Caries Immunology

Chapter 13
Background
Dental caries represents a health expenditure of several billion dollars per year
in the United States, even though water fluoridation has reduced caries by one-
half. Half the children (5-17 yo) in the United States experience caries, and about
half are caries free -- testimony to the tremendous impact that preventive dental
care has provided. Thus, although great progress has been made in preventive
dentistry, dental caries is still a major health problem, afflicting about 50% of our
children.
The National anticaries strategy has been four-pronged with goals to (1) to
combat the microbial agent; (2) to increase tooth resistance; (3) to modify diet;
and (4) to deliver anticaries measures to the public.
The first goal, combatting the microbial agent, is based on evidence that
specific microorganisms are an important part of the pathology of dental
caries.Therefore, we can target some, but not all, bacteria for immune regulation.
In 1924, Clarke (1924. Brit. J. Exp. Pathol. 5: 141-147) isolated an organism
that he felt to be from the earliest carious lesions in humans, Streptococcus
mutans.Although bacteria were widely accepted as the cause of dental caries, it
wasn't until 1945-46 that McClure and Hewitt (1946. J. Dent. Res. 25: 441-443)
showed that bacteria were indeed potential etiologic agents of dental caries.
Using penicillin, rats, and Lactobacillus acidophilus,these workers demonstrated
a positive correlation between microbial colonization and dental caries.
Subsequently, Orland et al. (1954.J. Dent. Res. 33: 147-174) used
gnotobiotic rats to prove that ingestion of a cariogenic diet alone was not
enough to produce dental caries; and in order for caries to occur, the animals had
to be infected by certain bacteria.
By the mid-1960s, the stage was set to combat cariogenic microbes
specifically, when a consensus regarding the target organism as well as the target
host defense system was reached. After languishing for a decade in the
shadows of Lactobacilli, Streptococcus mutans re-emerged as the prime
candidate for antimicrobial attack as a result of various epidemiological and
etiological studies. As discussed in the preceding chapter, Thomas B. Tomasi
and colleagues (1965. J. Exp. Med. 121: 10-24) provided an equally important
demonstration that the IgA system was the primary specific immunological
element in saliva. These two advances set the stage for dental vaccination
approaches targetting a specific pathogen (S. mutans) and manipulating a
specific humoral immune system (sIgA).
In this section on caries immunology, we will examine evidence that the
host mucosal immune system -- and sIgA in particular -- can be protective
against caries and discuss how HYPERimmunization may confer specific
immunity against dental caries.
Role of Innate
Factors in Caries
Dental caries is a multifactorial disease, as such, protection against dental
caries involves a number of factors. The teeth are protected by the mucosal
immune system discussed in Chapter 12, but for obvious reasons, lack some of
305
 the cellular components of that system. Thus, fluid phase factors secreted by
salivary glands are thought to be the most important of the mucosal immune
components. Persuasively, individuals with salivary hypofunction (especially,
xerostomia) often exhibit rampant decay. Although this is usually attributed to
water (and quite appropriately so), “nonspecific” innate factors also play a
number of functions in protecting the exposed, calcified tissues of the tooth from
dental caries (Mandel, 1979): this includes the buffering of bacterial acids,
clearance of organic waste (such as bacterial acids and ingested substrates),
reduction of surface free energy (resulting in a decrease in both physical and
chemical reactivity), clearance of bacteria, secretion of nonspecific immunefactors
such as acidic proline-rich proteins, lactoferrin, lysozyme, mucins, histatins, and
salivary peroxidase . Of these factors, we’ve mentioned all in the previous
chapter. Herein, let us reconsider the proline-rich proteins, as an illustrative
example of pleomorphism and dental caries.
Human Salivary
Proline-Rich Peptides
The human salivary acidic proline-rich proteins (PRP) are 150-170 amino
acid proteins comprising a major fraction of the protein in saliva. Importantly,
these proteins maintain saliva in a supersaturated state with respect to calcium
phosphate and constitute a significant fraction of the acquired enamel pellicle.
These two concepts suggests that the PRP play an important role in
mineralization processes near the surface of the tooth and may also modulate
microbial adherence prior to plaque formation.
The PRP are encoded by two genes, PRH1 and PRH2 which have been
localized to the short arm of chromosome 12 (Mamula et al., 1985. Cytogenet.
Cell Genet. 39: 279-284). PRH1 encodes three main PRPs, including “Db, PIF,
and Pa.” PRH2 encodes two PRPs, including “PRP-1 and PRP-2.” Interestingly,
some pleomorphism has been documented in the PRPs, and some variants
may be associated with greater susceptibility to dental caries, although this isnot
clear (Hay et al., 1994. J.Dent. Res. 73: 1717-1726. There was no strong
association between pleomorphism of acidic proline-rich proteins and dental
caries, and the same thing may be said about other nonspecific host factors and
dental caries (including lysozyme, mucins, histatins, salivary peroxidase, and
lactoferrin).
Natural Development of
sIgASecretory Immunity
Fetal development. As described in the previous chapter, the sIgA
system may be divided into two functional sites, the inducer site where B-cells
are exposed to antigen and develop IgA commitment, and the effector site,
where B-cells differentiate into plasma cells and where epithelium is involved in
the transcytosis of IgA into the secretion. During fetal development, inducersites
develop prior to effector sites. Peyers patches develop by week 11 of
gestation.
Salivary gland epithelial cells possessing the polymeric Ig receptor are not
observed in the fetus until week 19. At this time, the B-cells in the salivary glands
primarily form IgM, indicative of a naive condition. No sIgA is present in saliva at
birth. Shortly after birth, both IgA and IgM immunoglobulins appear to be
secreted in saliva.
Predentate infants (16-28 weeks). Secretory IgA antibodies to oral
streptococci have detected in the saliva of children as young as 6 weeks of age.
These antibodies are primarily directed against the “first wave” of streptococcal
invaders, ie., S. mitis and S. salivarius (Smith and Taubman, 1992. Crit. Rev.
Oral Biol. Med. 3: 109-133). These organisms initially colonize mucosalsurfaces.
No antibodies to S. mutans are detected.
Dentate children. As teeth erupt into the oral cavity, the microbiota
undergoes a second wave of change. Tooth colonizers such as S. sanguis and
306
 S. mutans begin to establish. Antibodies against S. mutans are usually
observed in 1 year old children. Antibody specificities are against the serotype-
specific carbohydrate, protein I/II, glucosyltransferase, glucans, and teichoicacids,
antigens of S. mutans which may be of future therapeutic significance (see
below). Within 10 years, the child exhibits levels of IgA which are comparable to
those of an adult. For the record, adult parotid saliva normally contains 30-160
µg/ml of IgA immunoglobulin.
The Relationship
Between Caries and sIgA
In early studies, no consistent correlation between salivary IgA levels and
resistance to caries was observed. Some hints were provided that suggested
the potential for a protective effect of sIgA. For example, low titers of parotid
sIgA appeared to correspond with higher rates of dental caries (Ørstavik and
Brandtzaeg. 1975. Arch. Oral Biol. 20: 701-704). However, since the levels of
IgA antibody rather than IgA immunoglobulin may be important, these
studies were not definitive.
