bacterial hemagglutination and hemolysis · 1956] bacterialhemagglutinationandhemolysis 167...

23
BACTERIAL HEMAGGLUTINATION AND HEMOLYSIS ERWIN NETER Statler Research Laboratories and Department of Pediatrics, Children's Hospital, Laboratory of Bacteriol- ogy, Roswell Park Memorial Institute, and Departments of Pediatrics and Bacteriology, University of Buffalo, School of Medicine, Buffalo, New York CONTENTS I. Introduction ........................................................ 166 II. Direct Bacterial Hemagglutination ....................................................... 168 Hemagglutinating Bacteria and the Hemagglutination Reaction ......................... 168 Related Reactions ....................... ................................. 169 III. Indirect, Conditioned, or Passive Hemagglutination and Hemolysis ........................ 170 Hemagglutinating Bacteria and the Hemagglutination and Hemolysis Reactions .......... 170 Related Reactions . .................................................................... 173 Biologic Significance .................................................................. 174 Applications ........................................................................... 175 Demonstration and titration of bacterial antibodies .................................. 175 The polyvalent hemagglutination and hemolysis tests ................................ 175 Demonstration and titration of bacterial antigens ................................... 176 Differentiation of bacterial antigens ................................................. 176 Serologic identification of unstable bacterial suspensions ............................ 176 Antigenic composition of microorganisms ............................................ 176 Removal of an antigen from a mixture .............................................. 176 Immunization with modified erythrocytes ............................................. 176 Aid in determination of pathogenicity ............................................... 176 Clinical applications for diagnosis of infections ....................................... 176 IV. Bacterial Hemagglutination and Hemolysis of Erythrocytes Treated with Antigen-Antibody Mixture ............................................................................. 177 V. Bacterial Hemagglutination of Tannic Acid Pretreated Erythrocytes ...................... 177 VI. Bacterial Hemagglutination by Antibodies Against Chemically Attached Protein Antigens.. 178 VII. Red Cell Linked Antigen Hemagglutination Test ......................................... 178 VIII. Bacterial Panagglutination The Thomsen-Friedenreich Hemagglutination Phenomenon and Its Biologic Significance. 179 IX. Bacteriogenic Hemagglutination ......................................................... 180 X. Nomenclature .......................................................................... 181 XI. Summary and Outlook .................................................................. 181 XII. References ............................................................................. 182 "Every part is disposed to unite with the whole, that it may thereby escape from its own incompleteness."-Leonardo Da Vinci. I. INTRODUCTION both for identification and titration of hemagglu- Investigations carried out during the past fif tinating viruses and for detection and titration teen years have revealed that the red blood cell of the corresponding antibodies. Also, this reac- is a valuable tool for microbiologic studies. In tion has been studied intensively as a model of 1941, Hirst (1) as well as McClelland and Hare virus infection of the host cell, including the (2) independently described viral hemagglutina- mechanism of virus attachment and the identity tion, when these authors observed clumping of of cell receptors. Viral hemagglutination is dis- erythrocytes in the presence of influenza virus. cussed in modem textbooks of bacteriology, This hemagglutination is specifically inhibited by virology, and microbiology and was reviewed in homologous viral antiserum. Since that time detail by Burnet (3). Therefore, this subject is not numerous other viruses have been found to pos- included here, although reference will be made to sess hemagglutinating properties. The viral differences and similarities between viral and hemagglutination test has been used extensively bacterial hemagglutination. 166 on December 29, 2019 by guest http://mmbr.asm.org/ Downloaded from

Upload: others

Post on 10-Sep-2019

12 views

Category:

Documents


0 download

TRANSCRIPT

Page 1: BACTERIAL HEMAGGLUTINATION AND HEMOLYSIS · 1956] BACTERIALHEMAGGLUTINATIONANDHEMOLYSIS 167 Bacteria, too, cause agglutination of erythro- ofbacterial hemagglutination. (3) Specifichemag-

BACTERIAL HEMAGGLUTINATION AND HEMOLYSIS

ERWIN NETER

Statler Research Laboratories and Department of Pediatrics, Children's Hospital, Laboratory of Bacteriol-ogy, Roswell Park Memorial Institute, and Departments of Pediatrics and Bacteriology, University of

Buffalo, School of Medicine, Buffalo, New York

CONTENTSI. Introduction........................................................ 166

II. Direct Bacterial Hemagglutination....................................................... 168Hemagglutinating Bacteria and the Hemagglutination Reaction......................... 168Related Reactions ....................... ................................. 169

III. Indirect, Conditioned, or Passive Hemagglutination and Hemolysis........................ 170Hemagglutinating Bacteria and the Hemagglutination and Hemolysis Reactions.......... 170Related Reactions..................................................................... 173Biologic Significance.................................................................. 174Applications........................................................................... 175Demonstration and titration of bacterial antibodies .................................. 175The polyvalent hemagglutination and hemolysis tests................................ 175Demonstration and titration of bacterial antigens................................... 176Differentiation of bacterial antigens................................................. 176Serologic identification of unstable bacterial suspensions............................ 176Antigenic composition of microorganisms ............................................ 176Removal of an antigen from a mixture.............................................. 176Immunization with modified erythrocytes............................................. 176Aid in determination of pathogenicity............................................... 176Clinical applications for diagnosis of infections....................................... 176

IV. Bacterial Hemagglutination and Hemolysis of Erythrocytes Treated with Antigen-AntibodyMixture............................................................................. 177

V. Bacterial Hemagglutination of Tannic Acid Pretreated Erythrocytes...................... 177VI. Bacterial Hemagglutination by Antibodies Against Chemically Attached Protein Antigens.. 178VII. Red Cell Linked Antigen Hemagglutination Test......................................... 178

VIII. Bacterial PanagglutinationThe Thomsen-Friedenreich Hemagglutination Phenomenon and Its Biologic Significance. 179

IX. Bacteriogenic Hemagglutination ......................................................... 180X. Nomenclature .......................................................................... 181XI. Summary and Outlook.................................................................. 181XII. References............................................................................. 182

"Every part is disposed to unite with the whole, that it maythereby escape from its own incompleteness."-Leonardo Da Vinci.

I. INTRODUCTION both for identification and titration of hemagglu-

Investigations carried out during the past fif tinating viruses and for detection and titrationteen years have revealed that the red blood cell of the corresponding antibodies. Also, this reac-is a valuable tool for microbiologic studies. In tion has been studied intensively as a model of

1941, Hirst (1) as well as McClelland and Hare virus infection of the host cell, including the(2) independently described viral hemagglutina- mechanism of virus attachment and the identitytion, when these authors observed clumping of of cell receptors. Viral hemagglutination is dis-erythrocytes in the presence of influenza virus. cussed in modem textbooks of bacteriology,This hemagglutination is specifically inhibited by virology, and microbiology and was reviewed inhomologous viral antiserum. Since that time detail by Burnet (3). Therefore, this subject is notnumerous other viruses have been found to pos- included here, although reference will be made tosess hemagglutinating properties. The viral differences and similarities between viral andhemagglutination test has been used extensively bacterial hemagglutination.

166

on Decem

ber 29, 2019 by guesthttp://m

mbr.asm

.org/D

ownloaded from

Page 2: BACTERIAL HEMAGGLUTINATION AND HEMOLYSIS · 1956] BACTERIALHEMAGGLUTINATIONANDHEMOLYSIS 167 Bacteria, too, cause agglutination of erythro- ofbacterial hemagglutination. (3) Specifichemag-

1956] BACTERIAL HEMAGGLUTINATION AND HEMOLYSIS 167

Bacteria, too, cause agglutination of erythro- of bacterial hemagglutination. (3) Specific hemag-cytes. Eight types of bacterial hemagglutination glutination follows exposure of erythrocytes toare known at this time: (1) The first type (direct antigen-antibody mixture. (4) In the fourth type,bacterial hemagglutination) was described more the erythrocytes are pretreated with tannic acid,than half a century ago by Kraus and Ludwig exposed to protein antigen, and thus become(4), who observed clumping of erythrocytes in specifically agglutinable in the presence ofthe presence of staphylococci and vibrios. Al- homologous protein antibodies. (5) Similarly,though the list of hemagglutinating bacteria has antigens may be attached to the surface ofbeen enlarged since that time, the mechanism of erythrocytes by chemical linkages. (6) Modifica-this reaction remains to be elucidated. (2) A tion of erythrocytes with proteins may be ac-second, fundamentally different type of bacterial complished, too, through the agency of incom-hemagglutination was reported in 1947 by Keogh, plete red blood cell antibodies, the latter servingNorth, and Warburton (5). These authors found as Schieppers (carriers) (6). (7) Bacteria may pro-that antigen obtained from Hemophilus influen- duce hemagglutination in yet another way: theyzae is readily adsorbed to the surface of red blood cause changes of the erythrocytes which resultcells and that agglutination takes place on addi- in the appearance on the surface of the cell of antion of homologous bacterial antiserum. This otherwise undetectable antigenic component.indirect, conditioned, or passive hemagglutina- Agglutination ensues upon addition of antibodiestion reaction and its hemolytic modification reacting specifically with this newly exposedsince have been demonstrated with numerous antigen. (8) Finally, certain bacteria produceother bacterial antigens, and the tests have changes in normal serum which result in a newlyproved to be extraordinarily sensitive for the acquired hemagglutinating capacity. Table 1detection and titration of bacterial antibodies. presents a summary of these eight types ofGenerally speaking, it is the polysaccharide rather microbial hemagglutination.than the protein antigens which are adsorbed by Certain fundamental similarities and dif-untreated red blood cells. However, even protein ferences between these types of bacterial hemag-antigens can be attached to erythrocytes by glutination should be pointed out. In type 1several methods, representing types 3, 4, and 5 hemagglutination is effected by a bacterial

TABLE 1Types of microbial hemagglutination

1. Bacterium (bacterial + RBC - HA Direct microbialantigen) or virus hemagglutination

2. Bacterial or viral an- + RBC i Modified + Bacterial * HA Indirect, condi-tigen RBC or viral tioned, or passive

antibody microbial hemag-glutination

3. Microbial antigen + RBC HAand antibody

4. Microbial antigen + Tannic acid -* Modified + Microbial . HAtreated RBC RBC antibody

5. Microbial antigen + RBC or pre- -* Modified + Microbial . HAand complexing treated RBC RBC antibodychemical

6. Microbial antigen- + RBC Modified + Homologous HA Red cell linked an-incomplete RBC RBC antibody tigen testantibody complex

7. Bacterium (enzyme) + RBC Altered + RBC HA Thomsen-Frieden-RBC antibody reich phenomenon

8. Bacterium + serum - Altered + RBC HA Bacteriogenic he-serum magglutination

RBC = red blood cells.HA = hemagglutination.

on Decem

ber 29, 2019 by guesthttp://m

mbr.asm

.org/D

ownloaded from

Page 3: BACTERIAL HEMAGGLUTINATION AND HEMOLYSIS · 1956] BACTERIALHEMAGGLUTINATIONANDHEMOLYSIS 167 Bacteria, too, cause agglutination of erythro- ofbacterial hemagglutination. (3) Specifichemag-

168 ERWIN NETER [VOL. 20

product and does not depend on the presence of this activity. Although all bacterial species havemicrobial antibodies; in fact, these antibodies in- not yet been tested, it is clear that certain species,hibit the reaction. On the other hand, antibodies e. g., shigellae, do not cause hemagglutination.bring forth hemagglutination in types 2 to 7. Obviously, there is no relationship in the taxo-In types 2 to 6 the antibodies are directed against nomic position between active and inactive bac-microbial and other antigens adsorbed to the teria. For example, most strains of Hemophilussurface of the red blood cells, whereas in type 7 influenzae are inactive, whereas many strainsthe hemagglutinating antibody reacts with of Hemophilus aegyptius are active. Similarly, inantigenic components of the erythrocytes them- the family Enterobacteriaceae, some members,selves, unmasked by bacterial action. Finally, such as the AD (alkalescens-dispar) group arehemagglutination of type 8 is due to bacterial active, and members of the genus ShigeUa arealteration of the serum, which endows the latter inactive. Further, even within species and sero-with a newly acquired hemagglutinating activity. groups (Escherichia coli) there are hemagglutinat-

It is the purpose of this paper to review ing and non-hemagglutinating strains [Kauff-critically for the first time the pertinent data, so mann (10)]. It should be stressed that negativewidely distributed in the literature, on bacterial findings may be due to the selection of the donorhemagglutination and hemolysis and to point out species of the red blood cells. It has been shownareas in need of future research. that, with a given strain, erythrocytes from one

animal species are agglutinated, whereas thoseII. DIRECT BACTERIAL HEMAGGLUTINATION from others are not. To illustrate: H. aegyptius

Hemagglutinating Bacteria and the Hemagglutina- agglutinates human red blood cells readily,tion Reaction chicken cells to a lesser degree, and monkey, rab-

bit, and rat cells not at all. Most strains of theA list of hemagglutinating bacteria together AD group are highly active with human cells,

with selected references are presented in table 2 and only a few strains agglutinate erythrocytesPerusal of ths table shows that a fairly large from rabbit, guinea pig, and sheep. On the othernumber of unrelated bacterial species possess hand, certain strains of E. coli, Hemophilus per-

tussis, and Clostridium septicum agglutinate redTABLE 2 cells from all animal species tested, although not

Bacteria causing direct hemagglutination necessarily to the same degree. Attention shouldBactei Referencea be called to Guyot's (8) observation to the effect

