Detection of Leukocyte Antibodies by Means of I131-Labelled, Purified Anti--Human-Globulin Antibody. Problems of Nonspecific Adsorption of Globulin by Leukocytes

Initial studies of a quantitative method for detecting sensitization of whiteblood cells by antileukocyte antibodies are presented. The method entailsisolating I¹³¹-labelled, anti—human-globulin antibody (prepared in rabbits) byadsorption and elution from insoluble antigen, globulinazobenzylcellulose.The degree of adsorption of the labelled antibody by presumably sensitizedas opposed to nonsensitized leukocytes is measured by radioactive isotopecounting technics. The vexing problem of nonspecific adsorption of human orrabbit globulin by leukocytes is considered, and procedures for partially circumventing this problem are set forth. Seligmann's solution with normal rabbit serum added to a final concentration of 10 per cent appeared to be themost effective washing fluid in most instances for removing nonspecifically adsorbed globulins from leukocytes. The use of purified anti—human-globulinantibody, as opposed to whole Coombs serum, also very favorably affectedthe ratio between immunologically specific adsorption and nonspecific adsorption. Results of the over-all I¹³¹ method as applied to selected positiveand negative human sera are detailed.

Submitted on April 28, 1960 Accepted on July 3, 1960     

Mixed agglutination of leukocytes and erythrocytes in relation to studies on leukocyte antigens

Serologic experiments are described in which was studied the agglutination ofleukocytes from donors representing various ABO, Rh, and Lewis erythrocytegroups an certain canine blood groups, in the presence of corresponding antisera.

Appropriate mixing of antisera with leukocytes and erythrocytes from donorsof different groups was seen to produce clumping of leukocytes which did notconform to the reactions of the erythrocytes from the leukocyte donor. Whenviewed under phase microscopy, certain of these leukocyte clumps, which appeared homogeneous with ordinary light microscopy were found to be clumps ofleukocytes mixed with ghosts of erythrocytes reacting with the antibody present.

Factors contributing to this apparently immunologically non-specific clumping were the presence of complement-fixing, potentially hemolytic antibody andthermolabile components of serum. A possible relationship with erythrophagocytosis is suggested.

These observations indicate that certain results of this and other "leukoagglutination" technics, which have been interpreted as demonstrating thepresence of A and B antigens on human leukocytes, deserve re-evaluation andemphasize the importance of developing methods of preparing homogeneousleukocyte suspensions.

     

Studies on Hemoglobin: II. The Effect of Anti-Hemoglobin Sera on the Red Blood Cells in Vitro and in Vivo

Immune sera produced against adult human, fetal, canine and rabbithemoglobin solutions were found to agglutinate specifically the papain-treated red cells of the respective donors to a high titer. Intact red cells werealso specifically agglutinated by the same antibody-containing sera in thepresence of dextran, suggesting that the antibodies produced are incompleteones.

A marked shortening of the life span of rabbit erythrocytes followed theintravenous injection of anti-rabbit hemoglobin serum into normal rabbits.

The possible implications of the above findings in human pathology arediscussed.

Submitted on February 26, 1963 Accepted on July 19, 1963     

The Lutheran blood groups: a second example of anti-Lu b and three further examples of anti-Lu a

1. An example of the blood group antibody, anti-Lu^b, was found in a patientwho had a mild hemolytic transfusion reaction. It was shown to possess thecharacteristics of an immune antibody and to be able to distinguish between asingle dose and a double dose of the Lu^b gene.

2. Three new examples of the antibody, anti-Lu^a, are presented. All of themwere found in normal blood donors and have properties which indicate that theyare naturally occurring antibodies.

Dr. R. R. Race and Dr. R. Sanger confirmed the presence of anti-Lu^b in Mrs. S.’sserum, and studied other members of her family and the three anti-Lu^a sera. We aregrateful to them for many favors and their kind encouragement.

