Blood Typing: Agglutination Reactions and Antibodies (ABO and MNSs Systems), Slides of Genetics

Download Blood Typing: Agglutination Reactions and Antibodies (ABO and MNSs Systems) and more Slides Genetics in PDF only on Docsity!

UNIT Vm

BLOOD GROUPS

Soumbook inForensic Serologv, Immunology, and Biochem&ry

Karl Landstdm (1869-1943) (^) Ludwig W e l d (1884-1954) Courtesy National Library of Medicine (^) Courtesy Col. Frank R. Camp, Jr. and A m y Mcdical Research Laboratory, Fort Knox, KY

Fritz Schiff (1889-1940) Courtesy Col. Frank R.Camp, Jr., and Army Medical Research Labora- tory, Fort Knox, KY

Alexander S. Wiener (1907-1976) Courtesy National Library of Medicine

254

Blood Groups-Applications of GeneticsMarker Systems

SECTION 18. INTRODUCTION TO THE FORENSIC APPLICATIONS

OF OENETIC MARKER SYSTEMS TO IDENTIFICATION AND

DISPUTED PARENTAGE PROBLEMS

There are two major application areas of forensic or medicolegal serology and biochemistry: ( 1 ) disputed par-

entage testing; and (2) identification testing of bloodstains

and secretions, usually in criminal matters. Although both these areas make use of the same genetic marker systems, they tend to be separate subspecialties in practice. It is ex- ceptional, at least in this country, to find laboratories en- gaged in both kinds of testing. In parentage and affiliation testing applications, any genetic marker system for which the mode of inheritance has been firmly established can be employed, at least in theory. In blood and secretion stain analysis, there is an additional dimension: the blood or body fluid materials in the stains or spots must be identified before genetic markers and typed. Analysis, therefore, consists of identification tests followed by individualization tests (gen- etic marker system typing). Typing of the various systems in dried blood and secretion materials, and in post mortem tissues and fluids, presents problems that are not enwun- tered with freshly taken samples. This book was prepared with stain analysis work foremost in mind, and it is or- ganized accordingly. Units I1 through IV have discussed identification issues. Units V through VIII are devoted to individualizing markers. Emphasis has been given to blood

and body fluid stain analysis, but parentage testing applica-

tions are discussed in connection with the various systems. In

this section, the general principles and considerations in-

volved in parentage testing and in stain identification analysis

are discussed briefly as an overall introduction to the units on

genetic marker systems. Genetic marker systems may be

divided into five major classes: blood groups, isoenzymes,

serum group systems, hemoglobin, and HLA. They are

discussed in the units which follow. A sixth class that would

include "other" genetic markers, such as the polymorphic

proteins of saliva and hair, could be added for completeness. Professor Dodd, in her presidential address to the British Academy of Forensic Sciences (Dodd, 1980), has given an excellent and thoroughly readable overview of both the civil and criminal aspects of forensic serology.

1 8.1 Dlsputd Parentage Testing Disputed parentage cases usually involve disputed pater- nity. The mother is assumed to be the genetic mother in

these cases.The legal issue is usually the support of a de-

pendent child. Disputed affiliation cases can also arise in connection with immigration matters and citizenship claims, inadvertent baby mix-ups, kidnapping, disputes over the in-

heritance of money or property, and divorce cases.

In theory, any genetic marker system for which the inher- itance patterns are straightforward and well established can

be employed in disputed affiliation cases. It must also be

clear that the genetic marker system is expressed in the child according to its own genotype. Probably over 60 systems could be listed as potentially applicable at the present time

(Chakraborty er al., 1974; Lee, 1975; Joint AMA-ABA

Guidelines, 1976). As a practical matter, far fewer systems are normally used in casework. Some systems are much more informative than others. Some require reagents which are not widely available. Others requre special equipment or special training and expertise which every laboratory does not have. Cost is a consideration as well, since it must usually be borne by the defendant. It is clear that laboratories must be highly skilled in the typing of all the systems they use in casework, and such skill is usually developed through a com- bination of background knowledge and considerable prac- tical experience. Lee (1975) reviewed the status of parentage testing, and gives comparisons of the various combinations of systems used in different countries. Polesky and Krause (1977) discussed the ca~abilitesof American laboratories. ~ e n e t i cmarker systems employed in disputed parentage tests must follow well known and established rules of inher- itance. Thus, a child cannot have a gene which both parents lack, and must inherit one of a pair of chromosomes from each parent (if the system is wntrolled by a single locus, the chromosome will carry one allele; if the system is controlled by a series of linked loci, the chromosome will carry a hap- lotype). Further, a child cannot have a pair of genes unless both parents have the gene, and a child must inherit a ge- netic marker allele for which either parent is homozygous. Apparent violations of these rules are the basis for exclusions of parentage. Apparent violations of the first two rules, i.e.

( I ) a child has an allele which is absent in both parents, or

(2) in a multiple allelic system, a child lacks both alleles which are found in the alleged father, are called direct exclu- sions (or "first order" or "primary" exclusions). With some extremely rare exceptions, these can be accepted with great confidence. Such exclusions are based upon the presence or absence of genetic markers demonstrable by direct exam- ination. Apparent violations of the second two rules, i.e., (1) a child is homozygous for a marker allele which is not present in both parents, or (2) a child lacks a marker for which the alleged father is homozygous, are known as indi- rect exclusions (or "second order" or "secondary" exclu- sions). They are based on the inference of homozygosity detected by a negative reaction in a particular test. These exclusions are interpreted with great caution, particularly if

there is a single second order exclusion (one system). Each system has its own rarities and peculiarities, and one must be familiar with them and take them into consideration when interpreting the results of a parentage investigation. The probability (PE) that a particular system will exclude

a falsely accused father a priori can be calculated, and de-

pends on the gene frequencies of the particular system in the population of interest. For most systems, the PE can be calculated quite easily (see Walker, in AABB, 1978). For

some of the more complex systems, like Rh, Gm and HLA,

the calculations are correspondingly more complicated. Ta-

bles of values have been published for many systems, howev-

er, for easy reference (e.g. AABB, 1978). The accuracy of

the PE value depends on the accuracy of the gene fre- quencies used to perform the calculation. The cumulative probability of exclusion (CPE) for a series of genetic marker systems can be computed from the individual PE values

according to: CPE = 1 - (1 - PI) (1 - P2) (1 - P,)