_________________that erythrocytes from different individuals be-Micrococcus pyogenes var. aureus.. 4 longing to the same animal species revealedSarcina 8p..................... 7 identical reactivity with strongly, weakly, orEscherichia coli ................. 7,8, 9, 10, 11 non-reacting strains of E. coli. The reason for theSalmonella 8p.................... 7 above mentioned differences in activity of dif-AD (alkalescens-dispar) Group.. 12, 13 ferent bacteria toward erythrocytes from variousVibrio comma......... 7, 12 animal species remains to be elucidated.Vibrio 8p ......4 At this time it seems unwise to this reviewer,Pseudomonas sp.......... 14Hemophilus aegyptius....... .15 because of inadequate data, to offer generaliza-Hemophilus influenzae.15, 16 tions on the identity of the hemagglutinatingHemophilus pertussis ............ 5, 12, 17, 18 principles and the mechanism of the direct hemag-Bacillus parapertussis ........... 5 glutination reaction. It is stated by DafaallaBrucella bronchiseptica .............5, 12 and Soltys (22) that the hemagglutinating factorCorynebacterium diphtheria.....7, 19 of C. septicum cannot be separated from the bac-Corynebacterium pseudodiph- terial cell. If this were so, one should expect that

theriticum............. 19 bacterial cells are attached selectively to redClostridium botulinum........... 20, 21 blood cells. So far as this reviewer is aware, thisClostridium septicum ............ 22 has not been demonstrated. Indeed, with several

* The present scientific names and not neces- species separation of the hemagglutinating factorssarily those of the authors are used in tables 2, from the bacteria has been accomplished.3 and 4. Extensively studied is the hemagglutinating

on Decem

ber 29, 2019 by guesthttp://m

mbr.asm

.org/D

ownloaded from

Page 4: BACTERIAL HEMAGGLUTINATION AND HEMOLYSIS · 1956] BACTERIALHEMAGGLUTINATIONANDHEMOLYSIS 167 Bacteria, too, cause agglutination of erythro- ofbacterial hemagglutination. (3) Specifichemag-

1956] BACTERIAL HEMAGGLUJTINATION AND HEMOLYSIS 169

antigen of H. pertussi8 (17, 18, 23, 24). It is heat confirmation of the single report (7) is required,labile, being destroyed at 56 C for 30 minutes. particularly in view of the antigenic complexicityIt is inhibited by lipid components obtained from of this species.red blood cells. It has been clearly demonstrated Only limited information is available regard-that the erythrocyte receptor, which may be ing the component (receptor) of the red bloodidentical with, or at least related to, the in- cell which interacts with the hemagglutinatinghibitory lipid, differs from the receptor for in- principle. That it is not identical with the in-fluenza virus. H. pertusis contains another anti- fluenza virus receptor is clearly evident from thegen which is readily adsorbed on red blood cells facts that neither the receptor destroying enzymeand does not cause direct hemagglutination (cf. (RDE) nor influenza virus interfere with hemag-p 170). The heat lability of hemagglutinating glutination by H. pertussis and H. aegyptius.principles has been established also for those ob- The observation that certain lipids inhibit thistained from E. coli, H. aegyptius, and C. septicum. reaction suggests that lipid components of theSo far as the hemagglutinating factor of red blood cell serve as receptors.

Clostridium botulinum is concerned, it is distinct It is obvious that much additional work needsfrom the toxin, as evidenced by the results of to be done, particularly with respect to theprecipitation studies, utilizing the Oudin serum- characterization of many bacterial hemagglu-agar technique, and by the observation that red tinating principles, the corresponding red bloodblood cells remove from a filtrate only the hemag- cell receptors, and the process of the hemaggluti-glutinating factor but not the toxin. Lowenthal nation reaction itself.and Lamanna (25, 26) presented evidence to theeffect that, depending upon the pH, the toxin Related Reactionsand hemagglutinin either form a complex or are The hemagglutinating capacity is not re-dissociated. The botulinal hemagglutinin is ad- stricted to bacteria nor are erythrocytes thesorbed by numerous animal tissue elements other only agglutinable cells. Since the early days ofthan erythrocytes. Its uptake by red blood cells modern microbiology it has been known thatappears to be a true adsorption reaction, because certain plant products, referred to as phytoag-the reaction has a negative temperature coeffi- glutinins, agglutinate red blood cells. Agglutinat-cient and the combination does not appear to ing substances are found also in higher fungifollow the law of simple multiple proportions. (27). Rickettsia orientis, too, agglutinates cer-Electrolytes are required. It is of interest to note tain red blood cells from individual fowl; thisalso that botulinal hemagglutinin is reversibly reaction is specifically inhibited by homologousattached to erythrocytes, and that the receptor is antiserum (28). Reference to viral hemagglutina-different from that for influenza virus. In this tion has been made in the introduction. Evenconnection attention should be called to another pleuropneumonia-like organisms (PPLO) canobservation of the reversibility of bacterial cause agglutination of erythrocytes (29).hemagglutination. Upon addition of homologous In connection with the foregoing data on bac-bacterial antiserum to a red blood cell suspension terial hemagglutination it is interesting that thewhich had been agglutinated by a pseudomonas agglutinability of erythrocytes from differentstrain the erythrocytes became disagglutinated animal species varies characteristically with dif-and the bacteria agglutinated (14). ferent plant agglutinins. For example, Land-The hemagglutinating factor obtained from steiner (30) found that lentil extract was con-

Corynebacterium diphtheriae is said to be an siderably more active toward rabbit than pigeonantigenic lipid (19). The antigenicity of many red blood cells; in contrast, bean extracts wereof the hemagglutinating principles has been far moreactive toward pigeon thanrabbit erythro-established by the observation of the specific cytes. Similar observations were made with ricininhibitory effects of homologous antisera on and abrin. More recently, efforts have beenhemagglutination by Micrococcus pyogenes, H. made to discover seed extracts for the differentia-aegyptius, H. pertussis, C. diphtheriae, and C. tion of blood groups of man. For example, Boydsepticum. A notable exception appears to be E. (31), who refers to these antibody-like substancescoli; however, before the active hemagglutinating from plants as lectins, isolated a substance whichfactor can be considered to be non-antigenic, agglutinates A and AB, but not 0 cells, and which

on Decem

ber 29, 2019 by guesthttp://m

mbr.asm

.org/D

ownloaded from

Page 5: BACTERIAL HEMAGGLUTINATION AND HEMOLYSIS · 1956] BACTERIALHEMAGGLUTINATIONANDHEMOLYSIS 167 Bacteria, too, cause agglutination of erythro- ofbacterial hemagglutination. (3) Specifichemag-

170 ERWIN NETER [VOL. 20

precipitates blood group A substance. A plant Silverman, personal communication, Februaryagglutinin for human erythrocytes containing 10, 1956). The second apparent exception per-the N antigen has been described (32). Nungester tains to the flagellar antigen of Salmonellaand Halsema (32a) observed that Flexner- typhosa. Since only a single experiment wasJobling carcinoma cells, in contrast to erythro- carried out, suggesting that the flagellar antigencytes, were not agglutinated by certain phytoag- was adsorbed on untreated red blood cells (41),glutinins, although the latter were absorbed by corroborative evidence is required. However, itboth cell suspensions. should be kept in mind that, as will be pointed

Bacteria agglutinate cells other than erythro- out (p 177), Boyden and Anderson (87) demon-cytes. Rosenthal (9) found that E. coli causes strated adsorption of a tuberculo-protein toclumping of leukocytes, thrombocytes, spermato- untreated red blood cells. Further studies on thezoa, spores of molds, and pollens. More recently, interaction of various bacterial proteins and un-Rolle and Kalich (33) reported that the ability treated erythrocytes are needed.or inability of E. coli strains to agglutinate Unheated fresh Escherichia coli cultures failspermatozoa is related to pathogenicity or lack to modify red blood cells or do so only to aof pathogenicity. However, before this conclu- minimal degree. Upon standing for several dayssion can be accepted, confirmatory evidence is at 37 C or following heating at 100 C for 1 hour,required. the cultures and their supernatants become

highly active. This result is explained by twoIII. INDIRECT, CONDITIONED, OR PASSIVE HEMAG- independent effects of heat, namely, release from

GLUTINATON AND HEMOLYSIS the bacterial cells of somatic antigen into solu-tion and activation of the soluble antigen (46).

Hemagglutinating Bacteria and the Hemaglutina- The latter effect can be shown also with purifiedtion and Hemolysis Reactions lipopolysaccharides. It is interesting to note,

A large variety of bacterial species contain however, that the extent of heat activation variesantigens which are readily adsorbed by untreated with different antigen preparations. For exam-red blood cells and thereby render these modified' ple, two unheated salmonella lipopolysaccharideserythrocytes specifically agglutinable by homolo- failed to modify red blood cells in concentrationgous bacterial antibodies. Table 3 gives data on of up to 1000 ,g per ml, while another, preparedthe bacterial species, the antigens involved in the by the identical method, was active in a concen-reaction, and selected references. It is likely that tration of 6 ,g per ml. Following heating, the twoother bacterial species, not yet tested, also pro- former preparations were effective in a concentra-duce erythrocyte modifying antigens. tion of 12 ,ug per ml and the latter in a concentra-

In all but two instances, chemical analysis of tion of 1.5 ug per ml (59). The activation from athe modifying antigens has revealed that they are non-modifying to a modifying antigen can bepolysaccharides or contain polysaccharides as accomplished also by treatment with sodiumthe serologically active determinant. However, hydroxide (51, 59, 88). However, it must bethe active material of Pasteurella pestis was pointed out that not all bacterial polysaccharidestentatively considered to be a protein, although become erythrocyte-modifying upon treatmentfull chemical and physical analyses were not with sodium hydroxide (51). Furthermore, theundertaken and no claim regarding the purity of mode of action of both heat and sodium hydroxidethe antigen was made (69, 70). The more recent treatment remains to be determined. It is clear,observation of Chen (72) suggests that the al- however, that these treatments do not solelycohol-precipitated envelope antigen contains cause the appearance of an increased number ofboth protein and polysaccharide. It is conceiv- antigen molecules, since they do not effect theable, therefore, that the hemagglutination reac- antibody neutralizing capacity of the materialstion involves the latter material. Alternatively, (89). Suggestive evidence indicates that theseit is possible that the polysaccharide may have treatments of S. typhosa 0 lipopolysaccharide re-facilitated the adsorption of the protein (S. J. sult in liberation of a large portion of lipid and

' The term modification, as used in this review, thus remove a component which conceivably in-connotes antigenic alteration of erythrocytes by terferes with the adsorption of the antigen by redattachment of bacterial antigens. blood cells (55). This concept finds support in the

on Decem

ber 29, 2019 by guesthttp://m

mbr.asm

.org/D

ownloaded from

Page 6: BACTERIAL HEMAGGLUTINATION AND HEMOLYSIS · 1956] BACTERIALHEMAGGLUTINATIONANDHEMOLYSIS 167 Bacteria, too, cause agglutination of erythro- ofbacterial hemagglutination. (3) Specifichemag-

1956] BACTERIAL IEMAGGLUTINATION AND HEMOLYSIS 171

TABLE 3Indirect, conditioned, or passive bacterial hemagglutination

Bacteia Antigem Reerences

Micrococcus 8p. Polysaccharide 34, 35, 36Diplococcus pneumoniae Polysaccharide (acidic polymer) 34, 36, 37, 38Streptococcus 8p. Crude, polysaccharide 34, 39Gram-positive bacteria Crude, common antigen 40Neisseria meningitidis Polysaccharide 34Neisseria gonorrheae Polysaccharide 41, 41aEscherichia coli Crude, lipopolysaccharide 42, 43, 44

Vi (crude, acidic polymer) 37, 45Enteropathogenic E. coli Crude, lipopolysaccharide 44, 46, 47, 48, 48aBallerup* Crude 49

Vi (crude) 47, 50Aerobacter aerogenes Crude, polysaccharide 43, 51, 52Aerobacter cloacae Polysaccharide 51Paracolobactrum sp. Crude 42, 43Salmonella typhosa 0 (crude, polysaccharide, lipopolysac- 37, 38, 47, 53, 54,

charide) 55, 56Vi (crude) 38, 47, 50, 54, 56,

57, 58H (protein) (?) 41

Salmonella sp. Crude, polysaccharide, lipopolysaccharide 42, 52, 53, 59, 60, 61Shigella sp. Crude, polysaccharide 36, 62, 63, 64Serratia marcescens Lipopolysaccharide 37Proteus sp. Crude 42, 43

Lipopolysaccharide 34, 37X1e, XK Crude, polysaccharide 65, 66

Pseudomonas aeruginosa Crude 43Lipopolysaccharide 37, 67

Pseudomonas sp. Crude 43Vibrio comma Crude 68Pasteurella pestis Protein (?) 69, 70

Polysaccharide 71, 72Pasteurella multocida Capsular antigen 73Pasteurella tularensis Lipopolysaccharide, polysac¢haride 37, 74, 75Brucella sp. Crude 42, 76, 77, 78

Lipopolysaccharide-polypeptide 78Malleomyces mallei Crude 79Hemophilus influenzae Crude, polysaccharide 5, 34Hemophilus pertussis Crude 18Corynebacterium diphtheriae Polysaccharide 36Mycobacterium tuberculosis Crude cf. 80

Fractions, polysaccharide 81, 82, 83Mycobacterium paratuberculosis Crude 84Mycobacterium phlei Crude 84Leptospira sp. Crude 85, 86

* Now included in Escherichia freundii.