IgA deficiency. IgA deficiency is a relatively common disease afflicting
1:1000 individuals which has been associated with dental caries. Unfortunately,
subjects with this condition suffer from chronic rhinitis and sinusitis (both infectious
and allergic). This tends to increase habitual mouth breathing, use of sucrose-
containing medicinal syrups, poor oral hygiene during acute infection, and bottle-
feeding to help them sleep. Against this background, it is difficult to control these
studies. With this caveat, it was found that subjects with IgA deficiency fell into
two groups in terms of oral antibody: ie., those with compensatory IgM
antibodies against S. mutans in saliva and those without. Only in the group
without compensatory IgM was the caries activity significantly higher than age-
sex matched controls (McGhee and Michalek.1981 Ann. Rev. Microbiol. 35:
595-638).
Panhypo-or agammaglobulinemia. These subjects are in worse shape
than IgA deficiency. It is even more difficult to control these studies. However,
cases of increased caries activity have been reported (which must be viewed as
impressive given the antibiotic regimen that these individuals are often
provided).
Secretory IgA antibody against S. mutans. Specific IgA antibodies
againstS. mutans have been measured and found to be significant in parotid
saliva against all of the major serogroups of S. mutans. The protective effect of
these antibodies has not been completely demonstrated. Recently, it has been
shown that human parotid sIgA antibodies against surface antigen I/II (see
below) of S. mutans could block S. mutans adhesion to saliva-coated
hydroxyapatite (Hajishengallis et al., 1992. Infect. immun. 60: 5057-5064),
suggesting that there is a mechanism of protection available to the host against
certaincariogenicbacteria.However, as yet, there has been no strong correlation
between such antibodies in saliva and resistance to dental caries.
Serum antibodies. There are conflicting reports of the correlationbetween
serum antibody and caries resistance. Some would suggest that conflict results
from (1) not testing the correct specificity, (2) not testing at the right time (ie., after
fillings or long after challenge has subsided), or more germane (3) that serum
antibody is a non-protective response. Serum antibodies, intragingival
antibodies, complement, and granulocytes are constantly extravasating from the
periodontal crevice and into the oral environment. These components may
confer modest protection to the tooth in the cervical area, but they are not likely to
be significance in coronal portions of the teeth.
Spcific Immunity
Against Dental Caries
Naturally-induced antibodies in children. As mentioned above, it is
clear that infants and young children rapidly develop sIgA antibodies against
307
 many oral antigens, presumably by the enterosalivary pathway (Smith and
Taubman. 1992. Crit. Rev. Oral Biol. Med. 3: 109-133). However, an
association between these sIgA antibodies and resistance to dental infection by
these pathogens has yet to be convincingly demonstrated. Some have
observed neither salivary IgA nor crevicular IgG corresponds with colonization
by cariogenic bacteria (Camling. 1991. Studies of naturally-occurring antibodies
to mutans streptococci in humans. Dissertation. Dept. of Cariology. University of
Göteborg.). The crevicular IgG antibody is produced locally and appears to
reflect caries experience rather than protection. These results do not mean that
naturally-induced antibodies are unable to interrupt the caries process. The
student should maintain this perspective, although bacterial colonization was not
impaired, the issue of bacterial metabolism and current caries activity was not
addressed. Caries has been correlated with elevated sIgA antibodies and
elevated serum IgM antibodies to S. mutans. This probably reflects the
elevation of antibody which occurs during and after infections. As such, it is not
surprising that it is difficult to make a case for a protective role for antibodies
based upon cross-sectional data.
Caries vaccination. Naturally-induced immunity is not the same as
artificially-induced “hyperimmunization,” as observed after vaccine
administration. Hyperimmunization results in the elevation of antibody to
therapeutic or preventive levels against a specific microorganism. Generally, the
aim of a vaccine is to reduce the numbers of an offending pathogen or to interfere
with its metabolic activity and pathogenic components. In order to use
hyperimmunization, several important things must be considered: (1) What will
be the microbial target (ie., what is the offending pathogen?); (2) which
component of the immune system should be targetted; and (3) prior to
designing a vaccine, is there any evidence that hyperimmunization will work?
Offending Pathogens,
the Lactic Acid Bacteria
Criteria for cariogenicity. Focus has been placed upon the lactic acid
bacteria as specific etiologic agents initiating dental caries: especially, the
“mutans-streptococci,”Streptococcus mutans and S. sobrinus). Cariogenicity
cannot be traced to any one property of these streptococci, but rather a
combination of biological and biochemical properties. To be cariogenic, an
organism must exhibit tropism for teeth and must be acidogenic and aciduric
(Newbrun, E. 1983. Cariology, 2nd Ed. The William & Wilkins Company,
Baltimore. pp50-85). Additionally, the organism should utilize refined sugar
(sucrose, a disaccharide of glucose and fructose) as part of the disease process,
in view of the direct (albeit, not necessarily linear) correlation between dietary
sucrose consumption and caries experience (Newbrun, E. 1983. Cariology,
2nd Ed. pp86-121).
The Lactic Acid Bacteria as Prime Suspects. The lactic acid bacteria
comprise a heterogenous family of microorganisms which possess the
necessary biochemical characteristics to initiate and perpetuate the caries
Table 1: The Lactic Acid Bacteria
Genera Morphology Good Bad
Streptococcus Gram + cocci, in chains Protect oral environment from Enamal caires
worse pathogens
Lactobacillus Gram + rods, in chains Protect oral environment from Dentinal caries
worse pathogens
Dairy Products
Leuconostoc Gram + cocci in chains Useful dextran products --
Pediococcus Gram + cocci in tetrads Pickles --
process. Lactic acid bacteria are Gram-positive cocci and rods which include four
308
 genera: Lactobacillus, Streptococcus, Leuconostoc, and Pediococcus (Table 1)
All of these organisms exhibit metabolic properties which may be classified as
indifferent facultative (Loesche, 1975). In the process of generating energy,
indifferent facultative organisms ferment hexoses and always utilize organicacids
as terminal electron acceptors regardless of the presence or absence of oxygen
(no oxidative phosphorylation). This is in distinction to true facultative organisms,
which utilize oxygen when it is available, forming water and carbon dioxide, rather
than acid. The indifferent facultative organism always produces acid (they are
acidogenic). The predominant acid produced by most lactic
Lactic Acid Bacteria are not bad acid bacteria is lactic acid, which exhibits a lower pKa and
guys! These organisms are useful in less volatility than most organic acids, and is therefore the
the dairy industry (about 1023 most destructive to enamel. Lactic acid may also form
lactobacilli per year are used chelates with calcium, which would facilitate the
globally for dairy products - dimineralization of enamel.