..... (1 - P,), where PI is PE for the first sysem, P2is PE for the second, etc., and P, is PE for the nIh system, and where n systems have been used. A list of genetic marker systems that are applicable in forensic serology is shown in Table 18.1. The table shows the approximate PE for each system for Black and White populations. To the extent that data were available, the values are applicable to most U.S. populations. HLA (section 46) is the most powerful system by far. The CPE for all the systems in Table 18.1 would exceed 99%. If a panel of systems has been employed for paternity testing, and no exclusions have been found, it may be of interest to calculate the probability of paternity. This value can be calculated by a number of different methods, some quite simple and some quite complicated (see Walker in AABB, 1978). The value will depend on which systems were used in the tests, and genotypic and phenotypic distributions. In some countries, this information is routinely calculated. In this county, the rules of evidence vary in different jurisdic- tions. In some states, blood grouping evidence is inadmis- sable in paternity cases unless it is exclusionary. In other states, courts will admit inclusionary evidence and estimates of the probability of paternity. Many genetic marker sys- tems have rare alleles. These are seldom encountered, but can be very informative in parentage cases if found in a child and a putative parent. The incidence of rare alleles at 43

enzyme loci was discussed by Harris et a /. (1974).

Sometimes, special situations arise that call for somewhat different approaches. PE values for cases involving relatives

were discussed by Salmon and Brocteur (1978). Asano et al.

(1978a and 1978b) discussed procedures and interpretation when all the parties to the case are not available for testing (e-g. deceased putative father). Dodd and Lincoln (1978 and 1979) presented results on

over 1,500 cases of disputed parentage investigated by blood

grouping.

18.2 Blood and Body Fluid Staln Indlviduallzatlon In criminal cases, individualizing genetic marker systems are used for comparisons of stains or spots with the known blood or body fluids of people known or thought to be involved. Firm conclusions can be drawn only in cases of nonidentity, that is, when the person suspected of having deposited the stain is excluded because he or she lacks a genetic marker found in the questioned material. If samples are compared and found to be identical in one or more systems, it means only that the person tested is included in the subset of the population having the particular set of types found. If the frequencies of occurence of the types for the systems employed are known for the population in question, then the expected frequency of occurrence of the type or set of types can be calculated, and may be informa- tive. The figure obtained in these calculationsdepends on the gene frequencies for the systems. The less accurate the gene frequency estimates, the less accurate will be the resulting estimate of frequency of occurrence of certain types. In gen- eral, the gene frequency estimate improves as more and

more people are typed. In a population as heterogeneous as

in the United States, it is sometimes difficult to decide how the "population in question" should be defined. The ability of a genetic marker system to distinguish be- tween individuals in a population is clearly related to how well the types are distributed. The probabilities of identity and discrimination can be calculated for genetic marker systems, and this subject was discussed in section 1.2.8. The

principles are also covered in Selvin et al. (1979). In

Table 18.1 is shown the DI for most of the systems. DI is a measure of the power, or value, of a system in distinguishing between individuals selected at random from a population. Comparisons of DI, and combined DI values for several systems, can help examiners and laboratories in choosing the panel of systems they will use in particular cases or in all cases. There are other important considerations in choosing systems as well.

In general, fewer systems can be used in stain analysis that

in the typing of fresh materials. Techniques for typing stains

are not worked out for every system. There are perhaps 20

systems that could theoretically be used in stain analysis. It

is unlikely that so many systems would ever be applicable at

a practical level because of the many other constraints im-

posed by casework samples. In many cases, the age of the

stains is a problem. In others, the sample size is very limited. There are cost considerations (in comparison with the bene- fit to be derived) in adding new systems to the routine anal- ysis scheme. In individual laboratories, resources may be

rather scarce, and this will be a decisive factor as well.

Nonetheless, a greater degree of individualization of stains is possible now than ever before. As new systems are added to the list of those applicable to this kind of work, the situ- ation promises to improve even further in the years ahead.

Blood Groups-ABO and Secretor System

SECTION 19. THE A S 0 AND SECRETOR SYSTEMS

19.1 Orlgins and Earlier Studies The ABO blood group system was discovered in 1900 by Karl Landsteiner. In a footnote to a paper, devoted primarily to other matters, he said:

The serum of healthy humans not only has an aggluti- nating effect on animal blood corpuscles, but also on human blood corpuscles from different individuals. It remains to be decided whether this phenomenon is due to original individual differences or to the influence of injuries and possible bacterial infection. I observed this behavior as especially pronounced in the case of blood from severely ill patients. This phenomenon could be related to the dissolving capacity of serum for blood corpuscles in the case of various diseases, as it was described by Maragliano (10th Congress of Internal Medicine, 1892) The full account of Landsteiner's observations appeared in