inhibitory activity of various lipids, mentioned be called to the observation of Middlebrook (90)on page 172. It will be interesting to learn whether to the effect that two polysaccharides werethe above SalmoneUa lipopolysaccharides, which present in a tuberculin fraction, one active, thediffer so markedly in erythrocyte modifying other inactive, in erythrocyte modification, andcapacity, contain different amounts of firmly yet both inhibiting hemagglutination specifically.bound lipid. In this connection attention should Modification of erythrocytes, demonstrable by

on Decem

ber 29, 2019 by guesthttp://m

mbr.asm

.org/D

ownloaded from

Page 7: BACTERIAL HEMAGGLUTINATION AND HEMOLYSIS · 1956] BACTERIALHEMAGGLUTINATIONANDHEMOLYSIS 167 Bacteria, too, cause agglutination of erythro- ofbacterial hemagglutination. (3) Specifichemag-

172 ERWIN NETER [VOL. 20

subsequent agglutination in the presence of tion of Pasteurella tularense polysaccharide thanhomologous antibodies, has been accomplished red blood cells of blood groups A and B (74). Nowith approximately 2000 molecules of S. typhosa information is presently available to explainantigen per red blood cell (88). This conclusion these differences.is based on the observation that 25 x 10' As yet little is known regarding the receptors onerythrocytes adsorb 100 ,ug of the antigen. the surface of the red blood cell to which bac-

Until very recently, all bacterial antigens terial antigens can be attached. Such specificoperative in the hemagglutination test were receptors do in fact exist. This is evident fromfound to be characteristic of the species, common the findings that neither the A, B, or Rh antigensto closely related species, or characteristic of are blocked in E. coli and P. tularense modifiedserogroups or serotypes. Rants and associates erythrocytes (44, 75) nor does treatment of red(40) demonstrated, by means of this technique, an blood cellswithreceptor destroying enzyme (RDE)antigen that appears to be common to many and periodate interfere with modification byunrelated gram-positive bacteria (among others, bacterial polysaccharide (36, 75). It is conceivablestaphylococcus and vqrious species of strepto- that lipid structures serve as receptors. This iscocci). suggested by the observation that certain lipids,

Surprisingly few studies have been reported on such as lecithin, cephalin, lipid extract from redthe process of attachment of bacterial antigen to blood cell stromata, and certain serum fractionsuntreated red blood cells. It has been shown (59, rich in lipids inhibit the attachment of certain91) with crude E. coli cultures as well as with bacterial antigens to erythrocytes (49, 79, 92,purified Salmonella abortus-equi and E. coli 93, 94). A non-specific factor inhibiting the ad-lipopolysaccharides that electrolytes are required, sorption of tuberculin antigen by sheep cells hasinasmuch as modification does not occur in 5 been described by Fine and associates (95). Thisper cent glucose and 5 per cent sucrose solutions. factor has not yet been identified. It is of interestIn contrast, these antigens become attached to that it was present more frequently in patientsthe erythrocytes in the presence of isotonic so- with various diseases than in healthy individual.dium chloride, potassium chloride, or sodium Of particular importance is the fact that severalcitrate solutions. The failure of calcium chloride bacterial antigens, either simultaneously or con-solution to effect erythrocyte modification (59) secutively, modify erythrocytes and that oneremains unexplained. Interestingly, electrolytes antigen apparently does not block the adsorptionare likewise required for the attachment of cer- of another (37, 44, 59, 60). Whether differenttain viruses to red blood cells and of bacterio- receptors are operative with different bacterialphage to bacteria. Mention may be made of the antigens remains to be determined in the future.fact that modification takes place very rapidly In contrast to the foregoing observations, an(within a few minutes) at 37 C and only to a apparent interference between two S. typhosaminimal or slight extent at 4 C. For these rea- antigens has been demonstrated by Spaun (56).sons, the term "adsorption", as used in this re- This author found that when a standard amountview and many papers on the subject, does not of 0 antigen was mixed with a relatively largeconnote a particular mode of attachment of the amount of Vi antigen for modification of erythro-bacterial antigens. cytes, agglutination occurred only in the presenceRed blood cells from man, horse, sheep, rab- of Vi antibodies; however, red blood cells became

bit, ox, and goat are equally well modified by S. agglutinable by either 0 or Vi antibodies whentyphosa Vi antigen (57), as are erythrocytes smaller amounts of Vi antigen were used to-from man, dog, rabbit, sheep, guinea pig, rat, gether with the same amount of 0 antigen.and chicken by E. coli antigen (44). On the other Spaun's observations have been confirmed re-hand, Boyden (79) found that different amounts cently by Landy and Ceppellini (95a), and it isof mallein were required for modification of clear therefore that there exist interfering anderythrocytes from various animal species; he non-interfering antigen combinations.noted furthermore that ox red blood cells adsorb With this modification red blood cells acquirethe antigen without being agglutinated by a new serological specificity. Homologous bac-homologous antiserum. Human erythrocytes of terial antibodies cause specific agglutination. Inblood group 0 are more suitable for the adsorp- the presence of complement and antibodies cer-

on Decem

ber 29, 2019 by guesthttp://m

mbr.asm

.org/D

ownloaded from

Page 8: BACTERIAL HEMAGGLUTINATION AND HEMOLYSIS · 1956] BACTERIALHEMAGGLUTINATIONANDHEMOLYSIS 167 Bacteria, too, cause agglutination of erythro- ofbacterial hemagglutination. (3) Specifichemag-

1956] BACTERIAL HEMAGGLUTINATION AND HEMOLYSIS 173

tain modified erythrocytes are iysed. However, it the observation indicating that modification andshould be noted that erythrocytes from different subsequent hemagglutination may be enhancedsources are not equally susceptible to lysis. For if papain or trypsin treated red blood cells areexample, sheep red blood cells modified with used (96). Finally, attempts have been made tocrude or purified E. coli antigens are readily increase the sensitivity of the hemagglutinationlysed, whereas human erythrocytes are almost reaction by the use of anti-globulin serumentirely resistant to lysis. This latter finding is (Coombs test) (38, 54) and by replacement ofnot due to lack of adsorption of the bacterial the hemolytic with the conglutination systemantigen, since the treated red blood cells are as (38).readily agglutinated by homologous bacterialantiserum as are sheep red blood cells. The ques- Red Raction arises as to whether lysis of modified human The attachment of antigenic material, resultingred blood cells may occur in the presence of anti- in the acquisition of a new serologic specificity,bodies and complement of human origin. This is restricted neither to bacterial antigens nor towas found not to be s0 (59). Similarly, Skillman erythrocytes. Modification of red blood cells andand co-workers (96) obtained lysis with sheep subsequent hemagglutination in the presence ofcells but not with human cells in a streptococcal homologous microbial antibodies has been demon-hemolysis test; hemagglutination was obtained strated with antigens obtained from rickettsiaewith erythrocytes from both species. Middle- (101, 102), trypanosome (103), schistosome (104),brook (97), however, observed lysis of human red candida (105), histoplasma (106), and tricho-blood cells modified by tuberculin, although monas (107).hemolysis occurred only with larger amounts of De Gregorio (108) made the interesting ob-antiserum than were required for hemagglutina- servation that leukocytes, splenic, hepatic, andtion. Further studies are needed with a large renal cells adsorb salmonella antigen in vitrovariety of bacterial antigens and erythrocytes and thus become specifically agglutinable in thefrom many animal species to determine the con- presence of homologous bacterial antibodies.ditions for, or refractoriness to, lysis. Working with a polysaccharide preparation madeAttention should be called to the failure of some from tuberculin, Boyden (109) showed that

serum specimens to cause hemagglutination of modified leukocytes and lymphocytes becometuberculin modified erythrocytes, when a frac- attached to modified erythrocytes upon additiontion of the same sera showed antibody activity of homologous antibodies, indicating that these(98). As yet, the interfering serum component(s) white cells, too, had adsorbed the antigen.have not been identified. A number of reactions Further studies on the possible modification ofinterfering with antibody activity have been dis- other tissue cells are needed. Such studies maycovered and may be briefly mentioned here. reveal information as to which cells have re-Serum albumin inhibits agglutination of red ceptors for one antigen or another. Since leuko-blood cells exposed to a mixture of tuberculo- cytes play a significant role in the generalizedprotein and homologous antibody (87). The S. Shwartzman reaction (110), the question maytyphosa R antibody interferes with the 0 antibody be raised as to whether the bacterial antigen, re-in causing death of the bacterial cell (99). Also, sponsible for this reaction, is adsorbed to whitein vivo interference of the tuberculo-polysac- blood cells; adsorption may lead to injury of thecharide antibody with the antibody present in a leukocytes and to subsequent formation ofnewly discovered plasma fraction (IV-10) re- "fibrinoid" material. In this connection one re-sponsible for the delayed type of tuberculin calls that particles other than cells, e. g., collo-(PPD) skin sensitivity has been reported (100). dion particles and ion exchange resins (111), areThe latter findings require confirmation. capable of adsorbing various antigens and thusThe observation of Staub (88) to the effect that become agglutinable in the presence of the cor-

non-precipitating antibody can be demonstrated responding antibodies. Also, it is interesting toin the S. tyhposa hemagglutination test is of in- note that Adler (112) succeeded in attachingterest. Whether this antibody is in fact totally bacterial antigens to heterologous bacteria anddevoid of precipitating activity, remains to be rendering them susceptible to the bactericidaldetermined. Also, reference should be made to activity of the antibody reacting with the ar-

on Decem

ber 29, 2019 by guesthttp://m

mbr.asm

.org/D

ownloaded from

Page 9: BACTERIAL HEMAGGLUTINATION AND HEMOLYSIS · 1956] BACTERIALHEMAGGLUTINATIONANDHEMOLYSIS 167 Bacteria, too, cause agglutination of erythro- ofbacterial hemagglutination. (3) Specifichemag-

174 ERWIN NETER [VOL. 20

tificially attached antigen. It has been known for large amounts (10-20 mg) of antigen are

many years that bacteria may adsorb to their injected. Further work needs to be done to deter-surface components of the culture or suspending mine whether in vivo modification occurs follow-media. ing injection of other antigens as well. Of par-An interesting reaction between bacteria and ticular interest will be information as to whether

erythrocytes was described by Nelson (113) adsorption of bacterial antigen to erythrocytesunder the name of the immune-adherence phe- and other cells occurs in natural infections of mannomenon. This author observed the attachment and animals. Skillman and co-workers (96)of spirochetes and pneumococci to red blood claimed to have demonstrated in vivo modifica-cells, provided that homologous antimicrobial tion of red blood cells with streptococcal antigenantibodies and complement were present. Under in patients with active rheumatic fever. Thisthe conditions of these experiments hemagglu- important observation requires confirmationtination did not occur. Following adherence of from other laboratories. If, indeed, modificationthe microorganisms to erythrocytes enhanced takes place during natural infections, it is con-phagocytosis was observed. It may be pointed ceivable that hemolysis and anemia occur without that Rieckenberg (114) as early as 1917 de- the appearance of homologous bacterial anti-scribed a similar immunologic reaction; namely, bodies, and the question arises as to whether thisthe adherence of trypanosomes to blood plate- modification contributes to sludging of blood. Inlets in the presence of homologous antibody. The this connection attention should be called to therelationship between these two reactions is not reports of Nungester and his associates (115a,entirely clear because of inadequate data in 115b). These authors showed that pneumococcusRieckenberg's paper. polysaccharide affects red blood cells, resulting in

Since completion of this review a report on enhanced sedimentation in vitro and altered flowthis subject has been published by Lamanna and in vivoHollander (114a). It is established that the somatic antigens of

Biologic Sgnifranw enterobacteriaceae are made up of polysaccha-ride, lipid, and protein (or a protein-like sub-

The bacterial hemagglutination and hemolysis stance) (cf. 116). It is the polysaccharide whichreactions are of interest to the bacteriologist, endows this complex with its serological specifi-virologist, immunologist, hematologit, and to city, demonstrable by means of the hemagglutina-the clinician. Of particular importance i the tion test and other methods. This somaticquestion whether bacterial modification also antigen also has both toxic and pyrogenic proper-occurs in vivo. To date, only limited information ties. Renewed interest in the pathogenesis ofis available. Ceppellini and De Gregorio (115) fever following injection of bacterial pyrogensfailed to induce in vivo modification by injection has been shown (117, 118). For these reasons,of a bacterial polysaccharide from a ballerup study of the interaction between these andstrain. However, they made the interesting ob- similar antigens and red blood cells may even-servation that erythrocytes modified in vitro tually throw light on the modes of action ofsurvived for 28 hours in non-immunized animals, these materials as toxins and pyrogens. It haswhereas in immunized rabbits clinical evidence of been shown (89) that, to some extent, treatmentin vivo hemolysis occurred within a few hours. In with heat, sodium hydroxide, and periodate hasaddition, by means of the Coombs test, the reac- different effects on toxicity, pyrogenicity, erythro-tion between bacterial antibodies and modified cyte modifying capacity, and antibody neutraliz-red blood cells was demonstrated. In a later re- ing ability of these lipopolysaccharides, suggest-port (108) De Gregorio stated that when large ing that different reactive groupings are operativeamounts (10 mg) of Vi antigen were injected in these activities.intravenously into guinea pigs, antigen became Attention should be called to the properdin-attached to erythrocytes in vivo but not to tissue binding capacity of zymosan and certain bac-cells. Similarly,' Boyden (109), working with a terial polysaccharide antigens (119). Properdin,polysaccharide fraction made from tuberculin, as is well known (120), is one of the non-specificshowed that erythrocyte modification takes place factors contributing to resistance to infection.in vivo in guinea pigs, provided that relatively To what extent, if any, various polysaccharide

on Decem

ber 29, 2019 by guesthttp://m

mbr.asm

.org/D

ownloaded from

Page 10: BACTERIAL HEMAGGLUTINATION AND HEMOLYSIS · 1956] BACTERIALHEMAGGLUTINATIONANDHEMOLYSIS 167 Bacteria, too, cause agglutination of erythro- ofbacterial hemagglutination. (3) Specifichemag-