We are obligated to Miss Marie Cutbush for making available the Lu^aLu^a cells fromMrs. R. and her sister, and for a supply of anti-Lu^b serum.

Thanks are due to Dr. A. E. Mourant who furnished our original supply of anti-Lu^aserum and to Dr. Philip Levine for the anti-Tj^a and anti-Vel sera.

We are indebted to Dr. J. M. Fine of Milwaukee for permission to study Mrs. S. andto the patient and her family for their cooperation.

The sera from 18,613 blood donors were studied by Betty McCarthy, Rosemary Polka,Pearl Lemke, Agnes Molnar, Jeannette Flagstadt and Betty Hutter.

Submitted on March 20, 1957 Accepted on July 1, 1957     

Specific and cross-reacting antibodies in ABO heterospecific twin pregnancy

A "study in depth" is reported concerning the case of hemolytic disease of agroup B infant born to a group O mother, whose group A twin was apparentlyunaffected. It was shown that the hemolytic disease of the affected infant wasdue to a specific anti-B antibody. The study included parallel examinations ofthe antibodies in the sera of the three individuals.

The specificity of the B-anti-B reaction was demonstrated in vitro and in vivo.The powerful anti-B antibody of the mother had no effect on the group A twin,in whose serum anti-B was present in large amounts.

In vitro, studied by the usual technics of cross absorption the maternaland the fetal anti-A and anti-B serum antibodies behaved as if strictly specific.By applying successive absorption, elution and neutralization techniques, however, it was possible to demonstrate additional cross-reacting antibodies in thematernal serum which could be separated from one another and from the specificanti-A and anti-B antibodies. From the erythrocytes of the normal group Atwin such cross-reacting antibody could be eluted.

The cross-reacting anti-B antibody, isolated in pure form in eluates, could beshown to be loosely attached to group A erythrocytes without producing visibleagglutination reactions while after elution from A cells it did visibly agglutinategroup B cells. It could be eluted from A cells, absorbed by fresh A cells and reeluted while retaining its anti-B effect. It was neutralized by group B saliva only.A separate anti-A antibody with similar properties was eluted from B cells andspecifically neutralized by group A saliva.

A partial affinity of these antibodies for heterologous erythrocytes but not forspecific soluble substances was thus demonstrated. These findings support neitherthe linkage hypothesis of cross reactions between anti-A and anti-B nor theC-anti-C hypothesis of hemolytic disease. They are in keeping with the view thatgroup O sera contain variable complexes of anti-A and anti-B antibodies, composed of multiple fractions with different partial specificities. It is suggested thatthe occurrence or non-occurrence of cross-reacting antibodies found in sera ofgroup O mothers whose infants develop hemolytic disease is best explained onthis basis. It is further stressed that the demonstration of an antibody in motheror child in ABO hemolytic disease does not necessarily indicate its pathologicsignificance.

Submitted on January 11, 1957 Accepted on April 19, 1957     

On the interaction of dead leukocytic nuclei,L. E. factor and living leukocytes in the L. E. cell phenomenon

1. Isolated leukocyte nuclei quantitatively adsorb lupus factor from serum.In the presence of viable leukocytes such nuclei which have adsorbed thelupus factor become transformed into typical L.E. cells.

2. Leukocytes with an intact cytoplasm fail to adsorb L.E. factor.

3. These studies suggest that the living leukocytes, apart from simple phagocytosis, serve other functions in the L.E. phenomenon:

(a) the disruption of the integrity of the cytoplasmic envelope of deadleukocytes so as to allow for the adsorption of L.E. factor to the cell nucleus.

(b) the transformation of the nuclei, to which L.E. factor has been adsorbed, into swollen homogenous lupus bodies.

Submitted on October 7, 1957 Accepted on February 10, 1958     

Selenoid (crescent) bodies

"Selenoid (Crescent) Bodies" are known in hematology as Cuerpos en MediaLuna in Spanish and Corps en demi-lune in French.

