are present in silent loci on the same or different linear plasmids. Antigenic switching to a new serotype occurs when one of the silentvmp genes is duplicated and replaces the existingvmp gene in the expression locus, or when intraplasmidic recombination occurs, sometimes augmented by the introduction of point mutations in rearranged vmp genes (5, 6). Each population of spirochetes associated with a single acute episode is composed almost entirely of one serotype producing the same Vmp (7). Different serotypes predominate in subsequent populations, and this multiphasic antigenic variation presumably allows evasion from the mammalian immune response (8). Yet nothing is known about the influence of the spirochete's serotype on infection in the arthropod host or whether spirochetes change phenotypically as they alternate between ticks and mammals.

To examine if ticks transmit the same serotype of the relapsing fever spirochete as was acquired during a previous blood meal, we first infected two cohorts of nymphal O. hermsi ticks with either serotype 7 or serotype 8 of B. hermsii (Fig. 1, A and B) (9). The ticks were allowed to molt and were then fed singly on individual mice to determine the serotype transmitted to the mice (10). Eighteen of 95 ticks (19%) transmitted spirochetes, and in every infection, the first detectable spirochetemia in the mice after tick-bite consisted of the same serotype ingested previously by the ticks, whether this was serotype 7 or serotype 8. In addition, polymerase chain reaction analysis of the spirochete's vmptelomeric expression locus in 22 infected ticks (11 ticks with serotype 7 and 11 ticks with serotype 8) revealed no apparent DNA rearrangement (11). These observations indicated that the vmpgene present at the telomeric expression site did not change during passage through ticks.

Figure 1

Borrelia hermsiiswitches between bloodstream-specific and tick-specific outer surface proteins. (A) Serotype 7 in mouse blood visualized with antibody to Vmp7 (anti-Vmp7). (B) Serotype 8 in mouse blood visualized with anti-Vmp8. (C) Spirochetes in a salivary gland of a tick that ingested serotype 7 is not detectable with anti-Vmp7. (D) Spirochetes in a salivary gland of a tick that ingested serotype 7 visualized with anti-Vmp33. (E) Spirochetes in a salivary gland of a tick that ingested serotype 8 visualized with anti-Vmp33. (F) Serotype 7 in mouse blood visualized with anti-Vmp7 after transmission by tick bite. Spirochetes were not detectable in blood with anti-Vmp33. (Bar = 25 μM).

To determine if spirochetes expressed the telomericvmp gene in the tick, we infected additional cohorts of ticks with either serotype 7 or serotype 8 and allowed the ticks to molt to the next developmental stage (12). From 33 to 144 days after infection, we examined 41 ticks, including 23 infected with serotype 7 and 18 infected with serotype 8. Salivary glands from all ticks and a lesser number of midgut (n = 33) and synganglion (n = 22) preparations were examined by indirect immunofluorescence assay (IFA) with one of four antibodies (13). Although all tissues were infected, we were unable to detect spirochetes in the salivary glands using antibodies specific for Vmp7 (10 examined) (Fig. 1C) or Vmp8 (5 examined). Yet, when we examined the other salivary gland from these same ticks and others from the same infected cohort by IFA with a monoclonal antibody to Vmp33 (marker of serotype C), a surface protein seen previously only in the culture-adapted strain HS1 (14, 15), we detected many fluorescent spirochetes (Fig. 1, D and E). This was true for ticks infected with either serotype 7 (n = 17) or 8 (n = 12). When peripheral blood was examined during the first spirochetemia in mice only 4 days after infection by tick bite, bacteria were visualized by IFA with anti-Vmp7 (Fig. 1F) and anti-Vmp8 but not with anti-Vmp33. Therefore, as these spirochetes cycle between ticks and mammals, their outer surface alternates between bloodstream- and tick-associated proteins.

One stimulus for such a switch might be the change in temperature as the spirochetes are transferred from a warm-blooded mammal to a much cooler tick. We tested this hypothesis with blood from a mouse infected with serotype 8 that we cultured at 37° or 23°C (16). SDS–polyacrylamide gel electrophoresis (SDS-PAGE) analysis of whole-cell lysates (17, 18) indicated that growth at 23°C induced expression of vmp33, whereas growth at 37°C maintained the expression of vmp8 (Fig. 2A). Immunoblot analysis with specific antibodies confirmed that the synthesis of Vmp8 and Vmp33 was influenced by temperature, whereas synthesis of flagellin, a structural protein of the spirochete's periplasmic flagella, was not (Fig. 2, B to D) (18, 19).

Figure 2

Shift of B. hermsiito lower temperature induces switch from a bloodstream Vmp to the tick-associated Vmp33. (A) Serotype 8–infected mouse blood was inoculated into medium and incubated at 37° or 23°C. Proteins in whole-cell lysates were separated by SDS-PAGE and stained with Coomassie brilliant blue. The primary isolate (P-0) and the subsequent two passages (P-1 and P-2) did not switch after growth at 37°C (lanes 2 to 4), whereas growth at 23°C induced the switch in the primary isolate (P-0) (lane 5). Three arrows indicate flagellin (top) and the induced switch from Vmp8 (middle) to Vmp33 (bottom). Also shown is high-passage (HP) culture of the same strain (DAH) that has stably switched to Vmp33 (lane 6). Molecular mass standards (lane 1) are shown in kilodaltons. In (B) to (D), whole-cell lysates of serotype 8 spirochetes grown at 37° or 23°C were fractionated by SDS-PAGE and stained with Coomassie brilliant blue. Immunoblots containing replicate lysates in lanes 4 and 5 were first incubated with (B) anti-Vmp8, (C) anti-Vmp33, or (D) anti-flagellin, and then with 125I-labeled protein A.

Bloodstream vmp genes that allow for antigenic variation ofB. hermsii HS1 during infection in mammals are expressed at a single telomeric locus on a 28-kb linear plasmid (5). However, vmp33 is expressed by its own promoter (15), mapped recently to a different linear plasmid of ∼53 kb (20). Vmp33 was identified previously in one strain ofB. hermsii only after prolonged cultivation in vitro (14), although others have speculated as to a possible role of this protein during tick infection (21). We have identified vmp33 in 23 additional isolates of B. hermsii from western North America (22). Also, genes homologous to vmp33 and proteins antigenically related to Vmp33 have been identified in many other species ofBorrelia, including outer surface protein (Osp) C of the Lyme disease spirochete Borrelia burgdorferi (15,23). Therefore, this family of proteins appears to be conserved among all members of the genus Borrelia.

The temporal expression of vmp33 and ospCby B. hermsii and B. burgdorferi, respectively, implicates a common biological function for these proteins associated with tick transmission or early colonization in mammals. Borrelia hermsii produce Vmp33 when the temperature cools after their acquisition by ticks and continue to produce this protein during persistent infection of the tick's salivary glands until they are rapidly transmitted by Ornithodoros ticks that feed in 15 to 90 min. In contrast, most Lyme disease spirochetes in unfedIxodes ticks are restricted to the midgut (24), produce OspC only after the temperature increases and ticks have fed for two or more days (25), and are then transmitted after their dissemination from the midgut to salivary glands (26). Therefore, Lyme disease spirochetes produce OspC only after tick feeding has begun, but while there is still ample time for transmission to occur by Ixodes ticks that require several days to feed. Our results demonstrating alternating phenotypes of B. hermsii cycling between the mammalian bloodstream and arthropod vector also have parallels to the salivary trypanosomes (27) and other parasitic protozoa that require blood-feeding arthropods for their biological transmission.

  • * To whom correspondence should be addressed. E-mail: tom_schwan{at}


View Abstract

Navigate This Article