1956] BACTERIAL HEMAGGLUJTINATION AND HEMOLYSIS 175

antigens combine with plasma factors, such as applied to the study of many problems. It is notproperdin, and with circulating and sessile cells the purpose of this review, nor does space permit,in vivo during experimental and natural infection to appraise critically all these applications.remains to be determined. Rather, a few representative examples will be

In 1947 Keogh and associates (5) made the cited to indicate the potential usefulness of thesesignificant observation that red blood cells which techniques.had adsorbed Hemophilus influenzae antigen Demonmtration and titration of bacreial anti-became refractory to agglutination by influenza bodies. In many instances the hemagglutinationvirus and by the hemagglutinin of Hemophilia and hemolysis tests proved to be considerablypertussis. A detailed study of the effects of more sensitive than other serological methodsbacterial polysaccharides on viral hemagglutina- used for comparison, for example, the conven-tion has been carried out by MacPherson and tional bacterial agglutination test. Differences inassociates (51). These authors observed that sensitivity of these methods may be due, in partpolysaccharide antigens prepared from Aero- at least, to differences in the amount of antigenbacter aerogenes and A. cloacae inhibit red blood present on the surface of bacteria and of modifiedcell agglutination by influenza, mumps, and red blood cells. At times, the superiority of theNewcastle disease viruses. Furthermore, they hemagglutination method is of such a magnitudenoticed that modified red blood cells adsorb PR 8 as to appear qualitative rather than quantitative.virus as readily as untreated erythrocytes; mumps For example, Gaines and Landy (67) found anti-virus, on the other hand, was poorly adsorbed bodies to Peudomonas aeruinosa in the sera ofby polysaccharide modified red blood cells. normal individuals exclusively by means of theTherefore, two mechanisms account for the in- hemagglutination test. In one instance the hemag-hibition of virus hemagglutination. In the former glutinin titer was as high as 1:960, and yet bac-instance, the increased charge on the surface of terial agglutinins could not be detected. Similarthe red blood cells inhibits clumping by, but not observations have been recorded by Neter andadsorption of, influenza virus. In the second in. co-workers (122) with respect to antibodies tostance, inhibition of virus adsorption readily enteropathogenic E. coli in the serum of healthyexplains the failure of hemagglutination. The adults. On the other hand, the hemagglutinincomplexity of the subject is also evident from titers in the sera of volunteers who had ingestedthe fact that one particular polysaccharide anti- E. coli were from 5 to 20 times higher than thegen of MacPherson et al. is adsorbed to red blood corresponding titers of bacterial agglutinins, al-cells, but does not inhibit virus hemagglutina- though the same serogroup of E. coli was usedtion. (123). An explanation of these differences mustHemolysis of erythrocytes modified by bac- await the results of future research. It is evident,

terial antigens is of interest also to the student of however, that in the search for antibodies againstantibody-mediated hemolysis, since it shows that bacterial polysaccharides the hemagglutinationanti-erythrocyte antibodies are not needed for and hemolysis tests may be used to advantage,this reaction. This observation parallels studies particularly should other methods yield negativeon hemolysis of chemically modified azo-erythro- results. In passing, mention should be made ofcytes in the presence of complement and anti- the fact that preliminary absorption with un-bodies to the attached groups (121). Similarly, treated red blood cells is a necessity, when serumantigenically modified bacteria are killed upon specimens are tested with heterologous red bloodaddition of complement and of antibodies di- cells. This difficulty can be circumvented by therected against an artificially attached antigen use of human erythrocytes (blood group 0, Rh(112). These data may eventually aid in the negative) in the examination of human serumelucidation of the mode of action of antibody specimens.and complement in lysis of erythrocytes and kill- The polyvalent hemagglutination and hemolysising of bacteria. tests. A particularly welcome feature of the bac-

terial hemagglutination test, at least so far asApplrtica~iO9 enterobacterial antigens are concerned, is the

The indirect, conditioned, or passive bacterial fact that several antigens can be simultaneouslyhemagglutination and hemolysis tests have been adsorbed on erythrocytes. Such a red blood cell

on Decem

ber 29, 2019 by guesthttp://m

mbr.asm

.org/D

ownloaded from

Page 11: BACTERIAL HEMAGGLUTINATION AND HEMOLYSIS · 1956] BACTERIALHEMAGGLUTINATIONANDHEMOLYSIS 167 Bacteria, too, cause agglutination of erythro- ofbacterial hemagglutination. (3) Specifichemag-

176 ERWIN NETER (VOL. 20

suspension thus becomes a useful tool as a poly- in studies on P. pestis (69) and certain strains ofvalent antigen for the detection of any homolo- E. coli (48).gous antibody. This polyvalent hemagglutination Antigenic composition of microorganisms. Thetest has been used in the study of the antibody hemagglutination inhibition test has been usedresponse of patients with infections due to entero- successfully for the determination of the locationpathogenic E. coli, salmonellae, and shigellae of bacterial antigens in bacterial cells (83), and(59, 60, 122). Further studies on the diagnostic the hemagglutination test for the existence ofpotentialities of this method are indicated. cros-reacting antigens in unrelated bacterial

Demonstration and titration of bacterial antigens. species (80, 125). In view of the well known anti-The hemagglutination or hemolysis inhibition genic relationship between Mycobacterium tuber-tests provide useful tools for identification culosis and Mycobacterium leprae it is not surpris-and determination of minute amounts of certain ing to find that patients with Hansen's diseasebacterial antigens. To this end, the antigen under give a positive tuberculin hemagglutination testinvestigation is first mixed with a suitable amount (126, 127).of a known bacterial antiserum; then, erythro- Removal of an antigen from a mixture. If one iscytes modified by the homologous antigen are dealing with a mixture of two antigens, only oneadded, and the inhibition of specific hemaggluti- of which is adsorbable on untreated erythrocytes,nation or hemolysis is noted. The sensitivity of the latter may then be used for the separation ofthis procedure can be illustrated by the fact that these antigens. It appears likely that this absorp-as little as 0.006 jsg of Shigella dysenteriae poly- tion technique will lend itself for this purpose,saccharide could be measured, although no efforts just as erythrocytes have been used for thehad been made to increase the sensitivity of the separation of botulinal hemagglutinin and toxinmethod (64). Similarly, Wright and Feinberg (26).(75) determined P. tularense antigen in amounts Immunization with modified erythrocytes.of 0.001 M&g. Demonstration of microbial antigens Erythrocytes modified by a bacterial antigenby means of the hemagglutination inhibition test may be employed for immunization purposes.in specimens from patients may eventually be- Indeed, a pure Vi antiserum was produced income a suitable method for early and specific diag- rabbits by the injection of appropriately modifiednosis of certain infectious diseases. This approach rabbit erythrocytes (50). Similarly, antibodieshasbeenusedby Warburton and co-workers (124) developed following the injection of erythrocytesfor the detection of antigen in cerebrospinal fluid coated with E. coli (44) antigen, and P. tularensefrom patients with H. influenzae meningitis and (75) hapten. A non-immunizing hapten thus be-by O'Connor and MacDonald (65) for the comes an immunizing antigen in the rabbit (75).diagnosis of scrub typhus. It is conceivable that Aid in determination of pathogenicity. It hasthis method may be applicable also to the early been suspected on epidemiological grounds thatdiagnosis of enteric infections (59). certain serogroups of E. coli cause epidemic and

Differentiation of bacterial antigens. The sporadic diarrhea of the newborn and young in-hemagglutination and hemagglutination inhibi- fant. Feeding studies have contributed support-tion tests have been used with success for the ing evidence. However, in the majority of infantsstudy of closely related antigens. For example, antibody formation could not be demonstratedsuch differences were noted among Vi antigens of by the conventional methods. The hemagglutina-S. typhosa and a ballerup strain (50). Similarly, tion test, however, yielded positive results andtwo different modifying substances could be thus provided additional evidence of the patho-demonstrated in tuberculin by means of hemag- genicity of these microorganisms, now referred toglutination and hemagglutination inhibition as enteropathogenic E. coli (46, 48a, 59).tests (81). Clinical applications for diagnosis of infections.

Serologic identification of unstable bacterial sus- Before any new serologic technique, such as thepensions. Use of soluble bacterial antigen in the bacterial hemagglutination and hemolysis tests,hemagglutination reaction obtained from strains are introduced into clinical practice, it is of ut-in unstable suspension, which are unsuitable for most importance that a full understanding ofagglutination tests, may make possible their their usefulness and limitations be available.identification. This approach has been fruitful Particularly, it is necessary to know whether the

on Decem

ber 29, 2019 by guesthttp://m

mbr.asm

.org/D

ownloaded from

Page 12: BACTERIAL HEMAGGLUTINATION AND HEMOLYSIS · 1956] BACTERIALHEMAGGLUTINATIONANDHEMOLYSIS 167 Bacteria, too, cause agglutination of erythro- ofbacterial hemagglutination. (3) Specifichemag-

19561 BACTERIAL HEMAGGLUTINATION AND HEMOLYSIS 177

presence of antibodies or antigens thus demon- made another significant observation; namely,strated is correlated with activity of the infec- that the antibody operative in this particulartious process. With the introduction of the Mid- hemagglutination test is inhibited by a humandlebrook-Dubos tuberculo-polysaccharide hemag- serum fraction and that seemingly negativeglutination test, many clinicians hoped that this immune sera obtained from immunized rabbitsmethod will give them the long awaited serologic produce hemagglutination if the albumin frac-test for the specific diagnosis of active tubercu- tion is first removed. Clearly, a new type oflosis, be it mild or severe, recent or of long stand- hemagglutination reaction has been discovered,ing, in infants or adults. As is well known today warranting further research.and emphasized by Middlebrook (80) himselfthis test does not deserve widespread routine V. BACTERIAL HEMAGGLUTINATION OF TANNICclinical application at this time. However, sug- ACID PRETREATED ERYTHROCYTE5gestive evidence of the potential clinical useful-ness of a variety of bacterial hemagglutination O h supinta oiiainonsfviobtierythrocytes by polysaccharide antigen (cf.test is at hand, although only adequate trials -

ttsection III) takes place in two stages, namely,the field will give the final answers. For example, alteration of the surface of the red blood cell fol-it is conceivable that the hemagklutination tech- lowed by attachment of the polysaccharide anti-nic may provide information of diagnostic sig-

g.1

nifcanewthegad t ifanilediarhe du ~ gen, Boyden (129) investigated a variety of sub-stances, in order to determine whether any ofenteropathogenic E. coli (59), sa'monellosis6 these materials produce the first effect and rendershigellosis (63), and other bacterial infections. red blood cells capable of adsorbing protein

The test has been applied also to the detection antigens. These attempts proved to be highlyof typhoid carriers (45, 128). Recently, Skillman successful. Certain inulin preparations wereand associates (96) reported that a hemagglutina- shown to be effective, but far superior resultstion test for the demonstration of streptococcal were obtained with tannic acid. Both materialsantibodies as well as demonstration of in vivo in higher concentration agglutinate erythrocytes.modification of patients' red blood cells with In the procedure, which has to be carried outstreptococcal antigen have proved to be of con- with great care, erythrocytes are first treated withsiderable ald in ascertaining activity of rheumatic nmic acid. These pretreated red blood cells arefever. Confirmation of these results is needed, then exposed to the protein antigen and thus ac-before the conclusions can be accepted. quire a new serologic specificity, inasmuch as they

are specifically agglutinated in the presence ofIV

IBAOERYALTHEMAGCLUTESATEONADWIHEO- homologous protein antibodies. In this way Boy-818OFENNTHBOYE TREATEDWITden could demonstrate hemagglutination of

tuberculo-protein modified red blood cells byBoyden and Anderson (87) made the significant tuberculin antibodies. Similarly, streptococcal

observation that exposure of untreated sheep protein antibodies could be detected. Boydenred blood cells to a mixture of tuberculoprotein also devised a hemagglutination inhibition testand homologous antibody results in hemagglu- for the titration of protein antigens. Boyden'stination and, in the presence of complement, in contribution represents a significant advance andhemolysis. This reaction differs in certain essen- served as a valuable stimulus.tial features from the polysaccharide hemaggluti- Table 4 summarizes data on various bacterialnation and hemolysis tests discussed in the protein antigens which can be adsorbed on tan-previous section. In the first place, the antigen is nic acid treated erythrocytes. The homologousa protein. Secondly, it becomes dissociated from antibodies can be titrated with modified erythro-erythrocytes upon repeated washings. Thirdly, cytes, the antigens by means of the hemagglutina-sensitization of erythrocytes with the antigen- tion inhibition test.antibody mixture and subsequent hemagglutina- To illustrate the sensitivity of this procedure,tion takes place readily at both 37 and 4 C. the following observations are cited. StavitskyApparently, Middlebrook (90) also observed ag- (130) showed that as little as 0.001 pg of diph-glutination of red blood cells exposed to a mixture theria toxoid can be measured accurately. Eryth-of antigen and antibody. Boyden and Anderson rocyte modification was accomplished with 0.125

on Decem

ber 29, 2019 by guesthttp://m

mbr.asm

.org/D

ownloaded from

Page 13: BACTERIAL HEMAGGLUTINATION AND HEMOLYSIS · 1956] BACTERIALHEMAGGLUTINATIONANDHEMOLYSIS 167 Bacteria, too, cause agglutination of erythro- ofbacterial hemagglutination. (3) Specifichemag-

178 ERWIN NETER [VOL. 20

TABLE 4Bacterial hemagglutination of tannic acid treated erythrocytes

Bacteria Antigen References

Streptococcus pyogene8 Protein 129Corynebacterium diphtheriae Toxin, toxoid 37, 130, 131, 132, 133Pasteurella peti8 Protein 37, 71, 134, 135, 136

Toxin, toxoid 37, 137Clostridium perfringens Alpha toxin lecithinase 131Clostridium tetani Toxoid 131Mycobacterium tuberculosis Protein 82, 100, 129, 138, 139

,ug per ml of diphtheria toxoid (130) and with 0.3 study of the role of the protein antigen of the,ug per ml of purified capsular protein of Pas- plague bacillus in immunity and to the diagnosisteureUa pestis (136). of this infection (134). Several protein antigensThe sensitivity of the tannic acid hemaggluti- have been detected in the tubercle bacillus (139),

nation test for the titration of bacterial protein and information has been gained on the locationantibodies becomes even more apparent on corm- of various antigenic components on the surfaceparison with other available methods. For exam- and in the depth of the bacterial cell (83).ple, the hemagglutination test proved to be 20 Finally, attention should be called to the observa-to 50 times more sensitive for the titration of anti- tion of Hinz and Pillemer (140) that lysis of tan-bodies to the capsular antigen of the plague nic acid treated cells, noted by several investiga-bacillus than the complement fixation test (136). tors, is mediated by the properdin system.

