CLINICAL MICROBIOLOGY REVIEWS, JUly 1989. p. 285-314 Vol. 2. No.

3
0893-8512/89/030285-30$02.00/0
Copyright © 1989, American Society for Microbiology

Streptococcal M Protein: Molecular Design and Biological Behavior

VINCENT A. FISCHETTI
The Rockefeller Universit, New York. Newt York 10021

INTRODUCTION ................................................................................. 286

ELECTRON MICROSCOPY ................................................................................. 286
Thin Sections ................................................................................. 286
Whole Streptococci ................................................................................. 286
ISOLATION AND PURIFICATION ................................................................................. 286
Pepsin, Detergent, and Lysin Extraction ................................................................................ 287
STRUCTURAL ANALYSIS ................................................................................. 288
Evidence for the Coiled-Coil Structure of M Protein ................................................................. 288
Structural Analysis of the Complete M Molecule ...................................................................... 290
Basic Model ................................................................................. 291
Analysis of the Wall-Associated Region ................................................................................. 293
Domain Structure ................................................................................. 293
IMMUNOCHEMISTRY ................................................................................. 293
Cross-Reactions ................................................................................. 294
Conserved, Variable, and Hypervariable Epitopes .................................................................... 294
EVIDENCE FOR TWO DISTINCT CLASSES OF M PROTEIN ................................................... 294
MAP and Opacity Factor ................................................................................. 294
MAP I Antigen and the M6 Protein C-Repeat Domain .............................................................. 295
C-Repeat Domain and Rheumatic Fever ................................................................................. 295
SIZE VARIATION IN THE M MOLECULES ........................................................................... 295
Isolation and Sequence Analysis of Size Mutants ...................................................................... 295
ANTIGENIC VARIATION AND ITS RELATION TO SIZE VARIATION ...................................... 297
Size Variation Leads to Antigenic Variation ............................................................................ 297
Natural Selection of Size Mutants ................................................................................. 297
Variation in the N-Terminal Nonrepeating Segment .................................................................. 298
COMPARISON OF M-PROTEIN SEQUENCES ........................................................................ 298
Alignment of Known M-Protein Sequences ............................................................................. 298
Alignment of the Variable Region of Two Different M Proteins ................................................... 299
HOMOLOGY OF M PROTEIN WITH OTHER MOLECULES OF GRAM-POSITIVE BACTERIA .... 300
Group C and G Streptococci ................................................................................. 300
Staphylococcal Protein A ................................................................................. 300
STRUCTURAL RELATIONSHIP OF M PROTEIN WITH MAMMALIAN PROTEINS ..................... 301
Tropomyosin ................................................................................. 301
Other Coiled-Coil Molecules ................................................................................. 301
IMMUNOLOGICAL CROSS-REACTIVITY BETWEEN M PROTEIN AND
MAMMALIAN PROTEINS ................................................................................. 301
Heart-Reactive Antibodies ................................................................................. 301
Immunodeterminants Shared with Myosin and Other Human Proteins ......................................... 301
Renal Tissue-Associated Epitopes Shared with M Protein .......................................................... 301
Monoclonal Antibodies Cross-Reactive with Coiled-Coil Proteins ................................................. 302
Conformational Determinants ................................................................................. 302
THE emm GENE ................................................................................. 303
Relationship of M-Protein Genes ................................................................................. 303
Conversion of M- to M+ Phenotype ................................................................................. 303
Number of emm Gene Copies ................................................................................. 303
Regulation of the emm Gene ................................................................................. 304
HUMAN IMMUNE RESPONSE TO THE M MOLECULE .......................................................... 304
Immunodominant Region of the M Molecule ...........................................................................304
ANTIPHAGOCYTIC ACTIVITY OF M PROTEIN ..................................................................... 304
N-Terminal Charge Domain ................................................................................. 305
Effect on Phagocytic Cells ................................................................................. 305
Complement ................................................................................. 305
Factor H ................................................................................. 305
Fibrinogen ................................................................................. 306
ROLE OF M PROTEIN IN STREPTOCOCCAL ADHERENCE ................................................... 306
M-PROTEIN VACCINES.306
Type-Specific Protection ................. 306
285
286 FISCHETTI
Non-Type-Specific Protection ......................
CONCLUDING REMARKS ......................
ACKNOWLEDGMENTS ......................
LITERATURE CITED ......................
INTRODUCTION
The group A streptococcus (Streptococcus pyogenes) is
responsible for a number of suppurative human infections, of
which acute pharyngitis and impetigo are the most common.
As a consequence of ineffective antibiotic therapy or no
therapy, as many as 3 to 5% of individuals who suffer a group
A streptococcal pharyngeal infection may develop acute
rheumatic fever (32,204), a disease often resulting in cardiac
damage. While not currently a major problem in developed
countries, rheumatic fever is the leading cause of heart
disease in school-age children in developing nations (1). For
instance, by one estimate, over 6 million school-age children
in India are afflicted with rheumatic heart disease (1). Acute
glomerulonephritis, another sequela of group A streptococ-
cal disease, is usually the consequence of infection by
specific strains of streptococci (nephritogenic strains) which
infect either the throat or skin (174). The ability of group A
streptococci to persist in infected tissues is primarily due to
the cell surface M protein, a molecule which confers to the
streptococcus the ability to resist phagocytosis by polymor-
phonuclear leukocytes in the absence of type-specific anti-
bodies. Resistance to a group A streptococcal infection
appears to be related to the presence of type-specific anti-
bodies to the M molecule (128, 129). Since there are >80
different serotypes of M protein (i.e., M5, M6, M24, etc.), an
individual may become infected by more than one group A
streptococcal type during a lifetime (129).
Streptococcal M protein was identified over 60 years ago
by Rebecca Lancefield (126). A review by Lancefield in 1962
(129) clearly describes the studies carried out for nearly 35
years, defining this molecule as a major virulence factor for
the streptococcus due to its antiphagocytic property. In
1974, a comprehensive review by Fox (81) delineated studies
since the Lancefield review and underscored the knowledge
to that time of the structure, function, and immunochemistry
of the M molecule.
The streptococcal M protein is now probably one of the
best-defined molecules of the known bacterial virulence
determinants. Its structure, function, immunochemistry, and
method of antigenic variation are unique among known
virulence molecules and may serve as a model for certain
microbial systems. To better appreciate the thinking which
led to the current understanding of the M protein, whenever
possible, attempts will be made to present this review in an
historical perspective, with a chronological account of the
studies over the years, emphasizing the time since 1974.
While the antiphagocytic activity and genetics of the M
molecule will be addressed in this review, the reader is also
directed to recent reviews by Manjula (131) and Scott (J. R.
Scott, in B. Iglewski, ed., The Bacteria, Vol. XIII, in press)
on these subjects.
In this review, the basic structure of the M molecule will
be developed. Once established, it then becomes a simple
matter to integrate this information into a better awareness
of the biology of the M molecule and its relation to type
specificity, antigenic diversity, and cross-reactivity. With
these concepts understood, a strategy to control group A
streptococcus-related diseases may become more apparent.
CLIN. MICROBIOL. REV.
307
309
309
309
ELECTRON MICROSCOPY
Thin Sections
Thin-section electron micrographs identifying the M pro-
tein on the surface of group A streptococci were first
published by Swanson et al. (193), using ferritin-labeled
M-specific antibodies derived from type-specific rabbit sera.
The data suggested that M protein appears as hairlike
projections on the surface of the streptococcal cell wall (Fig.
1) (13, 169). M- streptococci were found to be devoid of
these surface projections (193).
Whole Streptococci
To better appreciate the features of the proteins on the
surface of the streptococcal cell wall, organisms were pre-
pared for electron microscopy by a variety of other tech-
niques (68). Whole streptococci mounted on grids, critical-
point dried, and viewed under a transmission electron
microscope exhibit projections which appear to be the result
of interactions among several adjacent M-protein fibrils
forming tuftlike structures (Fig. 2a and b). Interactions
between M molecules on adjacent streptococci, previously
observed in thin sections (Fig. 1) (169), are also seen in these
preparations. Whether these surface structures are com-
posed exclusively of M protein or are the result of complexes
of M protein and lipoteichoic acid (LTA) or other surface
molecules was not determined. However, ferritin-labeled
anti-M antibodies reveal that the extensions do contain M
protein (Fig. 3). Surface replica preparations confirm the
presence of these types of projections on the cell surface of
whole organisms (Fig. 4). Streptococci treated with trypsin
to remove the M protein were shown to have a smooth
surface in all of these preparations (not shown).
In experiments designed to answer questions about the
placement of M protein on the cell wall, Swanson et al. (193)
found that, after trypsinization to remove existing M protein
on living streptococci, newly synthesized M protein was first
seen by electron microscopy on the cell in the position of the
newly forming septum. In similar experiments designed after
the classical experiments of Cole and Hahn and using
fluorescein-labeled anti-M antibodies (44), trypsinized strep-
tococci placed in fresh media for 10 min revealed M protein
first at the newly forming septum (Fig. 5a). On organisms
examined after extended incubation, M protein was not
observed in the position of the old wall (Fig. 5b), suggesting
that the M molecule is produced only where new cell wall is
synthesized (193).
ISOLATION AND PURIFICATION
For many years the method used to prepare crude M
protein for purification was based on the one developed by
Lancefield (126) for the preparation of extracts for strepto-
coccal typing (for references prior to 1974, see Fox [81] and
Lancefield [129]). According to this procedure, organisms
are suspended in dilute HCI at pH 2.0 and placed in a
boiling-water bath for 10 min. After neutralization, the
resulting extract is centrifuged and the supernatant is used as
starting material for further characterization of the M mole-
 of~ ~.
VOL. 2, 1989
,:'
FIG. 1. Electron micrograph of ultrathin sections of group A streptococci exhibiting M-protein fibrils on the cell surface. Junctions
between streptococci reveal M-protein fibers from one coccus interacting with those from the adjoining organism. Magnification. x56,000.
cule. While the procedure effectively removes M protein
from the cell, it is unlikely that the final purified product
represents the native M-protein molecule (82). Recognizing
this limitation, other procedures for extracting M protein
from whole streptococci or their isolated cell walls were
attempted. These included sonic oscillation (25, 157), alkali
(47, 84), group C bacteriophage-associated lysin (43, 79),
nonionic detergent (64, 65), nitrous acid (91), cyanogen
bromide (202), and guanidine hydrochloride extraction (178),
all of which were usually coupled to a particular purification
scheme. However, in nearly all cases the final purified M
protein exhibited extensive heterogeneity. These results led
to the general conclusion at that time that M protein was
composed of a heterogeneous group of molecules or of
multiple subunits (82).
Pepsin, Detergent, and Lysin Extraction
The biochemical features of the M-protein molecule be-
came more apparent with the development of an extraction
procedure which specifically cleaves the M protein from the
streptococcal surface. Pepsin digestion at a suboptimal pH
of 5.8 releases a biologically active fragment of the M protein
(termed PepM) from the cell wall (11, 46). After purification.
the fragment is found to be homogeneous by sodium dodecyl
sulfate-polyacrylamide gel electrophoresis, strongly suggest-
ing that the M molecule on the cell surface was also likely to
be homogeneous, contrary to previous belief. Depending on
STREPTOCOCCAL M PROTEIN 287
the serotype, the size of the extracted PepM molecule varies
from 20,000 to 40,000 in molecular weight (34, 116, 120, 133,
152, 169). When injected into rabbits, the pepsin-derived
fragment elicits antibodies capable of initiating phagocytosis
of streptococci of the homologous serotype. PepM can also
absorb type-specific opsonic antibodies from serum (11,
137). Thus, the pepsin-derived M-protein fragment retains
some of the biologically important determinants of the native
M molecule present on the streptococcal surface. Pepsin
extraction, therefore, proved to be one of the better methods
for obtaining a workable fragment of the M protein and was
used to provide the starting material in a variety of studies
over the ensuing years (14, 18, 22, 117, 133, 137, 152, 158,
169). Although it was believed early on that pepsin removed
the M protein in its entirety from the streptococcal surface
(13), the nature of the complete molecule was not as yet
understood.
Nonionic detergent was found to remove M protein from
isolated streptococcal cell walls (70). The M molecule re-
leased by this method is seen by sodium dodecyl sulfate-
polyacrylamide gel electrophoresis as several closely spaced
protein bands which range from 15,000 to 35.000 in molecu-
lar weight. Like pepsin-extracted M protein, the detergent-
extracted molecules are able to absorb opsonic antibodies
from rabbit antiserum (70) and evoke type-specific opsonic
antibodies when injected into rabbits (64). An examination of
the physicochemical properties of these M molecules re-
288 FISCHETTI CLIN. MICROBIOL. REV.

FIG. 2. Critical-point-dried whole group A type 6 streptococci viewed directly by transmission electron microscopy. (a) M, M-protein
projections on the cell surface. Interactions between cocci seen in Fig. 1 are also observed in these preparations. Magnification, x88,000. (b)
Enlargement of the surface of a coccus from part a. Magnification, x330,000.

vealed that they are not globular but asymmetrical, with a
frictional ratio of 2.2 (70). Thus, the detergent-extracted M
molecules, though not as homogeneous as the pepsin-de-
rived protein, display many of the properties of PepM.
It was not until 1981, when M protein was purified from
the extracellular supernatants of type 12 group A strepto-
coccal L forms and protoplasts, that it became apparent that
the size of the native M molecule could be as large as 58
kilodaltons (kDa) (199). Extraction of M protein from strep-
tococci by solubilization of the cell wall with phage lysin
(termed lysin, a muralytic enzyme produced during a group
C streptococcal phage infection [69]) also yielded a large M
molecule (43, 79); however, the association of this M protein
with cell wall fragments, resulting in its high molecular
weight, could not be ruled out.
STRUCTURAL ANALYSIS
Evidence for the Coiled-Coil Structure of M Protein
Though sequence and peptide data were limited up to
1979, the available information suggested that M molecules
of heterologous serotypes differed in primary structure in
spite of a common antiphagocytic activity (18, 65). How-
VOL. 2, 1989
FIG. 3. Critical-point-dried whole group A type 6 streptococci pretreated with anti-M6 antibody and identified with ferritin-labeled goat
anti-rabbit IgG. Cells were viewed directly by transmission electron microscopy. M. A representative of the M-protein projection on the cell
surface containing the ferritin-labeled antibody. Fibers between cocci also contain bound ferritin particles. Magnification. x96,000.
ever, despite these apparent differences, physical and chem-
ical data accumulated to that time revealed that all of the M
molecules examined shared certain noteworthy properties
(81). Their extreme thermal stability (126), elongated shape
(70, 84, 166), and appearance as a network of fibers on the
cell surface (193) suggested that M molecules resemble
previously characterized fibrous proteins. When the partial
sequence of the PepM24 (terminology used for the M protein
extracted from M24 streptococci with pepsin) protein (18)
was compared with that of known proteins, two of the M24
cyanogen bromide peptides exhibited up to 40%Y identity
with rabbit skeletal muscle tropomyosin (99). In addition, a
hexapeptide, repeated five times in the PepM24 molecule,
was identical to a sequence in tropomyosin. This was the
first direct evidence that M protein may be structurally
similar to tropomyosin, the prototype molecule for an ot-
helical coiled-coil fibrillar protein (145). The isoelectric point
(a reflection of amino acid composition), solubility, heat
stability, and stability in acid and base of tropomyosin were
similar to those found for M protein, further confirming the
chemical and structural relationship of these two molecules
(99).
Additional biochemical support for the probable coiled-
coil structure of M protein came when the partial sequences
of the M5, M6, and M24 proteins were compared with that of
tropomyosin (136). The sequences within these three mole-
STREPTOCOCCAL M PROTEIN 289
cules contained a repeating seven-residue periodicity of
nonpolar amino acids, a basic characteristic of x-helical
coiled-coil proteins like tropomyosin. A more detailed study
of the structure and conformation of the PepM6 and lysin-
extracted M6 proteins further established the close relation-
ship between M protein and tropomyosin (169). In this
analysis, ultracentrifugation studies revealed that the M
molecules exist as stable dimers, and circular dichroism
spectra indicated that the molecules are 70% a-helical.
Electron microscopic images of platinum-shadowed M mol-
ecules were seen as rods in which the pepsin- and detergent-
extracted forms were approximately half the length of the
lysin-extracted protein. Measurements also revealed that the
M molecules were similar in width to tropomyosin, and fiber
X-ray diffraction indicated that a coiled-coil structure is
present in the M molecule. Assessment of M protein on the
streptococcal surface compared with isolated M molecules
indicated that each fiber on the cell wall consists of a single
dimeric M-protein molecule of about 50 to 60 nm in length
(Fig. 1) and is not composed of multimers of smaller subunits
as is the case for pili of gram-negative organisms (33).
Though common in mammalian cells, the x-helical coiled-
coil design is unique for bacterial surface molecules (41, 42).
Studies by Phillips et al. (169) indicated that purified M
protein released from the streptococcal cell wall with phage
lysin was similar to the size of the native cell-wall-bound M
290 FISCH ETTI
.:
%. ..oif:
FIG. 4. Tantalum and tungsten surface replica of critical-point-dried streptococci revealing the distribution of the projections on the cell
surface. Magnification, x96,000.
molecule. M proteins extracted with pepsin are almost half
the size of the lysin-extracted molecule and are likely
cleaved products derived from the native M molecule. This
is further supported by electron micrographs of pepsin-
treated M-bearing (M F) streptococci depicting shorter sur-
face fibers than untreated cells (13, 169). Thus, after pepsin
digestion, roughly half of the M molecule remains associated
with the streptococcal cell.
Structural Analysis of the Complete M Molecule
The completion of the PepM5 sequence (133) allowed for
the first time a thorough analysis of this region of an M
molecule (141), i.e., the distal half of the native protein (169).
The results confirmed what previous partial sequence data
had suggested: except for the N-terminal 11 amino acids, the
remainder of the sequence exhibits a seven-residue periodic-
ity in the distribution of nonpolar amino acids. In addition, it
revealed that the N-terminal region of this fragment contains
a significantly higher net negative charge than the C-terminal
region. While sequence repeats were also identified in the
PepM5 molecule (139), they were not as extensive as those
found in the partial PepM24 sequence (18).
CLIN. MICROBIOL. REV.
.P~.~;
dm4, ..
"k
,f
A.9
cl.
,.24
..
km t:
f
'i .11
i.
Z"
I-W. .0.
A
Until 1986, the majority of structural information was
based on the pepsin fragment of the M molecule. No
information, besides the physicochemical analysis on the
lysin-extracted M protein (169), was available concerning
the C-terminal half. The successful cloning of the M6 gene
(enzin-6) (180) resulted in the ability to determine the com-
plete amino acid sequence of an entire M molecule (94). Like
the partial PepM24 sequence (18, 19), the M6 sequence was
composed of extensive amino acid repeats extending into the
C-terminal half of the protein. Approximately 80% of the
complete M6 molecule is composed of four sequence repeat
regions (Fig. 6) (94). Regions A and B contain five direct
tandem repeat blocks of 14 and 25 amino acids each,
respectively. Each A repeat may be further divided into two
seven-residue sub-blocks (not shown). The three central
repeats in the A and B regions are identical, while the
repeats at each end of the five repeat blocks diverge slightly.
Region C consists of two and one-half tandem repeats of 42
amino acids each, which are not as conserved as the A and
B repeats. Region D is composed of four seven-residue
"quasirepeats" (77, 94). Recent reports show that amino
acid sequence repeats also occur in the complete M24 (153),
VOL. 2, 1989 STREPTOCOCCAL M PROTEIN 291

FIG. 5. Fluorescein-labeled anti-M6 antibody analysis of the appearance of M6 protein on streptococcal cell walls. M6 streptococci were
treated with trypsin to remove surface M protein. Cells, reincubated at 370C, were removed at intervals, fixed, and stained with
fluorescein-labeled anti-M6 antibody. (a) After 10 min of incubation, the M protein is located within a thin band at the position of the newly
forming septum. Magnification, x4,000. (b) Location of the M protein after 40 min of incubation. Note that no fluorescein label is seen in the
position of the old wall. Magnification, x4,800.

M5 (148), and M12 (175) proteins, the M5 molecule being the
least repetitive.
A C-terminal segment, composed of 19 hydrophobic
amino acids, is predicted to have an a-helical conformation
which would be sufficient in length to traverse the cytoplas-
mic membrane (77). An adjacent region, rich in proline and
glycine, is predicted to be located within the cell wall
peptidoglycan (161).
Basic Model
In general, x-helical coiled-coil proteins are constructed
from a repeating seven-residue amino acid pattern (a-b-
c-d-e-f-g)n in which residues in positions a and d are hydro-
phobic and form the "core" residues in the coiled-coil and
the intervening residues are primarily helix promoting (145).
The arrangement of the amino acids of the M6 protein within
the seven-residue pattern is shown in Fig. 7. The seven-
residue periodicity found from amino acids 12 through 362
strongly suggests that this region is in a coiled-coil confor-
mation and forms the helical central rod region of the
molecule. Discontinuities in the heptad pattern (seen espe-
cially in repeat B), which have been seen in the sequence of
other M proteins (141) and other coiled-coil molecules (144,
163, 166), probably account for the flexibility of the M
molecules observed in electron micrographs (Fig. 1). Based
on these irregularities in the heptad pattern, the central rod
region is divided into three subregions which correlate with
the A-, B-, and C-repeat blocks (77).
Based on a detailed analysis of data of M-protein se-
quences, it was concluded that serologically different M
proteins may be built according to a basic scheme: an
extended central coiled-coil rod domain flanked by func-
tional end domains (77). Because the Q-helical coiled-coil is
a structure capable of tolerating a vast number of amino acid
sequence changes while preserving its conformation, a large

Wall-associated
Non Helical
-i
Region Helical Central Rod Region Anchor Region
u
L-LAAAAAAAAAAAA~~~~~~~~~~~~~~~nAA~~~~~nAAAAAAA~~~~~~~~nAAAAAA
--]ml &A &B &B kB &B &B ILA "%,A ILmmmmmwwuww Li W li w UULJ

14 25 7
.I
42
I
N 4- Aj4 B2 B3 B 4B-d;_ C-|- Pro/Gly | Memb. Sc
........... ~~~... ......