1901. The essential finding was that the sera of healthy persons agglutinated the red cells of certain other healthy persons. The data allowed the sera of the individuals tested to be placed into one of three groups: in one, called "A", the serum reacted with the cells of another group, called "B", but not with the cells of group "A"; the serum of group " B

reacted with group "A" cells; in the third group, called "C",

the serum agglutinated the cells of both groups "A" and "B", but the red cells of group "C" were not reactive with sera from either group "A" or "B". Landsteiner said that there must be at least two agglutinins present, one in the group "A" serum, another in the " B serum, and both to- gether in the serum of group "C". There was a body of opinion at the time which held that the phenomena were a function of certain disease states, but Landsteiner said that this thinking was not in accord with his findings. In some placental sera examined, the agglutinins appeared to be ab- sent. It was noted that sera dried for two weeks on linen still gave the observed reactions, and that this fact might very well be exploited for medico-legal purposes. Landsteiner concluded by noting that these characteristics of human blood permitted an understanding of the consequences of various kinds of transfusions. The phenomenon of aggluti- nation of human cells by human sera was called iso- agglutination, and the agglutinins in serum were referred to as isoagglutinins, after the suggestion of Ehrlich and Mor- genroth. In 1902, von Decastello and Sturli conducted fur- ther studies on the phenomenon, and could fully confirm all

of Landsteiner's observations. In addition, it was found that

the blood of 4 out of 155 persons behaved differently from the previously described three groups, i.e., the serum con- tained no isoagglutinins, and the cells were agglutinated by

the sera of all the other groups. The isoagglutination pat- terns showed no dependence on any pathological condition. Differences in the titer of isoantibodies in different individu- als were noted as well, along with individual differences in agglutinability of cells with a serum of constant strength. In some newborn blood samples, the isoagglutinins were not observed. In cases where they were observed, they were of lower titer than in adult bloods, and it appeared that there was variable development of receptors on the cells of new- borns as well. Halban and Landsteiner (1902) conducted a number of irnmunohematological studies comparing mater- nal with neonatal blood, and obtained results very like those of von Decastello and Sturli with respect to isoagglutination. In 1907, Jansky published a paper in a rather obscure Bohemian journal, written in the Czechoslovakian language, in which he examined isoagglutination using the cells of 99 persons with the sera of 30 others. The tests involved over 3100 individual agglutination tests, and he recognized the existence of four isoagglutination blood groups from his data. Peculiarly, he did not mention or cite Landsteiner's work, though he cited other German literature in the field, including von Decastello and Sturli's paper. Jansky classified the isoagglutination groups according to Roman numerals, Groups I through IV. In 1910, Moss examined blood from 100 persons and noted four isoagglutination groups as well. He, likewise, classified them according to a Roman numeral designation. Unfortunately, Moss group I corresponded to Jansky group IV, and conversely, (Moss, 1910a and 1910b). The adoption of one or the other systems of nomenclature by various different authors and institutions led to quite a bit of confusion in the literature for quite a few years. Von Dun- gern and Hirszfeld [sometimes spelled Hirschfeld] con- ducted extensive experiments on the isoagglutination groups beginning around 1910, and concluded that the four groups could be accounted for by the presence or absence of two isoagglutinogens, which they called A and B. The aggluti- nins were designated cu and (von Dungern, 1910; von Dun- gern and Hirschfeld, 1910a, 19lob and 1911). Guthrie and his colleagues in this country used upper case letters to des- ignate agglutinins and corresponding lower case letters to designate agglutinogens (Guthrie and Huck, 1923). Lattes (1932) said that such usage was not defensible, because it was clear by then that the agglutinogens were inherited as Mendelian dominant characteristics. There were periodic calls in the literature for standardization of the nomen- clature at the time (e.g. Verzhr, 1927; Aldershoff, 1927). In this country, the American Immunological, Pathological and Bacteriological Societies called for the universal adoption of the Jansky system in 1921. In 1928, Kennedy published a survey of the nomenclature usage in American

Soun:ebook in Forensic Serology, Immunologv, and Biochemistry

hospitals. At that time, 72% were using the Moss system exclusively, 16% the Jansky system, and the rest two or more systems simultaneously. The 1921 recommendation of uni-

versal adoption of the Jansky system had not resulted in any

significantly increased use of the system, the survey found. Finally, more than half the hospitals surveyed did not favor a standardized "compromise" system, and about 1/, felt that such a system would only lead to additional confusion. The

matter was finally settled in 1930 when the Permanent Stand-

ardization Commission of the League of Nations Health Organization reported the work of its Laboratory Confer-

ence on Blood Groups (League of Nations Health Organiza-

tion, 1930). A resolution to adopt essentially the scheme of von Dungern and Hirszfeld was adopted at an April 1928 ses- sion of the Permanent Standardization Commission held in

Frankfurt-am-Main. The test sera for groups A and B were to

be called anti-A and anti-B, respectively. A comparison of

the Moss, Jansky and International systems appears in Table

19.1. The use of a t o denote anti-A, and 6 to denote antiB is still encountered, and does not lead to any confusion. In 1931, Kennedy published a paper with the stated pur- pose of making available in the more accessible literature the information contained in the original paper of Jansky (1907). A careful analysis of the data was carried out, and it was argued that Jansky had been the first person to ob- serve and categorize all four blood groups from a single set of isoagglutination tests all in the same paper. He said that Shattock (1900) had actually been the first investigator to publish an observation of isoagglutination in humans. Most authorities think that Shattock was actually seeing rouleaux formation, and in any case, his experiments were carried out with other things in mind, and the notion of isoagglutination groups did not occur to him. Moreover, he was studying blood from sick patients. Zinsser and Coca (1931) published a few remarks on Kennedy's paper. They said that Shattock had undoubtedly been looking at rouleaux formation. More to the point, they seemed anxious to emphasize that the credit for the discovery of the blood groups blonged to Land- steiner, without detracting from Jansky's contribution. Kennedy had not made an explicit issue of the matter of primacy, though one could gain the impression from reading his paper that he regarded Jansky's report of all four groups in the same paper as more significant than the separate reports of Landsteiner, and of his collaborators, von Deca- stello and Sturli. Kennedy said that, as far as he could tell, Guthrie and Huck (1923) were among the few American workers who read Jansky's paper, and the only ones who had actually analyzed the data. It is quite clear, however, that Moss (1910a and 1910b) had seen at least an abstract of the paper after his own were in press. In an added note, he gave Jansky due credit, and noted the differences in their Roman numerical designations of the blood group. There is no longer any question about the significance and priority of Dr. Landsteiner's contribution. Landsteiner him- self did not regard the description of all four groups in one and the same paper as the principal issue. Von Decastello and Sturli were in fact his collaborators, and their report,

describing what was later called group AB, was considered a logical extension of the previous work. Speiser (1961) reproduced an interesting letter from Landsteiner to Sturli on this point. The letter, written February 12, 1921, was rendered in English in Prokop and Uhlenbruck (1969): 12.2.