Modification of erythrocytes is inhibited bycertain cations, particularly Mn. The presence of VI. BACTERIAL HEMAGGLUTINATION BY ANTIBODIESthese and other inhibiting substances in crude AGAINST CIEMICALLY ATTACHED PROTEINdiphtheria toxoid probably accounts for the ANTIGENSfailure to demonstrate erythrocyte modification In addition to the aforementioned tannic acidwith some of these preparations (133). Studies hemagglutination test, protein antigens may beon interfering materials may eventually aid in attached to erythrocytes also by chemical link-our understanding of the precise changes caused ages. Such a method, devised by Pressman,by tannic acid treatment which result in the Campbell, and Pauling (140a), was used by Sta-newly acquired capacity for protein adsorption. vitsky and Arquilla (141). The method employsTannic acid treated erythrocytes adsorb two bis-diazotized benzidine as the protein-conjugat-

protein antigens (diphtheria and tetanus tox- ing material. The authors showed that, in addi-oids) simultaneously (131), just as untreated tion to several non-bacterial proteins, diphtheriaerythrocytes adsorb several polysaccharide anti- toxoid can thus be attached to erythrocytes; thegens. Mention should be made that tannic acid latter become agglutinated in the presence oftreated erythrocytes can adsorb polysaccharide homologous diphtheria antitoxin. It is feasibleantigens as well. However, there are differences to use the method also as a hemolysis test, pro-in the temperature and time dependence of these vided that complement is added to the system.adsorptions. Protein is adsorbed readily at lower Titration of the antigens is possible by the use oftemperature and at a greater speed than poly- the hemagglutination or hemolysis inhibitionsaccharide antigen. For these reasons Chen and procedures. Cole and Farrell (142) succeeded inMeyer (134) use a temperature of 37 C for 1 attaching to formalinized erythrocytes tuber-hour for polysaccharide adsorption and room culin PPD via tetrazotized benzidine.temperature for a maximum of 15 minutes forprotein adsorption. VII. RED CELL INKED ANTIGEN HEMAGGLUTINA-The tannic acid hemagglutination test lends TION TEST

itself readily to the study of many bacteriologic, Coombs and his associates have made an im-serologic, and immunologic reactions. Non- portant contribution by devising a new approachflocculating antitoxin has been detected (131). for attachment of antigens to red blood cells.The method has been applied with success to the They have shown that non-bacterial protein

on Decem

ber 29, 2019 by guesthttp://m

mbr.asm

.org/D

ownloaded from

Page 14: BACTERIAL HEMAGGLUTINATION AND HEMOLYSIS · 1956] BACTERIALHEMAGGLUTINATIONANDHEMOLYSIS 167 Bacteria, too, cause agglutination of erythro- ofbacterial hemagglutination. (3) Specifichemag-

1956] BACTERIAL HEMAGGLUTINATION AND HEMOLYSIS 179

antigens may be chemically coupled to incomplete 3. The bacterial enzyme does not by itselfRh (143), ox red cell (144) and Forssman (6) cause agglutination of red blood cells. The en-antibodies. These erythrocyte antibodies by zyme is heat labile and filterable.themselves do not cause hemagglutination. They *n*4. The T antibody is not a bacterial agglutinin,thmserveas dcopr(tcarrier)fL utheprotein.They in as much as it does not cause agglutination ofserve as Schleppem (carriers) for the protein. the very cultures which produce a transformationErythrocytes treated with such an antigen- of red blood cells.antibody complex then become agglutinable in 5. The active principle of the bacterial culturethe presence of homologous antibodies to the is rapidly fixed by red blood cells and, as its ac-artificially attached antigen. Reference to these tivity progresses, eluted in an unaltered statemethods is made, because it seems plausible to from the bacterial cell.this reviewer that this ingenious approach of 6. Transformed red blood cells no longerCoombs and associates may lend itself to the adsorb the transforming principle.study of bacterial protein antigens as well. 7. The T agglutinin reacts optimally at 15-

20C.VIII. BACTERIAL PANAGGLUTINATION 8. The T agglutinin differs from the normally

occurring cold agglutinins, as evidenced by theThe Thomsen-Friedenreie~i Hemaggluination Phe- results of absorption studies.nomenon and its Biologic Signifuzance 9. Bacterial transformation can be accom-

In 1927, Thomsen (145) observed the apparent plished with red blood cells from animal specieschange of blood group of a blood specimen, i other than man. With guinea pig erythrocytes

hemolysis can be demonstrated in addition tomuch as the erythrocytes were agglutinated hemagglutination, provided that T antibodiesin the presence of any normal human serum. The and complement are present.study of this accidentally made observation ret 10. The panagglutinating activity has beenvealed that it was due to a propagating agent, demonstrated in two strains of diphtheroids, inaffecting all human erythrocytes, and rendering several strains of vibrios, and one strain of athem panagglutinable. The agglutination reac- luminescent cocco-bacillus.tion was due to an apparent change of a latent 11. One of the active bacteria also has thered blood cell receptor, referred to as L (latent) capacity of destroying the T antigen by means ofreceptor and its reactivity with a third blood the "receptor destroying principle".group agglutinin. An extensive investigation of 12. This hemagglutination reaction may leadgu . .agglutiniten ines onof to erroneous results in blood grouping. Trans-

Thomsenoutb emaggutination enom en fusion into guinea pigs of altered erythrocyteswas caried out by Fiedenreich (146), and the leads to shock associated with hemolysis. (Theresults were published in monograph form (in similarity of this reaction to paroxysmal hemo-English) in 1930. Reading this treatise a quarter globinuria is mentioned.)of a century later, this reviewer is impressed withthe excellence of Friedenreich's experimental All basic observations of Thomsen and ofstudies and his interpretation of the results. Friedenreich have been amply confirmed duringThis hemagglutination reaction is frequently and the ensuing 25 years. Friedenreich's concept re-justifiably referred to as Thomsen-Friedenreich garding the mode of action of transforming bac-hemagglutination phenomenon. teria has certain striking resemblances to theThe following facts have emerged from these action of influenza virus on erythrocytes, as

fundamental studies. elucidated by Hirst (1) in 1941. In both instances,the evidence strongly suggests that adsorption of

1. The transformation of the red blood cell is the microbial enzyme or the virus on the red blooddue to enzymatic activity of certain bacteria and cells represents the first step of microbe-erythro-uncovers a substrate on the surface of the eryth- cyte interaction; that a specific chemical com-rocyte. This altered substrate, referred to as T pound on the surface of the red blood cell servesreceptor or T antigen, has the ability to react pu on thensurfac e of the re celewith the corresponding T antibody. (The ab- as receptor; that enzymatic change of the recep-breviation T was introduced by Friedenreich in tor takes place; that elution of the microbialrecognition of Thomsen's discovery.) enzyme or virus follows this enzymatic modifica-

2. T agglutinins are present in normal human tion of the red blood cell receptor; and that thesera, although conspicuously absent in cord altered erythrocyte no longer adsorbs the enzymeblood and blood from young infants. or virus.

on Decem

ber 29, 2019 by guesthttp://m

mbr.asm

.org/D

ownloaded from

Page 15: BACTERIAL HEMAGGLUTINATION AND HEMOLYSIS · 1956] BACTERIALHEMAGGLUTINATIONANDHEMOLYSIS 167 Bacteria, too, cause agglutination of erythro- ofbacterial hemagglutination. (3) Specifichemag-

180 ERWIN NETER [VOL. 20

It should be mentioned that various bacteria the attachment of influenza and other virusescause somewhat different changes on erythro- and on the antigenic makeup of the erythrocyte.cytes. Micrococcus pyogenes var. aureus unmasks Enzyme (trypsin) treatment of red blood cells isan antigen different from the T receptor (147). now used routinely for the detection of "incom-It is noteworthy also that periodate treated eryth- plete" Rh antibodies.rocytes are agglutinated by a panagglutinin Recently, it was shown by Stewart, Petenyi,other than the T antibody (147). and Rose (151) that erythrocytes after treatment

Important studies have been carried out with with influenza virus have a greatly shortenedthe enzyme of the cholera vibrio, usually referred survival time and that the panagglutinin may beto as RDE (receptor destroying enzyme) (148). responsible for their rapid destruction. ThatThe following changes are effected by this en- bacterial infections can cause an abrupt decreasezyme in addition to panagglutinability. in the titer of T agglutinins has been shown to

1. Inagglutinability of red blood cells by in- occur in experimental anthrax infection (152).fluenza virus due to lack of virus adsorption. Friedenreich's observation that transfusion of

2. Increased agglutinability in the presence of altered erythrocytes causes in vivo hemolysis has"incomplete" Rh antibodies, provided that the not attracted the attention it deserves of stu-red blood cells contain the Rh antigen. dents of "auto-immune" (auto-antibody) hemol-

3. Altered immunogenic characteristics of the ysis and other disease states with a similarerythrocytes. pathogenesis. So far as the clinical significanceAmong other panagglutinating bacteria are of the bacterial and viral panagglutination reac-

Clostridium perfringens, Vibrio proteus, and tions is concerned, the following questionsDiplococcus pneumoniae (149, 150). A partially present themselves: (1) Are natural bacterial andpurified enzyme preparation obtained from C. viral infections associated with enzymaticperfringens type B showed identical or, at least, changes of red blood cells? (2) If so, are thesesimilar activities on erythrocytes as the cholera enzymatically altered cells agglutinated or lysedvibrio enzyme. According to Chu (149), many by panagglutinins or specific immune bodiesother bacterial species cause panagglutinability acting on unmasked antigenic components andwithout some of the other changes associated do they engender auto-antibodies? It is con-with RDE action. It should be emphasized that ceivable that these or similar events account forsome bacteria produce panagglutinability with- certain manifestations (enhanced erythrophago-out viral inagglutinability, and others produce cytosis, hemolytic anemia, sludging of blood) ofviral inagglutinability without panagglutinabil- various disease processes. These reactions,ity. related to the Thomsen-Friedenreich phe-

It is remarkable that influenza virus causes nomenon, cannot be discussed here in detail,almost identical changes on the red blood cell as however provocative the data and their impli-RDE, presumably because it contains a similar cations are.enzyme. Furthermore, it has been shown thatcertain of these changes can be effected by other Is BACTERIOGENC MAGLNATONwell known enzymes (trypsin, ficin, papain) and Another type of bacterial hemagglutination,by periodate. which is closely related to the Thomsen-Frieden-So far as the biologic significance and applica- reich reaction, was discovered by Davidsohn and

tions of this hemagglutination reaction are con- Toharsky (153, 154) and was described ascerned, it is obvious that contamination of a bacteriogenic hemagglutination. These authorsblood specimen with panagglutinating bacteria observed that a special bacterial strain belongingleads to erroneous results in blood grouping. This to the genus Corynebacterium produced changesis because such cells are agglutinated by blood in human plasma and serum which endow thesegrouping serum, even if the corresponding blood materials with panagglutinating capacity. Ofgroup antigens (A, B, etc.) are not present on the importance is that even cord serum, which doessurface of the red blood cells. Study of the ac- not contain T antibody, can be rendered ag-tivity of bacterial enzymes causing panagglutina- glutinating by the microorganism. The strainbility have revealed important informaton on was named Corynebacterium H in honor ofthe chemical identity of the receptors serving for Hektoen. The active principle can be separated

on Decem

ber 29, 2019 by guesthttp://m

mbr.asm

.org/D

ownloaded from

Page 16: BACTERIAL HEMAGGLUTINATION AND HEMOLYSIS · 1956] BACTERIALHEMAGGLUTINATIONANDHEMOLYSIS 167 Bacteria, too, cause agglutination of erythro- ofbacterial hemagglutination. (3) Specifichemag-