Pepsin
P
t tail
Charged
FIG. 6. Characteristics of the complete M6 protein sequence (77, 94). Blocks A, B. C, and D designate the location of the sequence repeat
blocks. Numbers above the block indicate the number of amino acids per block. Block C3 is half the size of blocks C1 and C2. Shadowed
blocks in A and B indicate those in which the sequence diverges from the central consensus sequence. In the C repeats, the sequence in the
center of C1 and C2 are identical, while the sequence at the ends and block C3 diverge. Pro/Gly denotes the proline-glycine-rich region likely
located in the peptidoglycan. Membrane anchor is a 19-hydrophobic-amino-acid region adjacent to a 6-amino-acid charged tail. Pepsin
identifies the position of the pepsin-susceptible site after amino acid 228. Wall-associated region designates the segment of the molecule
located within the cell (this begins at Ala-298 in the C2 block [161]).
292 FISCHETTI CLIN. MICROBIOL. REV.

I Arg Val Phe Pro Arg Gly Thr I NON HELICAL

Val Clu Asn Pro REGION NH2

a b e d e r 9 tr
12 Asp Lyo Ala Arg Glu Leu
Is Leu Ann Lys Tyr Asp Val Glu
25 Aon Ser Met Leu Gin Ala Asn T
32 Ann Asp Lys Leu Thr Thr Glu
39 Ano Ann Aon Leu Thr Asp Gin w
46 Ann Lys Ann Leu Thr Thr Glu iii
5:3 Ann Lys Ann Leu Thr Asp Gin T Hypervariable I
60 Ann Lys Ann Leu Thr Thr GCIu
n
67 Ann Lys Asn Leu Thr Asp Gin 4
74 Ann Lys Ann Leu Thr Ther Gl u
8I Ann Lys Glu Leu Lys Ala Glu
88 Glu Ann Arg Leu Thr Thr Glu
95 Ann Lys Gly Leu Thr Lys Lys
102 Leu Ser Glu Ala Glu Glu Glu
109 Ala
Ito
117
Ala Ann Lys Glu Arg
Lys Glu Aa lIle Gly
Glu Ann
Thr Leu
Vorioble
124 Lys Lys Thr Leu Asp Glu Thr
131 Val Lys Asp Lyn
135 Ile Ala Lys Glu Gin Glu Ser rIn
142 Lys Glu Thr Ile Gly Thr Leu r
149 Lys Lys Thr Leu Anp Glu Thr mn >
156 Val Lys Asp Lyn r0
160 Ile Ala Lys Glu Cln Glu Ser 1n
167 Lys Glu Thr Ile Gly Thr Leu iii
174 Lys Lys Thr Leu Asp Glu Thr w Z
181 Val Lys Asp Lys
185 Ile Ala Lys Glu Gin Glu Ser
192 Lys Glu Thr Ile Gly Thr Leu
199 Lys Lys I le Leu Asp Glu Thr 0
206 Val Lys Asp Lys o
210 Ile Ala Arg Glu Gin Lys Ser
217 Lys Gin Asp Ile Gly Ala Leu i
224 Lyns Gin Glu Leu Ala Lys Lys_n
231 Asp Glu Gly Z
Z
Z-Pepsin
234 Asn Lys Val Ser Glu Ala Ser
241 Arg Lys Gly Leu Arg Arg Asp
248 Leu Asp Ala Ser Arg Glu Ala
255 Lys Lys Gin Val Glu Lys Asp
262 Leu Ala Ann Leu Thr Ala G
Clu
269 Leu Asp Lys Val Lys Glu Glu 15
276
283
Lys Gin Ile
Arg Gin Gly
Ser Asp Ala
Leu Arg Arg
Ser
Asp
1D Conserved
>
290 Leu Asp Ala Ser Arg Glu Ala 4
297 Lys Lys Gin Val Glu Lys Ala Cs
304 Leu Glu Glu Ala Ann Ser Lys
311 Leu Ala Ala Leu Glu Lys Leu
318 Ann Lys Glu Leu Glu Glu Ser
325 Lys Lys Leu Thr Glu Lys Glu
332 Lys Als Glu Leu Gin
337 Ala Lys Leu Clu Ala Clu
343 Ala Lys Ala Leu Lys Glu Gin o
350
357
Leu Ala Lys
Leu Ala Lys
Gin Ala Glu
Leu Arg Ala
Glu Group
Carbohydrate
363 G4y Lys Ala Ser Asp Ser Gin Thr
371 Pro Asp Ala Lys
375 Pro Gly Ann lys Val Val
331 Pro Gly Ly s Gly Gin Ala Z WALL
387 Pro Gin Ala Gly Thr Lys C DOMAIN
393
369
Pro Asn Gin Asn lys Ala
Pro Met Lys Glu Thr Lys Arg Gin leu
0 Peptidoglycan
408 Pro Ser Thr Gly Glu Thr Ala Asn Pro i
417 Phe P'he Thr Ala Ala Ala leu T
MEMBRANE
Membrane
424 Thr VaI %let Ala Thr Ala Gly
431 Val Ala Ala \'aiI'ai _
COOH
436 Lys Arg Lys CGlu Glu Asn CHARGED
TAIL FIG. 8. Proposed model of the M6 protein from strain D471
FIG. 7. Complete M6 protein sequence arranged to highlight the based on current sequence and structural data (77, 94). The coiled-
seven-residue periodicity found in the helical central rod region. coil rod region extends about 50 nm from cell wall with a short
Region assignments are based on sequence and conformational nonhelical domain at the N terminus distal from the cell surface. The
analyses (77). Arrangement of the sequence is based on the position proline-glycine-rich region of the molecule is located within the
of amino acids in a seven-residue periodicity designated by letters a peptidoglycan, with a short segment of the coiled-coil rod region
through g beginning at residue 12 and continues, with interruptions positioned in the group carbohydrate portion of the wall (161). The
at residues 109, 131, 156, 181, 206, 231, and 337, through residue membrane anchor segment extends through the cell membrane, with
362. Alignment from residues 363 to 416 is used essentially to the charged tail projecting into the cytoplasm. Regions containing
highlight the regularity of the position of prolines in the sequence. conserved, variable, and hypervariable epitopes among heterolo-
No periodicity is found from residue 363 to the end. Three major gous M serotypes are indicated.
regions are indicated (nonhelical, helical, and anchor). The region
buried within the cell wall is delimited by the wall domain (161). The
pepsin-susceptible site is between Ala-228 and Lys-229. Because 3.6 residues are necessary to achieve one turn of
the helix (102), hydrophobic residues at positions a and d in
the heptad (Fig. 7) will form an inclined stripe around the
number of serologically distinct streptococcal M proteins helix. For thermodynamic stability, a dimeric, ropelike
may be constructed according to this basic scheme. Al- superstructure would be formed as a result of the internal-
though the sequence of the heptad may differ between M ization of the hydrophobic residues from two helical chains
types, even among strains of the same type, the presence of coiling around each other (75, 136, 145) (Fig. 8). The
predominantly helix-promoting residues will allow the dif- presence or absence of disruptions in the periodicity of
ferent molecules to attain a common cx-helical conformation. hydrophobic residues at positions a and d of the heptad
VOL. 2, 1989
allows these regions to become more or less flexible (detailed
in reference 77).
A short 11-residue nonhelical region is found at the
N-terminal end of the coiled-coil rod region, which is distal
to the cell surface (Fig. 8). In all M proteins examined thus
far, except M49, which lacks this segment (116), the se-
quence of this region is distinct from one M type to another
(94, 133, 148, 153, 175) but is nearly 100% identical within an
M type (95). It is not certain whether the tertiary conforma-
tion of this N-terminal region is similar among the different
M-protein types despite sequence variation. The N-terminal
segment plays an important role in the biological activity of
the molecule, since antibodies generated to this region
effectively opsonize the streptococcus of the M type from
which it was derived (12, 15, 34, 55, 57, 106, 107) (see
below). Thus, it seems likely that one purpose for the central
rod region may be to act as a shaft to position the N-terminal
region away from the cell surface (Fig. 8).
Beginning at the C-terminal end of the central rod region is
a segment characterized by the presence of regularly spaced
prolines and glycines (Fig. 7). This is followed by two
phenylalanine residues beginning at residue 417 which signal
the start of a 19-amino-acid stretch of hydrophobic residues
terminated by six charged amino acids. It was suggested that
the hydrophobic amino acids would serve as a membrane
anchor and the adjacent proline-glycine region would be
located in the peptidoglycan and act to stabilize the molecule
within the cell wall (77). The charged tail, typical of a stop
transfer signal seen in the membrane-associated segments of
several membrane-bound proteins (178a), would be expected
to be located on the cytoplasmic side of the membrane.
Sequence comparison of the C-terminal region of the M6
protein with the C-terminal end of other cell wall proteins
from gram-positive bacteria reveals significant homology
(see below), suggesting that a common mechanism may exist
for the attachment of proteins within the gram-positive cell
wall.
Analysis of the Wall-Associated Region
From the translated deoxyribonucleic acid (DNA) se-
quence of the M6 C-terminal region, it was implied that this
segment may be responsible for M-protein attachment (94);
however, it was not directly determined. Experiments were
therefore designed to isolate and identify the M-protein
region embedded within the streptococcal cell wall. Trypsin
was used to digest the exposed portion of M protein from the
cell wall, leaving behind the wall-associated region. After the
cell wall was solubilized with a muralytic enzyme, antibodies
to synthetic peptides of sequences within the C-terminal
region of the M molecule were used to identify the released
wall-associated M-protein fragment. A 16-kDa fragment was
identified as the smallest trypsin-resistant fragment of M
protein. Because of its location within the cell wall matrix,
this fragment was resistant to attack by trypsin despite the
presence of several lysine and one arginine residue within
the sequence. Amino acid composition of this fragment
revealed that it was rich in proline and glycine but lacked the
amino acids found in the mebrane anchor. Sequence analysis
of the purified fragment indicated that its N-terminal end was
at about residue 300, thus localizing a small segment of the
helical central rod region of the M molecule (residues 298 to
362) in addition to the proline-glycine region (residues 363 to
416), within the cell wall (Fig. 7).
Further evidence suggested that residues 298 to 362 of the
helical rod region are located within the group carbohydrate
STREPTOCOCCAL M PROTEIN 293
moiety of the cell wall, localizing the proline-glycine region
within the peptidoglycan. Since no detectable amino sugars
were found with the wall-associated M-protein fragment, it is
likely that the C-terminal region of the M6 molecule is
intercalated within the group carbohydrate and cross-linked
peptidoglycan and not covalently linked to them (161).
M protein is rarely detected in the spent growth media of
streptococci. Therefore, the incorporation of this molecule
into the cell wall and membrane appears to be closely
regulated. It is postulated that the C-terminal membrane
anchor may be necessary to keep the M molecule attached in
the right orientation within the cell membrane until such time
as the peptidoglycan is cross-linked (156) around the proline-
glycine region. Recent studies (V. Pancholi and V. A. Fis-
chetti, submitted for publication) have revealed that at pH
5.5 in the presence of 30% raffinose, the streptococcal cell
wall may be removed with a muralytic enzyme, allowing
nearly all the M protein to remain attached to the protoplast
membrane. Shifting the protoplasts to pH 7.4 results in the
release of the M molecule into the supernatant. A mem-
brane-bound thiol enzyme was found to be responsible for
the release. The biological significance of such an enzyme is
currently speculative. Since the released form of the M
molecule lacks the C-terminal 19 hydrophobic amino acids
and charged tail (161), its attachment to the cell appears to be
through the cytoplasmic membrane.
Domain Structure
Techniques using limited proteolysis have been effective
in defining functional and structural domains in a variety of
proteins (215), particularly immunoglobulins (88, 167, 185).
Similarly, the PepM fragment of the M protein may be
regarded as a functional domain and the pepsin-susceptible
site may be seen as a hinge region of the molecule. This
preferential site for pepsin at a suboptimal pH is found
between the B- and C-repeat blocks in M5 (147), M6 (94),
and M24 (153) proteins (Fig. 6) even though the B-repeat
block in M24 is not homologous with that found in M5 and
M6. PepM49, however, has a primary pepsin site after a
C-repeat block and a secondary site between the B and C
repeats (135). Thus, since the type-specific and antiphago-
cytic determinants are located within the PepM fragment,
this half of the M molecule may be considered to be a
functional domain of the protein. However, besides the wall
and membrane anchor domains, the functional attributes of
the C-terminal domain comprising the C- and D-repeat
blocks have yet to be clearly defined.
Thus, the streptococcus uses the coiled-coil design to
accomplish a variety of tasks. The rodlike conformation
allows the functional antiphagocytic domain to be positioned
away from the cell wall. The plasticity of the sequences
necessary to form the coiled-coil structure allows the dif-
ferent domains of the molecule to be tailored for specific
functions.
IMMUNOCHEMISTRY
The M-protein molecule is the basis of the Lancefield
serotyping scheme of group A streptococci (126, 177, 195).
The typing sera used for this purpose are a set of M-
protein-specific rabbit sera prepared against whole organ-
isms of specific serotypes. To remove cross-reactive anti-
bodies, sera are extensively absorbed with organisms of
selected heterologous serotypes. Nearly 80 to 85 different
serotypes have now been identified, but there exist a large
294 FISCHETTI
number of strains in streptococcal collections for which
typing sera are not as yet available. The identification of a
specific strain as being a particular serotype requires that its
M-protein extract react in a precipitation reaction with a
specific typing serum. Therefore, these procedures tend to
identify isolates which have not changed immunodetermi-
nants responsible for the type-specific precipitation reaction
with absorbed typing sera (177, 195).
Cross-Reactions
Immunochemical studies prior to 1985 were confined
primarily to either acid- or detergent-extracted M protein or
to the N-terminal pepsin fragment. Analysis of the reactivity
of antibodies raised to detergent-extracted M6 protein with
peptides representing nearly the complete M6 protein re-
veals that the detergent-derived M molecules (DetM) repre-
sent the N-terminal half of the native M6 protein (78; V. A.
Fischetti, unpublished data) and thus are similar to the
pepsin-derived molecule. It is speculated that the nonionic
detergent activates an enzyme within the cell wall prepara-
tion which cleaves the M molecule in the region of the
pepsin-susceptible site.
Acid extracts likely contain fragments of the M protein
derived from both N- and C-terminal regions. Shared anti-
genic determinants among M serotypes have been described
(81); however, based on the complexity of the acid M
preparation, the nature of the observed cross-reactions was
not clearly understood. One hypothesis was that cross-
reactive determinants were actually cellular contaminants
associated with or linked to the M protein (i.e., M-associated
protein [MAP]) (211, 212), whereas others believed that
some determinants of the M molecule were shared among
certain types (80, 83).
It was clear that immunization of rabbits with the PepM
fragment resulted in antibodies which were essentially type
specific and opsonic and were free of extensive reactivity to
non-type-specific epitopes found previously with acid-ex-
tracted M proteins (21). Peptide map analysis of M protein
(derived from the N-terminal half of the molecule) from
cross-reactive and non-cross-reactive serotypes revealed
that cross-reactions correlated with the extent of structural
similarity among the M molecules analyzed (65). Limited
cross-reactions were observed when 125I-labeled M6 protein,
representing the N-terminal half of the molecule, was used in
a radioimmunoassay with unabsorbed rabbit antiserum pre-
pared to heterologous M serotypes (64). Using monoclonal
antibodies, Dale and Beachey (52) found that some antibod-
ies cross-reacted with purified PepM proteins from certain
other serotypes; however, the location of these epitopes was
not defined. Thus, it was clear that, while cross-reactive
epitopes were present in the N-terminal half of the M
molecules, they were apparently limited in scope.
Conserved, Variable, and Hypervariable Epitopes
Using polyclonal rabbit antisera to M6 protein, Manjula et
al. (132) found cross-reactivity with the C-terminal region of
the PepM5 molecule (located near the center of the native
molecule), whereas no reactivity was observed with the
N-terminal region of the PepM5 molecule. The results sug-
gested that the N-terminal 25% of the native M molecule
may represent a hypervariable region of M proteins. In
support of this, Dale et al. (55, 57) established that synthetic
peptides derived from the N-terminal sequence (residues 1 to
20) of the M5 molecule stimulated type-specific antibodies
CLIN. MICROBIOL. REV.
which did not cross-react with M protein from other sero-
types. This was later proven also to be true for M6 (15, 106),
M19 (34), and Ml (120) proteins.
Monoclonal antibodies directed to the N- and C-terminal
halves of the native M6 protein were used by Jones et al.
(109) in an attempt to identify the location of the epitopes
accountable for limited and broad cross-reactions among M
serotypes. Results revealed that an epitope responsible for
cross-reactions among 5 of 56 serotypes was located in the
N-terminal half of the molecule (within the B repeat) while a
highly cross-reactive epitope (reactive with 20 of 56 sero-
types) was located in the C-terminal half (within the C
repeat) (Fig. 6). These studies were extended with two other
monoclonal antibodies, one reacting with an epitope at the
N-terminal end within the A-repeat region and one in the
C-terminal half within the C-repeat (106, 108). The C-
terminal antibody cross-reacted with 30 different serotype
strains, while the new N-terminal monoclonal antibody was
found to be type specific and reacted only with the M6
serotype strain. This monoclonal antibody was the only one
of the four studied that was opsonic for M6 streptococci.
Thus, it is clear from a variety of immunological tech-
niques that the C-terminal region of the M molecule is
antigenically more conserved among different serotypes. As
the N-terminal region is approached, the greater the anti-
genic variation (Fig. 8). Therefore, the basis of the M-protein
typing scheme is now more clearly understood. In the
preparation of typing sera, absorption of antistreptococcal
antisera with heterologous strains was necessary to remove
the antibodies reactive with the conserved regions of the M
molecule, in addition to antibodies to other common anti-
gens found on the cell surface. This resulted in what would
be expected to be a monospecific antibody population,
directed to the type-specific N-terminal region of the M
molecule (Fischetti, unpublished data).
EVIDENCE FOR TWO DISTINCT CLASSES OF
M PROTEIN
MAP and Opacity Factor
MAP were identified by Widdowson and co-workers in the
1970s (211, 213). The MAP antigens are present in acid
extracts derived from streptococci which express M protein.
These antigens lack type specificity but copurify with the
type-specific M protein. At the time MAP antigens were
characterized, other investigators had identified non-type-
specific M antigens in acid extracts which are similar to MAP
(83, 201). It was not until M protein was obtained by pepsin
digestion of whole streptococci that the type-specific sub-
stance could be clearly separated from non-type-specific M
antigens (46).
The MAP derived from one serotype shares antigenic
determinants with MAP derived from many heterologous
serotypes. Two serologically distinct MAP antigens, MAP I
and MAP II, have been identified, and most isolates of a
particular M serotype contain either MAP I or MAP II
antigen on their surface (211, 214). Expression of MAP II is
closely linked to production of opacity factor (OF), a li-
poproteinase whose activity is detected by serum opales-
cence; streptococci bearing MAP I lack OF activity. The
MAP I antigen is present on the surface of nearly all M
serotypes associated with outbreaks of acute rheumatic
fever. Thus, a link between MAP antigen and the virulence
properties of group A streptococci is suggested. However,
the biochemical composition of MAP I and II remained
elusive.
VOL. 2, 1989
MAP I Antigen and the M6 Protein C-Repeat Domain
In a recent antigenic analysis of surface-exposed M-
protein epitopes, it became apparent that two basic M-
protein structures exist (28). A panel of seven antibodies
directed to well-defined epitopes located in the B-repeat,
pepsin site, and C-repeat regions of M6 protein was tested
for immunoreactivity to over 130 streptococcal isolates
representing more than 60 different serotypes. Isolates of
only four serotypes (M5, M14, M19, and M36) shared
antigenic determinants located in the B-repeat and pepsin
site regions of M6 protein. In contrast, about half of the
serotypes tested displayed strong immunoreactivity with
antibody probes directed to C-repeat region epitopes of M6
protein. There is a striking correlation between those sero-
types possessing the C-repeat region epitopes of M6 and
those previously described as having MAP I antigen (211).
Furthermore, nearly all isolates which produce OF (and by
definition express MAP II antigen) lack immunological
cross-reactivity with antibodies to the C-repeat region. It is
likely that the MAP I antigen represents the C-repeat region
as defined for M6 protein. While it is possible that the MAP
I antigen is located on another related surface molecule, the
evidence that it is M protein is significant (28, 211, 213). In
summary, the MAP I and MAP II M types closely parallel
two major structural classes of M-protein molecules, termed
classes I and II, respectively. Streptococci bearing class I M
proteins share surface-exposed antigenic epitopes with the
C-repeat region of M6 protein and fail to produce OF,
whereas class II serotypes produce OF but are deficient in
epitopes cross-reactive with the M6-like C-repeat region
(28).
The close proximity of the highly conserved C-repeat
region to the poorly homologous B-repeat and pepsin sites
strongly suggests that the C-repeat region forms a distinct
antigenic domain (28) (Fig. 6). In support of the immuno-
chemical analysis, structural data suggest that the C-repeat
antigenic domain may have unique conformational charac-
teristics as well. The B-repeat region through residue 230
forms a more flexible coiled-coil structure than the C-repeat
region beginning at residue 231 (77). In addition, amino acid
sequence identity between maximally aligned sequences of
M6 (94) and M12 (175) is <27%Y between residues 121 and
231, but abruptly increases to 97% complete identity be-
tween residues 232 and 298 (extending from the beginning of
the first C-repeat subblock to the perimeter of the cell wall).
(Fig. 7). Thus, the data strongly suggest that the amino-
terminal end of the conserved class I M-protein antigenic
domain lies at the beginning of the first C-repeat block (28).
Class I and II M proteins appear to lack immunological
cross-reactivity in their surface-exposed regions, however.
antigenic relatedness between class I and II M proteins can
be demonstrated in their extracted forms (28). Whether the
shared antigenic epitopes reside in buried portions of the
M-protein molecule remains to be established. The degree of
homology between M-protein genes of different serotypes
has been addressed in nucleic hybridization studies. By
using DNA probes corresponding in positions to the C-
repeat region and to the cell wall and membrane anchor
regions of M6, a class I M protein (147, 182). distinctions can
be made between class I and II serotypes (28). Hybridization
to chromosomal DNA of most class II serotypes is observed
only under conditions of low stringency, whereas class I
serotypes hybridize under high-stringency conditions. These
findings suggest that there are fundamental differences be-
tween class I and II M proteins at the DNA level (28).
STREPTOCOCCAL M PROTEIN 295
C-Repeat Domain and Rheumatic Fever
Class I streptococcal serotypes are represented among
both skin and nasopharyngeal isolates. Likewise, both throat
and skin strains are found among the OF-producing class II
serotypes. Thus, the conserved C-repeat domain does not
appear to be involved in determining the tissue site of
infection. In addition, serotypes associated with acute glo-
merulonephritis are represented by both class I and II
streptococci. However, there is a strong correlation between
certain serological M types and outbreaks of acute rheumatic
fever (30). Of 16 streptococcal serotypes which have been
associated with one or more major outbreaks of rheumatic
fever, all but one type express the C-repeat region epitope
which defines class I M protein (28). Rheumatic fever
patients have high complement-fixing titers to MAP I antigen
(211). and MAP I-specific antibodies are absorbed by human
heart tissue (213). Thus, the surface-exposed conserved
repeat domain of class I serotypes may be a virulence
determinant for rheumatic fever (28).
SIZE VARIATION IN THE M MOLECULES
In this and following sections, a direct comparison of the
known M-protein sequences is presented. But, to fully
appreciate the sequence diversity found within M molecules,
it is important to understand a recently observed property of
the M-protein molecule, namely, size variation. By using a
broadly cross-reactive monoclonal antibody as a probe, the
size of lysin-extracted M protein from a number of strepto-
coccal strains was examined by Western blot (immunoblot)
(74). M protein derived from streptococcal isolates of 20
different serotypes exhibited variation in size ranging from
41 to 80 kDa, depending on the strain. Similar size variation
is observed when different M6 streptococcal strains isolated
over a period of 40 years are examined in this way. This
variation in size may be explained by the observation that
the M sequence contains extensive repeats at the DNA level
(94). Long reiterated DNA sequences are likely to serve as
substrates for recombinational events or for replicative
slippage (62, 192), generating deletions and duplications
within the M-protein gene and leading to the production of M
proteins of different sizes.
Isolation and Sequence Analysis of Size Mutants
To test this hypothesis, it was necessary to isolate clonally
related strains of M-protein size mutants in the laboratory,
so that the parent and mutants could be compared. A
Western blot system was devised so that M-protein size
mutants could be detected and isolated if they comprised at
least 5% of the culture. From type 6 strain D471, about 1 in
2,000 colony-forming units (CFU) were found to produce a
smaller size M protein, a rate that seemed too high to be the
result of spontaneous mutation (71). Analysis of the DNA
sequence of the M6 genes of four spontaneous M-protein
size mutants confirmed that they arose by recombinational
events between reiterated DNA sequences within the struc-
tural gene encoding M protein (95). Each of the mutants
bearing smaller-size M protein had a deletion of one or more
of the A- or B-repeat blocks (Fig. 9).
Comparison of aligned sequences of the M6 size variants
indicates that extensive changes can occur within this one
strain. While the region on the C-terminal side of the last
C-repeat sub-block remains highly conserved (95), the A-
and B-repeat segments have undergone duplications and
296 FISCHETTI CLIN. MICROBIOL. REV.