An original method of staining selenoid bodies and one which is suitablefor morphologic structural studies has been described. The method is alsosuitable for staining stromas of erythrocytes.

Selenoid bodies originate from erythrocytes and correspond to their stromas.Selenoid bodies can be produced in vitro by the spreading of blood on aslide. Two factors influence their formation: a mechanical factor by frictionof the erythrocytes against the surface of the slide and a chemical essentialfactor represented by the lipids of the blood, increasing the fragility of theerythrocytes.

The number of selenoid bodies in any smear of blood is directly proportional to its contents in lipids, either from an exogenous source (alimentarylipemia) or from an endogenous origin.

Selenoid bodies have also been found in the dog, sheep, rabbit, guineapig, rat and hen.

Submitted on August 8, 1957 Accepted on November 18, 1957     

An in vitro technic for quantitating and studying the dynamics of leukocyte agglutination

1. An in vitro technic for quantitating leukocyte agglutination has beendescribed.

2. With this technic the dynamics of leukocyte agglutination and the response of leukocytes to antigenic agents in both sensitive and nonsensitivesystems have been studied.

3. The relationship of this study to the problem of agranulocytosis isdiscussed.

Submitted on June 21, 1957 Accepted on November 25, 1957     

A rapid slide test for heterophile antibody in infectious mononucleosis

The sera of 473 individuals were examined for sheep cell agglutinins both by theslide test and the Paul-Bunnell method. In this group there were 46 patients withpositive slide tests and 35 of these individuals also had a diagnostic serum dilutiontest for heterophile antibody. In 11 cases the slide test was positive but the Paul-Bunnell test gave very low serum dilution values. However, when the slide testwas carried out at 37 C, it was negative in 9 of the 11 cases. In the remaining 2instances, one patient had a Forssman type of antibody which gave a 1:64 titer insaline and the slide test was positive at 37 C. In the other case no studies were madeon the effect of temperature and the nature of the agglutination reaction wasunfortunately not determined.

Using human and bovine albumen, sheep serum and human AB serum absorbedwith sheep cells as a diluent no evidence for blocking or hyperimmune antibodywas discovered in the cases of infectious mononucleosis studied in this series.Moreover, of the 6 patients with negative serology but with strong clinical andhematological evidence for the disease, no blocking or hyperimmune antibodywas disclosed by the slide test or by the use of absorbed human AB serum. Theconclusion seems justified that blocking, incomplete or hyperimmune heterophileantibody must be rather uncommon in infectious mononucleosis.

In the use of the rapid slide test it has been pointed out that cold agglutinins,(which may be abolished by warming to 37 C) and Forssman antibodies (whichmay be absorbed by guinea pig kidney) can give positive results. However, diseases in which cold agglutinins are strong enough to give a positive slide testare relatively rare and the occurrence of Forssman antibodies of a strength likely togive a positive slide test would be decidedly uncommon. In any event unless furtherexperience reveals more serious discrepancies, the rapid slide test as described inthis paper seems to offer a practical screening test to detect clinically significantamounts of heterophile antibody in cases of infectious mononucleosis.

     

Studies on the Mechanism of in Vitro Acid Hemolysis in Paroxysmal Nocturnal Hemoglobinuria

1. The mechanism of acid hemolysis of PNH erythrocytes appears to be adirect mechanism, with the production of an initial "hole" in the membranesufficiently large to permit the direct egress of hemoglobin.

2. PNH erythrocytes react like normal red cells with rabbit antihuman redcell antibody and human or guinea pig C' at pH 7.2 to produce a membranedefect which is less than 32.5 Å in effective diffusion radius.

3. Inhibition of acid hemolysis of PNH erythrocytes by high molecularweight dextrans (Dextran 150) is also associated with inhibition of K⁺ loss,indicating inhibition of complement.