Moineau et al., 2002. ASM News 68: The one other property peculiar to the lactic acid bacteria is
388-393), making pickles, as well as their extracellular utilization of sucrose. Species
protecting us f r o m t e r r i b l e representative of all four genera of the lactic acid bacteria
pathogens. T h e glucans from form extracellular glucose polymers (glucans) from sucrose
Leuconostoc mesenteroides have been via a glucosyltransferase enzyme system that will be
used for extending blood (dextran) discussed in detail later. In general, extracellularcarbohydrate
and for gel filtration chromato- polymers enable microorganisms to control their external
graphy (Sephadex). environment. Glucose is a precious commodity. Therefore,
Even in the mouth, they are expenditure of glucose to form glucans must be important
normally beneficial. Streptococcus to the lactic acid bacteria. Some of the lactic acid bacteria also
sanguis exerts antimicrobial effects form polyfructans from sucrose which may also participate in
against periodontal pathogens and the caries process.
Candida albicans. Of the lactic acid bacteria, two genera have been
associated with caries (because they colonize teeth). These
are Lactobacillusand Streptococcus. The former has been implicated in dentinal
caries, whereas the latter has been associated with initiating the caries processin
enamel.
Let us focus our discussion on
TABLE 2 the streptococci. There are many
STREPTOCOCCUS MUTANS SEROTYPES & SPECIES species or groups of streptococci
Species Serotypes G + C% Host Isolate inhabiting the mouth and the tooth
S. mutans c, e, f 36-38% Primates 90% surface, b u t the mutans-
streptococci have been most
S. sobrinus d, g, h 44-46% Primates 8-40%
closely associated with caries of
S. ferus c 43-45% Wild rats NA dental smooth surfaces, pits, and
S. macaque c 35-36% Monkeys NA fissures. There are six serotypes
S. cricetus a 42-44% Hamsters NA of mutans-streptococci which have
S. rattus b 42-44% Rats NA been described in man (table 2).
In man, the most prevalent
serotype of S. mutans associated with smooth surface dental caries is serotype
c. In smooth surface caries, S. mutans serotype c is the predominant group
associated with enamel caries. It is isolated in about 80-87% of cases in the
United States, most of Europe, and Japan (Shklair and Keene, 1976.In: H.M.
Stiles, W.J. Loesche, and T.C. O'Brien (eds.) Microbial Aspects of Dental
Caries. Information Retrieval, Inc., Washington, D. C. pp 201-210; Perch et al,
1974.Acta Pathol. Microbiol. Scand. [B]. 82: 357-370; Hamada et al., 1976.
Jpn. J. Microbiol. 20: 33-44; Bright et al., 1977. J. Dent. Res. 56:1421). In
Swedish children, 36% of all tooth surfaces harbored serotype c, and 54%
serotype d/g (Bratthall and Köhler, 1976. J. Dent. Res. 55: C15-C21).
Designing an
Anticaries Vaccine
Targetted immune systems. Two immune systems were targetted for
309
 Table 3: Effect of Immunization with S mutans on hyperimmunization: the secretory IgA
Caries Scores in Rats system and the crevicular (serum and
MEAN CARIES SCORE gingival) IgG-IgM-IgA system. “Cellular
(some penetration into dentin) immune” mechanisms were not targetted
GROUP Buccal Sulcal Proximal for several reasons: first, cells would have

IMMUNIZED AND 0.6 8.7 0.3

difficulty functioning in the mouth and
INFECTED
second, immunity against bacteria is
usually not handled by cellular immune
NOT IMMUNIZED 2.1 10.5 1.6
AND INFECTED
mechanisms unless they are chronic and
persistent (usually meaning that the host is
NOT IMMUNIZED, 0.1 1.8 0.0
NOT INFECTED
having a hard time handling the infection).
Most bacterial infections are handled by
Michalek et al., 1976. Infect. Immun. 1 2 : 782
secretory immunity (including secretory
IgA) or the antibody (IgG)-complement - neutrophil axis. The neutrophil is not
always necessary for the latter system to be effective.
Evidence that an anti-caries vaccine would be effective. A number
of studies performed in the 1970s indicated that it is possible to protect
laboratory animals against dental caries by using hyperimmunization. The results
of one of these studies is shown in table 3. As you can see, hyperimmunization
of rats (fed a cariogenic diet) lead to a high level of protection against smooth
surface caries (buccal and proximal) but less than impressive protection against
pit and fissure caries (sulcal). This data set is of interest because it not only shows
the protective effects of hyperimmunization, but also the differential protective
effects based upon location. That is, even if we could protect teeth by
vaccination, such vaccination would not protect the pit and fissure locations. As
future dentists, then, you realize that your patients would require additional
protection against pit and fissure lesions (such as a sealant).
Why we cannot immunize parenterally with whole cells of S.
mutans. Obviously, immunization against dental caries starts with the organism
most tightly associated with caries, S.
Figure 1: CANDIDATE ANTIGENS mutans. U n f o r t u n a t e l y , S . mutans
Insoluble possesses antigens which are cross-reactive
glucans
with heart muscle, especially, the cardiolipin
Glucan- Lipoteichoic Acids of the sarcolemma sheaths. Although patient
Binding death is certainly one form of caries control,
Protein
this means that whole cells of S. mutans are
Ribosomes
not likely to be viewed as acceptable
GTF-I Adhesins
SAI/II
parenteral antigens and they should be used
Serotype with caution per os (orally administered). In
carbohydrate monkeys receiving parenteral S. mutans, no
antigen GTF-S heart muscle damage was noted. This may
Peptidoglycan Soluble glucans
be just luck, since our species may respond
TABLE 4. LIST OF CANDIDATE ANTIGENS
differently to a cardiolipin challenge due to
subtle differences in our immune response
Class Sites genes, especially MHC class II molecules
Glucosyltransferases Enzymatic site and CD1. Therefore, it is of crucial
Glucan binding site importance to use an alternative means of
Adhesins SA I/II, B, P1 (S. mutans ) vaccination. Several alternatives include (1)
SpaA (S. sobrinus ) purification of the candidate antigens and
Dextranase Enzymatic site use of a subunit vaccine or (2) using
Glucan binding site recombinant DNA methods to place
Glucan-binding proteins Glucan binding site
virulence factors from cariogenic organisms
into a noncariogenic, non-crossreactive
Surface polymers Glucans bacterium.
Serotype-specific CHO Ag Candidate antigens have been
Lipoteichoic acid selected because they are believed to play
310
 some role in the pathogenic activities of S. mutans and S. sobrinus (Figure 1).
Extracellular protein targets include glucosyltransferases (GTF), dextranases,
adhesins (such as Spa A or SA I/II), and glucan-binding protein. Other non-
protein candidate antigens have also been proposed, including extracellular
glucans and the serotype-defining antigen (Table 4). Let’s consider these
antigen targets.
Glucosyltransferases

At least two types of glucosyltransferases are known to exist inS. mutans (GTF-
S and GTF-I, which synthesize water-soluble and water-insoluble glucans,
respectively). InS. sobrinus there are three (GTF-S Pd, GTF-S Pi, and GTF-I).