Cell-Associated

Non-Helical Anchor Region

Region Helical Central Rod Region I I
NV .o 0 0 04
1

M6.1 N Cl C2 C3 |I| Pro/Gly3 Anchor FC

M6.2 N Al A2 A3 B Cl C2 C3 1DD1314 Pro/Gly |ncr FC

M6.3 N A Cl C2 C3 DIDID Pro/Gly |emnhb.f .C

M6.8 N A Cl C2 C3 2i3 4 Pro/Gly | Anchor

FIG. 9. Alignment of M6 sequences of four different M6 strains (95). M6.1 from strain D471 is the prototype M6 protein. M6.2 is from a
deletion mutant derivative of M6.1 in which two A-repeat blocks have been deleted. M6.3 is a deletion mutant derivative of M6.2 in which
one B repeat has been removed. M6.8 from strain M01015 is a natural M6 isolate from a pharyngitis outbreak in the 1980s. This strain differs
from the prototype M6.1 in the nearly complete loss of A repeats. Region assignments are as those described in the legends to Fig. 6 and 7.

deletions (Fig. 9). The size mutant derivative of the M6.1

parent, M6.2, deleted two A-repeat blocks, while mutant
M6.3, a derivative of M6.2, was unchanged in the A-repeat
blocks but is deleted for one B repeat. However, in spite of
the extensive variations in the number of repeat blocks,
these strains were still typed as M6.
Size variation also occurred in M6 strains isolated from
single localized outbreaks of pharyngitis as well as in serial
weekly isolates from the throats of individual patients in the
1940s, prior to penicillin therapy (Fig. 10) (74, 95). Based on
DNA sequence analysis, the size variation of six clinical M6
isolates is the result of recombination events within the
repeat sequences (95), as was found with the mutants
isolated in the laboratory. For example, when compared
with M6.1, M6 strain MO1015 producing M protein M6.8
(isolated during an M6 outbreak in the 1980s) has only one
and one-half A-repeat segments (Fig. 9), whereas M6.9
(isolated in the 1940s) has five A repeats (two of which differ
in size from the others) and an extra C-repeat segment (95)
(data not shown). Sequence analysis of the M6 genes also
indicated that the isolates from each clinical group are
clonally related and not the result of separate acquisitions of
unrelated M6 strains during the course of the infection (95).
When the streptococcal strains isolated weekly from the
throats of 15 patients were examined for the size of the M
molecule (see Fig. 10 for an example of two such patients),
8 patients harbored streptococci which exhibited changes in
M-protein size but not serotype during the course of the
infection. The observation that size variants occur among
clinical isolates suggests that the size mutant constituted the
major organism in the streptococcal population at the time of
isolation and suggests that a given size mutant has a selective
advantage under clinical conditions.
Thus, it is quite clear that variations can occur in both the
number and size of the repeat blocks within the M-protein
sequence. Even within an outbreak resulting from a single
infecting strain, the M protein from the progeny of the
Is olIa te Isolate
2 34 5 67 2 3 4 5 6 7
I"
.::::.:: (W at. A* Oa j*
!I a *iA
i. a
GL53 GL44
FIG. 10. Western blot of lysin extracts of serial streptococcal
isolates from two patients with pharyngitis illustrating a change in
the size of the M protein. Lanes represent extracts of weekly
streptococcal isolates from the time the patients presented with
pharyngitis (week 1). Seven weekly isolates of M17 streptococci
were obtained without patient treatment in 1942. After week 3,
patients are usually asymptomatic and are considered to be strep-
tococcus carriers. The strains were grown and adjusted to compa-
rable optical densities, and the M protein was extracted with phage
lysin, which solubilizes the cell wall releasing M protein. After
sodium dodecyl sulfate-polyacrylamide gel electrophoresis and
Western blot, the nitrocellulose was probed with a monoclonal
antibody reactive with M protein. The blots represent the typical
multiple banding pattern of M protein extracted by this method (74).
It can be seen that in patient GL53 the size of the M protein
produced by the streptococci was the same to the fifth weekly
isolate. At week 6, the streptococci isolated from this individual
synthesized a smaller M protein which changed again in week 7
(though all still typed as M17). In patient GL44, the size of the M17
molecule remained the same for 4 weeks, at which time the
population of streptococci isolated produced a smaller M molecule
(weeks 5 to 7). At week 7, the amount of M protein on these
organisms was reduced, which is typical of organisms isolated in the
carrier state.
VOL. 2, 1989
[ Al ]C A2 ][
M6.1 ....MLQANND KLTTENNNLTDONKNLTTENKNLTDQNKNLTTENKNLTDQNKNLTTENKELKAEENRLTTENK....
M6. 1 ....MLQANND NLTTENKNLTDQNKNLTTENKELKAEENRLTTENK....
[ KKK l] A4 IL A5 ]
L N
T Q A3
Al T D
NE TL
N N
N KiN
L N
TD E
0
N L
TT
KN
A2
L Al/ A3 ]L A4
M6.5....MLQANND NLTTENKNLTDQNKNLTTENKELKAEENRLTTENK....
FIG. 11. Alignment of the amino acid sequence of the A-repeat region of the M6.1 protein and a deletion mutant derivative, M6.5 (74, 95).
The deletion of two A-repeat blocks in the emm-6.5 gene, which likely arose by a recombination event from the center of the divergent Al
repeat block to the center of the consensus A3 repeat block, has generated a new sequence for this M6 protein mutant (107); the sequence
DK (Asp-Lys) was converted to DN (Asp-Asn).
infecting streptococcus isolated from infected individuals
exhibits size differences.
ANTIGENIC VARIATION AND ITS RELATION TO SIZE
VARIATION
Antigenic diversity of surface antigens is a mechanism by
which successful pathogens have been able to survive in the
environment after infecting immunocompetent hosts (7, 45,
146, 194). In bacteria, the best-studied variable antigens are
the gonococcal pilin (146, 194) and the Borrelia variable
major protein (7). M protein differs from these other anti-
genically variable surface proteins in that they undergo gene
conversion events involving extragenic recombination be-
tween multiple genes, whereas the single copy of emm-6 in
streptococci (see below) must involve intragenic recombina-
tion events for changes to result.
Clues as to the mechanism of antigenic variation of the M
protein are based on two previously mentioned observa-
tions: size variation of the M molecule (74) and the presence
of repeat sequences within the M-protein gene (94). Based
on these observations, it was hypothesized that changes in
the amino acid sequence may arise as a result of recombi-
nation between inexact repeats. Because the end repeat
blocks within the A- and B-repeat regions of the M6 se-
quence diverge slightly from the consensus sequence of the
three internal blocks, recombination between one of these
terminal blocks and any other central block can generate a
new sequence. These changes may then be amplified by
recombination events within these repeated sequences. Se-
quences involved in antigenic determinants may be dupli-
cated or in some cases deleted by such recombination events
(95). Thus, intragenic recombination events would be ex-
pected to accelerate the rate at which alterations accumulate
in the M-protein gene. Such recombination events were
found to have occurred in the A- and B-repeat regions of
deletion mutants isolated in the laboratory from strain D471
(Fig. 9). In mutant M6.5, derived from M6.1, a deletion of
the equivalent of two repeat blocks occurred from a recom-
bination event beginning from the center of the divergent Al
STREPTOCOCCAL M PROTEIN 297
A3 ][ A4 1E A5 ]
A5 ]
repeat block to the center of the consensus A3 block. This
resulted in a sequence change from an Asp-Lys (D-K) to an
Asp-Asn (D-N) within this mutant (107) (Fig. 11).
Size Variation Leads to Antigenic Variation
To test the hypothesis that size changes in the emm-6 gene
can lead to antigenic variation of the M protein, monoclonal
and polyclonal antibodies directed to defined epitopes on the
parental M6 molecule were used in competition assays with
purified M protein isolated from the parental M6 strain and
size mutants (107). The results revealed that antigenic
changes occurred in epitopes located at or near the deletions
within mutant M6 proteins, but not at more distant epitopes.
The studies also revealed that an M6 monoclonal antibody
opsonic for the parental M6 strain lost its ability to opsonize
one of the M6 size mutants, supporting the view that size
changes may also result in changes in antigenic determinants
responsible for streptococcal survival in an immune environ-
ment. Thus, antibodies originally produced in response to
sequences present in the parental M6 molecule were unable
to bind to mutant M6 protein or opsonize certain M6 size
mutants. It was discovered that, in some size mutants in
which a direct sequence change had not occurred, changes in
conformation were predominant (107). Therefore, a change
in the size of the M molecule may be an attempt on the part
of the streptococcus to evade immune clearance. While this
may be true for changes in the opsonogenic A-repeat region,
antibodies directed to the B- and C-repeat regions have been
shown not to be opsonic for the streptococcus (106). Thus,
the pressure for these regions to change may be at a different
level, i.e., adherence and colonization.
Natural Selection of Size Mutants
Group A streptococcal pharyngitis is a self-limiting dis-
ease in which the production of opsonic antibodies to the M
molecule results in the elimination of clinical illness (128).
However, in some instances, after symptoms subside and in
the presence of opsonic antibodies, the infected individual
298 FISCHETTI
M6.1 N A
M5 N A13
M24 N Bi B2 B3 B4 B5
FIG. 12. Alignment of the sequences of three different serotype M proteins (M6.1 [strain D471], M5 [strain Manfredo], and M24 [strain
Vaughn]) (94, 148, 153). N terminal to the Pro-Gly region, each M protein differs in the size and number of repeat regions. M24 is lacking
A-repeat sequences. Amino acid identity of >98% is found within the Pro-Gly and membrane anchor regions among the three M proteins.
Sequence variation increases as the N terminus is approached.
may carry the streptococcus in the throat for several months
without clinical signs (termed carrier state) (86, 92, 111). It is
clear that more than one opsonogenic epitope is present
within the N-terminal region of the M molecule and antibod-
ies to only one epitope are adequate for the phagocytosis of
these organisms (15, 19, 106, 179). Whereas antigenic vari-
ation in the gonococcus and Borrelia surface proteins is the
result of the exchange of large genetic blocks, in the strep-
tococcus only a few amino acids are changed at a time.
Therefore, M-protein mutants created during a streptococcal
infection may not be suitably altered in all opsonogenic
epitopes to escape immune recognition and clearance. How-
ever, prior to clearance from the infective or carrier state,
the streptococcus has the ability to infect other human hosts
(203). Thus, M-protein mutants produced in the initial host
during clinical illness or during the carrier state are able to be
disseminated throughout the population, leading in time to
the accumulation of changes within the M molecule. Conse-
quently, through immunological pressure, the antigenic
structure of the M protein is continuously evolving by the
selection of organisms producing variant M molecules. This
process is likely to result in a continuum of antigenically
unique M molecules among the streptococci in the environ-
ment.
Variation in the N-Terminal Nonrepeating Segment
The N-terminal nonrepeating segment of the M molecule
contains protective epitopes. While variation within the
repeat segments may be attributed to intragenic recombina-
tion events, no information is currently available regarding
the method by which sequence changes occur within the
short nonrepeating segment. Sequence analysis of M6 from
several strains isolated from the 1940s to the 1980s revealed
that, while variation is seen in the A, B, and C repeats, the
sequence from the N terminus to the first A repeat (about 30
amino acids depending on the serotype) has been conserved
(95); S. Hollingshead, unpublished data). While current data
suggest that no other homologous regions exist outside the
emm-6 gene (182, 183), it has been suggested that this
N-terminal segment could arise from recombination between
extragenic non-emmn-6 segments (148). Attempting to explain
the mechanism of variation of this N-terminal region among
M serotypes, Haanes-Fritz et al. (90) examined the DNA
sequence of segments outside the emrn-6 structural gene of
CLIN. MICROBIOL. REV.
Cell-associated
Cl 1C3 |Regbn Memb. c
C2 Pro/Gl y Anchor
>98%
C2 1C31 D Region Pro/GIy
Memb.
Anchor C
>98%
|BH Cl C2 C3 D Region Pro/Gly Anchor C
Ml, M6, M12, and M24. They found that, although the 5'
sequences of the four mature M proteins exhibited consid-
erable variability, extensive similarities were observed in the
upstream flanking regions and signal peptides. The authors
proposed that these conserved flanking regions may be
involved in rare occurrences of homologous recombination
with regions outside the M gene, resulting in the insertion of
new variable segments. To date, however, the existence of
these extragenic regions has not been established, which
may be a consequence of the use of unsuitable probes.
An alternative interpretation of the apparent stability of
N-terminal sequences within a serotype is based on the
method used to type the strains used in sequence studies. As
mentioned earlier, the M typing sera are highly absorbed,
containing antibodies that are specific for a given serotype M
protein; i.e., the typing sera define the serotype. Since the
N-terminal end is the most variable region of the M mole-
cules, antibodies in these typing sera are naturally directed
to epitopes in the N-terminal extreme. Thus, when M6
strains are selected for M6 protein analysis, it stands to
reason that they all will likely have a similar N-terminal
sequence. An M6 strain which has diverged sufficiently in
the N-terminal sequence to prevent the typing serum from
reacting (a single amino acid change could accomplish this)
would not be typed as M6 and, thus, would not be analyzed
as an M6 protein. Since more than half of the streptococcal
strains isolated in nature are classified as nontypable (173), it
is conceivable that many of these organisms are derivatives
of typable strains having sequence changes within the non-
repetitive N-terminal region. Since the sequences of only a
few M molecules have been analyzed to date (none of which
are derived from nontypable streptococci), it is difficult to
determine which types, if any, are related to one another
with respect to the sequence in the N-terminal extreme.
COMPARISON OF M-PROTEIN SEQUENCES
Alignment of Known M-Protein Sequences
In addition to the complete M6 sequence from strain D471
(94), sequence data are now available for a number of other
streptococcal serotypes. The sequences are available for the
complete M5 (strain Manfredo) (148), M24 (strain Vaughn)
(153), and M12 (strain CS24) (175) proteins and the complete
pepsin fragments of M5 (strain B788) (133) and M49 (strain
B915) (116) and partial Ml (strain T1/195/2) (120, 152) and
VOL. 2, 1989 STREPTOCOCCAL M PROTEIN 299

d a d a d a d a d a d a d
M6 RVFPRUG-TVEN DKR SMYQANDKlTfTENNN[TDQKKN[TTNK N[
L 56
PepM5 TVT TISD QR KEA D tELE HDLKTK EG| NEG KTENEG KTEN|EG 55
a d a d a d a d a d a d a d aa d
TDQNK TITTENKNrTI1DQ NSNLTTENKE EAEENRLTTENKGLTKKLSEAEEEAANKER 114
KTEKS ERKTAE SE EHEA DKLK 84
a d a d d a d a d a d d a d
ENKEAIG L DETRKDKIAKEQESVETI[TWKI
F TEEVKDKIA1E 163
QQRDTLSWQ ETEREVQNTQYNNETLKIKNGDLTKELN TRQELAN 133

a d a d d a d a d a d d a d a
E S KE|T IGTKKT[DETVKDKIAKEQESKETI GTLKKILDETVKDKII REQKSKQD 219
ESKENEKALNELLEKTVKDKIAKEQEN|KETIGTLKKILDETVKDKLAKEQKSKQN 189
d a d
IGALKQELA 228
IGALKQEL 197
FIG. 13. Sequence comparison between the PepM5 protein (strain D788) (134) and the corresponding region of the M6.1 molecule (M6
N-terminal half) from strain D471 (77, 94). Identical residues are boxed. Lowercase a and d above the sequence designate the positions of
these residues in the seven-residue periodicity (a, b, c, d, e, f, and g) based on the alignment in Fig. 7. A 29-residue gap was introduced in
the MS sequence to maximize the number of identities since the MS sequence in this strain is shorter than that represented in Fig. 12.

M19 (unpublished strain) (34) proteins. In addition, the
C-terminal one-third of M5 (strain T5B/126/4), M19 (strain
(SS400), M24 (strain (C98/135/2), M30 (strain 6GL130), and
M55 (strain A92817311) is also available (96). Like the differ-
ences observed among the various M6 strains (Fig. 9),
alignment of the M-protein sequences from M6, MS, and
M24 serotypes reveals differences among the three in the
number and size of the repeat blocks (Fig. 12). In these three
serotypes, in addition to M19, M30, and M55 (96), a highly
conserved region (>98% identity) is found C terminal to the
D repeat (located within the cell wall), and, depending on the
serotypes being compared, sequence variation increases
among these types as the N terminus is approached (Fig. 8).
Strong sequence homology is observed among these three
types within the C-repeat region; however, this homology is
greatly reduced in the B-repeat region and nearly lost in the
A-repeat region. The extent of homology depends on the
serotypes being compared, some types being more related
than others. For instance, the C-repeat blocks between M6
and M5 and between M6 and M24 are about 80% identical;
however, the B blocks in M6 are not very homologous to
those in M24 (18% identity) but are significantly homologous
to those in M5 (54% identity) (Fig. 12). The homology of the
B blocks between M5 and M24 is <10%. In the M24
sequence, all of the A repeats have been deleted, leaving an
extended nonrepetitive region N terminal to the B repeat. A
similar pattern is seen in the sequence of the pepsin fragment
of the M49 sequence; however, in this sequence, remnants
of an A-repeat region still exist (not shown) (116).
The preceding discussion serves to illustrate the differ-
ences in the sequence and repeat blocks for the M protein
from these specific serotype strains. M proteins from dif-
ferent strains of the same serotype are likely to differ in size
from the prototype strain and will exhibit differences in
sequence as well as immunodeterminants depending on the
nature of the size variation. For instance, the major differ-
ence between the N-terminal half of M5 from strain B788
(112) and M5 from strain Manfredo (148) is a 14-residue
insertion in the A-repeat region of strain B788 and a 25-
residue insertion in the B-repeat region of strain Manfredo.
Therefore, strain designations are critical when comparisons
are to be made between different M proteins. Although very
few M sequences are currently known, it is clear that certain
serotypes are more closely related than others based on the
degree of homology within the B- and C-repeat segments (28,
108, 109). Perhaps, as more data are gathered on other M
proteins, the evolution of the M serotypes will become more
apparent.
Alignment of the Variable Region of Two Different
M Proteins
The sequence of PepM5 protein from strain B788 (133),
comprising the N-terminal half of the native M5 molecule (to
the C-terminal end of the B-repeat region), was compared
with the corresponding region of the M6.1 sequence (from
strain D471) (77, 94). Maximal homology is achieved if a gap
of 29 residues is introduced in the M5 sequence. This is
necessary since the M molecules from these two strains
differ in size, M5 being smaller (Fig. 13) (74). The first 85
residues of the M6 sequence show 33 identities (39%) with
the PepM5 molecule. Homology of 56% is observed between
residues 115 to 228 of the M6 sequence and 85 to 197 of
PepM5. This increases to 77% identity between residues 164
to 228 of M6 and 134 to 197 of M5. In all, 43% identity is
observed between these two segments.
Although the number of identities within the hypervariable
region of these two molecules (residues 1 to 163 of M6 and 1
to 133 of M5) is far less than those in the region C terminal
to these segments, half (22 of 47) of the identities are either
asparagine or hydrophobic amino acids that occupy core
residues at positions a and d of the heptad periodicity (Fig.
7). It is clear that hydrophobic residues in these positions are
required for the maintenance of the coiled-coil structure of
the M protein (136, 164). Therefore, in spite of considerable
sequence dissimilarity between these two M molecules
within the hypervariable region, identities are preserved in
critical structural positions. Thus, the maintenance of the
coiled-coil rod region appears to be a biologically significant
characteristic of the M molecule (76, 77, 141).
300 FISCHETTI CLIN. MICROBIOL. REV.