Submitted on April 13, 1966 Accepted on January 9, 1967     

Studies on the etiology of the elevated serum isomerase in chronic myelocytic leukemia

1. The phosphohexose isomerase activity of the serum of patients with chronicmyelocytic leukemia closely parallels the rise and fall in total granulocyte count.

2. Its separated leukocytes the isomerase activity of the granulocytes wasfound to be approximately three times that of the lymphocytes.

3. The serum isomerase activity was estimated in dogs with induced leukocytosis and leukopenia. The enzyme activity bore a closer relationship to leukocyte destruction than to leukocyte production.

4. These findings suggest that the elevated serum phosphohexose isomeraseactivity found in patients with chronic myelocytic leukemia probably originatesfrom disintegrating granulocytes.

Submitted on May 1, 1957 Accepted on September 17, 1957     

Influence of Granulocyte Antibodies on Granulocyte Function

The following substances were tested for their influence on granulocyte function: 8 sera that contained human granulocyte-specific alloantibodies against the antigens NA1, NA2 and NB1, two HLA antisera, and the monoclonal antibodies W6/32 and CLB-FcR-gran 1. The effects examined included spontaneous and directed migration, immune phagocytosis inhibition and the generation of oxygen radicals. Using the under-agarose technique, spontaneous migration of sensitized granulocytes was normal. For all antibodies tested, the chemotactic index for N-fMLP, LTB₄ and opsonized zymosan was greater than 1. Granulocyte immune phagocytosis of sensitized sheep red blood cells was strongly inhibited by all alloantisera and monoclonal antibodies. The generation of oxygen radicals after triggering the respiratory burst with sensitized sheep red blood cells was also strongly inhibited in the chemiluminescence assay. Immune phagocytosis and chemiluminescent response of granulocytes lacking the corresponding antigen of the tested alloantibodies were not affected. Since sensitization of neutrophils with F(ab')₂ fragments of the monoclonal antibodies W6/32 and CLB-FcR-gran 1 showed lower inhibition of generation of oxygen radicals after triggering, Fc-dependent interaction with the target cells seems to be necessary for inhibition. Our results suggest that binding of NA1-, NA2- or NB1-specific alloantibodies to granulocytes not only causes neutropenia, but also impairs granulocyte function.     

A summary of atypical antibodies,rare genotypes,and ABO hemolytic disease encountered in a one year survey

1. Among 13,000 selected blood specimens studied in 1954, there were 1882which contained atypical antibodies, mainly anti-D.

2. Of the 177 sera with atypical antibodies other than anti-D, 85 containedanti-c, anti-E or anti-K and 18 sera contained anti-Le^a.

3. There were 52 cases of intragroup hemolytic disease, more than half ofwhich were due to anti-c.

4. Intragroup isoimmunization with production of antibodies other thananti-D was detected in 47 cases as incompatibility in the crossmatching. In 31additional cases failure to detect incompatibility resulted in transfusion reactions.

5. There were 30 examples of very rare Rh genotypes due to the presence ofrare chromosomes dCe (r'), dcE (r'), DCE (R^z), and the rarest of all, dCE (R^y).All told, there were 7 bloods which contained chromosome dCE (r^y) includingone—the first of its kind—which was homozygous.

6. It is recommended that rare genotypes be registered on a world-wide basisunder the auspices of a suitable agency and that these bloods be made availableas test cells or as compatible donors.

7. Sixty-two cases of ABO hemolytic disease are reviewed briefly. In all butfour the mothers were in group O. Of the 55 group O mothers tested, all but 5showed enhancement of the specific antibody in the indirect Coombs reaction.