Pd refers to “primer-dependent” and Pi refers to “primer-independent.” A primer
is a short glucose-oligosaccharide. Primer dependency simply means that these
enzymes require a primer to catalyze glucose polymerization. Various molecular
weights have been assigned to these enzymes, but in general, they have a
molecular weight of about 160,000 kdal. The GTF enzymes may be
evolutionarily related to both the glucan-binding protein and to
fructosyltransferase. The GTF enzymes utilize the energy in the hemiacetal-
hemiketal bond between the glucosyl and fructosyl sugars in sucrose, to form
glucans (polymers of glucose in both α1-3 and α1-6 linkages). GTF enzymes
also have lectin activity and feature a glucan binding domain. Antibodies which
impede GTF function have been shown to be protective in animals.
Adhesins
Surface Protein Antigen (known as SA I/II, B, P1 in S. mutans and Spa
A in S. sobrinus ). Surface protein antigens are large proteins (185-210 kdal
[inS. mutans ] and160-180 kdal [inS. sobrinus ]) which constitute 35% of cell
surface protein, and are thereby the predominant cell-surface protein.
Interestingly, these are immunologically related to dextranase, an enzyme
(below). The main function of these surface proteins appear to be sucrose-
independent adherence (that is, the attachment of the bacterium to the tooth in
the absence of sucrose). A “fuzzy coat” on the bacterial surface, observed by
electron microscopy, is the ultrastructural appearance of these surface antigens.
Mutants lacking SAI/II also lack a fuzzy coat and bind poorly to experimental
pellicles (Harrington and Russell, 1993. J. Bacteriol. 175: 5925-5933). SA I/II
also appears to mediate the saliva-induced aggregation of S. mutans (Koga et
al., 1990. Infect. Immun. 58: 289-296). Antibodies against the surface protein
antigen are protective in monkeys. Refining these earlier observations is the
more recent observation that SA I/II possesses a saliva binding region (SBR).
Antibodies against SBR appear to protect against the colonization of S. mutans
to teeth in mice (Huang et al., 2001. Infect Immun 69:2154-2161).
Dextranases
Dextranases are 160-175 kdal protein enzymes which break down
polymers of glucose in α 1-6 linkage (dextrans). These dextranases are
probably used by oral streptococci to modify the glucan product of GTF,
clipping away the α 1-6 linked oligomers and thereby increasing the proportion of
α 1-3 linkages. Additionally, it is remotely possible that this may permit
extracellular glucans to serve as energy stores. Dextranases may function in
sucrose-independent adherence (via an SpaA-related epitope). Mutants lacking
both dextranase and SpaA are avirulent.
Surface
Carbohydrates
Glucans are tree-like homopolymers of glucose featuring gazillions of
branches but only one trunk (Figure 2). The tips of the gazillion branches are
311
 Figure 2. GLUCAN STRUCTURE called “non-reducing” and the tip of the
6 trunk is called “reducing” (an arcane
α 1-6 α 1-6 α 1-6 α 1-6 factoid). There are two physical types:
4 5 A C Water-soluble and water-insoluble,
1
and there are two main enzymes
Branch involved in their production (GTF-S
3 2
Point and GTF-I, described above). Table
α 1-3
Carbons 5 shows the results of methylation
numbered analysis of soluble and insoluble
glucans. In methylation analysis,
α 1-3 methyl groups are added to any un-
B
Methylation Analysis reacted hydroxyl; thus, a 2,4,6-
A ---> 2, 3, 4 methyl ether trimethyl ether of glucose would prove
α 1-3 B ---> 2, 4, 6 methyl ether that those carbons had free hydroxyls.
C ---> 2, 4 methyl ether We can the infer that sugar was linked
at it’s 1 and 3 carbons. Inasmuch as
you do not find any 1 glucosyl methyl
ethers (well, there is one, the “reducing” terminus, it is heavily outnumbered), we
are safe to say that the 1 carbon (the anomeric carbon) is always linked.
The results (Table 5) show that soluble and insoluble glucans derived from
Table 5: Methylation analysis of Glucans (GTF = enzyme preparation; CFF = cell free fluid)
Results tabulate percentage of linkages (Hare et al., 1978. Carbohyd. Res. 6 6 : 245-264)
Source α1,3 α1,6 Nonreducing α1,3,6
linked a linked b Terminic Branchpointsd
K1-R GTF-I insoluble glucan 85 4 9 2
OMZ-176 GTF-I insoluble glucan 88 7 4 1
OMZ-176 GTF-S soluble glucan trace 31 37 37
OMZ-176 CFF insoluble glucan 57 18 15 10
NSW47 CFF insoluble glucan 38 37 12 11
NSW47 CFF soluble glucan 28 41 15 16
a2,4,6-trimethyl ether; b2,3,4-trimethyl ether; c 2,3,4,6-tetramethylether; 2,4-dimethylether

bacteria (or their cell-free fluids [CCF]) are quantitatively distinct, but not
absolutely, with respect to linkage. Both contain α 1-3 and α 1-6 linkages in linear
array, and both contain α 1-3-6 linkage branch points. These results are reinforced
in studies using (i) linkage-specific enzymes or (ii) anomeric proton - NMR
spectral shift analysis for nearest neighboring linkages (the NMR doublet of the
anomeric proton shifts upward near α 1-3 linkages).
There are somewhat more α 1-6 (dextran) type linkages in the water-
soluble form and more α 1-3 (mutan) linkages in the water-insoluble form. The
quantitative distinction increases with enzyme purity. For example, note the
greater proportion of α 1-3 (mutan) linkages in purified GTF-I compared to GTF-
S from S. mutans OMZ-176; but realize that we’re usually dealing with whole
cells. Because we isolate soluble and insoluble glucans from cells, it is most likely
that both enzymes participate in the production of both products. Another main
difference between water-soluble and water-insoluble glucans appears to be
size (water-soluble forms are usually 20,000-50,000 daltons, whereas water-
insoluble forms are generally 10 6- 10 7 daltons). Anyhow, I only mention this stuff
since immunologists often tend to gloss over some really nice biochemistry
when discussing carbohydrates.
Can a carbohydrate be a good antigen? Well, yes! Compare a
polymer of hexose sugars with a polymer of amino acids. Two different amino
acids can be linked in only two ways (ie., A--->B or B--->A). The immune
system can distinguish such primary structure. Since we are talking about a
312
 homopolymer of glucose, let’s take a look at how homopolymers of glucose can
convey information. One sugar will be linked to the other sugar via the hydroxyl
of its anomeric carbon (C1) as either an α or β anomer. That alone would provide
two different structures, which the immune system can distinguish. Now, this α or
β anomer can form a linkage with the hydroxyl of the second, third, fourth, and
sixth carbon (C2, C3, C4, or C6). Actually, there are instances where two
anomeric carbons are linked to one another (ie., the α or β anomer C1 can form a
linkage with another α or β anomeric C1). We won’t even count that. Therefore,
two identical sugars can produce at least eight different antigens (just think if we
used one galactose and one glucose!).
