Mycoplasma Variable Surface Antigen (VSA)

How Some Mycoplasmas Evade Host Immune Responses

Variable surface antigen proteins are key for how these microorganisms
evade host immune responses

Warren L. Simmons and Kevin Dybvig

Warren L. Simmons is an assistant professor and Kevin Dybvig is a
professor at the University of Alabama at Birmingham, Department of
Genetics.

Summary

* The C-terminal variable region of the mycoplasma variable surface
antigen (Vsa) proteins contains up to 60 tandem repeat units, and this
variable length proves a major factor affecting adherence properties and
shielding these bacteria from host immune responses.

* The cell-surface Vsa proteins constitute as much as 10% of total
mycoplasma protein, helping to shield cells against environmental
factors.

* When grown on solid surfaces and producing short Vsa proteins, M.
pulmonis forms biofilms that confer partial resistance to immune system
components.

* Mycoplasmas form biofilm-like aggregates on host tissues, where they
might also form tower structures to protect against innate immune system
responses.

Even though free-living, mycoplasmas depend on their hosts for key
nutrients, including purines, pyrimidines, several amino acids, and
sterols, which are incorporated into the membranes of these wall-less
bacteria. These requirements reflect their relative genomic simplicity.
The 580-kb genome of Mycoplasma genitalium, for example, is comparable
in size to that of a large virus.

Nonetheless, this highly successful group containing both commensals and
pathogens withstands robust host immune responses. Mycoplasmal diseases
are generally inflammatory, including mycoplasma-induced asthma
episodes, pelvic inflammatory disease, and bronchiectasis. The Centers
for Disease Control and Prevention estimate that walking pneumonia,
caused by Mycoplasma pneumoniae, accounts each year for 2 million cases
and 100,000 hospitalizations in the United States. Mycoplasmas often can
be detected in individuals long after their symptoms disappear.

Although many factors contribute to disease chronicity, the mycoplasmas
appear to evade immune surveillance by varying their antigenic patterns
and by shielding themselves from components of the immune system. In
focusing on the murine pathogen M. pulmonis as a model, we learned that
one key feature for how this group of microorganisms avoids host immune
responses is the abundant variable surface antigen (Vsa) proteins that
undergo high-frequency phase and size variation. The length of these
proteins affects the adherence properties of the mycoplasma as well as
its degree of shielding from the immune system.

Mycoplasmas Avoid Adaptive Immune Responses through Vsa Phase Variation

HYPERLINK "http://www.asm.org/microbe/index.asp?bid=53844" FIGURE 1

Vsa lipoproteins have a 242-amino-acid N-terminal region and a
C-terminal variable region containing as many as 60 tandem repeat units
ranging in size from 10–19 amino acids (Fig. 1A and B). Each cell
transcribes only one vsa gene. Silent vsa genes are missing sequences
that code for the conserved N-terminal region. Each of these silent
genes contains little more than a vsa recombination site (vrs) followed
by a tandem repeat sequence. Through gene rearrangement between vrs
sequences, each of the silent vsa genes is capable of recombining into
the vsa expression site. All of the identified gene rearrangements are
DNA inversions that serve to replace the tandem repeat region of the
formerly expressed gene with the tandem repeat region of a newly
expressed gene.

The repertoire of Vsa proteins (Vsa types) available to M. pulmonis
strain CT is VsaA, C, E, F, G, H, and I. Although each cell has a single
vsa expression site and produces only one Vsa type at a time,
subpopulations of cells in culture or during animal infections produce
each of the alternative Vsa types. Phase switching occurs at a frequency
of about 10-3 per CFU per generation and likely contributes to disease
chronicity.

For example, when mice with normal immunity are infected with M.
pulmonis that predominantly consists of a single Vsa type (few phase
variants present in the inoculum), the Vsa population remains unchanged
early during infection (day 3) but many phase variants arise within 2 or
3 weeks. If the mice lack B and T cells because of a rag mutation, the
Vsa population remains unchanged throughout the experiment. These data
suggest that the immune system exerts selection pressure for Vsa phase
variants in ordinary mice. There is no evidence of tissue tropism as the
Vsa types are the same in various tissues and blood. Vsa phase variants
arise only after the onset of a specific antibody response, apparently
helping these bacteria to avoid that adaptive immune response.

The Mycoplasmal Shield

The cell surface Vsa proteins constitute as much as 10% of the total
cellular protein, helping to shield cells against environmental factors.
Variations in size modulate mycoplasmal cell surface properties.

Vsa size varies when slipped-strand mispairing occurs during DNA
replication (Fig. 1C), either increasing or decreasing the number of
tandem repeats. A short VsaA protein containing three tandem repeat
units is referred to as VsaA-R3 (an R3 protein). The long form of a Vsa
protein is often referred to as an R40—typically with 40 tandem
repeats but sometimes with as many as 60. Like phase variation, size
variation occurs stochastically at a frequency of about 10-3 per CFU per
generation.