The Hague van Slingelandstr. 39, Holland Dear Dr. Sturli, I was very pleased to receive your card. I should like to ask you this: Isoagglutinins, which we have already studied, have achieved great importance in America, as they do many transfusions and for this purpose a blood

group determination is always made. I do not know if

you are aware of this. You will find something about this, for example, in the first number of the Journal of the American Medical Assoc. of this year. The Ameri-

cans always say that I have found only 3 groups, as if

the few cases of the fourth group were the main thing. Only recently you and Decastello were cited as the authors of this fourth group, earlier it was usually Moss or Jansky who have not done anything new at all. It would be important for me to be able to say, when the occasion arises, that when you were working with me as my pupil or collaborator, you undertook with Decastello the continuation of my work. That was the way things happened. It does not harm, or rather it does not make

any difference to your position and I would appreciate

it, as other people could not then accuse me of an error

as they have until now. I myself saw at the time the

necessity for more numerous investigations and found, however, that it was desirable that you should do these. I should be glad if you would let me have a line about this. Have you seen the new book on Immunity by

Bordet?....

Yours, Landsteiner

In 1930, Landsteiner received the Nobel Prize for Phys- iology or Medicine "for his discovery of human blood groups". In his Nobel lecture, he mentioned the medico- legal applications of blood grouping. Until the blood groups

were discovered, he noted, "... .forensic medicine knew no

way of distinguishing between blood stains of different per- sons. Since the isoagglutinins and the corresponding aggluti- nogens will also keep for a considerable time in a dried condition, the problem can in certain cases be solved, in particular when the bloods in question, e.g. that of the ac- cused and that of the victim, belong to different

groups...... .according to a report by Lattes, who was the

first to use it in forensic cases, [the test] has proved useful in a number of cases, and has been the basis of court verdicts and of the acquittal of accused persons." Landsteiner's con- tributions to immunology and serology extend far beyond his discovery of the blood groups. Landsteiner was born in 1868, and died in 1943. He spent the last 21 years of his life at the Rockefeller University in New York. His scientific life was

Sourcebook in Forensic Serology, Immunology, and Biochemistty

absorption experiments by many workers, including Koeck- ert ( 1920), SchiStze ( 1921 ), Hooker and Anderson ( 1921 ) and Dyke (1922b), helped to establish the acceptance of the two agglutinogen-two agglutinin scheme. Group A cells ab-

sorbed only the a from group 0 serum, while group B cells

would selectively absorb the 8. Group AB cells absorbed

both agglutinins from group 0 serum, or a and 6 separately

from B or A sera. Hooker and Anderson (1921) prepared

heterologous antisera against human red cells in rabbits, and, by appropriate selective absorption of the antisera, could render some of them group specific. While the four group scheme was fairly widely accepted by 1920, apparent exceptions to the results expected on the basis of the theory had been observed. There were cases where agglutination between a serum and cells was expected but failed to occur, other cases in which cross-agglutination was observed between persons belonging to the same group,

and cases where 0 cells had been observed to agglutinate.

The experiments which may have attracted the most atten- tion in this regard were those of Guthrie and his collabo- rators. Based on extensive observations, they postulated at first one additional agglutinogen-agglutinin pair (Guthrie and Huck, 1923; Huck and Guthrie, 1924), and later several more agglutinogen-agglutinin combinations (Guthrie and Pessell, 1924a and 1924b; Guthrie et al., 1924). They pointed out that the number of theoretically possible combi- nations, even using two antigens and two antibodies, was not four but nine, without violating the principle that agglutinins never occur in the serum of a person whose cells have the corresponding agglutinogen (Landsteiner's Rule). Thus, one could imagine (using present-day nomenclature) OaB, Oa, 08, 00, AD, Ao, Ba, Bo and ABo, where "ow indicates no agglutinins. They had in fact observed a blood in which the cells were of group B, but the serum had no agglutinins, i.e.,

behaved as an AB serum. Coca and Klein (1923a and

1923b) accepted the interpretation of Guthrie and co-

workers that there were additional agglutinogen-agglutinin combinations and described one themselves which they called "X". Other workers confirmed the observations, and accepted the interpretation as well (e.g. Simson, 1926; Bun- ker and Meyers, 1927). Landsteiner and Witt (1924) at first thought that these observations might indeed indicate an additional pair of factors. They noted that two different types of group AB blood could be distinguished, one of which contained an agglutinin for certain group A cells, and the other of which did not. They later changed their opinion as to how this observation should be interpreted, as will be discussed below in connection with A subgroups. The Italian workers, Lattes and Cavazutti (1924) and Mino (1924) did not agree at all that it was necessary to invoke the existence of additional agglutinogen-agglutinin combinations to ex- plain the observations of atypical isoagglutination. These could be understood, they thought, on the basis of quan- titative differences in cell receptors and in agglutinin content

of sera, as well as of varying avidity of agglutinin for recep-

tor, while retaining the basic four group hypothesis. The

matter will be discussed somewhat more in connection with subgroups.