Glucans function in plaque accumulation, act as molecular sieves and
convection barriers, retainwater, and although they do not play a role in the initial
colonization of teeth by bacteria, they clearly may greatly strengthen the
attachment of the producing organism to the tooth surface. In general, glucans
enable the producing organism to control its microenvironment. The significance
of this is clearly shown by the increasing proportion of mutans-streptococci in
subjects with high dietary sucrose intake. Antibodies to dextran appear to
prevent the binding of GTF to the bacterial cell surface and have been
proposed as a possible means of conferring caries protection.
Serotype-defining carbohydrate antigens are complex carbo-hydrate
heteropolymers containing galactose and glucose. There are eight “serotypes”
of mutans-streptococci. These serotypes have been designated a-h. Only
serotypes cef (S. mutans) and dgh (S. sobrinus) are important in man. These
recognitive structures are probably used to specifically bind enzymes such as
GTF to cell surface, and thus, they have been proposed as a target for a caries
vaccine. Antibodies against the serotype-defining carbohydrate antigens are
protective and appear to prevent the binding of GTF to cell.
Lipoteichoic acids are amphipathic molecules found on the surface of
Gram-positive bacteria which are, in some ways, analogous to
lipopolysaccarides of Gram-negative bacteria. They consist of a carbohydrate
backbone of polyribose ± phosphate or polyglycerol ± phosphate. The
carbohydrates are covalently coupled lipid. It has been proposed that these
structures may be involved in adhesion by hydrophobic interaction (not likely).
The problem with LTA as a candidate antigen is that they feature epitopes which
may cross-react with host tissue antigens.
Active
Anticaries Immunization
Recently, there has been a resurgent interest in the development of a
caries vaccine which would confer active immunity. As discussed above, whole,
attenuated bacteria are not usually considered safe due to heart muscle cross-
reactivity. However, this may be a false concern; heart muscle cross reactivity
may not be a problem if the vaccine is administered in an oral manner and
stimulates the sIgA system rather than the IgG system. This is because IgA
antibodies would tend to be beneficial and suppress inflammatory reactions by
immune elimination at the mucosal surface (thereby decrease IgG responses)
and if systemic IgA is elicited it could bind to antigen and prevent complement
fixation.
Historical. The monoinfected gnotobiotic rat system has consistently
demonstrated the efficacy of vaccines against caries when those vaccines
stimulated sIgA. Less convincing (but generally favorable over a short term)
results are obtained when an immunization route favoring responses by (but not
limited to) the serum immunoglobulin system when conventional (not maintained
gnotobiotically) rats and monkeys are used. Some believe that crevicular fluid
IgG is elevated in gingival inflammation (pathotropic potentiation), and this can
lead to opsonization and phagocytic clearance of pathogenic bacteria.
Salient animal terminology. As future cariologists, you may someday
313
be asked to evaluate an animal study. As discussed briefly above, the results of
the animal studies vary with how the animal was reared and maintained.
Therefore, you need to know the terminology of animal studies. “Gnotobiosis” is
a term derived from Greek, which means “known life;” ie., that the microflora of
the animal is known. Gnotobiotic animals may be “germfree,” a conditionreferred
to as “axenic” (ie., “axenic” means “germfree”). Germfree animals are animals
originally derived by Caesarian section, and subsequently bred in a germfree
environment. Not all gnotobiotic animals are germfree, nor do they need to start
life as germfree animals. Gnotobiotic animals may also be “monoinfected,” ie.,
infected with a single microbe or multiply infected, provided that the microflora is
defined. Colonies of animals can also harbor one or several microbes, thus the
colony produces gnotobiotic offspring which begin their life infected rather than
germfree. Axenic animals will not experience dental caries even when fed on an
extremely cariogenic diet. Animals monoinfected with cariogenic bacteria will
suffer much dental destruction on the same diet. Vaccines in the monoinfected
model are always much more successful than vaccines in animals reared in a
conventional manner. Thus, you should be wary of all such monoinfected studies.
For this reason, some investigators attempted studies using “specific
pathogen-free” (SPF) animals. This is a mealy-mouthed term indicating that the
animals are raised under conditions in which a specific pathogen is eliminated.For
example, in a caries vaccine study, it would be important to maintain an SPF
colony in the absence of S. mutans, such that the pathogen can be introduced in
a controlled manner. One problem with the SPF model is that investigators never
really state with certainty that their SPF animals are indeed missing the the
pathogen. Either germfree or SPF animals may be “Caesarian-obtained, barrier-
sustained,” or COBS. Actually, any animal can be a COBS animal, so this is a
bit of a useless term.
Routes of vaccination. Currently, there are two favored potential
routes of vaccine administration:Peroral (per os, p.o.) and intranasal (nasal)
(Michalek et al., 2001. BioDrugs 15:501-508). Whole cells of S. mutans
encapsulated within gelatin were used to immunize human volunteers p.o. The
question was whether sIgA antibodies can be elicited by oral ingestion of whole
S. mutans. A priori, this form of vaccination requires the activity of the “enteric
pathway,” since gelatin capsules preclude intraoral immunization. It has been
reported that peroral immunization by S. mutans can elevate sIgA antibodies
(Gregory and Filler. 1987. Infect. Immun. 55: 2409-2415). Individuals were
administered gelatin-capsules (10 consecutive days) containing killed S. mutans
whole cells which was isolated from the volunteer themselves. SIgA, specific
against GTF and SA I/II were detected in all cases, and in each case, there was a
reduction in viable S. mutans isolated from dental plaque, but it was unclear
whether this was of any value in terms of caries prevention. SIgA antibodies
were detected in the saliva and tears, and in the colostrum/milk of mothers giving
birth. The question which was not addressed was whether potentially harmful
serum IgG antibody against LTA were elicited by this protocol. Further, other
studies suggest that peroral immunization is not always this successful (Linzer et
al., 1981. Infect. Immun. 31:345-351; Cole et al., 1984. Infect. Immun. 46:
702-709).
Subunit vaccination. There are several “twists” to the subunit vaccine
approach which embellish the traditional methods of using purified bacterial
antigens. Some of these are discussed below.
Synthetic peptides. It may also be possible to use only a piece of a
large protein, such as GTF, as an immunogen (Smith et al., 1993. Infect.Immun.
61: 2899-2905; Smith et al., 1994. Infect. Immun. 62: 5470-5476). Synthetic
peptides derived from a glucan-binding domain of glucosyltransferase,
TGAQTIKGQKLYFKANGQQVKG, or a multimeric synthetic peptide derived
from an amino-terminal sequence, DANFDSIRVDAVDNVDADLLQ, have
314
 been used as potential immunogens. Antiserum against
TGAQTIKGQKLYFKANGQQVKG raised in rats was found to inhibit GTF by
30% (not that impressive), and a monoclonal antibody against
DANFDSIRVDAVDNVDADLLQ was 80% inhibitory. By the way, you are not
expected to know the
Figure 3: MOLECULAR APPROACH TO A CARIES VACCINE primary sequences used in
Bacillus globigii Streptococcus sobrinus synthetic peptide vaccines,
A GATCT just that synthetic peptides
offer an alternative.