19561 BACTERIAL HEMAGGLUTINATION AND HEMOLYSIS 181

from the bacterial culture by filtration. Among a bacterial product responsible for bacterial hemag-large number of other bacterial strains tested, glutination have been called "antihemagglu-only a strain of Pseudomonas aeruginosa was tinins" (4).found to be active. The filtrate also alters red 2. Agglutination of red blood cells resultingblood cells themselves, rendering them panag- from the action of bacteria or bacterial productsglutinable. By means of absorption studies it was has been referred to as "direct bacterial hemag-shown that the antigen-antibody system in- glutination" (46) in order to differentiate thisvolved in bacteriogenic hemagglutination is reaction from that requiring bacterial antibodiesdifferent from that of the Thomsen-Friedenreich ("indirect bacterial hemagglutination").reaction and entirely independent of blood 3. Antigens which become attached to un-groups. In conformity with the nomenclature of treated red blood cells have been referred to asthe Thomsen-Friedenreich reaction (T antigen "erythrocyte modifying antigen" (94), "hemo-and T antibody), Davidsohn and Toharsky gave sensitin" (81), "erythrocyte sensitizing sub-their antigen-antibody system the designation stance" (ESS) (85), and "erythrocyte coatingof H. The identity of the agglutinating principle antigen" (35). The term "erythrocyte sensitiza-appearing in plasma and serum remains to be tion" has been used also for the reaction betweenestablished, and no claim has been made as to erythrocytes and antibodies (complete andwhether it is an antibody in the strict sense of incomplete) directed against the former's com-the word. It has been shown by Spielmann (155) ponents. The term "coating" must not implythat certain bacteria render human serum coverage of the entire surface, since the attachedspecimens containing "incomplete" Rh anti- antigens leave certain surface antigens un-bodies capable of causing Rh hemagglutination. covered.From a practical point of view it is obvious that 4. Antibodies causing agglutination of modi-contamination of blood grouping serum with a fied erythrocytes have been called "hemag-strain causing bacteriogenic hemagglutination glutinins" (43).renders this reagent unsuitable. No information 5. The following terms have been used withis at hand as to the possible in vivo occurrence of regard to hemagglutination or hemolysis oftransformation of plasma and bacteriogenic modified erythrocytes: "hemagglutination", "pas-hemagglutination. sive hemagglutination" (38), "conditioned hemag-

glutination" (103), "indirect hemagglutination"X. NOMENCLATUJR1E (46), "antigen-coated red cell technic" (6),It is only natural that, with the rapid develop- "polysaccharide specific hemagglutination" (134),

ment in a new field, such as bacterial hemag- "Keogh antigen adsorption test" (69), "Keoghglutination and hemolysis, a number of terms method" (83), and "polysaccharide lysis test"are being introduced, some of them being (41). The term "indirect bacterial hemagglutina-synonyms. In a recent editorial (156) the present tion" was coined to differentiate it from that dueauthor has pointed out that a logical, generally to direct bacterial action in the absence of anti-acceptable nomenclature should be devised. body (46).Such a task should not be undertaken by a 6. The hemagglutination method, employingreviewer of the subject, and recommendations, tannic acid treated erythrocytes, is referred totherefore, will not be offered here. Rather, some as "Boyden method" (83), "protein-specificof the terms used (but not necessarily coined) hemagglutination test" (134), and "hemag-by investigators in this field are herewith pre- glutination test". It may be pointed out that thesented with some comments. term "protein-specific hemagglutination test"

1. Bacterial products or components which, may be misleading, since tannic acid treatedwithout participation of antibodies, cause ag- erythrocytes also adsorb polysaccharides.glutination of red blood cells are referred to as"hemagglutinins" (4). It should be kept in mind XI. SUXMMARY AND OUTLOOKthat this term is used also for antibodies acting Perusal of the numerous papers dealing directlyon antigenic components of red blood cells as or indirectly with bacterial hemagglutinationwell as for antibodies causing agglutination of and hemolysis clearly reveals that striking ad-modified erythrocytes. Antibodies against the vances have been made, particularly during the

on Decem

ber 29, 2019 by guesthttp://m

mbr.asm

.org/D

ownloaded from

Page 17: BACTERIAL HEMAGGLUTINATION AND HEMOLYSIS · 1956] BACTERIALHEMAGGLUTINATIONANDHEMOLYSIS 167 Bacteria, too, cause agglutination of erythro- ofbacterial hemagglutination. (3) Specifichemag-

182 ERWIN NETER [VOL. 20

last decade. Many related reactions are of HAmagglutination (Bakterio-Haemoagglu-interest to bacteriologists, virologists, hematolo- tination). Centr. Bakteriol. Parasitenk.,gists, and immunologists alike. It is also obvious, 47, 640-653.as mentioned during the foregoing discussion, 9. ROSENTHAL, L. 1943 Agglutinating prop-that numerous problems await elucidation by erties of Escherichia coli. J. Bacteriol.,thatnuerousroblem await lucidaion by45, 546-650.competent and interested workers. It is note- ,KAUFFMANN, F. 1948 On haemagglutina-worthy that far more is known regarding indirect tion by Escherichia coli. Acta Pathol.than direct bacterial hemagglutination, whereas Microbiol. Scand., 25, 502-506.the reverse is true regarding viral hemaggluti- 11. COLLIER, W. A., AND JACOEB, M. 1955nation. The study of the interaction of bacteria Experiments with a hemagglutinatingand their antigens with red blood cells has strain of E. coli. Antonie van Leeuwen-yielded valuable information on the antigenic hoek. J. Microbiol. Serol., 21, 113-123.composition of various microorganisms and of 12. GRIFFIITS, J. J. 1948 Hemagglutinationerythrocytes, and the hemagglutination and by bacterial suspensions with specialhemolysis reactions have been useful as tools for reference to Shigella alkalescens. Proc.the detection and titration of numerous bacterial Soc. Exptl. Biol. Med., 67, 358-362.

antigens and antibodies. If some yearsfrom w13. GUPTA, N. P. 1950 A note on the haemag-antigens and antibodies. It some years from now glutination by organisms of the alkales-

another review be written on bacterial hemag- cens-dispar group. Acta Pathol. Micro-glutination and hemolysis, it can be confidently biol. Scand., 27, 300-303.expected that many of the gaps in our knowledge, 14. DRIMMER-HERRNHEI8ER, H. 1953 Haemag-now so clearly apparent, will have been filled. glutination by a pseudomonas of faecal

origin. Bull. Research Council Israel, 2,REFERENCES 445.

1. HiRsT, G. K. 1941 The agglutination of 15. DAVIs, D. J., PITTMAN, M., AND GRIFFITTS,red cells by allantoic fluid of chick em- J. J. 1950 Hemagglutination by thebryos infected with influenza virus. Sci- Koch-Weeks bacillus (Hemophilus ae-ence, 94, 22-23. gyptius). J. Bacteriol., 59, 427-431.

2. MCCLELLAND, L., AND HARE, R. 1941 The 16. SMITH, C. H. 1954 Relationship betweenadsorption of influenza virus by red cells haemagglutination and pathogenicity inand a new in vitro method of measuring strains of Haemophilus isolated from theantibodies for influenza virus. Can. eye. J. Pathol. Bacteriol., 68, 284-287.Public Health J., 32, 530-1538. 17. WARBURTON, M. F., AND FISHER, S. 1951

3. BURNET, F. M. 1952 Haemagglutinationin The haemagglutinin of Haemophilus per-relation to host cell-virus interaction. tussis. Australian J. Exptl. Biol. Med.In Ann. Rev. Microbiol., 6, pp 229-246. Sci., 29, 266272.Edited by C. E. Clifton, S. Raffel, and 18. FISHER, S. 1950 The haemagglutinin ofH. A. Barker. Annual Reviews, Stanford, Haemophilus pertussis. Australian J.Calif. Exptl. Biol. Med. Sci., 28, 509-516.

4. KRAus, R., AND LUDWIG, S. 1902 tJber 19. LEARNED, G. R., AND METCALF, T. G. 1952Bacteriohbmagglutinine und Antihimag- A study of the hemagglutinative behaviourglutinine. Wien. klin. Wochschr., 16, of the lipid antigens of Corynebacterium120-121. diphtheriae. Trans. Kansas Acad. Sci.,

5. KEOGH, E. V., NORTH, E. A., AND WAR- 55, 431-438.BURTON, M. F. 1947 Haemagglutinins 20. LAMANNA, C., AND LOWENTHAL, J. P. 1951of the Haemophilu group. Nature, 160, The lack of identity between hemag-63. glutinin and the toxin of type A botulinal

6. CoomBs, R. R. A., AND FIsET, M. L. 1954 organism. J. Bacteriol., 61, 751-752.Detection of complete and incomplete 21. STERNE, M. 1954 Hemagglutination byantibodies to egg albumin by means of a Clostridium botulinum type D. Science,sheep red cell-egg albumin antigen unit. 119, 440-441.Brit. J. Exptl. Pathol., 35, 472-477. 22. DAFAALLA, E. N., AND SOLTYS, M. A. 1951

7. FUKUHARA, Y. 1909 lJber Himaggluti- Studies on agglutination of red cells bynierende Eigensohaften der Bakterien. clostridia. Brit. J. Exptl. Pathol., 32,Z. Immunitatsforsch., 2, 313-322. 510-515.

8. GUTOT, G. 1908 Vber die bakterielle 23. FISHER, S. 1948 The inhibition of per-

on Decem

ber 29, 2019 by guesthttp://m

mbr.asm

.org/D

ownloaded from

Page 18: BACTERIAL HEMAGGLUTINATION AND HEMOLYSIS · 1956] BACTERIALHEMAGGLUTINATIONANDHEMOLYSIS 167 Bacteria, too, cause agglutination of erythro- ofbacterial hemagglutination. (3) Specifichemag-

19561 BACTERIAL HEMAGGLUTINATION AND HEMOLYSIS 183

tussis haemagglutinin by extracts of bacterial polysaccharides. Australian J.erythrocytes. Brit. J. Exptl. Pathol., Exptl. Biol. Med. Sci., 29, 51-62.29, 357-363. 37. LANDY, M. 1954 On hemagglutination pro-

24. FISHER, S. 1949 The erythrocyte receptor ceduresutilizingisolatedpolysaccharide andfor pertussis haemagglutinin. Brit. J. protein antigens. Am. J. Public Health,Exptl. Pathol., 30, 185-189. 44, 1059-1064.

25. LOWENTHAL, J. P., AND LAMANNA, C. 1951 38. BIER, 0. 1951 Observations preliminairesFactors affecting the botulinal hemag- sur l'hMmagglutination, l'hemolyse et laglutination reaction, and the relationship conglutination "passives". Ann. inst.between hemagglutinating activity and Pasteur, 81, 650-656.toxicity of toxin preparations. Am. J. 39. KIRBY, W. M. M. 1951 HemagglutinationHyg., 54, 342-353. reaction in streptococcal infections and

26. LOWENTHAL, J. P., AND LAMANNA, C. 1953 acute rheumatic fever. Proc. Soc. Exptl.Characterization of botulinal hemag- Biol. Med., 78, 519-522.glutination. Am. J. Hyg., 57, 46-59. 40. RANTZ, L. A., RANDALL, E., AND ZUCKERMAN,

27. BERNHEIMER, A. W., AND FARICs, M. E. A. 1956 Hemolysis and hemagglutination1953 Hemagglutinins among higher fungi. by normal and immune sera of erythro-J. Immunol., 70, 197-198. cytes treated with a nonspecies specific bac-

28. O'CONNOR, J. L. 1945 Hirst's haemag- terial substance. J. Infectious Diseases,glutination phenomenon exhibited by 98, 211-222.Rickettsia orientalis (Syn. Tsutsugamushi). 41. Tnoit~s, J. C., AND MENNIE, A. T. 1950Med. J. Australia, 2, 459-460. Bacterial polysaccharides in the diagnosis

29. FAHEY, J. E. 1954 A hemagglutination of infections. The polysaccharide lysisinhibition test for infectious sinusitis of test. Lancet, II, 745-746.turkeys. Proc. Soc. Exptl. Biol. Med., 41a. CHANAIN, I. 1954 An investigation of86, 38-40. Neisseria gonorrhoeae by a red cell sen-

30. LANDSTEINER, K. 1945 The specificity of sitization technique. J. Hyg., 52, 425-443.serological reactions, p. 5. Revised ed. 42. FREEMAN, N. L., FELSENFELD, O., ANDHarvard University Press, Cambridge, EVELAND, W. C. 1955 Slide hemag-Mass. glutination tests with 0 antigens of enteric

31. BOYD, W. C., SHAPLEIGH, E., AND Mc- organisms and Brucella. Am. J. Clin.MASTER, M. 1955 Immunochemical be- Pathol., 25, 332-335.havior of a plant agglutinin (lectin). 43. NEEDELL, M. H., NETER, E., STAUBITZ, W.Arch. Biochem. Biophys., 55, 226-234. J., AND BINGOHA, W. A. 1955 The

32. OTTENsOOSER, F., AND SILBBRscEMUDT, K. antibody (hemagglutinin) response of pa-1953 Haemagglutinin anti-N in plant tients with infections of the urinary tract.seeds. Nature, 172, 914. J. Urol., 74, 674-682.

32a. NUNGESTER, W. J., AND HALsEmA, G. 44. NETER, E., BERTRAM, L. F., ZAX, D. A.,1953 Reaction of certain phytoagglu- MURDOCK, M. R., AND AaBESAN, C. E.tinins with Flexner-Jobling carcinoma 1952 Studies on hemagglutination andcells of the rat. Proc. Soc. Exptl. Biol. hemolysis by Eacherichia coli antisera.Med., 83, 86-. J. Exptl. Med., 96, 1-15.