19.2 Inherlbnce of the ABO Blood Oroups Langer in 1903 and Hektoen in 1907 noted in passing that mothers and their children could have the same blood groups. If they entertained the notion that the blood groups were inherited, they were not explicit on the point. In 1908, Epstein and Otterberg reported the blood groups of two families at a meeting of the New York Pathological Society, and suggested that the blood groups might be inherited. The clear and unequivocal demonstration of the heredity of the blood groups, however, came from the work of von Dungern and Hirszfeld (von Dungern, 1910; von Dungern and Hirsz- feld, 191Ob). The first experiments were carried out in dogs with immune isoagglutinins, but they soon looked at a col- lection of data from 71 human families, including 342 peo- ple, in Heidelberg. The data were interpreted to mean that the agglutinogens, and not the agglutinins, were inherited. A blood group agglutinogen could not be found in children if it was lacking in both parents. If both parents have an aggluti- nogen, most of the children have it, but a few may not. And in a family where the parents are AB and 0 , A and B children are possible. The hereditary factors determining the presence of the agglutinogen were believed to be dominant, those determining their absence, recessive. The factors be- haved strictly according to Mendelian principles, and it was thought that A and B were determined at separate, indepen- dent genetic loci. This work formed the basis of hundreds of studies by many workers, including themselves. According to the von Dungern and Hirszfeld notion that the genetics of the system could be explained on the basis of two, indepen- dently inherited allelomorphic pairs, A being dominant to

non-A, and B being dominant to non-B, nine genotypes and

four phenotypes could be represented as shown in Table 19.2, where A and B represent the genes for agglutinogens

A and B, and a and b represent the recessive genes for their

respective absence. Hirszfeld (1928) reviewed in detail the experimental foundations for scheme, as well as numerous family and population studies with which it was consistent. In 1924, the mathematician Felix Bernstein published the first of his several papers in which he formulated a theory of inheritance for the ABO groups based upon statistical ge- netic considerations, and inferences drawn from the rather extensive body of population data which existed by that time. He introduced the notion of multiple alleles at a single genetic locus for the system. There was precedent for multi- ple allelic systems in Drosophila, but not in human genetics. According to Bernstein, there are three genes, A, B and R, which can give rise to the genotypes RR, RB, BB, RA, AA and AB. The agglutinins corresponding to A and B were

called a and B in conformance with the accepted usage, and

he said that an agglutinin p would correspond to R. Accord-

ing to this scheme, a child inherits only one blood group

1

Blood Groups-ABO and Secretor System

Table 19.2 Blood Group Genetics According to

vonDungern and Hirszfeld

Genotype

aabb

AAbb. Aabb

aaBB. aaBb

AABB, AABb,

AaBb. AaBB

genetic factor from each parent, instead of two as postulated

by von Dungern and Hirszfeld. Genotypically RR people

belonged to group 0, RB and BB people were of group B,

RA and AA people were group A, and AB people were of

group AB. Bernstein analyzed a considerable amount of the

available population data using a statistical genetic model,

and found that his model fit the observations much better

than the two allelic pair idea. If p, q and r are taken to stand

for the gene frequencies of genes A, B and R, respectively,

then the occurrences of the genotypes AA, AB, AR, BB, BR

and RR in the population are given by p2, 2pq, 2pr, q2, 2qr

and rZ, respectively, where p + q + r = 1 and p2 + 2pq + 2pr 4- q2 + 2qr +? = 1 (Bernstein, 1924 and

Bernstein elaborated upon his ideas somewhat further in

1930 (Bernstein, 1930a and 1930b). His work is consid-

erably more involved both in bulk and in computational

complexity than has been indicated here. Much of it consists

of detailed statistical and probabilistic analyses of the then-

existing population data to demonstrate the correctness of

Phenotype

0

A

B

AB

his genetic hypothesis in comparison with the two allelic pair

idea, and to answer various objections that had been raised

to his ideas.

The alternative theories of Furuhata (1927 and 1929) and

of Bauer (1928 and 1929) may be mentioned for the sake of

completeness. Furuhata's theory postulated two pairs of al-

leles, but said that they were closely linked. He thought,

further, that the genes exerted control over the presence or

absence of the agglutinins as well as of the agglutinogens.

The chromosomal composition could be ab, Ab or aB, but

never AB. The idea was similar to Bernstein's in that there

were three "hereditary units", and to von Dungern and

Hirszfeld's in that two pairs of genes were involved. He

would not allow that AB could come about in one and the

same chromosome, and thus in a single gamete, as the result

of a crossover. For all practical purposes, Furuhata's theory

turns out to yield results identical with that of Bernstein,

except that the explanatory mechanism is different, of

course. Bauer's idea was similar to Furuhata's, except that

he allowed for crossing over in certain cases, giving rise to an

Blood G m p - A B O and Secretor Systems

cussion though, in part because the resulting classification be repeated. It is best, with all these methods, to choose a

can depend on the method used, but particularly because, as B-serum which had a high anti-Al titer to begin with. The

will be discussed below, there are some cases which do not data of Olbrich and Walther (1941) indicated that the anti-A

clearly fall into one of the two categories. Von Dungern and titer of a B-serum tends to be higher as the anti-A, titer is

Hirszfeld (1911) used a saturation method for the prepara- higher also. So-called "irregular" antibodies can be used as

tion of anti-A, strum, in which B-serum was simply satur- anti-A, reagents. These are anti-Al which occur in the sera

ated with A2 cells, and the remaining antibody constituted of A2 and A2B people (Lauer, 1928; Friedenreich, 1931).