Molecular genetics
Ampicillin
resitance
and the enteric pathway.
gene
ENTERO-
Molecular genetic
Cosmid
clone
SALIVARY approaches now offer one of
"Harmless" Enteric Bacteria PATHWAY the most exciting means of
OF sIgA
Expressing SpA
delivering a “subunit” vaccine
which would be cost
effective. The problem with subunit vaccines has been the inability to maintain
sufficiently high levels of antigen in the gut to stimulate antibody production in a
cost-effective manner. Recently, candidate antigen genes have been introduced
into “harmless” enteric bacteria (Figure 3). These bacteria proliferate for some
time and exhibit considerable greater staying power in the gut than simple
gelatin capsules filled with antigen. This method of immunization is currently under
investigation. But think about it, no microbe which can colonize a human should
be considered totally “harmless.” Also, some of the plasmid vectors used are
marked with genes encoding antibiotic resistance (but this is just a minor
problem).
Gingival swabs and the “local pathway.” The gingiva is an area in
which local immune responses can be elicited. The swabbing of gingiva with a
3800 kdal low molecular weight component of S. mutans has been found to
elicit both increases in IgG in the crevicular fluid and sIgA in the saliva of monkeys
(Lehner et al., 1986.Infect. Immun.52:682-687). From a didactic point ofview,
it is difficult to ascribe the sIgA response to local (gingival) immunization rather
than the enteric pathway, since some antigen must be ingested. From a
therapeutic point of view, the method itself may be useful: ie., the swabbing was
administered only ten times over a year period and resulted in a reduction in S.
mutans as well as caries.
Liposomes. Liposomes are artificial membrane vesicles which can be
prepared to contain both aqueous-phase solutes internally or intramembranous
molecules within their membranes. Liposomes represent a relatively benign
mechanism of increasing immune responses to antigens (ie., they are
“adjuvants”). One method of increasing antibody responses by gingival
immunization has been the sequestration of candidate antigens (GTF, in this
case) into liposomes, permitting the liposomes to dessicate, and administering
the dehydrated liposomes to humans. This resulted in salivary IgA2 antibodies
againstGTF, suggesting that dehydrated liposomes may be useful in generating
specific salivary immunity against target antigens in the oral cavity (Childers et al.,
1994.Oral Microbiol. Immunol. 9: 146-153).
Coupling. Another method for enhancing immune responses to antigens is
to a “poor” antigen is to couple the “poor” antigen to a “good” antigen. For
example, polysaccharide antigens are usually poor antigens: they tend to be T-
independent and therefore, sustain primary immune response characteristics
(without T-cells, you usually can't get isotype switching or hypermutation). To
circumvent this problem, polysaccharide antigens may be coupled to a protein
(proteins are T-dependent, usually). This will result in increased specificity and
isotype switching. Intragastric administration of liposomes containing
polysaccharides of S. mutans coupled to a protein has been used in the rat
315
 model (Wachsmann, D. et al, 1986. Infect. Immun. 52:408-413).
Antiidiotype vaccine. One potential method for eliciting antibodies
against any target includes the use of an antiidiotypic vaccine. In this approach,
antibodies which possess an idiotope that resemble a bacterial epitope (ie.,
“internal image” antibodies) are injected into a host. If these antibodies are the
same allotype as the host, the host will form antibodies against only the internal
image. These antibodies against the internal image can then stimulate
antiidiotypic antibodies which also can bind to bacteria (Jackson et al., 1990.
Infect. Immun. 58: 1932-1936). Interestingly, the study cited also used
liposomes as a mechanism of delivery of the antiidiotypic antibodies by gastric
intubation. The antiidiotypic vaccine lead to greatly reduced caries in the
gnotobiotic rat model.
Adjuvants. Many of the peptide antigens described above would be
poorly immunogenic were it not for the use of adjuvants. The dentist should be
aware that many traditional adjuvants used in animals (such as complete Freund’s
adjuvant, a mixture of mycobacterial components and mineral oil) are too toxic for
human use. An inexpensive adjuvant approved for use in humans is “alum,” an
inorganic salt of aluminum. Liposomes, mentioned above, may offer an attractive
adjuvant system.
The most promising adjuvant stimulating mucosal sIgA responses appears
to be cholera toxin, which is under intense investigation by Michael Russell’s
group at the University of Alabama. It appears to stimulate persistently high
levels of sIgA after a single boost (Hajishengallis et al., 1996.Infect. Immun.
64:665-667). Cholera toxin is a heterodimer featuring a toxic CTA-subunit and a
nontoxic CTB-subunit. Adjuvanticity is associated with the nontoxic CTB-
subunit, and a clever approach has been to replace the CTA-subunit with
antigens -- such as SA I/II -- derived from S. mutans (Wu and Russell, 1993.
Infect. Immun. 61:314-322; Katz et al., 1993. Infect. Immun. 61: 1964-1971;
Toida et al., 1997. Infect. Immun. 65: 909-915). And indeed, as you may have
predicted, they have even constructed an enteric bacterial clone which
expresses SAI/II-CTA2/CTB. The enteric bacterium selected for this was an
‘avirulent’ strain of Salmonella typhimurium (Harokopakis et al., 1997. Infect.
Immun. 65:1445-1454).
Also of great interest to you budding clinical dentists out there is that the
intragastric admininistration of fluoride, at concentrations attainable from
inadvertant ingestion of fluoride gels or prescribed fluoride supplements, has
been shown to be a potent adjuvant of mucosal immunity in rats (Butler et al.,
1990. Immunol Lett. 26:217-220). Intragastric NaF caused increases in the size
and cellularity of the Peyer's patches and mesenteric lymph nodes as well as the
number of plasma cells secreting IgG and IgA antibodies to various antigens
concurrently administered in the drinking water. The frequency of CD4+ T cells in
these lymphoid tissues was elevated while that of CD8+ T cells was significantly
decreased. How fluoride does this has never been elucidated, nevertheless, it
raises the possibility that one protective affect of fluoride administration is the
stimulation of immunity to concurrently ingested oral bacteria, and argues in favor
of administration of fluoride as part of a caries vaccine program.
Potential for a Passive
Immunization Approach
When antibodies are passively administered to monoinfected gnotobiotic
animals, as expected, a reduction in disease occurs. Monoclonal antibodies
against S. mutans can also prevent the colonization of human teeth by S.
mutans (Ma et al., 1987. Infect. Immun. 55: 1274-1278). Thus, passive
immune approaches may reasonably be expected to be effective. However,
“cost effectiveness” is another issue. Dental scientists have developed a
number of fairly clever strategies which may see future application.