PROTEIN A ANHADANKAQALPETGEENPLIGTTVFGGLSLALGAALLAGRRREL
22% (52%)
M PROTEIN NKAPMKETKRQLPSTGETANPFFTAAALTVMATAGVAAVV-KRKEEN
56% (63%)
G PROTEIN AKKDDAKKAETLPTTGEGSNPFFTAAALAVMAGAGALAVASKRKED
: : : :. : .. . . .::...: .:.: 35% (60%)
PROTEIN A ANHADANKAQALPETGEENPLIGTTVFGGLSLALGAALLAGRRREL
FIG. 14. Alignment of the C-terminal 46 amino acids of staphylococcal protein A, group A streptococcal M protein, and group G
streptococcal protein G. The protein A sequence is positioned on top and bottom to allow comparison with protein G and M protein. Double
dots designate identities, while single dots designate charge or hydrophobic conservation. A gap was introduced in the M-protein sequence
to maximize identities with protein G. Values at right are percent identities (percent identities plus conservative substitutions).

HOMOLOGY OF M PROTEIN WITH OTHER
MOLECULES OF GRAM-POSITIVE BACTERIA
Group C and G Streptococci
M protein or M-like proteins of group A streptococci have
been reported in group B (142), C (142, 217), E (58), and G
(142, 143) streptococci. These conclusions are based either
on serological cross-reactions with typing sera developed to
M protein from group A organisms or on the ability of these
organisms to resist phagocytosis. DNA hybridization analy-
sis was performed to define better the relationship of the M
protein from group A streptococci with that of other strep-
tococcal groups (183). By using nearly the complete M6
protein gene as a probe, homologous DNA is found in group
C and G streptococci but not other groups (183). Group C
streptococci are primarily animal pathogens but have been
implicated in serious human infections (151). Group G or-
ganisms are found as part of the normal flora of the skin,
upper respiratory tract, and female genital tract. However,
these organisms are responsible for an increasing number of
human diseases (4, 6, 142).
Monoclonal antibodies prepared to the M6 protein from
group A streptococci cross-react with a protein found on
group G clinical isolates (105). More than half of the group G
strains examined are resistant to phagocytosis in human
blood. When surface protein is extracted from group G cells
with the muralytic enzyme mutanolysin and reacted in
Western blots with M-protein-specific monoclonal antibod-
ies, the protein pattern resembles that seen with M protein
from group A organisms. A highly purified, pepsin-derived
protein fragment from a group G strain was capable of
inducing antibodies in rabbits which opsonized group G cells
in a phagocytosis assay (105). In a separate study, group G
streptococcal clinical isolates were found to survive in
human blood and shown to exhibit surface fibers by electron
microscopy. Rabbit antisera to either whole group G strep-
tococci or partially purified pepsin extracts were also op-
sonic for homologous isolates (31). Using DNA probes
representing the C-terminal and leader sequences of the M12
protein gene, Simpson et al. (187) found that DNA from
group G strains which infect humans is related to the group
A streptococcal M-protein DNA. This relationship was not
seen in DNA from group G strains which cause animal
diseases. Thus, M proteins on group A and G streptococci,
especially group G organisms involved in human disease,
appear to be functionally and structurally similar.
A second protein located on the surface of group G
streptococci (protein G) has sequence homology with group
A M protein. Protein G is both an immunoglobulin G (IgG)-
binding protein that binds all four classes of human IgG and
animal IgG classes and an albumin-binding protein (2).
Analysis of the DNA sequence reveals that the C-terminal
end of protein G is highly homologous to the M-protein
C-terminal membrane anchor and a small portion of the wall
domain, with 56% identical residues (61, 159) (Fig. 14).
Although there is some sequence homology between the
remaining portion of protein G and M protein, the most
striking feature lies in the presence of sequence repeat
blocks. Like M protein, the repeat blocks divide protein G
into distinct domains. A repeat domain located within the
C-terminal half of the molecule, proximal to the cell wall, is
responsible for the IgG-binding activity of protein G (89),
while the albumin-binding activity is located in the N-
terminal repeat domain (3). Although protein G is rich in
a-helix-forming residues, a detailed analysis concerning its
conformation as a coiled-coil structure has not as yet been
performed.
S. equi, a Lancefield group C organism, is the causative
agent of "strangles," a disease of horses resembling human
pharyngitis. Reports of an M-like molecule with the ability to
render these streptococci resistant to phagocytosis are well
documented (189, 196, 217). The interesting characteristic of
the M or M-like protein on S. equi is the lack of different
serotypes (8, 9). For instance, Bazeley reports that an
antiserum prepared against one strain of S. equi protected
mice from challenge by each of 32 different strains isolated
from various parts of Australia (8). Thus, antigenic diversity
does not appear to exist in this group C antiphagocytic
molecule as it does in the group A M protein, even though
many of the characteristics of the cloned S. equi M-like
protein are shared with group A M protein (85).
Staphylococcal Protein A
Protein A is an IgG-binding protein from staphylococci
which also has features in common with streptococcal M
protein. Like protein G, the region of sequence homology
with M protein is located at the C terminus of protein A,
where it exhibits 22% identity with M protein (89a) and 35%
with protein G (Fig. 14) (61, 159). The homology among
these three proteins is even more striking when conservation
of charge or hydrophobicity is considered. The conservation
of sequence and chemical properties of amino acids within
the C-terminal region suggests that there likely exists a
common mechanism by which protein A, protein G, and M
protein are anchored to the gram-positive cell wall. Although
protein A does not exhibit (x-helical coiled-coil potential
(159), the repetitive nature of the sequence (198) is reminis-
cent of the repeats found in both M protein (94) and protein
G (61). While serological differences between protein A
molecules on various staphylococcal strains has not yet been
reported, the presence of size variation among clinical
isolates has recently been described (38).
VOL. 2. 1989
TABLE 1. Sequence homology of M protein
with mammalian proteins
M protein Mammalian protein
M24 Myosin heavy chain (rabbit cardiac)
Tropomyosin (human cardiac)
Keratin (type Il bovine)
Lamin (human)
MS Tropomyosin (human fibroblast)
Myosin heavy chain (rabbit cardiac)
Keratin (type 1. bovine)
Tropomyosin (human cardiac)
M6 Myosin (rat cardiac)
Keratin (type 1. bovine)
Tropomyosin (human cardiac)
M12 Keratin (type 1. human)
Myosin heavy chain (rabbit cardiac)
Identity based on fastp analysis of known sequences (130) and inspection.
STRUCTURAL RELATIONSHIP OF M PROTEIN WITH
MAMMALIAN PROTEINS
Tropomyosin
M protein resembles muscle oQ-tropomyosin on the basis of
physicochemical properties and sequence homologies (99).
As described previously, N-terminal sequences of certain
M24 peptides have up to 40% identity with tropomyosin, and
a hexapeptide of tropomyosin appears five times in the M24
protein (99). In subsequent studies, homologies with tropo-
myosin were also observed with other M sequences, namely.
MS and M6 (136). Many of the observed identities corre-
spond to the core hydrophobic amino acids in positions a and
d in the seven-residue repeat pattern necessary to maintain
the coiled-coil (refer to Fig. 7 and section on structure) (77.
136, 164). However, in certain regions, identities are also
found in externally positioned amino acids. It was suggested
that the correlation between sequences in M protein and
tropomyosin provides direct evidence, at a molecular level,
of a structural similarity between a streptococcal and a
mammalian protein, and this similarity may be relevant in
the pathogenesis of streptococcal diseases (99, 150).
Other Coiled-Coil Molecules
An analysis of the PepM5 sequence revealed that the
periodicity and sequence of this M protein more closely
resemble the coiled-coil rod region of myosin than tropomy-
osin (138, 141). Sequence comparison of the complete M5
(148), M6 (77, 94), M12 (175), and M24 (153) proteins with
known protein sequences in a data base reveals the best
homology to be primarily with myosin and tropomyosin
(Table 1). Other coiled-coil proteins such as keratin and
lamin also show significant homology with M proteins.
The basic structure of the M6 protein (77) is remarkably
close to that of eukaryotic intermediate filaments (keratin.
desmin, and vimentin), whose elemental structures are also
based on a coiled-coil central rod region with structurally
different end regions (166, 190). Its superstructure is com-
posed of two to three ot-helices wrapped around each other,
and, like M protein, intermediate filaments from different
sources vary immunologically as well as in size and compo-
STREPTOCOCCAL M PROTEIN 301
sition (190). While it would be tempting to speculate on the
possible role of these homologies in the pathogenesis of
(4 Identity certain streptococcus-related diseases, it would be best to
(amino acid await the results of future experiments.
overlap)"
34 (44) IMMUNOLOGICAL CROSS-REACTIVITY BETWEEN M
33 (42) PROTEIN AND MAMMALIAN PROTEINS
40 (30)
36 (45) Heart-Reactive Antibodies
32 (28) Cross-reactions have been observed between streptococ-
39 (41) cal components and mammalian tissue. Kaplan and Meyes-
43 (30) erian (113) described a component of M5 and M19 strepto-
33 (45) coccal cell walls which induces antibodies reactive with
38 (50) muscle tissue and correlates with M protein. Zabriskie and
44 (32) Freimer (219) observed that purified streptococcal mem-
38 (50) branes evoked cross-reactive antibodies which reacted with
human cardiac muscle sarcolemma. While subsequent at-
41(32) tempts were made to identify these cross-reactive antigens
31 (32) from the streptococcus (200). they have not been character-
ized sufficiently to determine their identity.
Immunodeterminants Shared with Myosin and Other
Human Proteins
Dale and Beachey (54) discovered that rabbits immunized
with PepM5 protein produced antibodies which cross-react
with myosin but not tropomyosin or actin. By enzyme-linked
immunosorbent assay, the cross-reactive epitope is also
found in the M6 and M19 proteins and is exposed on the
surface of the streptococcus. Furthermore, these investiga-
tors found that the M5 molecule contains at least three
epitopes which stimulate antibodies to human sarcolemmal
membranes and are partially shared with M6 and M19
proteins (53). One of the myosin cross-reactive epitopes was
localized to amino acids 84 to 116 of the M5 sequence (56).
More recently. Sargent et al. (179). using antibodies to
eight synthetic peptides representing the entire PepM5 pro-
tein, determined that peptide 164-197 stimulates antibodies
in rabbits which react with isolated cardiac sarcolemmal
membranes but not with purified myosin. The reactivity is
localized to a 40-kDa protein in the sarcolemmal membrane.
The results indicate that the epitope located in the 164-197
sequence is shared with M6 but not M24 or M19. This
sequence is located close to the center of the native M5
molecule (within the B-repeat region), N terminal to the
pepsin-susceptible site. This segment is highly conserved
between the M5 and M6 proteins (77) (Fig. 13) but not M24.
Synthetic peptides representing the N-terminal 20 amino
acids of M5 (57) and M6 (15) proteins were found to evoke
type-specific protective but not heart-reactive antibodies.
These studies suggested that, by using sequences from the
N-terminal end of the M molecule, heart-reactive epitopes
might be avoided. However, subsequent experiments (34)
revealed that PepM19 and synthetic peptides representing
residues 1 to 24 of M19 evoked both opsonic antibodies and
heart-reactive antibodies to a sarcolemmal protein, indicat-
ing that epitopes stimulating autoimmunity may also be
localized within the type-specific N-terminal region.
Renal Tissue-Associated Epitopes Shared with M Protein
Sera from patients with poststreptococcal glomerulone-
phritis are shown to have antibodies against basement mem-
brane antigens (63). The antigens involved in these reactions
are glomerular heparin-sulfate proteoglycan and basement
302 FISCHETTI CLIN. MICROBIOL. REV.

bc e ef c e bc
11I 310
r 320
II 1
I 330 1 340
abcdefgabcdefgabcdefgabcdefgabcdefgabcdefg
M6 .1 LEEANSK KLEKLNKELEESKKLTEKELQAKLEAEAA
Tropomyosin YEEVARKLVIIESDLERAEERAELSEGKCAELEEELKTVTNN
170 180 190 200

c
q

e 9
FIG. 15. Alignment of the sequence of M6 protein and human cardiac tropomyosin in a region exhibiting significant homology. Lowercase
letters a through g directly above the sequence designate the positions of these amino acids within the seven-residue pattern (a-b-c-d-e-f-g)
in both segments based on previous analysis (77, 191). Lowercase letters b, c, g, e, and f identify these external locations in the heptad repeat.
Double dots indicate identities and single dots are conservative substitutions. Below, a cross sectional view of a coiled-coil (looking down the
helices from the N terminus) revealing the relative position of residues a to g in the structure (adapted from McLachlan and Stewart [145]).
Positions a and d are nonpolar and, thus, internalized in the formation of the coiled-coil. This reveals that the identities in positions b, c, g,
e, e, f, c, e, b, c would be clustered nearly contiguously along the outer face of the coiled-coil.