Submitted on April 9, 1956 Accepted on June 25, 1956     

The Unitary Nature of "Complete" and "Incomplete" Pathologic Cold Hemagglutinins

Several human pathologic sera containing high titered cold agglutinins werestudied to determine whether the serologic activity ascribed to an "incompletecold antibody" could be separated from the "complete" cold agglutinin activity.Separation was not achieved by physicochemical methods, including zoneelectrophoresis, density gradient ultracentrifugation, and anion exchangechromatography. Both activities were susceptible to destruction by mercaptans.Neither activity could be differentially absorbed from the sera. Using "Bombay" and I-negative ("i") red cells, a difference in specificity of the two activities for the H or I antigen of human erythrocytes could not be demonstrated.The simplest interpretation of these findings is that there is only one antibodyinvolved, the cold agglutinin, and that the serologic manifestation usually attributed to an additional "incomplete cold antibody", i.e. the production of apositive antiglobulin reaction of the "non--globulin" type, results from aninteraction of complement components with the cold agglutinin-erythrocytecomplex.

Three of these cold agglutinating sera were unreactive with I-negativeerythrocytes, in keeping with the reported anti-I specificity of these antibodies. A fourth serum retained moderate, though greatly reduced, activityagainst these cells, and the interpretation of this finding is discussed.

The anti-H specificity of the incomplete cold antibodies in normal humansera was confirmed by their failure to sensitize "Bombay" erythrocytes. Thiswas in sharp contrast to the excellent reactivity of the pathologic sera withthese cells, demonstrating that pathologic cold agglutinins are unrelated tothe incomplete cold antibodies present in most normal sera.

Submitted on October 23, 1961 Accepted on December 2, 1961     

Absence of dengue 2 infection enhancement in human sera containing Japanese encephalitis antibodies

Sera from 52 young adults resident in a rural area in North Thailand were studied for plaque-reducing neutralizing antibodies against dengue (DEN) viruses types 1-4 and Japanese encephalitis (JE), and for DEN-2 infection-enhancing antibodies using a newly described microtest in the human monocyte cell line, U-937. Infection-enhancing antibody titers in U-937 cells using a simplified micromethod were similar to results obtained by published methods using human peripheral blood leukocytes and a macrotest using U-937 cells. In the sample, there were 23 with antibodies to one or more DEN viruses with or without accompanying JE antibodies; 16 sera demonstrated antibodies only to JE and 13 had no detectable antibodies to any flavivirus. All but two DEN antibody-containing sera enhanced DEN-2 infections in U-937 cells, often to titers of 1:10,000 or greater. By contrast, only one of 16 JE-immune sera enhanced DEN-2 infection in monocytes, and that at a dilution of 1:100. None of the flavivirus-negative sera had DEN-2 enhancing activity. The failure of human anti-JE contrasts with the ability of rabbit anti-JE to enhance DEN-2 infections, but correlates with the absence of recorded instances of dengue shock syndrome in human beings sequentially infected with JE and then a DEN virus. This report seemingly reconciles in vitro and in vivo phenomena, and may provide an opportunity to study mechanisms involved.     

The specificity of antibodies to the F(ab′)2 fragment of human IgG

The specificity of IgG anti-F(ab′)₂ antibodies was examined in unfractionated sera of patients with rheumatoid arthritis and also with affinity-purified antibody preparations. Examination of the sera by an enzyme-linked immunosorbent assay, using pooled human F(ab′)₂ fragments absorbed to microtiter plates, revealed that IgG anti-F(ab′)₂ antibodies cross-react with human and rabbit IgG and rabbit F(ab′)₂. IgG anti-F(ab′)₂ antibodies were purified by affinity chromatography and, when tested by a fluid-phase inhibition enzyme-linked immunosorbent assay, were found to be of 2 types. One fraction, similar to pepsin agglutinator, reacted with human F(ab′)₂ fragment alone. The other fraction was cross-reactive with human IgG and yet failed to react with idiotopes on Fab or epitopes on Fc fragments. The IgG anti-F(ab′)₂ antibodies we purified had no reactivity toward a human immune complex prepared from tetanus toxoid and antitoxoid.     