the desired reagent. Thomsen et al. (1930) used a saturation The anti-Al titer varies with the temperature, the antibodies

technique as well, but in this case a B-serum was absorbed being detectable in more of the sera at lower than room

with a particular amount of unknown blood cells, and the temperature. Additionally, anti-Al is more often found in

titer of the remaining antibody against A cells was then A2B than in A2 sera. Taylor et al. (1942) found a, in about

determined. The process is repeated several times, and the &(26.15%)ofA2B andin about &(1.6%)A2 bloods in the

data can be plotted to yield so-called "saturation curves". (^) U.K. Juel(1959) reported 12 A2Bbloods out of 40 (30%) as These can then be compared with the results obtained using having anti-A, a t So, and 9 (22.5%) at 18'. Lenkiewicz and

measured amounts of known Al and A2 cells. The finding Sarul (197 1) said that they found about 14% of A, bloods

that pepsin possesses A blood group characteristics has been and 51% of A2B bloods to have anti-A, at 10' in a large

exploited for the production of anti-Al serum (Ottensooer number of cases examined in the Warsaw area. Prokop and

and Zurukzoglu, 1932; Schiff, 1933a). Prokop and Uh- Uhlenbruck (1969) cite various values from several authors,

lenbruck (1969) noted that not all examples of pepsin are the numbers ranging from about 21 to^ 40%^ of A2B^ bloods

suitable for use with B sera in the production of anti-Al and from about 2 to 8% of A2 bloods. Other techniques have

serum, and that the specificityof the reagent does not persist been proposed for discriminating AI from A2, based on blood

beyond a few days, at which time the pepsin treatment must group substances in saliva (see in subsequent section) and on

Table 19.3 ABO System Genetics with A , and A , Subgroups

4 Reaction With

Genotype Phenotype Anti-A Anti-A , Anti-B

A l A ,

AIA2 A , + + -

A 10

*,A,

A2 +

A 2 0

66

B - -^ +

BO

A 1B A ,B + + +

A 2= A 2B + - +

7

Sourcebook in Forensic Serology, Immunologv, und Biochemistry

the use of various animal sera. Perhaps the simplest tech- nique is the use of Al specific phytagglutinins, one of the most common being from Dolichos bij7orus (Bird, 1951 and 1952). Phytagglutinins are discussed more fully in sub- sequent sections. It should be pointed out that Lattes in 1932 still did not accept the notion of qualitatively different subgroups of A as having been established. The data could be explained, he thought, on the basis of avidity differences in antibodies. The issue is still apparently not fully settled (Juel, 1959; Makela er al., 1969; Race and Sanger, 1975). 19.3.1.2 Subgroup A,. In 1936, Friedenreich described an example of A cells which reacted more weakly than A, cells (Friedenreich, 1936a and 1936b).This type of behavior was characteristic of certain families, was apparently rare, and no transitions could be observed between this weak A and A2. Six unrelated people, out of about 4000, were found to have the weak A cells, and were investigated along with members of their families. Since the results showed that the cells were of a distinct type of A, they were designated A3. A3cells reacted with anti-A reagents at least as weakly as do A2B cells (which react more weakly than Az cells), ifmot more so. Sera which gave strongly positive reactions even with A2Bcells gave only weak reactions with the A3 cells. It was noted that these A3 cells formed a few rather large, fragile agglutinates amongst a large number of free cells. This characteristic agglutination picture was later referred to by Dunsford (1959) as "mixed field agglutination." With weak antisera, the A, would be missed altogether, i.e.,

grouped as 0. With sera of too high titer, the A, might be

misclassified as an A2, but Gamelgaard (1942) noted that, because of the characteristic and unusual agglutination pat- tern, this should not be much of a problem. The sera of A, people contain a normal quantity of Bagglutinin, and no unusual a,-agglutinins, except for a trace of "cold w a g - glutinin" in a few cases. The absorption capacity of A, cells for anti-A was found to be intermediate between A2 and A2B.Three A3Bbloods were noted, their absorption capacity being less than that of A2B.Friedenreich suggested that the A, characteristic was almost surely the result of an addi- tional allele at the ABO locus, recessive to both Al and A,. Gammelgaard (1942), in his extensive and important study on the weak A receptor types, looked at 170 persons of type A, and 33 of type A3B.The agglutination picture with A,, as originally described by Friedenreich, was confirmed. The absorption characteristics, and the fact that the weak A, agglutination does not decrease as rapidly upon serum dilu- tion as does A2B agglutination with anti-A reagents, were also confirmed. In the material studied by Gammelgaard, the incidence of A3was about 1: 1000 A persons. The A3cells appeared to consist of a spectrum of red cells of varying receptor strength, from relatively agglutinable to non- agglutinable. Subsequent studies on Aj have not un- equivocally proven, but have not excluded the possibility that A, is an allele of the system (Race and Sanger, 1975). Race and Sanger also note that the diagnosis of A3B is not an easy one, and that family evidence is really needed for

confirmation. Dguchi et al. (1978) could isolate a population of A, cells by affinty chromatography on a Sepharose col-

umn, to which was bound Lima bean anti-A lectin, which

were completely inagglutinable by anti-A. Studies with eel serum anti-H showed that these cells had a large number of H sites, but only about 9% of the A sites that would be found on an AI cell. The A structure was believed to be the same on those A, cells as on A, cells, but the density of A sites too low to permit agglutination by lima bean anti-A lectin. 19.3.1.3 Further subgroups of A. The other weak A recep tors which have been described are principally weaker than A,. They have been given a variety of designations, and the relationships between the different ones are not always clear. They are all quite rare. In 1935, Fischer and Hahn described a blood from a patient which had a very weak A receptor.