Maternal immunization. Passive immunization can occur by oral
316
 immunization (secretory IgA is stimulated) of pregnant rats. The milk from
immunized rat mothers confers protection to the weanlings. It is possible that any
mammal can be protected in this fashion.
Xenogeneic immunization. It has been shown that cows can be
Figure 4: A PASSIVE IMMUNIZATION APPROACH
immunized against cariogenic bacteria and that
antibodies against those bacteria appear in the
cow's milk (Figure 4). The cow's milk (or
whey) can then confer protection immunity in a
passive manner. This type of immunization is
shown in the diagram. The antibodies were of
the IgG1 subclass, indicative of the parenteral
immunization used. In cows milk andcolostrum,
IgG1 is the major secreted immunoglobulin
isotype. Both S. mutans and caries scores
were reduced (Michalek et al., 1987. Infect.
Immun. 55:2341-2347) in gnotobiotic rats. Of
course, gnotobiotic rats are easy to protect
compared to conventional animals and humans;
however, whey from immunized cows, used as
a mouthrinse, appeared to decrease S.
mutans in volunteers. Recently, a Finnish group has initiated studies on the
potential therapeutic aplication of bovine whey IgG1 and report that such IgG1
interferes with the intacke of carbohydrate, formation of glucans, and adherenceof
S. mutans (Loimaranta et al., 1997. Vaccine 15: 1261-1268) and a Japanese
group has found that mouth-rinses with Holstein cow milk (funny, huh?)
immunized with a fusion protein of S. mutans (a piece of a “cell surface protein
antigen [probably SAI/II]” fused with the glucan-binding domain of GTF-1) has
been effective in preventing the recolonization of eight human volunteers by S.
mutans (Shimazaki et al., 2001. Clin Diagn Lab Immunol 8:1136-1139).
One other source of edible xenogeneic antibodies is also under
investigation -- chicken eggs. A Japanese group working with Susan Michalek
has begun to explore the potential therapeutic capacity of chicken egg IgY in a
mouthrinse (Hatta et al., 1997, Caries Res. 31: 268-274). Rocky Balboa has
been experimenting eating dozens of raw chicken eggs.
Did you ever want to know about cow’s milk? Milk, as you budding
dentists may know, consists of the following ingredients (in order of
weight/volume amount): water; lactose (4-O-β-D-galactopyranosyl-D-glucose);
triacylglycerols; phosphoglycoproteins (α 1, α 2, β and κ -casein); β-lactoglobulin;
α -lactalbumin; proteose-peptone; immunoglobulins; serum albumin; K+; Na+;
Ca ++; phosphorus; Cl -; diglycerides; cholesterol; phosphotidylcholine;
phosphotidylethanolamine; sphingomyelin; monoacylglycerols; carboxylic acids;
enzymes; hormones; and other lipids, proteins, carbohydrates, and minerals
(Partridge. 1997. Dept. Food Sci. & Human Nutr.). Before getting too bored with
that list, remember how we dentists are always extolling the virtues of milk
calcium -- well, there it is, right after sodium. The caseins (phosphoglycoproteins)
are the major ingredients of cheese, and fractionate as the main structural
component of curds during the production of cheese. Whey is the fluid part of
milk which separates from the casein-enriched curds. Within the whey is the
lactose sugar (1/7 as sweet as sucrose), once considered waste but now quite
useful as a carrier in pharmaceuticals, soups, and spices. The whey proteins (β-
lactoglobulin, α -lactalbumin, proteose-peptone, immunoglobulins, serum
albumin and other various minor proteins) are those components of milk which
are of interest to us.
Pasteurization. Bovine milk is heat-pasteurized in two ways: (i) a
standard holder method (63.5 degrees C for 30 min) and (ii) a high-temperature,
short-time (HTST) method (71.7 degrees C for 15 s). An experimental method
317
 of “pasteurization” avoids heat altogether, and instead puts milk under extremely
high pressures, drops the pressure and forces the milk through a tiny orifice at the
speed of sound. Milk processed in this manner tastes better and can last for
months at room temperature in unopened containers (this is actually better than
standard pasteurization protocols). At any rate, heat pasteurization probably will
not affect most immunoglobulins (except for IgE), and if it did, there are other
experimental pasteurization methods on the horzon.
Cows are different. With respect to immunoglobulin levels, cow’s milk is
rather different than human milk. The main immunoglobulin in bovine milk whey,
as mentioned above, is IgG1 (0.5 mg/mL), followed by IgA, IgM, and IgG2
(0.08, 0.08, and 0.06 mg/mL, respectively). The precision of these
measurements has been questioned a bit, since some IgA and IgM may
associate with fat and be underestimated, especially after refrigeration
(Honkanen-Buzalsk and Sandholm,1981. Comp. Immun. Microbiol. Infect. Dis.
4: 329-342). From the perspective of the cow, it would appear that chronic
inflammatory immunoglobulins (IgG1, IgM, IgG2) are an important part of a
defense axis which also may include secreted macrophages, which are the
predominant cell type found in bovine milk (there are about 25 times more
macrophages than neutrophils) and which possess Fc receptors for both IgG1
and IgG2 (Lee et al., 1980. J. Dairy Res. 47: 39-50). Anyhow, I just wanted you
to sense that other mammals have found other defenses more useful than the
IgA system to defend some secretions.
Why are cows different? Well, this could be because other animals may
need to defend both their infants and their udders (fairly unique structures) against
different types of enemies and in different ways. Additionally, in animals in which
placental transfer of IgG does not occur -- such as sheep, cows, pigs, and horses
-- milk is the only source of maternal antibodies. Unlike humans, such animals
have specific mechanisms for the absorption of IgG in colostrum and milk within
the intestines (Jensen et al., 2001. J. Nutr. 131: 3259-3265). Finally, the
observation that there is nothing sacred about IgA in the secretions of other
animals lends credence to the notion that other immunoglobulins may be able to
compensate in IgA deficiency in humans.
Other Potential Xenogeneic
Passive Immunization Strategies
Monoclonal antibodies. Monoclonal antibodies (MAb), are antibodies of
a single specificity which are produced by cells derived from a single B-cell
clone. In most cases, the Mab are derived from mice (therefore, they are
xenogeneic). Fusion of a normal mouse plasma cells and myeloma cells results
in the formation of “hybridomas” with the antibody-forming properties of the
plasma cell and the proliferative properties of the myeloma cell. When grown in
tissue culture, hybridomas are capable of almost unlimited production of MAb.
At present, MAb are used diagnostically to assess immunocompetence, to
identify infectious diseases, and to monitor the concentrations of hormones and
chemotherapeutic agents in the plasma. Some are used as immunosuppressive
Monoclonal Antibody Chimaeric MAb CDR-Graft MAb
(MAb)
L chain
H chain Human
H chain Murine
L chain
Murine Murine Murine Human Human Human
V C V C V C
domain domains domain domains Framework domains
Murine
Figure 5. Xenogeneic antibodies
CDR

318
 agents. Their exquisite specificity also make them ideal guidance systems as
immunotoxins.