Improvement of Leukocyte Bactericidal Activity in Chronic Granulomatous Disease

Chronic granulomatous disease (CGD) is characterized by an inability ofpatients’ leukocytes to generate hydrogen peroxide and to kill non-peroxide-forming bacteria, such as staphylococci and serratiae. We have introducedglucose oxidase into CGD leukocytes in order to generate peroxide and therebykill S. aureus and S. marcescens. These results support the concept that a leukocyte oxidative deficiency is primary to the pathogenesis of CGD.

Submitted on June 23, 1969 Accepted on September 8, 1969     

The efficiency of oxidative phosphorylation by normal and leukemic human leukocytes

1. A new modification of existing methods has been described for the separation of leukocytes from whole blood which provides a procedure for therapid isolation of uninjured cells suitable for the study of oxidative phosphorylations.

2. This method has been employed in a study of the relative efficiencyand yield of oxidative-linked phosphorylations mediated by normal and leukemic or immature leukocytes. The maximum aerobic phosphorylating capacitywas exhibited by chronic lymphocytic leukemic leukocytes, followed in decreasing order of activity by acute monocytic leukemic leukocytes and chronicmyelocytic leukemic leukocytes. Oxidative phosphorylation was not demonstrated with normal leukocytes.

3. Results of this study suggest that expression of leukocyte metabolicdata on a unit nitrogen basis more accurately reflect the morphologically obvious size differences among the various leukocytes than presentation of dataon a unit cell basis.

4. The aerobic phosphorylations mediated by leukemic leukocytes werefound to be dependent upon substrate addition and were depressed by lowlevels of dinitrophenol. Under the experimental conditions employed in thisstudy, glucose-6-phosphate was formed in stoicheiometric amounts. These results indicate that leukemic leukocytes are capable of the aerobic esterification of inorganic phosphate accompanying the oxidation of selected Kreb'scycle intermediates.

Submitted on August 2, 1957 Accepted on November 2, 1957     

Immunologically Specific Antigens in Leukemic Tissues

The injection of extracts of leukemic and Hodgkin’s disease tissues of manelicit an antibody reaction in both man and rabbit. This response is similarto that elicited by leukemic mouse tissues when injected into rabbits. Theantibody reaction may be demonstrated by the technics of passive cutaneousanaphylaxis, immunodiffusion, microprecipitin and immunofluorescence. Tissue extracts from non-leukemic individuals do not elicit a similar response.Rabbits immune-tolerant to normal human tissues produce antibodies specificto leukemic human antigens. Antibodies develop in those individuals who areexposed for a long time to either human or mouse leukemia.

These immunologic studies demonstrate specific antigenic differences between normal and leukemic tissue extracts. It is postulated that the differencebetween normal and leukemic extracts is the consequence of the presence ofviruses or the alterations caused by them.

Accepted on February 1, 1963     

The effect of Chlamydia trachomatis on luminol-dependent chemiluminescence of human polymorphonuclear leukocytes: requirements for opsonization

The factors involved in the in vitro interaction of two strains of Chlamydia trachomatis with polymorphonuclear leukocytes were studied by employing the technique of luminol-dependent chemiluminescence. Unopsonized, partially purified elementary bodies of chlamydia failed to induce a significant chemiluminescence response when compared with serum-activated zymosan (less than 90%). Opsonization of the chlamydia with human sera greatly enhanced the chemiluminescence response. This enhancement was independent of the presence or absence of antibody specific for chlamydia or of complement. Primate serum had 77% of the activity of human serum; nonprimate sera (sheep, cow, horse, and rabbit) demonstrated substantially less activity. The magnitude of the chemiluminescence response observed with opsonized chlamydia was also dependent on the chlamydia-to-polymorphonuclear leukocyte ratio, with the greatest effect seen at 10:1. Chlamydia opsonized with human sera containing less than 100 mg of IgG/dl did not stimulate chemiluminescence more than did unopsonized chlamydia. These results suggest that human IgG may interact with C. trachomatis independent of specific antibody-binding sites.