When observed by transmission electron microscopy, mycoplasmal cells
appear to be surrounded by a nap (Fig. 2, Panel A). The length of the
Vsa proteins correlates with the thickness of the surface of the
mycoplasma. For example, for mycoplasmas that produce a R40 Vsa protein,
that nap is about 26 nm thick, whereas it is about 16 nm thick on
mycoplasmas producing a R3 protein.

We believe that the shield is contained within the nap. When mycoplasmas
produce a long Vsa protein, the thick nap partly blocks access to the
outer membrane of mycoplasmal cells. When mycoplasmas produce a short
Vsa protein, however, there is ready access to the cell surface. Vsa
length and shielding affect phenotype. For instance, M. pulmonis strains
that produce a long Vsa have a longer doubling time than do strains the
produce a short Vsa. Cells producing a long Vsa protein may lack access
to key nutrients because the shield interferes with cell surface
degradative enzymes, preventing them from breaking down host
macromolecules.

Vsa length correlates with the ability of mycoplasmal colonies to adsorb
sheep erythrocytes (hemadsorb) and to grow on polystyrene or glass.
Mycoplasmas that produce a short VsaA, hemadsorb well and grow readily
on polystyrene or glass. Mycoplasmas that produce a long Vsa protein do
nothemadsorb or attach to plastic or glass. These findings suggest that
molecules involved in adherence extend beyond the shield and promote
adherence only if Vsa is short.

The shield modulates the susceptibility of individual cells to killing
by complement (Fig. 2, Panel A). When Vsa is short, the mycoplasmas are
highly sensitive to the complement membrane attack complex (MAC), while
cells producing long Vsa are protected, even though complement component
C3 deposits on the mycoplasma surface and MAC forms in both cases. Thus,
M. pulmonis activates the pathway yet resists lysis when the shield is
thick. Similarly, the length of the tandem repeat region of the Vlp
lipoprotein is associated with shielding M. hyorhinis from
growth-inhibiting antibodies, according to Kim Wise and his group at the
University of Missouri.

However, the shield does not protect individual mycoplasmas against the
pore-forming antimicrobial peptide gramicidin. Independent of the length
or type of Vsa protein, individual mycoplasmal cells are killed
efficiently by gramicidin at 100-fold-lower levels than are needed to
kill Mycoplasma mycoides. The shield apparently allows small molecules
to pass through it.

Hemadsorption, adherence to polystyrene, and susceptibility to lysis by
MAC may seem unrelated, but these phenotypes all reflect interactions
that are taking place at the surface of the mycoplasma. The finding that
Vsa shielding affects these properties independent of the Vsa type
indicates that the modulation of the surface interactions is a
generalized, nonspecific process.

In our view, slipped-strand mispairing within the vsa repeat region
generates short-Vsa subpopulations with minimal shielding and long-Vsa
subpopulations with maximal shielding. Minimally shielded cells are
highly adherent and grow well. Maximally shielded cells resist
complement and perhaps other components of innate immune defenses.
However, even a thick shield does not fully protect mycoplasmas against
adaptive immunity, necessitating phase variation of the shield.

Mycoplasmas Form Encapsulating Biofilms

When grown on solid surfaces, mycoplasmas form biofilms that confer
partial resistance to innate host defenses. In the case of M. pulmonis,
biofilms form only if the Vsa protein is short and the shield thin.
Typical of many biofilms, the M. pulmonis biofilm contains an
extracellular matrix containing protein, lipid, and polysaccharide. M.
pulmonis biofilms have honeycombed regions containing numerous cavities
that allow cells free access to the surrounding medium. In this region,
the mean distance between a cell and its three closest neighbors is
about 9 мm.

The biofilms also contain tower structures, and within the towers are
channels. The cells in the towers are so densely packed that the
distances between the mycoplasmas cannot be measured. Lipid and
polysaccharide appear to be most abundant in the tower regions of the
biofilm, filling the channels within the towers.

The polysaccharide composition of the matrix is complex, reacting with
lectins WGA and GS-II that bind to N-acetylglucosamine, and GS-I, which
binds to б-linked galactose. Lipophilic fluorescent probes identify
membrane within the matrix between cells. VsaA epitopes within the
matrix of the honeycombed region suggests that Vsa lipoprotein is
anchored to the membrane material in the matrix. Antibodies to the VsaA
protein detect epitopes on the external surfaces of the tower structures
but not within the internal regions, suggesting that the density of
material is too high for the antibody to penetrate.

Long Vsa-producing mycoplasmas form free- floating biofilms that are
held together by a flexible extracellular matrix containing lipid and
polysaccharide in addition to Vsa. The volume of this extracellular
matrix is much greater than that of attached biofilms, and the mean
distance between cells in the microcolonies is 21 мm, more than twice
the distance between cells of ordinary biofilms.

A plausible explanation for the differences in mean distances between
cells within a microcolony and a biofilm is that the cells of the
biofilm are restricted to close positions by virtue of their attachment
to glass or plastic surfaces. However, when biofilm-forming cells are
grown in polypropylene tubes, a surface to which they do not attach, the
cells are too dense to measure distances between them. These results
suggest that long Vsa proteins interact with other components of the
extracellular matrix in such a way that greater volumes of matrix
material accumulate between cells.