membrane type IV collagen and laminin. It is speculated that
the cross-reactivity may be related to structural similarity
between the streptococcal M protein and basement mem-
brane antigens such as laminin. More recently, it has been
shown that monoclonal antibodies prepared to renal cortex
cross-react with M proteins from type 6 and 12 streptococci
(87). These studies were extended to type 1 M protein, in
which antisera raised in rabbits to a synthetic peptide
representing the N-terminal 1 to 26 amino acids and a similar
peptide lacking residues 20 to 22 both were able to opsonize
Ml streptococci and react with human renal glomeruli by
immunofluorescene (119). The cross-reactive epitope was
localized to the tetrapeptide Ile-Arg-Leu-Arg located at
residues 23 to 26 in the Ml peptide, while the region
responsible for eliciting the opsonic antibodies was residues
1 to 20. The Ile-Arg-Leu-Arg sequence has not been found in
other M-protein sequences such as MS, M6, and M24, which
are not considered nephritogenic serotypes. However, it was
also not part of the M49 sequence, a nephritis-associated
serotype (116). The authors state that this result indicates
that the observed tetrapeptide may not be the only sequence
defining a renal autoimmune epitope (119).
Monoclonal Antibodies Cross-Reactive with Coiled-Coil
Proteins
Cunningham and Russell (49) prepared monoclonal anti-
bodies to streptococcal membranes from M5 streptococci
which cross-react with detergent-extracted human heart
sarcolemmal antigens. The monoclonal antibodies react
more strongly with heart antigens than with kidney or
skeletal muscle antigens. Some of the monoclonal antibodies
are directed to skeletal muscle myosin, and one is reactive
with ventricular myosin. The myosin-reactive antibody
could be absorbed with either myosin or whole streptococci
(121). Two monoclonal antibodies reactive with M5 protein
and a streptococcal membrane protein were found to be
polyspecific. One reacts with DNA, polyinosinic acid,
polydeoxythymidylic acid, and cardiolipin, and its reactivity
with the cytoskeleton could be blocked with antivimentin
(50); the other reacts with the cytoskeleton and is inhibited
by the a-helical proteins myosin, actin, and keratin. Further
characterization of these monoclonal antibodies revealed
that they are specific for epitopes in the heavy chain of
myosin: one for light meromyosin and the other for both light
and heavy meromyosin. Both of these myosin fragments
contain a portion of the coiled-coil rod region. The authors
speculate that the induction of antibody-producing clones
against myosin by a streptococcal antigen may explain at
least some of the cross-reactive responses seen during
streptococcal sequelae. Recently, Fenderson et al. (62a)
described the production of monoclonal antibodies which
react with M5 and M6 proteins and human and rabbit
tropomyosin. At least one of these cross-reactive monoclo-
nal antibodies also exhibits strong reactivity with human
heart tissue by immunofluorescence.
Conformational Determinants
Since many of the M-protein cross-reactive antibodies (48,
50, 56, 119, 121) also react with other coiled-coil proteins
(i.e., myosin, tropomyosin, keratin, vimentin, etc.), it is
likely that in several cases the immunological recognition is
based on a common conformational determinant. Sequence
alignments between M protein and tropomyosin (Fig. 15), as
well as between other cross-reactive coiled-coil molecules,
do not reveal a common linear sequence (Fischetti, unpub-
lished data). Identities are usually distributed among several
nonadjacent amino acids within a localized region of these
molecules. This is not surprising based on the molecular
constraints imposed on a coiled-coil structure, i.e., where
every first and fourth amino acid in the heptad repeat are
VOL. 2, 1989
internalized in the formation of the coiled-coil, leaving the
intervening residues in external positions (145). For exam-
ple, an analysis of the position of the identities within the
seven-residue repeats found in an alignment between M6
protein and tropomyosin. reveals that they appear primarily
in positions b, c, g, e, and f, all of which are located nearly
contiguous to one another on the outer surface of the
coiled-coil structure (Fig. 15). Thus, it is possible that the
spatial distribution of these residues in two conformationally
similar molecules could form the basis of a common immu-
nodeterminant.
Consequently, since streptococci are capable of varying
the M-protein sequence (95, 107), certain sequences found
within this changing molecule are likely to be homologous
with sequences located in relatively stable mammalian
coiled-coil proteins. Thus, as seen with the analysis of the
M5, M6, M12, and M24 proteins, each shares homology with
myosin, tropomyosin, and other coiled-coil proteins to dif-
fering degrees as well as to different regions of the respective
molecules (Table 1). While some homologous regions may
form cross-reactive determinants, the significance of these
determinants in the pathogenesis of streptococcal diseases is
just beginning to be understood.
THE emm GENE
As noted previously, no information regarding the C-
terminal half of the M molecule was available until the gene
for M6 protein was first cloned from streptococcal strain
D471 (emm-6.1) into Esclewrhia (/oli (94, 180). The M
protein produced in E. coli (ColiM6) was shown to react with
both monoclonal and polyclonal antibodies prepared to the
streptococcus-derived M6 protein. The cloned molecule
accumulates in the periplasmic space of E. coli, and no M
protein is detected on the surface of E. co/i or in the growth
supernatant (73). A small amount of M protein is present in
the cytoplasm, where its molecular size is larger than the
periplasmic form. This is consistent with the idea that the
larger molecule retains the hydrophobic leader sequence
prior to export to the periplasm.
The M-protein genes from M5 (115), M12 (175), and M24
(153) streptococcal serotypes have also been cloned and
shown to produce immunologically active M proteins. On
sodium dodecyl sulfate-polyacrylamide gel electrophoresis,
the cloned M6 molecule appears as a close triplet of bands.
a characteristic also seen with the cloned M5 protein (171).
The reason for the multiple banding pattern is not entirely
understood, but seems to be associated with the C-terminal
region, since pepsin cleavage of ColiM6 results in a C-
terminal fragment which retains the triple banded structure
(72) while the N-terminal half (or PepM) is homogeneous (21,
34, 72, 137).
Relationship of M-Protein Genes
The relationship of the emmn-6 gene in M6 streptococci to
emm genes in other streptococcal serotypes was determined
by hybridization experiments, using the enmm-6.1 gene as a
probe (183). The results revealed that DNA from all of the 56
different serotypes tested contain homologous cmiui genes.
Among the group A streptococci tested in these studies were
several which were functionally M- (i.e., those phagocy-
tized in human blood). An e(mm gene could be identified in all
but one M- strain. The reason for the M- phenotype in
emm1117-containing strains is not known, but may reflect a
defect at the level of translation or transcription of M
protein.
STREPTOCOCCAL M PROTEIN 303
Localization of the region in the emin gene responsible for
hybridization with all tested serotypes was accomplished by
hybridizing DNA from 10 different serotypes with smaller
probes derived from the emm-6.1 gene. This investigation
revealed that the 3' region of the emim-6. / gene is conserved
among these serotypes while the 5' segment is variable (182).
A similar result was also reported when DNA probes corre-
sponding to the 5' and 3' ends of the einmn-5 gene (their snp5
notation for the gene) were used in hybridization studies
with DNA from heterologous M serotypes (114). These
findings corroborated earlier results of studies which used
monoclonal antibodies with defined epitopes to identify the
variable and conserved segments of the M molecule (109).
Ribonucleic acid sequence analysis of emm gene tran-
scripts primed with specific oligonucleotide probes from
emnm-6.I was used to determine the degree of homology
within the 3' end of the emmn-5, emm-19, emmn-24, emnm-30.
and emmn-55 genes (96). The results revealed that the C-
terminal one-third of these genes are >95% identical to
emnmn-6.1. This was later confirmed with the reported se-
quence from other strains of M5 (148) and M24 (153)
proteins.
Conversion of M- to M+ Phenotype
DNA hybridization studies demonstrated that M- strain
T28/51/4 lacked an emm gene (183), and electron microscopy
revealed that this strain has no surface fibrils associated with
the presence of M protein (193). In addition, this strain is
easily phagocytized in normal human blood (193). With a
specially constructed shuttle vector, experiments were de-
signed to move the cloned emm-6.I gene into this emim-
deficient streptococcus (181). The resultant transconjugant
was converted to M6', resistant to phagocytosis in human
blood, and opsonized by anti-M6 antiserum. Immunofluo-
rescence studies revealed that M6 protein was located on the
cell surface of the transconjugant. When inoculated into
rabbits, the transconjugant induced opsonic antibodies to
M6 streptococci. In a similar set of experiments, the emnin-5
gene was transformed into S. sanguis, and M5 protein was
demonstrated on the cell surface by immunofluorescence
(171). Although producing less M protein than authentic M5
streptococci, these organisms removed opsonic antibodies
from anti-M5 serum, indicating that the antiphagocytic de-
terminants are expressed on the surface of S. sanguis. Thus.
the emm-6.1 and emnm-5 genes contain all information nec-
essary to convert an M- streptococcus to a biologically and
immunologically MX organism, further emphasizing the role
played by M protein in the antiphagocytic capacity of
streptococci.
Number of emm Gene Copies
As stated earlier, one copy of the emin gene is detected in
an M6 strain (183) and in the chromosome of nine different
serotypes as determined by Southern hybridization with M6
gene probes (182). This supports the hypothesis that recom-
bination leading to size and antigenic variation is an intra-
genic event. However, in contrast, two bands were found to
hybridize with a digest of the M5 chromosomal DNA. This
result may be due to restriction site heterogeneity for this M5
strain or the presence of two einmn-5 gene copies in the M5
organisms. The latter possibility is supported by Kehoe et al.
(115). who identified two copies of an emmn-5 gene in a strain
different from the one used in the studies by Scott et al.
(182). Since these were the only two reported studies of
304 FISCHETTI
emmn-5 genes, it is not certain whether the presence of two
einin genes in M5 streptococci is the rule or whether these
strains are unusual.
Regulation of the emm Gene
Lancefield observed that streptococcal strains maintained
in the laboratory for extended periods of time (several weeks
to a few months) lose their ability to survive in human blood
(129). Quantitative measurements of the M-protein content
suggest that these streptococci do not contain sufficient M
protein for survival (24, 129). Passage of these organisms
through mice or human blood results in the surviving strep-
tococci producing higher quantities of M protein (24), sug-
gesting that a regulatory mechanism may be responsible for
M-protein expression.
Cleary et al. (40) found that M12 strain CS44 produced M-
mutants at high frequency which correlated with a variation
in colony morphology. Spanier et al. (188) isolated two M-
mutants from this population, which, by Southern blot
analysis, were found to contain a small deletion in a region
upstream from the M12 structural gene promoter. In a recent
analysis of these strains and derivatives thereof, it was
determined that the deletion was about 500 base pairs
upstream of the emm-12 gene (186). Although this charac-
teristic could not be confirmed in other M serotypes, these
results suggest that a positive regulator may be located
upstream from the M-protein structural gene.
A positive regulator for the production of M protein was
also suggested by the experiments of Scott et al. (181), in
which a plasmid containing the emm-6.1 gene and its pro-
moter was transferred to a strain lacking an emm gene (T28/
51/4). Even though the plasmid was present in a high copy
number, the production of M6 protein was about 30- to
50-fold less than in the parental M6 strain from which the
emm-6.1 gene was derived. They suggested that the emrn-
deleted strain may also be deleted for a positive regulator
necessary for wild-type expression. By using Tn9l6 muta-
genesis in D471, the strain from which the emm-6.1 gene was
derived, a mutant was found which produced 50-fold less M
protein than the parental strain. By Southern blot analysis, it
was shown that the Tn insertion was about 2 kilobases
upstream from the emm-6.1 structural gene. Northern (ribo-
nucleic acid) blot analysis revealed that the mutant did not
produce any detectable emm-6.1-specific messenger ribonu-
cleic acid. Thus, the authors named the gene responsible for
this effect mry, for M-protein ribonucleic acid yield.
In support of the hypothesis that the emrn-deleted strain
T28/51/4 was unable to sustain full emm expression, South-
ern blot hybridization revealed that this strain was in fact
deleted for the may gene (35). Thus, a positive regulator
appears to be necessary for high M-protein expression.
Whether environmental factors play a role in this expression
or whether, like certain positive regulators in other patho-
gens (149, 206), mnry controls other virulence determinants in
the streptococcus requires further study.
HUMAN IMMUNE RESPONSE TO THE M MOLECULE
In earlier experiments designed to examine the human
immune response to the M protein as a result of a strepto-
coccal infection, it was clear that, once infected with a
specific streptococcal serotype, an individual remained pro-
tected from subsequent infection by strains of that serotype
but was susceptible to other streptococcal types (see review
by Lancefield [1291). This limited protection was based on
CLIN. MICROBIOL. REV.
the development of type-specific opsonic antibodies to the M
molecule which were found to persist years after a strepto-
coccal infection (128). However, after such an infection,
measurable amounts of opsonic antibodies are usually de-
layed when compared with the response to other streptococ-
cal antigens (124, 125, 176). In more recent studies with
competitive inhibition assays with purified M protein, it was
discovered that type-specific opsonic sera contained anti-
bodies directed to a majority of the immunodeterminants on
the M molecule while nonopsonic sera did not (64). Thus, it
was suggested that, for some reason, opsonic antibodies are
produced late during an immune response to the M protein.
Immunodominant Region of the M Molecule
To define further the development of an immune response
to the total M6 protein, synthetic peptides representing 82%
of the native molecule were examined for reactivity with
rabbit and human sera by kinetic enzyme-linked immunosor-
bent assay (78). The results reveal that rabbits immunized
with the native M protein or whole streptococci respond by
reacting first and predominantly to one of the three sequence
repeat regions of the molecule, namely, the B repeat. Anti-
bodies to this region have recently been shown to be
nonopsonic (106). Antibodies to peptides representing the
variable N-terminal and adjacent A-repeat regions appear
only when opsonic antibodies are detected in the serum.
Antibodies to peptides located within the conserved C-
terminal half of the molecule are limited even after several
immunizations. By the same assay, an examination of hu-
man sera revealed that those sera opsonic for M6 strepto-
cocci contain antibodies reactive predominantly to the N-
terminal and A-repeat regions. Nonopsonic human sera to
M6 streptococci show low reactivity to most peptides. By
Western blot analysis of the N- and C-terminal halves of the
M6 molecule, all human sera tested contain antibodies to the
conserved C-terminal half, while only sera opsonic for M6
streptococci react with the N-terminal variable half. The
authors suggest that antibodies to the C-terminal half are
primarily directed to conformational determinants and thus
may not react with small synthetic peptides (78).
The immunodominant nature of the B-repeat region could
explain the observed delay in the appearance of opsonic
antibodies in human sera after exposure to M-protein-pos-
itive streptococci (124, 125, 176). It is possible that an
exaggerated response to the nonopsonogenic B repeat may
in some way affect the response to the adjacent regions. It
also explains the observed presence of high levels of binding
antibodies to the M molecule in the absence of opsonic
activity. That M6 antibodies to the N-terminal and A-repeat
segments were elevated in the M6 opsonic human sera is in
keeping with the observation of Lancefield (128) that lifelong
type-specific immunity usually occurs after a streptococcal
infection.
ANTIPHAGOCYTIC ACTIVITY OF M PROTEIN
The capacity of the group A streptococcus to cause
disease may be attributed partially to its ability to resist
phagocytosis by human neutrophils. This capacity can easily
be demonstrated in the laboratory, using a bactericidal or
opsonophagocytosis assay developed by Lancefield (127) in
which a few log-phase M t streptococci are added to a small
amount of heparinized human blood from a donor lacking
antibodies to the streptococcal serotype being tested. The
tube is sealed and rotated end over end at 370C to allow
VOL. 2, 1989
maximal contact between phagocytes and streptococci. At
the end of 3 h. the contents of the tube are analyzed for
surviving streptococci. In this assay, the M+ streptococcus
grows unabated, whereas streptococci with insufficient
quantities of M protein are easily cleared from the system. In
the case of an M- organism, several thousand may be added
to this system and be phagocytized effectively, while <10
streptococcal chains of an M+ strain are adequate to allow
growth to several hundred chains at the end of the incubation
period. The addition to the assay tubes of rabbit antiserum.
prepared against either purified M protein or whole strepto-
cocci of a specific serotype, neutralizes the antiphagocytic
activity of the M molecule, allowing streptococci of that
same serotype to be phagocytized. Antisera prepared to
other streptococcal serotypes which cross-react with the M
protein of the organisms in the assay have no opsonic effect
on these streptococci.
A different assay has been used by Beachey and co-
workers (21) in which streptococci are mixed with human
blood much the same as described in the Lancefield assay
(127); however, a higher bacterial inoculum is used. After
3-h rotation, a smear is made of the blood and the phago-
cytes are observed under the microscope and scored for
their association with streptococci. This assay does not
distinguish between live attached organisms and ingested
killed streptococci. In many cases, however, good correla-
tion is seen with the Lancefield method (127); nevertheless,
a different parameter is likely being measured by this assay
(i.e., at the level of attachment and ingestion).
N-Terminal Charge Domain
The mechanism by which the M molecule exerts its
antiphagocytic effect has been elusive since its discovery
nearly 60 years ago. Despite sequence variation within the
N-terminal hypervariable region of the M molecules, a
common feature found in the sequence of M5 (139), M6 (77),
M24 (153), and M49 (116) is a concentration of acidic amino
acids within this segment. These residues occur predomi-
nantly in external positions within the coiled-coil structure
(77). Some of these amino acids likely enter into salt bridge
formation between the two M-protein chains and function to
stabilize the coiled-coil (165), while others become hydrated
and contribute to the net negative charge in this region.
While the exact function of this conserved charge domain is
not known, it has been suggested that it may contribute to
the antiphagocytic property of the molecule (66, 136, 139) by
hampering contact between the streptococcus and the phago-
cyte by electrostatic repulsion. In support of this idea, it was
found that, when opsonic IgG was fractionated into acidic
and basic components, the opsonic activity was localized in
the IgG population that was most basic in charge while
nonopsonic binding antibodies were located in the other
charged IgG populations (66). This suggests that the basic
antibodies bind to the more acidic regions of the M molecule
(i.e., the N-terminal region) and thereby neutralize the
negative charge.
Effect on Phagocytic Cells
In attempts to determine whether the M protein on the
whole streptococcus impairs the phagocytic capacity of
neutrophils, M streptococci were rendered nonviable with
mitomycin and used at 150-fold excess over phagocytes in a
bactericidal assay containing a small number of viable phago-
cytosis-susceptible streptococci (140). The results revealed
STREPTOCOCCAL M PROTEIN 305
that the phagocytes could effectively seek out and destroy
the relatively few susceptible streptococci in the presence of
the excess M' cells. Thus, the nonviable M+ streptococci
did not interfere with the phagocytic activity of the neutro-
phils. When these experiments were repeated with live
streptomycin-susceptible M + and streptomycin-resistant
M- streptococci, to select for M- survivors on streptomycin
plates, similar results were obtained (Fischetti, unpublished
data). This helped to rule out the possible effects of mitomy-
cin or nonviable organisms in the previous experiments.
Hence, in the presence of M ' streptococci, the phagocytes
are not preoccupied by attempting to phagocytize the M+
cells: they simply ignore them.
Complement
The alternative complement pathway is considered to be
an important part of the nonspecific defense system against
infection (155). The deposition of a large number of comple-
ment component C3b fragments on the surface of microbial
particles, termed opsonization, results in the uptake of these
particles by phagocytic cells bearing C3b receptors. Several
lines of evidence have indicated that M protein exerts its
antiphagocytic effect by interfering with the alternative
complement pathway. It has been shown that M- strepto-
cocci are quickly phagocytized after being opsonized via the
alternative pathway while M+ streptococci resist phagocy-
tosis in the absence of type-specific antibodies (29, 168).
That M streptococci fail to bind effectively to phagocytic
cells suggests that C3b was either insufficiently deposited on
the streptococcal surface or inacessible to the C3b receptors
of the phagocyte. When Jacks-Weis et al. (103) quantitated
the amount of C3 deposited on the streptococcal surface,
they found that M- organisms bound up to four times more
C3 than M+ cells. They concluded that, while reduced, the
amount of C3 deposited on the M+ cell should be sufficient
for phagocytosis and, thus, could not be the sole reason for
the antiphagocytic effect of the M molecule.
Factor H
The deposition of C3b on the surface of particles is
amplified by particle-bound C3 convertase, also termed
C3b,Bb, which is a labile enzyme complex that activates C3
(154). Specific inhibitors such as decay-accelerating factor
and factor H prevent C3 activation by rendering C3b sus-
ceptible to inactivation by factor I. Factor H is a 150-kDa
,3-globulin present in serum (500 jg/ml) and functions to
control fluid-phase as well as surface-bound C3b (205).
While the studies of Jacks-Weis et al. (103) demonstrate
that M' streptococci bind less C3b than M- cells, the
deposition on M' cells was found to be uneven compared
with the more uniform distribution of C3b on M- organisms.
Horstmann et al. (98), using purified complement compo-
nents instead of serum, did not find any difference in either
the deposition or the inactivation of C3b or the dissociation
of the C3b,Bb complex on the surface of M+ and M-
streptococci when complement regulatory proteins were
omitted. In fact, both strains exhibited a spontaneous decay
of the C3b,Bb complex similar to zymosan-bound C3b,Bb
which served as a control. However, when purified factor H
was added to streptococcus-bound C3b, the cofactor activity
of factor H for the cleavage of C3b by factor I was six to
eight times higher when C3b was bound to M+ streptococci
instead of M- organisms. M+ streptococci were found to
bind factor H selectively. Furthermore, M+ streptococci of
306 FISCH ETTI
five different serotypes, as well as purified M protein, were
found to bind factor H with dissociation constants as high as
6 x 10' M, emphasizing the functional significance of this
binding reaction.
It was proposed that the antiphagocytic action of the M
protein may be explained by this finding. When a strepto-
coccus contacts serum, the M protein specifically binds
factor H to its surface. Any C3b fixed in the vicinity of this
complex is controlled by factor H, which inhibits or reverses
the formation of C3b,Bb complexes and serves as a cofactor
in the conversion of C3b to iC3b by factor I. Thus, the low
numbers of C3b molecules deposited on the M streptococ-
cus will prevent C3b-dependent phagocytosis. The regula-
tion of complement deposition by binding factor H to a
bacterial surface appears to be a novel route by which a
pathogen is able to evade alternative pathway activation and
phagocytic clearance.
Fibrinogen
Kantor (110) found that a component in partially purified
acid extracts of group A streptococci precipitated fibrinogen.
In a number of studies, the active component was subse-
quently found to be M protein (101, 110, 152, 207, 209).
Similar binding characteristics for fibrinogen were found for
other streptococcal groups (C and G), suggesting that M-
protein analogs are also present on these organisms (122). In
attempts to examine the role of this interaction with regard
to the antiphagocytic action of the M protein, Whitnack and
Beachey (207) measured the association of streptococci with
polymorphonuclear leukocytes in the presence of serum and
plasma. In this study, bound and ingested streptococci were
not differentiated. Their results indicate that fibrinogen
blocks the association of streptococci with the polymorpho-
nuclear leukocytes. Immunofluorescence analysis with anti-
C3 revealed that streptococci pretreated with fresh serum
bound this complement component evenly on their surfaces
whereas streptococci treated with plasma exhibited limited
complement deposition. The authors suggested that the
restricted deposition of C3 prevents complement-mediated
opsonization of these organisms. This finding differed from
that of Jacks-Weis et al. (103), who detected an uneven
deposition of C3 on streptococci pretreated with nonimmune
serum. Chhatwal et al. (39) also demonstrated, by immuno-
fluorescence, that fibrinogen inhibited C3 fixation on the
surface of group A, B, C, and G streptococci.
Subsequent experiments by Whitnack et al. (210) were
designed to examine the influence of fibrinogen on opsoniza-
tion of streptococci by type-specific and cross-reactive anti-
sera. Rabbit antisera against PepM protein of the homolo-
gous type overcame the antiopsonic effect of the fibrinogen.
Using cross-reactive sera, these authors found that, in the
two M proteins (M5 and M6) studied in detail, antibodies
inhibiting fibrinogen binding to M protein accounted for a
large part of the cross-reacting anti-M antibody in the sera.
In all, these studies indicate that fibrinogen binds to both
type-specific and cross-reactive regions of the M molecule.
To explain the common antiphagocytic function of M pro-
teins from different serotypes, it was suggested that the
fibrinogen binding site may be conserved among different
serotypes. That the CR3 receptor on human polymorphonu-
clear leukocytes binds both complement C3b and fibrinogen
(218) may explain why fibrinogen-coated streptococci (208)
adhere to but are not ingested by these phagocytic cells.
Thus, since the assay used in many of these studies mea-
sures phagocyte association, the role of fibrinogen in strep-
tococcal clearance is not completely defined.
CLIN. MICROBIOL. REV.
R. D. Hortsmann and V. A. Fischetti (manuscript in
preparation) examined the effect of fibrinogen-bound M
protein on factor H binding and on the regulation of C3b on
the streptococcal surface. Results revealed that, while the
presence of fibrinogen on the streptococcal surface slightly
reduced the binding of factor H to M protein, it did not
interfere with complement inhibition. They found that C3b is
equally well regulated by factor H when it is bound to
fibrinogen. Thus, factor H reduces the deposition of C3b on
the streptococcal surface with or without fibrinogen bound
to the M protein.
ROLE OF M PROTEIN IN STREPTOCOCCAL
ADHERENCE
Group A streptococci colonize the nasopharyngeal mu-
cosa of humans, leading to an inflammatory response fol-
lowed by clinical illness or an asymptomatic carrier state. It
was first shown by Ellen and Gibbons (60) that streptococci
bearing the M-protein molecule on their surface can adhere
better to epithelial cells in vitro than M-deficient organisms.
There is also strong evidence that LTA secreted from the
streptococcus mediates direct contact with the epithelial cell
surface. M protein or another surface molecule might facil-
itate this process by anchoring and orienting the LTA
adhesin so that the lipid moiety is able to make contact with
the epithelial cell (20). Studies also suggest that M-positive
streptococci can attach to epithelial cells by a mechanism
independent of LTA. When M+ and M- streptococci are
tested for their binding capacity to human pharyngeal,
buccal, and tongue epithelial cells, the M ' bacteria attach in
signficantly higher numbers to the pharyngeal than to the
buccal or tongue cells (197). The M- organisms bind in
equivalent low numbers to all three oral cell types. While
M I streptococci could be inhibited from binding to pharyn-
geal cells by using the complete M6 protein molecule as well
as LTA, 1,000-fold more LTA than M protein is required to
attain equivalent inhibition.
M protein binds specifically to F actin, forming regular
parallel bundles of actin filaments in solution (36). In addi-
tion to isolated M protein, whole M' streptococci also bind
specifically to F-actin filaments. While M protein does not
compete with tropomyosin for binding to actin, it does
compete with myosin subfragment 1 as a result of this
interaction. While the biological implication of this binding is
unclear at this time, it is becoming more apparent that actin
is present on the surface of a number of cell types such as
fibroblasts (184), lymphocytes (160), and human B cells (5).
Therefore, the interaction of streptococcal M protein with
surface actin on a host cell may be a possible event. Thus,
while there may be multiple mechanisms by which strepto-
cocci can attach to epithelial cells, it seems apparent that M
protein provides the organism with an adherence advantage;
consequently, it is considered to be an attachment factor.
M-PROTEIN VACCINES
Type-Specific Protection
Since the observation by Lancefield that antibodies to the
M protein have the capacity to opsonize streptococci in
preparation for phagocytic clearance (126), the M protein
has been a prime candidate for a vaccine to prevent group A
streptococcal infections. Fox has reviewed the trials and
tribulations of the early attempts in M-protein vaccine de-
velopment (81). Since the early 1970s, few human trials have
VOL. 2, 1989
been attempted (recently reviewed [D. Bessen and V. A.
Fischetti, in M. M. Levine and G. Wood, ed., The Molecular
Approach to Newt and Improved Vaccines, in press]). This is
partially based on previous problems with hypersensitivity
reactions found with the acid-extracted M-protein prepara-
tions (81); essentially, only type-specific protection was
observed. In addition, attempts to separate heterologous
protein contaminants from the type-specific determinants
were disappointing. Except for one investigation (22), all
vaccine development since the 1970s has been based on
animal studies and on an analysis of the immune response to
a variety of M-protein preparations with and without adju-
vants (10, 26, 27, 37, 78, 104, 216) and in combination with
other antigens (37).
In 1979, Beachey et al. (22) used pepsin-extracted M24
protein (the N-terminal half of the native M molecule) to
immunize human volunteers. This highly purified prepara-
tion was found to be free of non-type-specific reactivity.
Unlike earlier acid-extracted products (81), skin delayed-
type hypersensitivity tests in 37 adults proved to be negative
with this PepM protein. Immunization with alum-precipi-
tated PepM24 protein led to the development of type-specific
opsonic antibodies in 10 of 12 volunteers and the formation
of positive delayed cutaneous hypersensitivity to PepM24 in
11 individuals. None of the volunteers developed heart-
reactive antibodies as determined by immunofluorescence.
These studies clearly indicated that M-protein vaccines free
of sensitizing antigens could be produced, but also illustrated
the type specificity of the immune response.
Encouraged by these studies, Beachey and co-workers
embarked on a systematic approach to develop methods by
which a type-specific epitope-based vaccine could be devel-
oped to protect against streptococcal disease. It was quickly
learned that peptides representing the first 20 or so N-
terminal amino acids of M24 protein evoked type-specific
opsonic antibodies to M24 streptococci (17). Experiments
with synthetic peptides of the N terminus of MI, MS, M6,
and M19 proteins resulted in the same conclusions (12, 15,
23, 34, 107, 179). When a hybrid peptide was synthesized
from the sequence of the M24 and MS proteins and injected
into animals in complete Freund adjuvant, opsonic antibod-
ies were produced to both MS and M24 streptococci (12).
Opsonic antibodies to three M proteins were obtained when
the N-terminal sequences of three M-protein sequences
(MS-M6-M24) were synthesized in tandem and injected into
rabbits (16).
Attenuated Salmonella typhimarium has also been used
effectively for delivering the M-protein antigen in a mouse
model (174). The cloned gene for the complete MS protein
was introduced into the aroA strain of S. tvphimarium by
transformation with a plasmid vector. The transformed
salmonella expressed the MS protein in the cytoplasm and
had the capacity to induce serum and salivary IgA, IgG, and
IgM to MS protein when fed orally to BALB/c mice. The
orally immunized mice were protected from challenge by MS
but not heterologous M24 streptococci.
An important factor to consider for the development of a
type-specific epitope-based vaccine is the potential of the
streptococcus to generate new M serotypes by changing the
amino-terminal portion of M protein. In the type 6 M-protein
molecule of strain D471, type-specific opsonizing antibodies
are directed against epitopes located both at the amino-
terminal end (residues 1 through 21) and within the A-repeat
block (106), which begins at amino acid 27 and continues to
residue 96 (77, 94). As mentioned previously, high-frequency
intragenic recombinational events within the A-repeat block
STREPTOCOCCAL M PROTEIN 307
region can lead to a significant loss in the opsonizing ability
of monospecific antibody (107). However, to date little
information is available on mechanisms for generating anti-
genic variation at the extreme amino-terminal nonrepeat
region. A second major concern for type-specific epitope-
based vaccines centers on the antigenic diversity of the more
than 80 serological types of M protein. Certain serotypes are
more likely than others to be associated with severe cases of
pharyngitis or outbreaks of rheumatic fever; thus, the num-
ber of distinct type-specific determinants incorporated into a
vaccine can be limited to some extent (28). A type-specific
streptococcal vaccine against upper respiratory infection
and acute rheumatic fever would require a multivalent
antigen corresponding to stable immunodeterminants of
those serotypes which together account for the majority of
nasopharyngeal isolates.
A clear understanding of the molecular basis of antigenic
variation within the hypervariable nonrepeating segment of
the M protein should enable investigators to make decisions
on how best to circumvent the evasive mechanisms used by
group A streptococci.
Non-Type-Specific Protection
The incidence of group A streptococcal respiratory infec-
tion rises sharply at age 4, peaks at age 6, and rapidly
declines above age 10, reaching adult levels by 18 years (32).
At its peak incidence, 50% of children between the ages of 5
and 7 suffer from streptococcal infection each year. The
decreased incidence of streptococcal pharyngitis in adults
might be explained by an age-related host factor. Alterna-
tively, protective antibodies directed to antigens common to
all group A streptococcal serotypes might arise as a conse-
quence of multiple infections experienced during childhood,
resulting in an elevated response to conserved protective
epitopes of the M molecule. In exploring the second hypoth-
esis, it was found that sera of most adults examined have a
strong antibody response to the conserved regions of M6
protein (27a, 78). While antibodies directed to conserved
epitopes fail to neutralize the antiphagocytic property of M
protein (106), they might afford protection by an alternative
mechanism.
Kurl et al. (123) found that whole-cell vaccines consisting
of type 12, 18, 30, 49, 50, or 55 streptococci were equally as
efficient at protecting mice immunized intranasally from
lethal infection after challenge with M50 streptococci. The
authors suggest that local immunity against M50 strepto-
cocci could be achieved by a nonspecific mechanism; how-
ever, the molecular components of the streptococcus in-
volved in eliciting this protection were not defined.
It should be emphasized at this juncture that humans are
the only reservoir for group A streptococci, and except for
M type 50, which is naturally virulent for mice (97), all other
serotypes need to be adapted to the animal being considered
as a possible model for streptococcal disease. For instance,
group A streptococci isolated fresh from human infection are
not virulent for mice, and nearly 107 CFU need to be
administered intraperitoneally to cause death within 1 to 2
days (129). However, after such a passage in the mouse
(usually several passages are required), a strain is isolated
which will cause death with a dose of <10 CFU. For some
group A strains, a mouse-virulent organism may not be
isolated. Although the amount of M protein increases after
mouse passage, its role in mouse virulence is not completely
defined (93). However, it is clear that immunization with M
protein or passively administered type-specific M-protein
308 FISCHETTI CLIN. MICROBIOL. REV.