33. ROLLE, M., AND KAICH, J. 1954 Zur 45. LANDY, M., AND LAmIB, E. 1953 EstimationDiagnose der toxischen Kolikeime durch of Vi antibody employing erythrocytesAgglutination der Spermazellen. Munch. treated with purified Vi antigen. Proc.med. Wochschr., 96, 579-580. Soc. Exptl. Biol. Med., 82, 593-598.

34. KEOGH, E. V., NoRTH, E. A., AND WARBUR- 46. NETER, E., GORZYNsKI, E. A., ZALEWsEI,TON, M. F. 1948 Adsorption of bacterial N. J., RACEMAN, R., AND GINO, R. M.polysaccharides to erythrocytes. Nature, 1954 Studies on bacterial hemagglutina-161, 687-688. tion. Am. J. Public Health, 44, 49-54.

35. ROUNTREE, P. M., AND BARBOUR, R. G. H. 47. LE MINOR, L., Ls MINOR, S., AND GRABAR,1952 Antibody to the erythrocyte-coating J. 1952 Reaction d'hemagglutinationpolysaccharide of staphylococci: Its oc- passive et d'hemolyse directe au moyen decurrence in human sera. Australasian globules rouges sensibilis6s par des sub-Ann. Med., 1, 80-83. stances solubles 0 et Vi d'ent6robacteria-

36. HATEs, L. 1951 Specific serum agglutina- c6es. Ann. inst. Pasteur, 83, 62-70.tion of sheep erythrocytes sensitized with 48. OGAWA, T., ONArA, N., MIYAO, K., OGAWA,

on Decem

ber 29, 2019 by guesthttp://m

mbr.asm

.org/D

ownloaded from

Page 19: BACTERIAL HEMAGGLUTINATION AND HEMOLYSIS · 1956] BACTERIALHEMAGGLUTINATIONANDHEMOLYSIS 167 Bacteria, too, cause agglutination of erythro- ofbacterial hemagglutination. (3) Specifichemag-

184 ERWIN NETER [VOL. 20

J., MIZUNO, Y., AND KONDO, M. 1954 onstration of antibodies to Enterobac-Investigation on the hemagglutination of teriaceae. Ann. N. Y. Acad. Sci., inEscherichia coli 0111, 0,5, and 026 from the press.view point of a practical value. Nagoya 60. NETER, E., GORZYNSKI, E. A., GINO, R. M.,Med. J., 2, 101-109. WESTPHAL, O., AND LUDERITZ, 0. 1956

48a. STULBERG, C. S., ZUELZER, W. W., AND The enterobacterial hemagglutination testPAGE, R. H. 1956 Escherichia coli 0127: and its diagnostic potentialities. Can. J.B8, a serotype causing infantile diarrhea. Microbiol., 2, 232-244.J. Immunol., 76, 281-287. 61. FULTHORPE, A. J. 1954 Agglutination of

49. DE GREGORIO, M., AND Di NARDO, A. 1955 sheep erythrocytes sensitized with Sal-Emoagglutinazione condizionata (S. bal- monella polysaccharides. J. Pathol. Bac-lerup). Bull. ist. sieroterap. milan., 34, teriol., 68, 315-325.123-130. 62. CHUN, D., AND PARK, B. 1956 Demon-

50. CHU, D. C.-Y., AND HOYT, R. E. 1954 stration of Shigella flexneri antigens byReactions of T Vi and ballerup Vi haptenes means of hemagglutination test. J. In-with antisera against Vi-coated eryth- fectious Diseases, 98, 82-87.rocytes. J. Hyg., 52, 100-104. 63. NETER, E., AND WALKER, J. 1954 Hemag-

51. MACPHERSON, I. A., WILKINSON, J. F., AND glutination test for specific antibodies inSWAIN, R. H. A. 1953 The effect of dysentery caused by Shigella sonnei. Am.Klebsiella aerogenes and Klebsiella cloacae J. Clin. Pathol., 24, 1424-1429.polysaccharides on haemagglutination by 64. NETER, E., AND GoRzYNSEI, E. A. 1954and multiplication of the influenza group Erythrocyte-modifying capacity of Shigellaof viruses. Brit. J. Exptl. Pathol., 34, dysenteriae (Shiga) antigen and its poly-603-615. saccharide component. Proc. Soc. Exptl.

52. RABE, E. F., AND PLONKO, M. 1954 The Biol. Med., 85, 503-506.antibody response to gram negative or- 65. O'CONNOR, J. L., AND MACDONALD, J. M.ganisms: An explanation of the differences 1950 Excretion of specific antigen in thebetween bacterial and hemagglutinating urine in Tsutsugamushi disease (scrubantibody titers. Pediatrics, 14, 351-356. typhus). Brit. J. Exptl. Pathol., 31,

53. TAKEDA, Y., WATANABE, T., KURIBAYASHI, 51-64.K., AND KIUCHI, K. 1952 Studies on the 66. HAN, E. S. 1951 Hemagglutination testhemagglutination of erythrocytes sen- for epidemic and murine typhus feversitized with endotoxins. Japan. J. Exptl. using sheep erythrocytes sensitized withMed., 22, 273-284. proteus OX19 extracts. Am. J. Trop.

54. CEPPELLINI, R., AND DE GREGORIO, M. Med., 31, 243-251.1953 Emagglutinazione ed emolisi con- 67. GAINES, S., AND LANDY, M. 1955 Prev-dizionate mediante gli antigeni della alence of antibody to pseudomonas inSalmonella typhi. Boll. ist. sieroterap. normal human sera. J. Bacteriol., 69,milan., 32, 429-444. 628-633.

55. LANDY, M., TRAPANI, R-J., AND CLARK, W. 68. FELSENFELD, O., FREEMAN, N. L., ANDR. 1955 Studies on the 0 antigen of MOORING, V. L. 1955 Tube and slideSalmonella typhosa. Am. J. Hyg., 62, technic in the hemagglutination of Vibrio54-65. comma. Am. J. Trop. Med. Hyg., 4,

56. SPAUN, J. 1952 Determination of Sal- 318-320.monella typhi 0 and Vi antibodies by 69. AmEs, C. R. 1951 The envelope substancehemagglutination. Acta Pathol. Micro- of Pasteurella pestis. Brit. J. Exptl.biol. Scand., 31, 462-469. Pathol., 32, 259-273.

57. CORVAZIER, P. 1952 Etude de l'antigane 70. SILVERMAN, S. J. 1954 The isolation ofVi a l'aide d'une technique d'hemagglu- fractions from Pasteurella pestis for use intination passive. Ann. inst. Pasteur, 83, a hemagglutination test. J. Lab. Clin.173-179. Med., 44, 185-193.

58. SPAUN, J. 1951 On the determination of 71. NEEL, R., TASLIMI, H., AND EFTEXHARI, M.Vi - antibodies by haemagglutination. 1955 Valeur pratique comparee des reac-Acta Pathol. Microbiol. Scand., 29, 416- tions d'agglutination de conglutination418. directe d'hdmagglutination polyosidique

59. NETER, E., WESTPHAL, O., LUYDERITZ, O., et proteinique pour le diagnostic de laAND GORZYNSKI, E. A. 1956 The bac- peste. Arch. inst. d'Hessarek., 9, 85-107.terial hemagglutination test for the dem- 72. CHEN, T. H. 1952 The method of the

on Decem

ber 29, 2019 by guesthttp://m

mbr.asm

.org/D

ownloaded from

Page 20: BACTERIAL HEMAGGLUTINATION AND HEMOLYSIS · 1956] BACTERIALHEMAGGLUTINATIONANDHEMOLYSIS 167 Bacteria, too, cause agglutination of erythro- ofbacterial hemagglutination. (3) Specifichemag-

1956] BACTERIAL HEMAGGLUTINATION AND HEMOLYSIS 185

hemagglutination test and some observa- five strains of leptospirae. Am. J. Trop.tions on the antigen. J. Immunol., 69, Med., 3, 481-489.587-596. 86. Cox, C. D. 1955 Hemolysis of sheep

73. CARTER, G. R. 1955 A hemagglutination erythrocytes sensitized with leptospiraltest for the identification of serological extracts. Proc. Soc. Exptl. Biol. Med.,types. Am. J. Vet. Research., 16, 481- 90, 610-615.484. 87. BOYDEN, S. V., AND ANDERSEN, M. E. 1955

74. ALExANDER, M. M., WRIGHT, G. G., AND Agglutination of normal erythrocytes inBALDWIN, A. C. 1950 Observations on mixtures of antibody and antigen, andthe agglutination of polysaccharide-treated haemolysis in the presence of complement.erythrocytes by tularemia antisera. J. Brit. J. Exptl. Pathol., 36, 162-170.Exptl. Med., 91, 561-566. 88. STAUB, A.-M. 1954 Role des anticorps

75. WRIGHT, G. G., AND FEINBERG, R. J. 1952 antipolyosidiques dans l'agglutination desHemagglutination by tularemia anti- bacilles typhiques. Ann. inst. Pasteur,sera. J. Tmmunol., 68, 65-71. 86, 618-635.

76. BRODHAGE, H., AND FEY, H. 1955 Beo- 89. NETER, E., WESTPHAL, O., LUDERITZ, O.,bachtungen bei der Hfimagglutination- GORZYNSKI, E. A., AND EICHENBERGER, E.reaktion auf Bang und Tuberkulose mit 1956 Studies of enterobacterial lipopoly-menschlichen und tierischen Sera. Z. saccharides: Effects of heat and chemicalsHyg. Infectionskrankh., 141, 76-81. on erythrocyte-modifying, antigenic, toxic,

77. HIRSCHBERG, N., AND YARBROUGH, M. E. and pyrogenic properties. J. Immunol.,1952 Fractions of Brucella for adsorbed 76, 377-385.antigens for collodion agglutination and 90. MIDDLEBROOK, G. 1952 Antigens of tu-hemagglutination tests. J. Infectious Dis- bercle bacillus involved in hemagglutina-eases, 91, 238-245. tion and hemolysis reactions. Bull. N. Y.

78. CARRERE, L., AND Roux, J. 1952 Hemag- Acad. Med., 28, 474-475.glutination passive d'hematies sensibi- 91. NETER, E., AND ZALEWSKI, N. J. 1953 Thelis6es par antiganes brucelliques ou des requirement of electrolytes for the ad-substances solubles sp6cifiques. Ann. inst. sorption of Escherichia coli antigen byPasteur, 83, 810-813. red blood cells. J. Bacteriol., 66, 42442.

79. BOYDEN, S. V. 1950 Adsorption by eryth- 92. NETER, E., WESTPHAL, O., AND LtDERITZ,rocytes of antigens of Pfeifferella mallei 0. 1955 Effects of lecithin, cholesterol,and Pfeifferella whitmori. Proc. Soc. and serum on erythrocyte modificationExptl. Biol. Med., 73, 289-291. and antibody neutralization by entero-

80. MIDDLEBROOK, G. 1954 Laboratory aids bacterial lipopolysaccharides. Proc. Soc.to diagnosis and therapy. Ann. Rev. Exptl. Biol. Med., 88, 339-341.Med., 5, 339-3. 93. BOrDEN, S. V., AND GRABAR, P. 1954 Rble

81. SORKIN, E., AND BOYDEN, S. V. 1955 A des lipides dans la sensibilisation desstudy of antigens active in the Middle- erythrocytes par les constituants de labrook-Dubos hemagglutination test pres- tuberculine. Ann. inst. Pasteur, 87, 257-ent in filtrates of culture of Mycobacterium 267.tuberculosis. J. Immunol., 75, 22-27. 94. NETER, E., ZALEwsxx, N. J., AND ZAx, D. A.

82. GRABAR, P., BOYDEN, S., TAQuET, A., AND 1953 Inhibition by lecithin and choles-BORDUAS, A. 1952 Mise en evidence de terol of bacterial (Escherichia coli) hemag-deux anticorps diff6rents dans les s6rums glutination and hemolysis. J. Immunol.,tuberculeux par h6magglutination pas- 71,145-151.sive. Compt. rend., 234, 899-901. 95. FINE, M., FLOCKS, M., AND FEICHTMEIR, T.

83. MEYNELL, G. G. 1954 The antigenic struc- V. 1955 Nonspecific hemagglutinationture of Mycobacterium tuberculosis, var. inhibiting factor in human serum. Am. J.hominis. J. Pathol. Bacteriol., 67, 137- Med. Sci., 229, 670-675.150. 95a. LANDY, M., AND CEPPELLINI, R. 1955

84. FISHER, S. 1951 Antigenic relationships Production of 'O inagglutinability' inof erythrocyte adsorbable fractions of erythrocytes coated with typhoid Vi and 0some mycobacteria. Australian J. Exptl. antigens. Nature, 176, 1266-1267.Biol. Med. Sci., 29, 1-8. 96. SKILLMAN, R. K., SPURRIER, W., FRIEDMAN,

85. CHANG, R. S., AND McCoomB, D. E. 1954 I. A., AND SCHWARTZ, S. 0. 1955 Rheu-Erythrocyte sensitizing substances from matic fever activity determination by

on Decem

ber 29, 2019 by guesthttp://m

mbr.asm

.org/D

ownloaded from

Page 21: BACTERIAL HEMAGGLUTINATION AND HEMOLYSIS · 1956] BACTERIALHEMAGGLUTINATIONANDHEMOLYSIS 167 Bacteria, too, cause agglutination of erythro- ofbacterial hemagglutination. (3) Specifichemag-

186 ERWIN NETER [VOL. 20

two correlative methods. Arch. Internal on the generalized Shwartzman reaction.Med., 96, 51-60. J. Exptl. Med., 96, 605-624.