Chimaeric MAb. The most important factor which limits the therapeutic
potential of MAb is that they are xenogeneic, and clinical testing of these
reagents has led to some disappointment. One approach to circumvent this
problem has been the combination of antigen-specific portions of MAb with
human constant or framework domains. Chimaeric MAb represent the second
generation of MAb which partially circumvents this problem by genetic
engineering. Chimaeric MAb graft rodent immunoglobulin V regions to human
immunoglobulin C regions (Figure 5). V-genes are cloned from hybridoma
mRNA using the polymerase chain reaction and linked to preformed expression
vectors containing human heavy and light chain C-regions. Because the C region
of an immunoglobulin also confers function to an antibody, chimaeric MAb permit
the engineering of functional attributes. For example, a chimaeric MAb
possessing an IgG1 isotype C region will be most effective in complement
activation and antibody-dependent cell-mediated cytotoxicity; whereas a
chimaeric antibody of the IgA subclass may exhibit anti-inflammatory effects.
CDR-grafted MAb. Although chimaeric MAb seem rather exotic, even
they are being replaced by a third generation MAb, the complementarity-
defining region (CDR)-grafted MAb. CDR refers to those areas of an antibody
which bind to an antigen. The variable region of immunoglobulin light and heavy
chains actually contains 3-4 hypervariable regions and intervening framework
regions. In general, the hypervariable regions are equivalent to the CDR. A
CDR-grafted MAb contains rodent hypervariable sequences, human framework
sequences, and human constant regions. There is some loss of affinity in CDR-
grafted MAb, but it is usually less than one order of magnitude, and minor
tweaking of the framework regions can often minimize loss of affinity. CDR-
grafted MAb have already been used in therapy in organ transplant immune
suppression (targetting CD3, CD4, or the IL-2 receptor), rheumatoid arthritis
(CD4 and CDw52), Crohn’s Disease (CD4), systemic vasculitis (CDw52),
leukemia and lymphomas (CDw52 and the IL-2 receptor), septic shock (TNFα ),
neoplasm (Lewis-Y, p185HER2 [human epidermal growth factor receptor 2],
placental alkaline phosphatase, carcinoembryonic antigen), and viral infection
(HIV, herpes simplex virus) (Winter and Harris, 1993. Immunol. Today 14:243).
Xenomic mice. A rather exotic form of allogeneic antibody therapy has
been developed against a xenogeneic background. In this case, xenomic
mice are genetically engineered which make human immunoglobulins (Lewin.
1997. J. NIH Res. 9: 31-33). I guess in some ways, animal rights advocates
would be pleased that xenomic mice do not suffer from tumors, but on the other
hand, may be somewhat upset that xenomic mice need to be chronically
exposed to antigen (immunized, usually with inflammatory consequences). It just
doesn’t pay to be a mouse. And of course there is some hysterical concern
regarding the exchange of genetic material from humans to mice. These
considerations aside, presently, there are several xenomic mice. One group
possesses genes for many human variable germline genes (66 out of 95 for the
heavy chain locus and 32 out of 76 for the k light chain locus), and another has
fewer variable region genes but all the constant region genes. In the latter case,
purely human immunoglobulins can be produced in mice. The advantage of
xenomic mice is that, theoretically, one strain of mice may be able to make
polyclonal human antibodies against any number of antigenic challenges, thus
circumventing the need to constantly form new hybridomas to combat new
antigens and providing polyclonal specificity, which may have functional
advantages over monoclonal specificity.
Root Surface
Caries
Pathogenesis. We have discussed primarily the pathogenesis and host
319
 defense of pit/fissure and smooth surface caries. The organisms associated with
cervical caries includes the Streptococcus mutans, but is much more significantly
associated with Gram-positive filamentous organisms, such as Actinomyces
viscosus, A. naeslundii, A. odontolyticus, A. eriksonii, and Rothia dentocariosa
(reviewed in Newbrun E. 1983. Cariology, 2nd Ed. William & Wilkins.
Baltimore, pp50-85).
Neutropenia. Some evidence suggests that at least partially, cervical
caries are controlled by the immune system. As one may suspect, given the
gingival localization of these lesions, it is not the sIgA system, but instead the
complement-IgG-neutrophil axis. The most suggestive evidence that a different
arm of the immune system is involved in the protection against cervical caries is
the rather sporadic observation that root surface caries may be associated with
neutropenia (Mishkin et al., 1976. O. Surg. O. Med. O. Pathol., 42:738-745;
Pernu et al., 1996. J. Periodontol. 67:454-459). If this is true, then is it justifiable
to focus on the sIgA system in caries vaccinations? The answer is “yes,” since
root surface caries is not a major problem in the USA (especially for children...
and really, who cares whether ‘gramps’ gets root surface caries anyway?).
Summary
In this chapter, there are several things which the student should try to learn.
First, innate immunity (including mainly secreted factors) probably plays an
important role in dictating caries resistance; however, pleomorphisms in no one
factor have been associated with dental caries. Second, specific secretory
antibodies conferring a “natural immunity” against dental bacteria follows a clear
time course in the developing infant, with several waves of antibody production
whose specificities match the changes which occur within the mouth, most
notably, the eruption of teeth. Third, we identified the bacteria which were
associated with dental caries, and found that they belonged to an extended
family of “lactic acid bacteria.”S. mutans and S. sobrinus are those bacteria which
are most often associated with smooth surface and pit and fissure lesions. These
organisms possess two attributes which make them pathogenic in the presence
of dietary sucrose -- formation of lactic acid and formation of extracellular glucans.
Fourth, we examined the potential of hyperimmunization against these
cariogenic bacteria. In general, we noted that vaccination resulting in elevation of
IgA may be effective against smooth surface caries, but probably provide little
effectiveness against pit and fissure lesions. We also noted that root surface
caries may be protected by the IgG-neutrophil mechanism, not the secretory
IgA system.
Fifth, the two broad categories of immunization were “active” and “passive.”
Active immunization protocols may require subunit vaccines, in order to avoid the
problem of heart muscle cross-reactivity. Potential subunits were identified and
included glucosyltransferases, adhesins, dextranases, and extracellular
carbohydrate polymers. Oligomeric components (ie., peptides) refine the
specificity of the subunit vaccines. Since peptides are poorly immunogenic, one
aspect of the caries vaccine program has been to examine various adjuvant
strategies (liposomes, for example). Two forms of passive immunization which
may prove useful include maternal and xenogeneic immunization. These
xenogeneic antibodies can be administered to the child as food (mother’s or
cow’s milk, or raw chicken eggs...er, not that I would recommend raw chicken
eggs.... I used to eat raw chicken eggs, though... so it is possible!). Other
xenogenic antibodies include rather exotic “chimaeric” or “CDR-grafted
antibodies”.

SELF-HELP QUESTIONS

320