CTB Only (n=52) 106, 108, 109), it should be possible to generate antibodies
Peptide-CTB (nm52) reactive to the majority of serological types by using only a

VO 3- rP 0u 1.05 * * *
few distinct antigens for immunization. If a conserved-region
0

30-
vaccine is to be effective in preventing nasopharyngeal
(N) infection by group A streptococci, it may require stimulation
25- * s * of the secretary immune response.
0
Secretory IgA is a major first-line defense against bacterial
° 20-
infections of the mucosal epithelium (118), and it has been
X 15- $ 9 9 6 ~~~demonstrated that M-protein-specific secretary IgA can pro-
vide significant protection against streptococcal infection in
10- a mouse model (26, 27a, 67). Passive immunization at the
mucosa allowed for a more precise evaluation of the role
-
played by secretary IgA in preventing streptococcal infec-

1 2 4 6 8 9/10 14/15 human anti-M-protein secretary IgA administered intrana-

sally protected mice against systemic infection after intrana-
Days Post-Challenge sal challenge with group A streptococci. In contrast, anti-
FIG. 16. Phaaryngeal colonization of peptide-immunized mice M-protein-specific IgG administered intranasally had no

challenged withl group A streptococci. Mice (52 per group) were protective effect at this site, although it was highly opsonic
immunized intraanasally with either cholera toxin B-subunit (CTB) or and promoted streptococcal phagocytosis in whole blood.
synthetic peptic les derived from sequences in the conserved region An oral or intranasal route of immunization has been used
.l
of the M6 mole(cu.le cope to
coupled to chlr oi Bsubunit.
cholera toxin -uui. Phrnga
Pharyngeal
cultures were taiken up to 15 days following intranasal challenge with for several group A streptococcal vaccines, with the intent of
5 x 105 strepttococci. Results are expressed as the number of evoking a strong secretary IgA response (27, 51, 123, 170,
culture-positive plus dead mice at each time point from three 172). To determine whether the conserved exposed epitopes
separate experiiments. Although colonization differed, nearly the of M protein influence the course of mucosal colonization by
same number olf mice died each day in each group beginning on day group A streptococci, peptides corresponding to these re-
4, with a total f 15 dead per group by day 15. Statistical difference gions were used as immunogens in a mouse model (27).
in colonized anm d dead mice was significant (*P < 0.05) on days 1, 2, Synthetic peptides corresponding to residues 216 to 235, 248
6, 8, 9/10, and: 14/15 by chi-square analysis. to 269, and 275 to 284 of the M6 sequence (Fig. 7) were

covalently linked to the mucosal adjuvant cholera toxin B

antibodies wil11 protect mice from lethal streptococcal chal-
lenge with the homologous serotype (59).
The C-repe at region of MG protein is highly conserved
among strept(coccii of many distinct serological types (27a,
28, 106, 108,: 109). Monoclonal antibodies directed to deter-
minants localiized to the C-repeat region of M6 protein bind
to the surface of whole streptococci which represent more
than half of th e serotypes examined (108, 109). Despite these
findings, Mill' er et al. (147) suggest that antibodies to con-
served M-prostein epitopes located on the C-terminal side of
the pepsin cksavage site do not bind to intact streptococci
and, thus, arte of no value in a vaccine. Their findings are
based on antibodies prepared to a synthetic peptide whose
sequence is lo)cated within the conserved region immediately
C terminal to the pepsin site. Perhaps the conformation of
the peptide usied to raise antibodies is not sufficiently homol-
ogous to the s;ame region in the native molecule to enable the
antibodies to react with the intact molecule. This is sup-
ported by thieir finding that, after cleavage of the native
molecule witlh pepsin (resulting in a conformational change
in the M mc)lecule [1071), the peptide-specific antibodies
become react ive.
In view of overwhelming evidence for the presence of
exposed cons;erved epitopes on the native M molecule (28,
subunit and administered intranasally to the mice. About 1
month later, animals were challenged intranasally with live
streptococci and pharyngeal colonization was monitored for
15 days. It was found that mice immunized with the cholera
toxin B subunit-peptide complex showed a significant reduc-
tion in colonization compared with mice receiving cholera
toxin B subunit alone (Fig. 16). Thus, despite the fact that
conserved-region peptides are unable to evoke an opsonic
antibody response (106), these peptides have the capacity to
induce antibodies capable of influencing the colonization of
group A streptococci at the nasopharyngeal mucosa in this
model system.
The successful cloning of the emm-6.1 gene into vaccinia
virus and its high expression in mammalian cells infected
with the recombinant vaccinia virus may prove to be a
powerful vector for delivery of the M molecule to mucosal
surfaces (100). In an effort to express only the conserved-
region epitopes located in the C-terminal half of the mole-
cule, genetic engineering methods were used to truncate the
N-terminal half of the M6 gene and recombine the C-terminal
gene fragment into the vaccinia genome (VV:M6'). Western
blot analysis revealed that the cells infected with the
VV:M6' recombinant produce a molecule of about 30 kDa
that is reactive with a monoclonal antibody specific for the

TABLE 2. Throat culture results of mice vaccinated intranasally with either wild-type or recombinant vaccinia virus
No. culture positive"/total no. challenged (%. colonized or dead) at given days postchallenge"
Vaccinia virus vaccine
1 2 3 6 8 10
Wild type 19/25 (76) 18/25 (72) 19/25 (76) 18/25 (72) 21/25 (84) 19/25 (76)