The cells were agglutinated by some group 0 sera, but only

weakly or not at all by group B sera. Sheep immune anti-A agglutinated the cells weakly. The cells were designated A,. They absorbed less anti-A than either Al or A2cells, but the anti-A which was absorbed could be more easily eluted from the A, cells than from the A, or A2cells. The patient's serum contained anti-B, an anti-A, activity at 6-So, and an anti-A activity at "ice bath" temperature. In 1940, Gammelgaard and Marcusson described a blood with a very weak A recep tor, which was not A,, and which they called A,. They sup- posed that it was due to an additional allele in the ABO system. The investigated family of 64 persons had 24 Aq members. The incidence of A, was put at 1:60,000. The original designation was changed by Gammelgaard in 1942 to As, because some other cases had been found which were better referred to as A,. These were slightly weaker than AzB but did not give the characteristic A, agglutinates. A trace of A substance could be identified in the saliva of some of the A, persons, but not enough to try to distinguish be- tween secretors and non-secretors (see in subsequent sec-

tion). The AS cells were weaker than &, and this was the

reason for the redesignation. Gammelgaard speculated that Fischer and Hahn's A, was most likely an Aq. Gammelgaard described one odd blood, which he called A,, and which did not have the characteristics that would indicate it should be

called "&",that is, the cells did not agglutinate with "anti-

serum (see below) nor was there a decreased secretion of A substance in saliva, as is observed Al through AS. In 1948, Jonsson and Fast reported a blood in which the cells had a very weak A, but it differed somewhat from

previously reported variants, and was called &.Some of the

& sera had an a,-agglutinin, but no A was found in saliva.

Dunsford (1952) reported a blood which behaved very much Like the A, of Gammelgaard. Further examples of A, were reported in an English family, one of whose members also

had a very rare Rhesus type in addition to the & (Dunsford

and Aspinall, 1952). Estola and Elo (1952) found a weak A blood that was similar to Gammelgaard's AS, and this was called A,. Grove-Rasmussen et al. (1952) reported a case of a weak A receptor, which they called A,,. Ellis and Cawley found another case of an A,, in 1958. Salmon et al. (1965) reviewed A, and presented studies on 42 cases in 8 families.

Sowrebook in Forensic Setologv, Immunology, andBiochemistry

approaches will be discussed subsequently. Gibbs and Ake- royd (1959) used a technique for evaluation of the A recep- tor strength which had been developed by Wilkie and Becker (1955). The latter had devised the procedure for the study of B antigen-antibody reactions, based on the studies of the

Fiitti-Wurmser et al. group (see below). If one quantitates

agglutination by measuring the number of cells remaining unagglutinated under a standard set of conditions with different concentrations of antisera, and plots the percentage agglutination against the log of the antiserum concentration, a sigmoidal curve is obtained. With suitable computational transformations, the sigrnoidal curve can be fitted to a line, the slope of which is related to antigen strength. Gibbs and Akeroyd (1959) found that the slope decreased (antigen strength decreased) as A, >AIB>A2>AiB>A2B>A3. Grundbacher (1964 and 1965) devised a hemolytic assay for antigen strength, using an immune anti-A hemolysin serum

which had been absorbed with B and 0 cells. Under care-

fully controlled test conditions, hemolysis, which can be readily quantitated colorimetrically, becomes a measure of antigen strength. The results of the application of this tech- nique to the study of Al, A2 and Ai were mentioned in 19.3.1.4 (Grundbacher and Summerlin, 1971). Cohen and Zuelzer (1 965) used an immunofluorescence technique, and found that there was a continuum of antigen strength, run- ning from Al to the weakest forms. The data, they said, supported a quantitative basis for the differences in receptor strength, and they suggested that most of the weak forms be grouped under a single designation (they suggested A,). Wiener and Karowe (1944) suggested that the difference between At and Az might have a structural basis. If the antigenic receptor was the same in both cases, but if in the case of A2 it was attached to the cell surface by a shorter "stalk", as it were, i.e., not as accessible to the antibody, this might explain the weaker reactivity. The strength of the

receptor is thus explained by steric hindrance to antibody

binding. Carrying the argument a step further, A2B has a weaker reaction than A2 because the neighboring B recep- tors are attached by the longer (Al-like) stalks, and can thus interfere sterically with antibody molecule approach to the A2 receptor. This view was expressed again quite recently

(Wiener and Socha, 1974). The studies of Bar-Shany er al.

(1970) were interpreted to mean that the Al and A2 reac- tivity differences might very well have a steric basis. As expected, steric hindrance would give rise to greater differences in reactivity on red cell surfaces than in the solu-

ble, secreted blood group substances. Economidou et al.

(1967) used t2SI-labelledantibodies to determine the number

of A and B receptors in Al, A2, AIB and A2B cells. The

number of A sites decreased as AI>AIB>A2>A2B. B cells had more B sites than AIB cells. The differences in equi- librium constant and dissociation rate with A, and A2 cells using anti-A were consistent with slight differences in the molecular structure of the antigen.

Cartron et al. (1974) determined the A antigen site den-

sity on a series of samples representing a number of the "weak A" phenotypes, including A3, &, L, A,,, and &,.

They used a rabbit IgG anti-A which was labelled with IzI. Site density was found to decrease in the order A1>Az> A3>Ax>Aend>Am>&l.While A3 cells had about 30,

sites per cell, 41 cells had only about 700, compared with

850,000 for Al and 240,000 for A2.

19.3.2 Variants of B Variants of B are rarer than those of A, and there is

considerably less literature on them. Many of the weak B

types have been observed in only one or a few kindreds. As with weak A types, the relationships between the different weak-B types, given a variety of designations by different authors, are not always clear. One of the earliest obser- vations of weak B forms was that of Moskow (1935). He distinguished what he called BI and B2. Following his work, the area was not very active for a number of years, but a number of papers began appearing in the mid-1950's. The variants have been well reviewed in Prokop and Uhlenbruck (1969) and in Race and Sanger (1975), and references to the individual observations may be found in these classic works.