The accumulation of the Vsa protein and lipid in the extracellular
matrix may result from blebbing of mycoplasmal cell membrane into this
space. When observed by phase and transmission electron microscopy
(TEM), mycoplasmal cells form filaments and other cellular protrusions
(Fig. 2B). Membrane constrictions could lead to shedding of membranous
blebs from the mycoplasma. At some point along this continuum, the size
of the bleb becomes too small to contain a chromosome. Using the size of
the Eschericia coli chromosome (4.6 Mbp) and the volume of the E. coli
cell as a guide, we estimate that the minimal mycoplasmal cell has a
diameter of about 0.2 мm, which is close to the smallest pore size
through which mycoplasmal CFU can be filtered. Below this value, a
self-replicating mycoplasmal cell is not likely to exist.

An abundance of small vesicles, some with diameters of 50 nm or less,
are observed in cultures of M. pulmonis. Such vesicles, which may
account for some of the Vsa and lipid in the extracellular environment,
might function as decoys that tie up components of the immune system.
Vesicles may fuse with mycoplasmal cells, transferring nutrients and
other material between cells. If such vesicles sometimes contain DNA as
found for outer membrane vesicles of some species of gram-negative
bacteria, they may contribute to gene transfer between mycoplasmal
cells.

Biofilms Protect Mycoplasmas against Various Agents, Treatments

Bacteria within biofilms resist antimicrobial agents for several
reasons. For one, lower growth rates in biofilms reduce the
effectiveness of antibiotics that require high growth rates to kill. For
another, biofilms reduce the diffusion of some antibiotics, such as
aminoglycosides. Further, biofilms protect against antimicrobial
peptides. Less is known about how biofilms protect mycoplasmal cells
against antimicrobial agents and physical treatments. For example,
although M. bovis biofilms survive desiccation or 40- minute exposure to
50°C, both M. bovis and M. putrefaciens biofilms remain susceptible to
fluoroquinolones and oxytetracycline, according to Laura McAuliffe and
colleagues at the Veterinary Laboratories Agency in Surrey, United
Kingdom.

However, we find that biofilms partly protect M. pulmonis from molecules
that form pores in membranes. Whereas complement and gramicidin kill
individual, dispersed cells of M. pulmonis, those cells are protected
when they are encased in biofilms. For instance, more CFU are recovered
from biofilms that are left intact and incubated with complement or
gramicidin than are recovered from biofilms that are dispersed prior to
incubation. Perhaps cells in towers within biofilms acquire resistance
to complement by producing long Vsa proteins.

In any case, when complement-exposed mycoplasmas from a biofilm are
dispersed and again exposed to complement, 96% are killed. Thus
biofilmconferred resistance to complement is not an acquired trait.
Moreover, only those mycoplasmas in the towers of the biofilm are
protected, based on analyzing complement-treated biofilms with Hoechst
dye, which stains all cells, and propidium iodide, which stains only
dead cells.

These studies prove that the structure of the biofilm is protective. We
believe that the high density of cells and matrix material in biofilm
towers partly blocks access of complement and gramicidin to cell
membranes. An alternative possibility is that those agents reach the
membrane but cannot assemble poreforming complexes.

How Biofilms Might Help Mycoplasmas Evade Host Immune Responses

How well does the shield hypothesis apply to survival in hosts? We know
that mycoplasmas form biofilm-like aggregates and heavy layers on
mucosal tissue of the trachea and the genitourinary tract. Further, we
identified structures that resemble mycoplasmal tower structures on the
luminal side of tracheal explants (Fig. 3). Experiments involving
confocal scanning laser microscopy will soon help to determine whether
these structures constitute genuine biofilms.

Aggregates of biofilm-forming mycoplasmas attach securely to host
epithelium and grow to establish tower structures, thus protecting those
cells from the innate immune system, according to our model for how the
shield and the biofilm affect mycoplasmal pathogenesis (Fig. 4). Random
Vsa phase and size variation produce subpopulations of nonadherent
mycoplasmal cells.

Although many nonadherent cells remain embedded in towers, some detach
to establish other foci of infection. Even though vulnerable when
disseminating, the long Vsa proteins could partly protect individual
cells and clusters within microcolonies.

A striking feature of Vsa proteins is their extensive proline-rich,
tandem repeat region, much like surface antigens in other pathogens,
including other mycoplasma species and in gram-positive bacteria
phylogenetically related to mycoplasmas. Bacterial proteins with
extensive tandem repeats that could extend beyond the wall and interact
with host molecules and possibly contribute to shielding include the
alpha C proteins of group B streptococci, several proteins of
Mycobacterium tuberculosis, and various proteins of African trypanosomes
and the malaria pathogen Plasmodium falciparum. The tandem repeats of
these proteins generally are also proline-rich. Many of these antigens
are thought to be protective against host defenses. For group B
streptococci, the tandem domains of the alpha C proteins likely
contribute to evasion of host immunity.