Recombinant (VV:M6')" 3/28 (11) 9/28 (32) 7/28 (25) 8/28 (29) 8/28 (29) 8/28 (29)
Animals culture positive for group A streptococci or dead after streptococcal challenge.
' Significant difference (P < 0.005) at all time points between wild-type and recombinant virus-vaccinated mice by chi-square analysis.
Recombinant vaccinia virus contains the gene for the C-terminal half (conserved region) of the M6 molecule.
VOL. 2, 1989
C-terminal half of the M protein (V. A. Fischetti, W. M.
Hodges, and D. E. Hruby, Science, in press). Mice were
immunized intranasally with either the VV:M6' recombinant
or wild-type vaccinia virus and challenged intranasally and
orally 1 month later with 5 x 106 streptococci. Pharyngeal
cultures taken up to 10 days postchallenge revealed that the
VV:M6'-immunized animals differed significantly from con-
trols. Of the VV:M6'-immunized animals, only 16% of 152
total swabs taken were streptococcus positive, with 10 (6%)
yielding >100 CFU, whereas 69% of 115 swabs were posi-
tive in the wild-type group and 40 (35%) displayed >100
CFU. On average, >70% of the animals immunized with
wild-type virus were culture positive for group A strepto-
cocci at every swab day up to 10 days after challenge (Table
2). This is compared with <30% colonization of mice immu-
nized with the VV:M6' recombinant. These results confirm
and extend the previous studies that used synthetic peptides
from conserved regions of the M molecule to protect against
streptococcal colonization (27).
The achievement of non-type-specific protection in animal
models (27, 123, 216) and the evidence that protection can be
mediated by nonopsonic antibody (26) hold promise for an
alternative to type-specific M-protein vaccines. The greatest
challenge will lie in identifying those common peptides
which elicit the broadest protection without stimulating
antibodies cross-reactive with host tissue.
CONCLUDING REMARKS
The group A streptococcus utilizes a coiled-coil molecular
design for the M protein, a pattern commonly found in
eucaryotes and viruses (41, 42) which has not been thor-
oughly explored in bacteria. This type of structure offers a
number of distinct advantages to the organism. For instance,
the rodlike coiled-coil structure allows functional domains to
be located sufficiently away from the difficulties inherent in
microbial surfaces which in many cases actively stimulate
the alternative complement pathway. Because the coiled-
coil design may tolerate a large number of sequence changes,
yet maintain a common conformation, the molecule has the
capacity to change its immunodeterminants readily, allowing
the organism to avoid or delay immune recognition. Further-
more, the rodlike motif coupled with the plasticity of the
sequence allows this type of structure to be tailored for
several specific functions. This is accomplished by dividing
the molecule into discrete functional domains directed by
immunological and colonization pressures. Because of these
wide-ranging characteristics, the coiled-coil pattern is likely
to be found as commonly in bacteria as it is in eucaryotes.
ACKNOWLEDGMENTS
For their help in many of the studies on the structure, function.
and genetics of the M molecule, I acknowledge all of my collabora-
tors over the years. Discussions with Maclyn McCarty. John
Zabriskie, Emil Gotschlich, and the late Rebecca Lancefield have
helped shape my current understanding of the group A streptococ-
cus. Their time and patience are greatly appreciated. I thank Debra
Bessen, Emil Gotschlich. Kevin Jones. Belur Manjula, Vijaykumar
Pancholi, Olaf Schneewind. and June Scott for critical review and
suggestions for the manuscript.
This work was supported by Public Health Service grant A111822
from the National Institutes of Health.
LITERATURE CITED
1. Agarwal, B. L. 1981. Rheumatic heart disease unabated in
developing countries. Lancet i:910-911.
2. Akerstrom, B., T. Brodin, K. Reis, and L. Bjorck. 1985. Protein
STREPTOCOCCAL M PROTEIN 309
G: a powerful tool for binding and detection of monoclonal and
polyclonal antibodies. J. Immunol. 135:2589-2592.
3. Akerstrom, B., E. Nielsen, and L. Bjorck. 1987. Definition of
lgG- and albumin-binding regions of streptococcal protein G. J.
Biol. Chem. 262:13388-13391.
4. Aukenthaler, R., P. E. Hermans, and J. A. Washington. 1983.
Group G streptococcal bacteremia: clinical study and review of
the literature. Rev. Infect. Dis. 5:196-204.
5. Bach, M. A., D. E. Lewis, J. E. McClure, N. Parikh, H. M.
Rosenblatt, and W. T. Shearer. 1986. Monoclonal anti-actin
antibody recognizes a surface molecule on normal and trans-
formed human B lymphocytes: expression varies with phase of
cell cycle. Cell Immunol. 98:364-374.
6. Baker, C. J. 1974. Unusual occurrence of neonatal septicemia
due to group G streptococcus. Pediatrics. 53:568-570.
7. Barbour, A. 1988. Molecular mechanisms of antigenic varia-
tion in Borrelia. p. 365-377. In M. A. Horwitz (ed.). Bacteria-
host cell interaction. Alan R. Liss. Inc.. New York.
8. Bazeley, P. L. 1942. Studies with equine streptococci. 4.
Cross-immunity to Stcrpococcits equi. Aust. Vet. J. 18:189-
194.
9. Bazeley, P. L. 1943. Studies with equine streptococci. 5. Some
relations between virulence of Streptococcus equi and the
immune response of the host. Aust. Vet. J. 19:62-85.
10. Beachey, E. H. 1986. Type-specific opsonic antibodies evoked
with a synthetic peptide of streptococcal M protein conjugated
to polylysine without adjuvant. Infect. Immun. 51:362-364.
11. Beachey, E. H., G. L. Campbell, and I. Ofek. 1974. Peptic
digestion of streptococcal M protein. 11. Extraction of M
antigen from group A streptococci with pepsin. Infect. Immun.
9:891-896.
12. Beaches, E. H., H. Gras-Masse, A. Tarter, M. Jolivet, F.
Audibert, L. Chedid, and J. M. Seyer. 1986. Opsonic antibodies
evoked by hybrid peptide copies of types 5 and 24 streptococ-
cal M proteins synthesized in tandem. J. Exp. Med. 163:1451-
1458.
13. Beachey, E. H., and 1. Ofek. 1976. Epithelial cell binding of
group A streptococci by lipoteichoic acid on fimbriae denuded
of M protein. J. Exp. Med. 143:759-771.
14. Beachey, E. H., and J. Seyer. 1982. Primary structure and
immunochemistry of group A streptococcal M proteins. p.
401-410. In J. B. Robbins. J. C. Hill. and J. C. Sadoff (ed.).
Seminars in infectious disease. vol. 4. Bacterial vaccines.
Thieme-Stratton Inc.. New York.
15. Beachey, E. H., and J. M. Seyer. 1986. Protective and nonpro-
tective epitopes of chemically synthesized peptides of the
NH2-terminal region of type 6 streptococcal M protein. J.
Immunol. 136:2287-2292.
16. Beachey, E. H., J. M. Seyer, and J. B. Dale. 1987. Protective
immunogenicity and T lymphocyte specificity of a trivalent
hybrid peptide containing NH2-terminal sequences of types 5.
6. and 24 proteins synthesized in tandem. J. Exp. Med. 166:
647-656.
17. Beachey, E. H., J. M. Seyer, J. B. Dale, W. A. Simpson, and
A. H. Kang. 1981. Type-specific protective immunity evoked
by synthetic peptide of Streptococcus pvogenes M protein.
Nature (London) 292:457-459.
18. Beachey, E. H., J. M. Seyer, and A. H. Kang. 1978. Repeating
covalent structure of streptococcal M protein. Proc. Natl.
Acad. Sci. USA 75:3163-3167.
19. Beachey, E. H., J. M. Seyer, and A. H. Kang. 1980. Primary
structure of protective antigens of type 24 streptococcal M
protein. J. Biol. Chem. 255:6284-6289.
20. Beachey, E. H., A. Simpson, I. Ofek, D. L. Hasty, J. B. Dale,
and E. Whitnack. 1983. Attachment of Streptococcus pvogenes
to mammalian cells. Rev. Infect. Dis. 5(Suppl.):670-677.
21. Beachey, E. H., G. Stollerman, E. Y. Chiang, T. M. Chiang,
J. M. Seyer, and A. H. Kang. 1977. Purification and properties
of M protein extracted from group A streptococci with pepsin:
covalent structure of the amino terminal region of type 24 M
antigen. J. Exp. Med. 145:1469-1483.
22. Beaches, E. H., G. H. Stollerman, R. H. Johnson, I. Ofek, and
A. L. Bisno. 1979. Human immune response to immunization
310 FISCHETTI
with a structurally defined polypeptide fragment of streptococ-
cal M protein. J. Exp. Med. 150:862-877.
23. Beachey, E. H., A. Tartar, J. M. Seyer, and L. Chedid. 1984.
Epitope-specific protective immunogenicity of chemically syn-
thesised 13-, 18-, and 23-residue peptide fragments of strepto-
coccal M protein. Proc. Natl. Acad. Sci. USA 81:2203-2207.
24. Becker, C. G. 1964. Selection of group A streptococci rich in
M-protein from populations poor in M-protein. Am. J. Pathol.
44:51-60.
25. Besdine, R. W., and L. Pine. 1968. Preparation and description
of high-molecular-weight soluble surface antigens of a group A
streptococcus. J. Bacteriol. 96:1953-1958.
26. Bessen, D., and V. A. Fischetti. 1988. Passive acquired mucosal
immunity to group A streptococci by secretary immunoglobu-
lin A. J. Exp. Med. 167:1945-1950.
27. Bessen, D., and V. A. Fischetti. 1988. Influence of intranasal
immunization with synthetic peptides corresponding to con-
served epitopes of M protein on mucosal colonization by group
A streptococci. Infect. Immun. 56:2666-2672.
27a.Bessen, D., and V. A. Fischetti. 1988. Role of nonopsonic
antibody in protection against group A streptococcal infection,
p. 493-502. In L. Lasky (ed.), Technological advances in
vaccine development. Alan R. Liss, Inc., New York.
28. Bessen, D., K. F. Jones, and V. A. Fischetti. 1989. Evidence for
two distinct classes of streptococcal M protein and their
relationship to rheumatic fever. J. Exp. Med. 169:269-283.
29. Bisno, A. L. 1979. Alternate complement pathway activation
by group A streptococci: role of M protein. Infect. Immun. 26:
1172-1176.
30. Bisno, A. L. 1980. The concept of rheumatogenic and nonrheu-
matogenic group A streptococci, p. 789-803. In S. E. Read and
J. B. Zabriskie (ed.), Streptococcal diseases and the immune
response. Academic Press, Inc., New York.
31. Bisno, A. L., D. E. Craven, and W. R. McCabe. 1987. M
proteins of group G streptococci isolated from bacteremic
human infections. Infect. Immun. 55:753-757.
32. Breese, B. B., and C. B. Hall. 1978. Beta hemolytic streptococ-
cal diseases. Houghton Mifflin, Boston.
33. Brinton, C. C. 1965. Structure, function, synthetic and genetic
control of bacterial pili and molecular model for DNA and
RNA transport in Gram negative bacteria. Trans. N.Y. Acad.
Sci. 27:1003.
34. Bronze, M. S., E. H. Beachey, and J. B. Dale. 1988. Protective
and heart-crossreactive epitopes located within the N-terminus
of type 19 streptococcal M protein. J. Exp. Med. 167:1849-
1859.
35. Caparon, M. G., and J. R. Scott. 1987. Identification of a gene
that regulates expression of M protein, the major virulence
determinant of group A streptococci. Proc. Natl. Acad. Sci.
USA 84:8677-8681.
36. Chalovich, J. M., and V. A. Fischetti. 1986. Crosslinking of
actin filaments and inhibition of actomyosin subfragment-1
ATPase activity by streptococcal M6 protein. Arch. Biochem.
Biophys. 245:37-43.
37. Chedid, L., M. Jolivet, F. Audibert, G. Przewlocki, E. H.
Beachey, H. Gras-Masse, and A. Tartar. 1983. Antibody re-
sponses elicited by a polyvalent vaccine containing synthetic
diphtheric, streptococcal, and hepatitis peptides coupled to the
same carrier. Biochem. Biophys. Res. Commun. 117:908-915.
38. Cheung, A. L., A. S. Bayer, J. Peters, and J. I. Ward. 1987.
Analysis by gel electrophoresis, Western blot, and peptide
mapping of protein A heterogeneity in Staphylococcus aureus
strains. Infect. Immun. 55:843-847.
39. Chhatwal, G. S., I. S. Dutra, and H. Blobel. 1985. Fibrinogen
binding inhibits the fixation of the third component of human
complement on surface of groups A, B, C, and G streptococci.
Microbiol. Immunol. 29:973-980.
40. Cleary, P. P., Z. Johnson, and L. W. Wannamaker. 1975.
Genetic instability of M protein and serum opacity factor of
group A streptococci: evidence suggesting extrachromasomal
control. Infect. Immun. 12:109-118.
41. Cohen, C., and D. A. D. Parry. 1986. Alpha-helical coiled-
coils-a widespread motif in proteins. Trends Biochem. Sci.
11:245-248.
CLIN. MICROBIOL. REV.
42. Cohen, C., and G. N. Phillips. 1981. Spikes and fimbriae:
alpha-helical proteins form surface projections on microorgan-
isms. Proc. Natl. Acad. Sci. USA 78:5303-5304.
43. Cohen, J. O., H. Gross, and W. K. Harrell. 1977. Immunoge-
nicity and characteristics of M protein released by phage-
associated lysin from group A streptococci types 1 and 23. J.
Med. Microbiol. 10:179-194.
44. Cole, R. M., and J. J. Hahn. 1962. Cell wall replication in
Streptococciis pyogenes. Science 135:722-724.
45. Connell, T. D., W. J. Black, T. H. Kawula, D. S. Barritt, J. A.
Dempsey, K. Kverneland, Jr., A. Stephenson, B. S. Schepart,
G. L. Murphy, and J. G. Cannon. 1988. Recombination among
protein II genes of Neisseria gonorrhoeae generates new
coding sequences and increases structural variability in the
protein II family. Mol. Micorbiol. 2:227-236.
46. Cunningham, M. W., and E. H. Beachey. 1974. Peptic digestion
of streptococcal M protein I. Effect of digestion at suboptimal
pH upon the biological and immunological properties of puri-
fied M protein extracts. Infect. Immun. 9:244-248.
47. Cunningham, M. W., and E. H. Beachey. 1975. Immunochem-
ical properties of streptococcal M protein purified by isoelec-
tric focusing. J. Immunol. 115:1002-1006.
48. Cunningham, M. W., N. K. Hall, K. K. Krisher, and A. M.
Spanier. 1986. A study of anti-group A streptococcal monoclo-
nal antibodies cross reactive with myosin. J. Immunol. 136:
293-299.
49. Cunningham, M. W., and S. M. Russell. 1983. Study of
heart-reactive antibody in antisera and hybridoma culture
fluids against group A streptococci. Infect. Immun. 42:531-
538.
50. Cunningham, M. W., and R. A. Swerlick. 1986. Polyspecificity
of antistreptococcal murine monoclonal antibodies and their
implications in autoimmunity. J. Exp. Med. 164:998-1012.
51. D'Alessandri, R., G. Plotkin, R. M. Kluge, M. K. Wittner,
E. N. Fox, A. Dorfman, and R. H. Waldman. 1978. Protective
studies with group A streptococcal M protein vaccine. III.
Challenge of volunteers after systemic or intranasal immuniza-
tion with type 3 or type 12 group A Streptococcus. J. Infect.
Dis. 138:712-718.
52. Dale, J. B., and E. H. Beachey. 1984. Unique and common
protective epitopes among different serotypes of group A
streptococcal M proteins defined with hybridoma antibodies.
Infect. Immun. 46:267-269.
53. Dale, J. B., and E. H. Beachey. 1985. Multiple, heart-cross-
reactive epitopes of streptococcal M proteins. J. Exp. Med.
161:113-122.
54. Dale, J. B., and E. H. Beachey. 1985. Epitopes of streptococcal
M proteins shared with cardiac myosin. J. Exp. Med. 162:583-
591.
55. Dale, J. B., and E. H. Beachey. 1986. Localization of protective
epitopes of the amino terminus of type 5 streptococcal M
protein. J. Exp. Med. 163:1191-1202.
56. Dale, J. B., and E. H. Beachey. 1986. Sequence of myosin-
cross-reactive epitopes of streptococcal M protein. J. Exp.
Med. 164:1785-1790.
57. Dale, J. B., J. M. Seyer, and E. H. Beachey. 1983. Type-specific
immunogenicity of a chemically synthesized peptide fragment
of type 5 streptococcal M protein. J. Exp. Med. 158:1727-1732.
58. Daynes, R. A., and C. H. Armstrong. 1973. An antiphagocytic
factor associated with group E streptococci. Infect. Immun. 7:
298-304.
59. Dillon, H. C. 1972. Streptococcal infections of the skin and
their complications: impetigo and nephritis, p. 571-587. In
L. W. Wannamaker and J. M. Matsen (ed.), Streptococci and
streptococcal diseases. Academic Press, Inc., New York.
60. Ellen, R. P., and R. J. Gibbons. 1972. M protein-associated
adherence of Streptococcus pyogenes to epithelial surfaces:
prerequisite for virulence. Infect. Immun. 5:826-830.
61. Fahnestock, S. R., P. Alexander, J. Nagle, and D. Filpula. 1986.
Gene for an immunoglobulin-binding protein from a group G
streptococcus. J. Bacteriol. 167:870-880.
62. Farabaugh, P. J., V. Schmeissner, M. Hofer, and J. H. Miller.
1978. Genetic studies of the lac repressor VII on the molecular
nature of spontaneous hot spots in the lacdI gene of Escherichia
VOL. 2, 1989
coli. J. Mol. Biol. 126:847-863.
62a.Fenderson, P. G., V. A. Fischetti, and M. W. Cunningham.
1989. Tropomyosin shares immunologic epitopes with group A
streptococcal M protein. J. Immunol. 142:2475-2481.
63. Fillit, H., S. P. Damle, J. D. Gregory, C. Volin, T. Poon-King,
and J. B. Zabriskie. 1985. Sera from patients with poststrepto-
coccal glomerulonephritis contain antibodies to glomerular
heparin sulfate proteoglycan. J. Exp. Med. 161:227-287.
64. Fischetti, V. A. 1977. Streptococcal M protein extracted by
nonionic detergent. II. Analysis of the antibody response to the
multiple antigenic determinants of the M-protein moleule. J.
Exp. Med. 146:1108-1123.
65. Fischetti, V. A. 1978. Streptococcal M protein extracted by
nonionic detergent. Ill. Correlation between immunological
cross-reactions and structural similarities with implications for
antiphagocytosis. J. Exp. Med. 147:1771-1778.
66. Fischetti, V. A. 1983. Requirements for the opsonic activity of
human IgG directed to type 6 group A streptococci: net basic
charge and intact Fc region. J. Immunol. 130:896-902.
67. Fischetti, V. A., and D. Bessen. 1988. Effect of mucosal
antibodies to M protein on colonization by group A strepto-
cocci, p. 128-142. In L. Switalski, M. Hook, and E. H.
Beachey (ed.), Molecular mechanisms of microbial adhesion.
Springer-Verlag, New York.
68. Fischetti, V. A., and M. Fazio-Zanakis. 1985. Electron micros-
copy of the group A streptococcal cell surface, p. 180-183. In
Y. Kimura, S. Kotani, and Y. Shiokawa (ed.), Recent ad-
vances in streptococci and streptococcal diseases. Reedbooks,
Ltd., Chertsey, England.
69. Fischetti, V. A., E. C. Gotschlich, and A. W. Bernheimer. 1971.
Purification and physical properties of group C streptococcal
phage-associated lysin. J. Exp. Med. 133:1105-1117.
70. Fischetti, V. A., E. C. Gotschlich, G. Siviglia, and J. B.
Zabriskie. 1976. Streptococcal M protein extracted by non-
ionic detergent. 1. Properties of the antiphagocytic and type
specific molecules. J. Exp. Med. 144:32-53.
71. Fischetti, V. A., M. Jarymowycz, K. F. Jones, and J. R. Scott.
1986. Streptococcal M protein size mutants occur at high
frequency within a single strain. J. Exp. Med. 164:971-980.
72. Fischetti, V. A., K. F. Jones, S. K. Hollingshead, and J. R.
Scott. 1987. Molecular and antigenic architecture of the anti-
phagocytic group A streptococcal M protein, p. 123-131. In R.
Rott and W. Goebel (ed.), Molecular basis of viral and micro-
bial pathogenesis. Springer-Verlag. Berlin.
73. Fischetti, V. A., K. F. Jones, B. N. Manjula, and J. R. Scott.
1984. Streptococcal M6 protein expressed in Escherichia /oli.
Localization, purification and comprison with streptococcal-
derived M protein. J. Exp. Med. 159:1083-1095.
74. Fischetti, V. A., K. F. Jones, and J. R. Scott. 1985. Size
variation of the M protein in group A streptococci. J. Exp.
Med. 161:1384-1401.
75. Fischetti, V. A., and B. N. Manjula. 1982. Biologic and immu-
nologic implications of the structural relationship between
streptococcal M protein and mammalian tropomyosin, p. 411-
418. In J. B. Robbins, J. C. Hill. and J. C. Sadoff (ed.).
Seminars in infectious disease. vol. 4. Bacterial vaccines.
Thieme-Stratton, Inc.. New York.
76. Fischetti, V. A., B. N. Manjula, M. Fazio, P. F. Flicker, C.
Cohen, and G. N. Phillips. 1982. The alpha-helical coiled-coil
structure of streptococcal M protein, p. 85-87. In S. E. Holm
and P. Christensen (ed.), Basic concepts of streptococci and
streptococcal diseases. Reedbooks, Ltd., Chertsey. England.
77. Fischetti, V. A., D. A. D. Parry, B. L. Trus, S. K. Hollingshead,
J. R. Scott, and B. N. Manjula. 1988. Conformational charac-
teristics of the complete sequence of group A streptococcal M6
protein. Proteins Struct. Func. Genet. 3:60-69.
78. Fischetti, V. A., and M. Windels. 1988. Mapping the immuno-
determinants of the complete streptococcal M6 protein mole-
cule: Identification of an immunodominant region. J. Immunol.
141:3592-3599.
79. Fischetti, V. A., J. B. Zabriskie, and E. C. Gotschlich. 1974.
Physical, chemical, and biological properties of type 6 M-
protein extracted with purified streptococcal phage-associated
lysin, p. 26-37. In M. J. Haverkorn (ed.). Streptococcal
STREPTOCOCCAL M PROTEIN 311
disease and the community. Excerpts Medica. Amsterdam.
80. Fox, E. N. 1972. Discussion of chapters 13 and 15-17, p. 301-
309. In L. W. Wannamaker and J. M. Matsen (ed.). Strepto-
cocci and streptococcal diseases. Academic Press. Inc.. New
York.
81. Fox, E. N. 1974. M proteins of group A streptococci. Bacteriol.
Rev. 38:57-86.
82. Fox, E. N., and M. K. Wittner. 1965. The multiple molecular
structure of the M proteins of group A streptococci. Proc.
Natl. Acad. Sci. USA 54:1118-1125.
83. Fox, E. N., and M. K. Wittner. 1968. Antigenicity of the M
protein of group A hemolytic streptococci. IV. Cross-reac-
tivity between serotypes. J. Immunol. 100:39-45.
84. Fox, E. N., and M. K. Wittner. 1969. New observations on the
structure and antigenicity of the M proteins of the group A
streptococcus. Immunochemistry 6:11-24.
85. Galan, J. E., and J. F. Timoney. 1987. Molecular analysis of
the M protein of Streptococcus equi and cloning and expres-
sion of the M protein gene in Escheric hiai (0oi. Infect. Immun.
53:3181-3187.
86. Gastanaduy, A. S., E. L. Kaplan, B. B. Huwe, C. McKay, and
L. W. Wannamaker. 1980. Failure of penicillin to eradicate
group A streptococci during an outbreak of pharyngitis. Lancet
ii:498-502.
87. Goroncy-Bermes, P., J. B. Dale, E. H. Beachey, and W.
Opferkuch. 1987. Monoclonal antibody to human renal glomer-
uli cross-reacts with streptococcal M protein. Infect. Immun.
55:2416-2419.
88. Goto, Y., and K. Hamaguchi. 1987. Role of amino-terminal
residues in the folding of the constant fragment of the immu-
noglobulin light chain. Biochemistry 26:1879-1884.
89. Guss, B., M. Eliasson, A. Olsson, M. Uhlen, A.-K. Frej, H.
Jornvall, J.-I. Flock, and M. Lindberg. 1986. Structure of the
IgG-binding regions of streptococcal protein G. EMBO J. 5:
1567-1575.
89a.Guss, B., M. Uhlen, B. Nilsson, M. Lindberg, J. Sjoquist, and J.
Sjodahl. 1984. Region X. the cell-wall-attachment part of
staphylococcal protein A. Eur. J. Biochem. 138:413-420.
90. Haanes-Fritz, E., W. Kraus, V. Burdett, J. B. Dale, and E. H.
Beachey. 1988. Comparison of the leader sequences of four
group A streptococcal M protein genes. Nucleic Acids Res. 16:
4667-4677.
91. Hafez, K., A. M. ElKholy, and R. R. Facklam. 1981. Extraction
of group A streptococcal M protein with nitrous acid. J. Clin.
Microbiol. 14:530-533.
92. Hamburger, M., M. J. Green, and V. G. Hamburger. 1945. The
problem of the 'Dangerous Carrier" of hemolytic streptocci.
I. Number of hemolytic streptococci expelled by carriers with
positive and negative nose cultures. J. Infect. Dis. 77:68-81.
93. Hirsch, J. G., and A. B. Church. 1960. Studies of phagocytosis
of group A streptococci by polymorphonuclear leucocytes in
vitro. J. Exp. Med. 111:309-322.
94. Hollingshead, S. K., V. A. Fischetti, and J. R. Scott. 1986.
Complete nucleotide sequence of type 6 M protein of the group
A streptococcus: repetitive structure and membrane anchor. J.
Biol. Chem. 261:1677-1686.
95. Hollingshead, S. K., V. A. Fischetti, and J. R. Scott. 1987. Size
variation in group A streptococcal M protein is generated by
homologous recombination between intragenic repeats. Mol.
Gen. Genet. 207:196-203.
96. Hollingshead, S. K., V. A. Fischetti, and J. R. Scott. 1987. A
highly conserved region present in transcripts encoding heter-
ologous M proteins of group A streptococcus. Infect. Immun.
55:3237-3239.
97. Hook, E. W., R. R. Wagner, and R. C. Lancefield. 1960. An
epizootic in Swiss mice caused by a group A streptococcus.
newly designated type 50. Am. J. Hyg. 72:111-119.
98. Horstmann, R. D., H. J. Sievertsen, J. Knobloch, and V. A.
Fischetti. 1988. Antiphagocytic activity of streptococcal M
protein: selective binding of complement control protein Fac-
tor H. Proc. Nati. Acad. Sci. USA 85:1657-1661.
99. Hosein, B., M. McCarty, and V. A. Fischetti. 1979. Amino acid