97. MIDDLEBROOK, G. 1950 A hemolytic modi- 111. EVANS, E. E., AND HAINEs, R. F. 1954fication of the hemagglutination test for The agglutination of ion exchange resinantibodies against tubercle bacillus anti- particles coated with polysaccharide. J.gens. J. Clin. Invest., 29, 1480-1485. Bacteriol., 68, 130-131.

98. GERSTL, B., DAVIS, W. E., JR., KIRSH, D., 112. ADLER, F. L. 1952 Bactericidal actionHOLLANDER, A. G., BARBIERI, M., AND mediated by antibodies specific for het-WEINSTEIN, S. B. 1955 Detection of ap- erologous antigens adsorbed to bacterialparently absent circulating antibodies in cells. Proc. Soc. Exptl. Biol. Med., 79,tuberculous sera. Am. Rev. Tuberc., 72, 590-593.345-355. 113. NELSON, R. A., JR. 1953 The immune-

99. ADLER, F. L. 1953 Studies on the bac- adherence phenomenon. Science, 118, 733-tericidal reaction. J. Immunol., 70, 79- 737.88. 114. RIECKENBERG, H. 1917 Eine neue Im-

100. COLE, L. R., AND FAVOUR, C. B. 1955 munitftsreaktion bei experimenteller Try-Correlations between plasma protein panosomen-Infektion: die Blutplittchen-fractions, antibody titers, and the passive probe. Z. Immunititsforsch., 26, 53-64.transfer of delayed and immediate cutane- 114a. LAMANNA, C., AND HOLLANDER, D. H. 1956ous reactivity to tuberculin PPD and Demonstration of particulate adhesion oftuberculopolysaccharides. J. Exptl. Med., the Rieckenberg type with the spirochete101, 391-420. of syphilis. Science, 123, 989-990.

101. CHANG, S. 1953 A serologically-active 115. CEPPELLINI, R., AND DE GIIGoRIo, M.erythrocyte-sensitizing substance from 1953 Crisi emolitica in animali batterio-typhus rickettsiae. J. Immunol., 70, 212- immuni trasfusi con sangue omologo214. sensibilizzato in vitro mediante l'antigene

102. CIANG, R. S., Mumv, E. S., AND SNYDER, batterico specifico. Boll. ist. sieroterap.J. C. 1954 Erythrocyte-sensitizing sub- milan., 32, 445-453.stances from rickettsiae of the Rocky 115a. NUNGESTER, W. J., AND KLEIN, L. F. 1937Mountain spotted fever group. J. Im- Effect of pneumococcus type III specificmunol., 73, 8-15. polysaccharide on sedimentation of blood

103. MUNIZ, J. 1950 On the value of "condi- cells. Proc. Soc. Exptl. Biol. Med., 36,tioned hemolysis" for the diagnosis of 315-317.American trypanosomiasis. "O Hospi- 115b. YOUNGNER, J. S., AND NUNGESTER, W. J.tal", 38, 685-691. 1944 The effect of type III pneumo-

104. KAGAN, I. G. 1955 Hemagglutination after coccus polysaccharide and gelatin on theimmunization with schistosome antigens. circulation and sedimentation rate ofScience, 122, 376-377. erythrocytes in mice. J. Infectious Dis-

105. VOGEL, R. A., AND COLLINS, M. E. 1955 eases, 74, 247-253.Hemagglutination test for detection of 116. WESTPHAL, O., AND LtDERITZ, 0. 1954Candida albicans antibodies in rabbit Chemische Erforschung von Lipopoly-antiserum. Proc. Soc. Exptl. Biol. Med., sacchariden gramnegativer Bakterien.89, 138-140. Angew. Chem., 66, 407-417.

106. NORDEN, A. 1949 Agglutination of sheep's 117. WOOD, W. B., JR. 1955 The pathogenesiserythrocytes sensitized with histoplasmin. of fever. Am. J. Med., 18, 351-353.Proc. Soc. Exptl. Biol. Med., 70, 218-220. 118. TODD, J. P. 1955 Bacterial pyrogens. J.

107. McENTEGART, M. G. 1952 The applica- Pharm. and Pharmacol., 7, 625-641.tion of a haemagglutination technique to 119. PILLEMER, L., SCHOENBERG, M. D., BLUM,the study of Trichomona vaginalis infec- L., AND WuRZ, L. 1955 Interaction oftions. J. Clin. Pathol., 5, 275-280. the properdin system with polysaccharides.

108. Dim GREaIuo, M. 1955 Fissazione di Science, 122, 545-549.antigeni sulla superficie cellulare. Boll. 120. PILLEMER, L., BLUM, L., LEPow, I. H.,ist. sieroterap. milan., 34, 118-122. Ross, 0. A., TODD, E. W., AND WARDLAW,

109. BOYDEN, S. V. 1953 Fixation of bacterial A. C. 1954 The properdin system andproducts by erythrocytes in vivo and by immunity. Science, 120, 279-285.leucocytes. Nature, 171, 402-403. 121. SILVERSTEIN, A. M., AND MALTANER, F.

110. THOMAS, L., AND GOOD, R. A. 1952 Studies 1952 Hemolysis with complement of

on Decem

ber 29, 2019 by guesthttp://m

mbr.asm

.org/D

ownloaded from

Page 22: BACTERIAL HEMAGGLUTINATION AND HEMOLYSIS · 1956] BACTERIALHEMAGGLUTINATIONANDHEMOLYSIS 167 Bacteria, too, cause agglutination of erythro- ofbacterial hemagglutination. (3) Specifichemag-

1956] BACTERIAL HEMAGGLUTINATION AND HEMOLYSIS 187

intact azo-erythrocytes sensitized with intradermal neutralization test for theantisera homologous to the attached azo- assay of diphtheria antitoxin in humangrouping. J. Immunol., 69, 197-200. sera. Am. J. Hyg., 61, 143-154.

122. NETER, E., WESTPHA, O., LVDERiTZ, O., 134. CHUN, T. H., AND MEYER, K. F. 1954 AGINO, R. M., AND GORZYNSKI, E. A. 1955 hemagglutination test with the proteinDemonstration of antibodies against en- fraction of Pasteurella pestis. J. Im-teropathogenic Escherichia coli in sera of munol., 72, 282-298.children of various ages. Pediatrics, 16, 135. MCCRBum, F. R., JR., MERCIER, S., CH1EN,801-808. T. H., MEYER, K. F., AND GOODNER, K.

123. NETER, E., ZALEwsKi, N. J., AND EzRGUSON, 1955 Studies on the antibody patterns inW. W. 1953 Escherichia coli hemag- pneumonic plague patients. J. Infectiousglutinin response of adult volunteers to Diseases, 96, 88-94.ingested E. coli 055 Be. Proc. Soc. Exptl. 136. LANDY, M., AND TRAPANI, R. J. 1954 ABiol. Med., 82, 215-219. hemagglutination test for plague antibody

124. WARBURTON, M. F., KEOGH, E. V., AND with purified capsular antigen of Pa8-WILUAMS, S. W. 1949 A haemagglutina- teurella peetis. Am. J. Hyg., 59, 150-156.tion test for the diagnosis of influensal 137. WARN, J., WALZ, U., REEDAL, J. S., ANDmeningitis. Med. J. Australia, 1, 135-137. AJL, S. J. 1955 Immunological proper-

125. MAILARD, E. R., AND GAGLARDO, F. J. ties of purified Pasteurella pestis toxin.1952 The value of absorption in serologic J. Bacteriol., 70, 170-176.tests for tuberculosis. Am. Rev. Tubero., 138. MEYNULL, G. G. 1954 The antigenic struc-66, 762-764. ture of Mycobacterium tuberculosis, var.

126. GERNEZ-RIUux, C., MONTESTRUC, E., AND hominis. J. Pathol. Bacteriol., 67, 137-TACQuET, A. 1951 Les r6actions d'- 150.h6magglutination et d'hdmolyse condi- 139. BOYDEN, S. V., AND SOWN, E. 1955 Ationnee dans les diffdrentes formes de la study of antigens active in the tanniclpre. Ann. inst. Pasteur Lille, 4, 3-15. acid hemagglutination test present in

127. LEVINE, M. 1951 Hemagglutination of tu- filtrates of culture of Mycobacterium tu-berculin sensitized sheep cells in Hansen's berculosis. J. Immunol., 75, 15-21.disease (leprosy). Proc. Soc. Exptl. Biol. 140. HINZ, C. F., JR., AND PILLEMER, L. 1955Med., 76, 171-173. The requirement for the properdin system

128. 5ANcK, H. H., AND SPAUN, J. 1953 Sero- in the hemolysis of human erythrocyteslogical diagnosis of chronic typhoid car- treated with tannic acid. J. Clin. Invest.,riers by Vi haemagglutination. Acta 34, 912.Pathol. Microbiol. Scand., 32, 420-4M. 140a. PRtssiuN, D., CAMPBELL, D. H., AND

129. BOYDEN, S. V. 1951 The adsorption of PAULING, L. 1942 The agglutination ofproteins on erythrocytes treated with intact azo-erythrocytes by antisera ho-tannic acid and subsequent hemagglutina- mologous to the attached groups. J.tion by antiprotein sera. J. Exptl. Med., Immunol., 44, 101-105.93, 107-120. 141. STAVIxTKY, A. B., AND ARQUILLA, E. R.

130. STAVITSKY, A. B. 1954 Procedure and 1955 Procedure and applications of he-general applications of hemagglutination magglutination and hemagglutination-in-and hemagglutination-inhibition reactions hibition reactions with bis-diazotizedwith tannic acid and protein-treated red benzidine and protein-conjugated redblood cells. J. Immunol., 72, 360-367. blood cellos. J. Immunol., 74, 306-312.

131. STAVITSKY, A. B. 1954 Specific applica- 142. COLE, L. R., AND FARLL, V. R. 1955 Ations of hemagglutination and hemag- method for coupling protein antigens toglutination-inhibition reactions with tan- erythrocytes. J. Exptl. Med., 102, 631-nic acid and protein-treated red blood 645.cells. J. lmmunol., 72, 368-375. 143. CooMBs, R. R. A., HowARD, A. N., AND

132. FISHER, S. 1952 The estimation in vitro WILD, F. 1952 Titration of antisera toof small amounts of diphtheria antitoxin soluble proteins on the basis of an agglu-by means of a haemagglutination tech- tination reaction. Brit. J. Exptl. Pathol.,nique. J. Hyg., 50, 445-456. 33, 390-397.

133. LANDY, M., TRAPANI, R-J., Forw", R., 144. CoomBs, R. R. A., HOWARD, A. N., ANDAND KLUGLER, I. 1955 Comparison of a MYNORS, L. S. 1953 A serological pro-hemagglutination procedure and the rabbit cedure theoretically capable of detecting

on Decem

ber 29, 2019 by guesthttp://m

mbr.asm

.org/D

ownloaded from

Page 23: BACTERIAL HEMAGGLUTINATION AND HEMOLYSIS · 1956] BACTERIALHEMAGGLUTINATIONANDHEMOLYSIS 167 Bacteria, too, cause agglutination of erythro- ofbacterial hemagglutination. (3) Specifichemag-

188 ERWIN NETER [VOL. 20

incomplete or non-precipitating antibodies Australian J. Exptl. Biol. Med. Sci., 25,to soluble protein antigens. Brit. J. 127-136.Exptl. Pathol., 34, 525-534. 151. STEWART, W. B., PETENYI, C. W., AND ROSE,

145. THOMSEN, 0. 1927 Ein vermehrungsfAhiges H. M. 1955 The survival time of canineAgens als Veranderer des isoagglutina- erythrocytes modified by influenza virus.torischen Verhaltens der roten Blutkor- Blood, 10, 228-234.perchen, eine bisher unbekannte Quelle 152. WRIGHT, G. G., AND SLEIN, J. B. 1951der Fehlbestimmung. Z. Immunitits- Variation in the serum T-agglutininforsch., 52, 85-107. during anthrax-infection in the rabbit.

146. FRIEDENREICH, V. 1930 The Thomeen he- J. Exptl. Med., 93, 99-106.magglutination phenomenon. Levin and 153. DAVIDSOHN, I., AND TOHARSKY, B. 1940Munksgaard, Copenhagen, Denmark. The production of bacteriogenic hemag-

147. MOSKOWITZ, M., AND TREFFERS, H. P. 1950 glutination. J. Infectious Diseases, 67,An agglutinin in normal sera for periodate- 25-41.treated red cells. Science, 111, 717-719. 154. DAVIDSOHN, I., AND TOHARSKY, B. 1942

148. BURNET, F. M., AND ANDERSON, S. G. 1947 Bacteriogenic hemagglutination. J. Im-The "T" antigen of guinea-pig and human munol., 43, 213-225.red cells. Australian J. Exptl. Biol. 155. SPIELMANN, W. 1954 Versuche zur Modi-Med. Sci., 25, 213-217. fizierung von anti-Rh-testseren durch

149. CHu, C. M. 1948 Enzymic action of viruses Bakterien und deren Kulturfiltrate, Z.and bacterial products on human red Immunititsforsch., 111, 460-470.cells. Nature, 161, 606-7. 156. NETER, E. 1956 Microbial hemagglutina-

150. MCCREA, J. F. 1947 Modification of red- tion as an immunologic technique. (Edi-cell agglutinability by Cl. welchii toxins. torial) Am. J. Clin. Pathol., 26, 165-168.

on Decem

ber 29, 2019 by guesthttp://m

mbr.asm

.org/D

ownloaded from