sequence and physicochemical similarities between strepto-
coccal M protein and mammalian tropomyosin. Proc. Natl.
312 FISCHETTI
Acad. Sci. USA 76:3765-3768.
100. Hruby, D. E., W. M. Hodges, E. M. Wilson, C. A. Franke, and
V. A. Fischetti. 1988. Expression of streptococcal M protein in
mammalian cells. Proc. NatI. Acad. Sci. USA 85:5714-5717.
101. Hreyniewicz, W., B. Lipinski, and J. Jeljaszewicz. 1972. Nature
of the interaction between M protein of streptococcus pyo-
genes and fibrinogen. J. Infect. Dis. 125:626-630.
102. Huxley, H. E., and M. F. Perutz. 1953. New X-ray evidence on
the configuration of polypeptide chains. Nature (London) 167:
1053-1054.
103. Jacks-Weis, J., Y. Kim, and P. P. Cleary. 1982. Restricted
deposition of C3 on M+ group A streptococci: correlation with
resistance to phagocytosis. J. Immunol. 128:1897-1902.
104. Jolivet, M., F. Audibert, E. H. Beachey, A. Tartar, H. Gras-
Masse, and L. Chedid. 1983. Epitope specific immunity elicited
by a synthetic streptococcal antigen without carrier or adju-
vant. Biochem. Biophys. Res. Commun. 117:359-366.
105. Jones, K. F., and V. A. Fischetti. 1987. Biological and immu-
nochemical identity of M protein and group G streptococci
with M protein on group A streptococci. Infect. Immun. 55:
502-506.
106. Jones, K. F., and V. A. Fischetti. 1988. The importance of the
location of antibody binding on the M6 protein for opsonization
and phagocytosis of group A M6 streptococci. J. Exp. Med.
167:1114-1123.
107. Jones, K. F., S. K. Hollingshead, J. R. Scott, and V. A.
Fischetti. 1988. Spontaneous M6 protein size mutants of group
A streptococci display variation in antigenic and opsonogenic
epitopes. Proc. Natl. Acad. Sci. USA 85:8271-8275.
108. Jones, K. F., S. A. Khan, B. W. Erickson, S. K. Hollingshead,
J. R. Scott, and V. A. Fischetti. 1986. Immunochemical local-
ization and amino acid sequence of cross-reactive epitopes
within the group A streptococcal M6 protein. J. Exp. Med.
164:1226-1238.
109. Jones, K. F., B. N. Manjula, K. H. Johnston, S. K. Hollings-
head, J. R. Scott, and V. A. Fischetti. 1985. Location of
variable and conserved epitopes among the multiple serotypes
of streptococcal M protein. J. Exp. Med. 161:623-628.
110. Kantor, F. S. 1965. Fibrinogen precipitating by streptococcal
M protein. I. Identity of the reactants and stoichiometry of the
reaction. J. Exp. Med. 121:849-859.
111. Kaplan, E. L., A. S. Gastanaduy, and B. B. Huwe. 1981. The
role of the carrier in treatment failures after antibiotic therapy
for group A streptococci in the upper respiratory tract. J. Lab.
Clin. Med. 98:326-335.
112. Kaplan, M. H. 1963. Immunologic relations of streptococcal
and tissue antigens. I. Properties of an antigen in certain strains
of group A streptococci exhibiting an immunologic cross
reaction with human heart tissue. J. Immunol. 90:595-606.
113. Kaplan, M. H., and M. Meyeserian. 1962. An immunological
cross-reaction between group A streptococcal cells and human
heart tissue. Lancet i:706.
114. Kehoe, M. A., L. Miller, T. P. Poirier, E. H. Beachey, H. Lee,
and D. Harrington. 1987. Genetics of type 5 M protein of
Streptococcus pyogenes, p. 112-116. In J. J. Ferretti and R.
Curtiss III (ed.), Streptococcal genetics. American Society for
Microbiology, Washington, D.C.
115. Kehoe, M. A., T. P. Poirier, E. H. Beachey, and K. N. Timmis.
1985. Cloning and genetic analysis of serotype 5 M protein
determinant of group A streptococci: evidence for multiple
copies of the M5 determinant in the Streptococcus pvogenes
genome. Infect. Immun. 48:190-197.
116. Khandke, K. M., T. Fairwell, A. S. Acharya, B. L. Trus, and
B. N. Manjula. 1988. Complete amino acid sequence of strep-
tococcal PepM49 protein, a nephritis associated serotype:
conserved conformational design among sequentially distinct
M protein serotypes. J. Biol. Chem. 263:5075-5082.
117. Khandke, K. M., T. Fairwell, and B. N. Manjula. 1987.
Difference in the structural features of streptococcal M pro-
teins from nephritogenic and rheumatogenic serotypes. J. Exp.
Med. 166:151-162.
118. Kilian, M., J. Mestecky, and M. W. Russell. 1988. Defense
mechanisms involving Fc-dependent functions of immunoglob-
CLIN. MICROBIOL. REV.
ulin A and their subversion by bacterial immunoglobulin A
protease. Microbiol. Rev. 52:296-303.
119. Kraus, W., and E. H. Beachey. 1988. Renal autoimmune
epitope of group A streptococci specified by M protein tet-
rapeptide Ile-Arg-Leu-Arg. Proc. Natl. Acad. Sci. USA 85:
4516-4520.
120. Kraus, W., E. Haanes-Fritz, P. P. Cleary, J. M. Seyer, J. B.
Dale, and E. H. Beachey. 1987. Sequence and type-specific
immunogenicity of the amino-terminal region of type 1 strep-
tococcal M protein. J. Immunol. 139:3084-3090.
121. Krisher, K., and M. W. Cunningham. 1985. Myosin: a link
between streptococci and heart. Science 227:413-415.
122. Kronvall, G., C. Schonbeck, and E. Myhre. 1979. Fibrinogen
binding structures in beta hemolytic streptococci group A, C
and G: comparisons with receptors for IgG and aggregated
beta-2 microglobulin. Acta Pathol. Microbiol. Scand. Sect. B
87:303-310.
123. Kurl, D. N., A. Stjernquist-Desatnik, C. Schalen, and P. Christ-
ensen. 1985. Induction of local immunity to group A strepto-
cocci type M50 in mice by non-type-specific mechanisms. Acta
Pathol. Microbiol. Scand. Sect. B. 93:401-405.
124. Kuttner, A. G., and E. Krumwiede. 1941. Observations on the
effect of streptococcal upper respiratory infections on rheu-
matic children: a three year study. J. Clin. Invest. 20:273-279.
125. Kuttner, A. G., and T. F. Lenert. 1944. The occurrence of
bacteriostatic properties in the blood of patients after recovery
from streptococcal pharyngitis. J. Clin. Invest. 23:151-158.
126. Lancefield, R. C. 1928. The antigenic complex of Streptococ-
culs hinolvticius. 1. Demonstration of a type-specific substance
in extracts of Streptococcus Iiemnol'ticus. J. Exp. Med. 47:91-
103.
127. Lancefield, R. C. 1948. Differentiation of group A streptococci
with a common R antigen into three serological types with
special reference to the bactericidal test. J. Exp. Med. 106:
525-544.
128. Lancefield, R. C. 1959. Persistence of type specific antibodies
in man following infection with group A streptococci. J. Exp.
Med. 110:271-292.
129. Lancefield, R. C. 1962. Current knowledge of the type specific
M antigens of group A streptococci. J. Immunol. 89:307-313.
130. Lipman, D. J., and W. R. Pearson. 1985. Rapid and sensitive
protein similarity searches. Science 227:1435-1441.
131. Manjula, B. N. 1988. Molecular aspects of the phagocytosis
resistance of group A streptococci. Eur. J. Epidemiol. 4:289-
300.
132. Manjula, B. N., A. S. Acharya, T. Fairwell, and V. A. Fischetti.
1986. Antigenic domains of the streptococcal Pep M5 protein:
localization of epitopes crossreactive with type 6 M protein
and identification of a hypervariable region of the M molecule.
J. Exp. Med. 163:129-138.
133. Manjula, B. N., A. S. Acharya, S. M. Mische, T. Fairwell, and
V. A. Fischetti. 1984. The complete amino acid sequence of a
biologically active 197-residue fragment of M protein from type
5 group A streptococci. J. Biol. Chem. 259:3686-3693.
134. Manjula, B. N., A. S. Acharya, S. M. Mische, T. Fairwell, B. L.
Trus, and V. A. Fischetti. 1985. The complete amino acid
sequence of streptococcal Pep M5 protein: presence of two
structurally distinct domains and implications to biological
functions, p. 185-187. In Y. Kimura, S. Kotani, and Y.
Shiokawa (ed.). Recent advances in streptococci and strepto-
coccal diseases. Reedbooks, Ltd., Chertsey. England.
135. Manjula, B. N., T. Fairwell, A. S. Acharya, and K. M.
Khandke. 1988. Domain structure of streptococcal M protein:
Correlation with sites of limited proteolysis with pepsin. J.
Protein. Chem. 7:262-263.
136. Manjula, B. N., and V. A. Fischetti. 1980. Tropomyosin-like
seven residue periodicity in three immunologically distinct
streptococcal M proteins and its impliation for the antiphago-
cytic property of the molecule. J. Exp. Med. 151:695-708.
137. Manjula, B. N., and V. A. Fischetti. 1980. Studies on group A
streptococcal M proteins: purification of type 5 M-protein and
comparison of its amino terminal sequence with two immuno-
logically unrelated M-protein molecules. J. Immunol. 124:261-
VOL. 2, 1989
267.
138. Manjula, B. N., and V. A. Fischetti. 1986. Sequence homology
of group A streptococcal PepM5 protein with other coiled-coil
proteins. Biochem. Biophys. Res. Commun. 140:684-690.
139. Manjula, B. N., S. M. Mische, and V. A. Fischetti. 1983.
Primary structure of streptococcal pep M5 protein: absence of
extensive sequence repeats. Proc. Natl. Acad. Sci. USA 80:
5475-5479.
140. Manjula, B. N., M. L. Schmidt, and V. A. Fischetti. 1985.
Unimpaired function of human phagocytes in the presence of
phagocytosis-resistant group A streptococci. Infect. Immun.
50:610-613.
141. Manjula, B. N., B. L. Trus, and V. A. Fischetti. 1985. Presence
of two distinct regions in the coiled-coil structure of the
streptococcal Pep M5 protein: relationship to mammalian
coiled-coil proteins and implications to its biological proper-
ties. Proc. Natl. Acad. Sci. USA 82:1064-1068.
142. Maxted, W. R. 1948. Occurrence of the M substance of type 28
group A in streptococci of Lancefield groups B, C, and G. J.
Gen. Microbiol. 3:1-6.
143. Maxted, W. R., and E. V. Potter. 1967. The presence of type 12
M-protein antigen in group G streptococci. J. Gen. Microbiol.
49:119-125.
144. McLachlan, A. D., and J. Karn. 1983. Periodic features in the
amino acid sequence of nematode myosin rod. J. Mol. Biol.
164:605-626.
145. McLachlan, A. D., and M. Stewart. 1975. Tropomyosin coiled-
coil interactions: evidence for an unstaggered structure. J.
Mol. Biol. 98:393-308.
146. Meyer, T. F., N. Malwer, and M. So. 1981. Pilus expression in
Neisseria gonorrhoeae involves chromosomal rearrangement.
Cell 30:45-51.
147. Miller, L., V. Burdett, T. P. Poirier, L. D. Gray, E. H. Beachey,
and M. A. Kehoe. 1988. Conservation of protective and non-
protective epitopes in M proteins of group A streptococci.
Infect. Immun. 56:2198-2204.
148. Miller, L., L. Gray, E. H. Beachey, and M. A. Kehoe. 1988.
Antigenic variation among group A streptococcal M proteins:
nucleotide sequence of the serotype 5 M protein gene and its
relationship with genes encoding types 6 and 24 M proteins. J.
Biol. Chem. 263:5668-5673.
149. Miller, V. L., and J. J. Mekalanos. 1984. Cholera toxin is
positively regulated at the transcriptional level by ToxR. Proc.
Natl. Acad. Sci. USA 81:3471-3475.
150. Mische, S. M., B. N. Manjula, and V. A. Fischetti. 1987.
Relation of streptococcal M protein with human and rabbit
tropomyosin: the complete amino acid sequence of human
cardiac alpha tropomyosin, a highly conserved contractile
protein. Biochem. Biophys. Res. Commun. 142:813-818.
151. Mohr, D. N., D. J. Feist, J. A. Washington, and P. E. Hermans.
1979. Infections due to group C streptococci in man. Am. J.
Med. 66:450-456.
152. Moravek, L., 0. Kuhnemund, J. Havlicek, P. Kopecky, and M.
Pavlik. 1986. Type 1 M protein of streptococcus pyogenes.
N-terminal sequence and peptic fragments. FEBS Lett. 208:
435-438.
153. Mouw, A. R., E. H. Beachey, and V. Burdett. 1988. Molecular
evolution of streptococcal M protein: cloning and nucleotide
sequence of type 24 M protein gene and relation to other genes
of Streptococcus pyogenes. J. Bacteriol. 170:676-684.
154. Muller-Eberhard, H. J., and 0. Gotze. 1972. C3 proactivator
convertase and its mode of action. J. Exp. Med. 135:103-112.
155. Muller-Eberhard, H. J., and R. D. Schreiber. 1980. Molecular
biology and chemistry of the alternative pathway of comple-
ment. Adv. Immunol. 29:1-53.
156. Munoz, E., J.-M. Ghuysen, M. Leyh-Bouille, J.-F. Petit, and R.
Tinelli. 1966. Structural variations in bacterial cell wall pepti-
doglycan studied with Streptomyces. F1 endo-N-acetylmuram-
idase. Biochemistry 5:3090-3098.
157. Ofek, I., S. Berger-Rabinowitz, and A. M. Davies. 1969. Op-
sonic activity of type-specific streptococcal antibodies. Isr. J.
Med. Sci. 5:293-299.
158. Ofek, I., W. A. Simpson, and E. H. Beachey. 1982. Formation
STREPTOCOCCAL M PROTEIN 313
of molecular complexes between a structurally defined M
protein and acylated or deacylated lipoteichoic acid of Strep-
tococcus pyogenes. J. Bacteriol. 149:426-433.
159. Olsson, A., M. Eliasson, B. Guss, B. Nilsson, U. Hellman, M.
Lindberg, and M. Uhlen. 1987. Structure and evolution of the
repetitive gene encoding streptococcal protein G. Eur. J.
Biochem. 168:319-324.
160. Owen, M. J., J. Auger, B. H. Barber, A. J. Edwards, F. S.
Walsh, and M. J. Crumpton. 1978. Actin may be present on the
lymphocyte surface. Proc. Natl. Acad. Sci. USA 75:4484-
4488.
161. Pancholi, V., and V. A. Fischetti. 1988. Isolation and charac-
terization of the cell-associated region of group A streptococ-
cal M6 protein. J. Bacteriol. 170:2618-2624.
162. Pappenheimer, A. M., J. W. Williams, and C. A. Zittle. 1942.
The antigenic structure of hemolytic streptococci of Lancefield
group A. IX. Some physical properties of the M protein. J.
Immunol. 43:61-63.
163. Parry, D. A. D. 1981. Structure of rabbit skeletal myosin: an
analysis of the amino acid sequences of two fragments from the
rod region. J. Mol. Biol. 153:459-464.
164. Parry, D. A. D. 1982. Coiled-coils in alpha-helix-containing
proteins: analysis of the residue types within the heptad repeat
and the use of these data in the prediction of coiled-coils in
other proteins. Biosci. Rep. 2:1017-1024.
165. Parry, D. A. D., W. G. Crewther, R. D. B. Fraser, and T. P.
MacRae. 1977. Structure of alpha keratin: structural implica-
tions of the amino acid sequences of the type I and type 11
chain segments. J. Mol. Biol. 113:449.
166. Parry, D. A. D., and R. D. B. Fraser. 1985. Intermediate
filament structure. 1. Analysis of IF protein sequence data. Int.
J. Biol. Macromol. 7:203-213.
167. Peterson, P. A., L. Rask, K. Sege, L. Klareskog, H. Anundi,
and L. Ostberg. 1975. Evolutionary relationship between
immunoglobulins and transplantation antigens. Proc. Natl.
Acad. Sci. USA 72:1612-1616.
168. Peterson, P. K., D. Schmeling, P. P. Cleary, B. J. Wilkinson, Y.
Kim, and P. G. Quie. 1979. Inhibition of alternative comple-
ment pathway opsonization by group A streptococcal M pro-
tein. J. Infect. Dis. 139:575-585.
169. Phillips, G. N., P. F. Flicker, C. Cohen, B. N. Manjula, and
V. A. Fischetti. 1981. Streptococcal M protein: alpha-helical
coiled-coil structure and arrangement on the cell surface. Proc.
Natl. Acad. Sci. USA 78:4689-4693.
170. Poirier, T. P., M. A. Kehoe, and E. H. Beachey. 1988. Humoral
and mucosal immunity against group A streptococci evoked by
attenuated aroA Salmonella typhimurium expressing cloned
streptococcal M protein. J. Exp. Med. 168:25-32.
171. Poirier, T. P., M. A. Kehoe, E. Whitnack, and E. H. Beachey.
1987. Surface expression of type 5 M protein from Streptococ-
cus pyogenes in Streptococcus sanguis, p. 117-123. In J. J.
Ferretti and R. Curtiss III (ed.), Streptococcal genetics. Amer-
ican Society for Microbiology, Washington, D.C.
172. Polly, S. M., R. H. Waldman, P. High, M. K. Wittner, A.
Dorfman, and E. N. Fox. 1975. Protective studies with a group
A streptococcal M protein vaccine. II. Challenge of volunteers
after local immunization in the upper respiratory tract. J.
Infect. Dis. 131:217-224.
173. Quinn, R. W., R. Vander Zwaag, and P. N. Lowry. 1985.
Acquisition of group A streptococcal M protein antibodies.
Pediatr. Infect. Dis. 4:374-378.
174. Rammelkamp, C. H., and R. S. Weaver. 1953. Acute glomer-
ulonephritis. The significance of the variation in the incidence
of the disease. J. Clin. Invest. 320:345-358.
175. Robbins, J. C., J. G. Spanier, S. J. Jones, W. J. Simpson, and
P. P. Cleary. 1987. Streptococcus pyogenes type 12 M protein
regulation by upstream sequences. J. Bacteriol. 169:5633-
5640.
176. Rothbard, S. J. 1945. Bacteriostatic effect of human sera on
group A streptococci. I. Type-specific antibodies in sera of
patients convalescing from group A streptococcal pharyngitis.
J. Exp. Med. 82:93-105.
177. Rotta, J., R. C. Krause, R. C. Lancefield, W. Everly, and H.
314 FISCHETTI
Lackland. 1971. New approaches for the laboratory recogni-
tion of M types of group A streptococci. J. Exp. Med. 134:
1298-1315.
178. Russell, H., and R. R. Facklam. 1975. Guanidine extraction of
streptococcal M protein. Infect. Immun. 12:679-686.
178a.Sabatini, D. D., G. Kreibich, T. Morimoto, and M. Adesnik.
1982. Mechanism for the incorporation of proteins in mem-
branes and organelles. J. Cell Biol. 92:1-22.
179. Sargent, S. J., E. H. Beachey, C. E. Corbett, and J. B. Dale.
1987. Sequence of protective epitopes of streptococcal M
proteins shared with cardiac sarcolemmal membranes. J. Im-
munol. 139:1285-1290.
180. Scott, J. R., and V. A. Fischetti. 1983. Expression of strepto-
coccal M protein in Escherichia coli. Science 221:758-760.
181. Scott, J. R., P. C. Guenther, L. M. Malone, and V. A. Fischetti.
1986. Conversion of an M- group A streptococcus to M' by
transfer of a plasmid containing an M6 gene. J. Exp. Med. 164:
1641-1651.
182. Scott, J. R., S. K. Hollingshead, and V. A. Fischetti. 1986.
Homologous regions within M protein genes in group A
streptococci of different serotypes. Infect. Immun. 52:609-612.
183. Scott, J. R., W. M. Pulliam, S. K. Hollingshead, and V. A.
Fischetti. 1985. Relationship of M protein genes in group A
streptococci. Proc. Natl. Acad. Sci. USA 82:1822-1826.
184. Shearer, W. T., and E. B. Gilliam. 1984. PGEL not PGF2 alpha
reverses the anti-actin antibody stimulation of protein phos-
phorylation and DNA synthesis in L cells. Biochem. Biophys.
Res. Commun. 124:23-28.
185. Shinoda, T., N. Takahashi, T. Takayasu, T. Okuyama, and A.
Shimizu. 1981. Complete amino acid sequence of the Fc region
of a human delta chain. Proc. Natl. Acad. Sci. USA 78:785-
789.
186. Simpson, W. J., and P. P. Cleary. 1987. Expression of M type
12 protein by a group A streptococcus exhibits phaselike
variation: evidence for coregulation of colony opacity determi-
nants and M protein. Infect. Immun. 55:2448-2455.
187. Simpson, W. J., J. C. Robbins, and P. P. Cleary. 1987.
Evidence for group A-related M protein genes in human but
not animal-associated group G streptococcal pathogens. Mi-
crobiol. Pathol. 3:339-350.
188. Spanier, J. G., S. J. C. Jones, and P. P. Cleary. 1984. Small
DNA deletions creating avirulence in Streptococcus pvogenes.
Science 225:935-938.
189. Srivastava, S. K., D. A. Barnum, and J. F. Prescott. 1985.
Production and biological properties of M-protein of Strepto-
coccus equi. Res. Vet. Sci. 38:184-188.
190. Steinert, P. M., A. C. Stevens, and D. R. Roop. 1985. The
molecular biology of intermediate filaments. Cell 42:411-419.
191. Stone, D., J. Sodex, P. Johnson, and L. B. Smillie. 1975.
Tropomyosin: correlation of amino acid sequence and struc-
ture, p. 125-136. In E. N. A. Biro (ed.), Proceedings of the
IXth Federation of European Biochemical Societies Meeting.
Elsevier/North-Holland Publishing Co., New York.
192. Streisinger, G., Y. Okada, J. Emrich, J. Newton, A. Tsugita, E.
Terzaghi, and M. Inouye. 1966. Frameshift mutation and the
genetic code. Cold Spring Harbor Symp. Quant. Biol. 31:77-
84.
193. Swanson, J., K. C. Hsu, and E. C. Gotschlich. 1969. Electron
microscopic studies on streptococci. I. M antigen. J. Exp.
Med. 130:1063-1091.
194. Swanson, J., K. Robbins, 0. Barrera, D. Corwin, J. Boslego, j.
Ciak, M. Blake, and J. M. Koomey. 1987. Gonococcal pilin
variants in experimental gonorrhea. J. Exp. Med. 165:1344-
1357.
195. Swift, H. F., A. T. Wilson, and R. C. Lancefield. 1943. Typing
group A hemolytic streptococci by M precipitin reaction in
capillary pipettes. J. Exp. Med. 78:127-134.
196. Timoney, J. F., and D. Eggers. 1985. Serum bactericidal
responses to Streptococcus equi of horses following infection
or vaccination. Equine Vet. J. 17:306-310.
197. Tylewska, S. K., V. A. Fischetti, and R. J. Gibbons. 1988.
Binding selectivity of Streptococcus pyogenes and M-protein
to epithelial cells differs from that of lipoteichoic acid. Curr.
CLIN. MICROBIOL. REV.
Microbiol. 16:209-216.
198. Uhlen, M., B. Guss, B. Nilsson, S. Gatenbeck, L. Philipson, and
M. Lindberg. 1984. Complete sequence of the staphylococcal
gene encoding protein A. J. Biol. Chem. 259:1695-1702.
199. van de Rijn, I., and V. A. Fischetti. 1981. Immunochemical
analysis of intact M protein secreted from cell wall-less strep-
tococci. Infect. Immun. 32:86-91.
200. van de Rijn, I., J. B. Zabriskie, and M. McCarty. 1977. Group
A streptococcal antigens cross-reactive with myocardium.
Purification of heart-reactive antibody and isolation and char-
acterization of the streptococcal antigen. J. Exp. Med. 146:579.
201. Vosti, K. L. 1975. Characterization of a non-type-specific
antigen(s) associated with group A type 12 M protein. Infect.
Immun. 11:1300-1306.
202. Vosti, K. L., and W. K. Williams. 1978. Extraction of strepto-
coccal type 12 M protein by cyanogen broide. Infect. Immun.
21:546-555.
203. Wannamaker, L. W. 1954. The epidemiology of streptococcal
infections, p. 157-165. In M. McCarty (ed.), Streptococcal
infections. Columbia University Press, New York.
204. Wannamaker, L. W. 1973. The chains that link the throat to the
heart. Circulation 48:9-18.
205. Weiler, J. M., M. R. Daha, K. F. Austen, and D. T. Fearon.
1976. Control of the amplification convertase of complement
by plasma protein betalH. Proc. Natl. Acad. Sci. USA 73:
3268-3272.
206. Weiss, A. A., and S. Falkow. 1984. Genetic analysis of phase
change in Bordetella pertusis. Infect. Immun. 43:263-269.
207. Whitnack, E., and E. H. Beachey. 1982. Antiopsonic activity of
fibrinogen bound to M protein on the surface of group A
streptococci. J. Clin. Invest. 69:1042-1048.
208. Whitnack, E., and E. H. Beachey. 1985. Inhibition of comple-
ment-mediated opsonization and phagocytosis of streptococ-
cus pyogenes by D fragments of fibrinogen and fibrin bound to
cell surface M protein. J. Exp. Med. 162:1983-1997.
209. Whitnack, E., and E. H. Beachey. 1985. Biochemical and
biological properties of the binding of human fibrinogen to M
protein A streptococci. J. Bacteriol. 164:350-358.
210. Whitnack, E., J. B. Dale, and E. H. Beachey. 1984. Common
protective antigens of group A streptococcal M proteins
masked by fibrinogen. J. Exp. Med. 159:1201-1212.
211. Widdowson, J. P. 1980. The M-associated protein antigens of
group A streptococci, p. 125-147. In S. E. Read and J. B.
Zabriskie (ed.), Streptococcal diseases and the immune re-
sponse. Academic Press, Inc., New York.
212. Widdowson, J. P., W. R. Maxted, D. L. Grant, and A. M.
Pinney. 1971. The reltionship between M antigen and opacity
factor in group A streptococci. J. Gen. Microbiol. 65:69-80.
213. Widdowson, J. P., W. R. Maxted, and A. M. Pinney. 1971. An
M-associated protein antigen (MAP) of group A streptococci.
J. Hyg. 69:553-570.
214. Widdowson, J. P., W. R. Maxted, and A. M. Pinney. 1976.
Immunological heterogeneity among the M-associated protein
antigens of group A streptococci. J. Med. Microbiol. 9:73-88.
215. Williams, K. R., E. K. Spicer, M. B. LoPresti, R. A. Guggen-
heimer, and J. W. Chase. 1983. Limited proteolysis studies on
the Escherichia coli single-stranded DNA binding protein.
Evidence for a functionally homologous domain in both the
Escherichia coli and T4 DNA binding proteins. J. Biol. Chem.
258:3346-3355.
216. Wittner, M. K., and E. N. Fox. 1977. Homologous and heter-
ologous protection of mice with group A streptococcal M
protein vaccines. Infect. Immun. 15:104-108.
217. Woolcock, J. B. 1974. Purification and antigenicity of an M-like
protein of Streptococcus equi. Infect. Immun. 10:116-122.
218. Wright, S. D., J. I. Weitz, A. J. Huang, S. M. Levin, S. C.
Silverstein, and J. D. Loike. 1988. Complement receptor type
three (CD11b/CD18) of human polymorphonuclear leukocytes
recognize fibrinogen. Proc. Natl. Acad. Sci. USA 85:7734-
7738.
219. Zabriskie, J. B., and E. H. Freimer. 1966. An immunological
relationship between the group A streptococcus and mamma-
lian muscle. J. Exp. Med. 124:661.