Race and Sanger (1975) classify the B variants into three

categories, based on whether the individual secretes B sub- stance or H substance in saliva or not, and on whether there is an anti-B in the serum. Table 19.4 gives a summary of the weak B variants.

A

19.4 Antlbodles of the AS0 System

19.4.1. Anti-A and anti-B Ordinarily, human serum contains anti-A, anti-B, or both or neither, corresponding to the antigen(s) that is (are) ab- sent. Absence of an expected agglutinin nearly always means that there is something unusual about the person's blood. The isoantibodies develop in most cases around the age of 3-6 months, and the titer increases for a number of of years, after which it declines steadily (Morville, 1929; Thomsen and Kettel, 1929). Infant serum contains some 10% of the IgM levels found in adults, and occasionally anti-A or anti-B is present in this fraction. It was thought for a long time that the isoagglutinins of newborns were primarily derived from the mother, but this need not always be the case, because isoagglutinins against maternal cells can be detected in some cases (Toivanen and Hirvonen, 1969a). Infant serum con- tains IgG antibodies derived from the maternal circulation. The initiation of IgM synthesis in the newborn precedes that of IgG by several weeks, IgM biosynthesis beginning within

a few days of birth (West et al., 1962).

The origin of anti-A and anti-B have occupied much at- tention, because their "natural" occurrence in human sera appears, on the face of it, to be in violation of the funda- mental immunological principle: No antibodies without an antigenic stimulus. A major discussion of this subject was given by Thomsen in 1936. Furuhata (1927) thought that the isoantibodies were under direct genetic control, and that the genes controlling them were allelic to the genes re- sponsibie for the isoagglutinogens (see in section 19.2). This

Blood Group-ABO and Serretor Sysrems

Table 19.4 Summary of Some Weak B Types

Anti-B in B in H in Race and Sanger Notation Serum Saliva Saliva Category

Bv + ( + I 1

• (^) 1

Bw; Bx; Bm - 2 + - 2

B3;Bx; Bweak - -3 + 3

l Some kind of B in saliva Or weak cold anti-B Or doubtful B in saliva

idea is no longer widely accepted, but the concept of genetic

control of isoantibody biosynthesis has persisted in the Ja-

panese literature for a long time, and the issue is not com-

pletely settled. Grundbacher (1976) recently determined the

anti-A and anti-B agglutinin (and hemolysin) titers in 401

group 0 people from 74 families. The titers were generally

higher in Blacks than in Whites. Whites also generally

showed higher a than @ titers, and antibody levels were

generallyhigher in women than in men. In Blacks, the agglu-

tinin titers tended to be equal, and sex differences were not

apparent. Statistical analysis of data from related persons

suggested that 20 to 30 percent of the variation could be

accounted for by genetic variation. Schii and Adelsberger

(1924) believed that group A people had a small amount of

B present, just enough to stimulate anti-B production, but

not so much as to absorb the antibody formed. The stimu-

lating agents were termed "isoreagins". Dupont (1934) first

suggested that isoantibodies do arise in response to imrnu-

nization, not with red cell antigens, but with closely related

antigens from various things in the environment. The A-like

and B-like antigenic stimuli could come from a number of

different sources, e.g., bacteria inhabiting the gut, or on dust

particles which are inhaled, or on various materials ingested

as food. This notion has been called the "cryptogenic immu-

nization" theory. There is some experimental support for the

idea from the work of Springer et al. (1959). White Leghorn

chickens form an anti-B-like agglutinin for human red cells

when they are quite young, but this agglutinin does not form

if the chicks are kept in a germ-free environment. The idea

that some external stimulus is required for the production of

these antibodies is more satisfying in terms of present-day

immunology. Kabat (1956) was inclined to accept the view

that a stimulus is needed. Burnet's clonal selection theory

(Burnet, 1959) can explain isoantibody formation. In these

terms, the antibodies are understood as an immunological

tolerance phenomenon. The plasma cells forming molecules

that can act as anti-A or anti-B (or anti-H) are influenced

very early in embryonic life by the presence of an antigen. If

an antigen is present, it has the effect of destroying those

cells producing antibodies to it, i.e., the antigen is recognized

as "self'. Thus, in an A person, the anti-A machinery is shut

down very early on, and only anti-B develops. There have

been other theories proposed too, which are reviewed by

Prokop and Uhlenbruck (1969). The work of Filitti-

Wurmser et al. (see below) has strongly suggested that iso-

antibody formation does have some sort of genetic basis, a

position with which a number of authorities agree.

A distinction is often made between "natural" and "im-

mune" antibodies, the former being present in the serum of

a person who has had no known exposure to the correspond-

ing antigen. The best examples of socalled "naturally occur-

ring" antibodies are the iso-anti-A and anti-B, although

1

conducted a number of experiments using "Mrs. G" serum and a few other sera which had become available. In addi- tion, a cattle serum absorbed with A,B cells was studied, along with a rabbit immune serum prepared by injecting rabbits with a purified human H substance. This last had been obtained by Morgan and van Heyningen (1944) from

ovarian cyst fluid from group 0 women and purified by

Morgan and Waddell (1945). This purified material

inhibited the reaction of 0 cells with a rabbit antiserum

prepared against it, and it also inhibited the reaction of

absorbed cattle serum with 0 cells. The experimentsof Mor-

gan and Watkins (1948) indicated that certain of the human

anti-0 sera were, in fact, reacting with the 0 gene product.

The serum could be inhibited in its reaction with 0 cells by soluble substances from human ovarian cyst fluids and erythrocytes. These soluble substances were suggested to be

0 gene products, and were said to be secreted only rarely.

Those sera which detect H, and the cattle and immune rab-

bit and goat sera, are inhibited by the so-called "0 sub-

stance" secreted in saliva and body fluids of all secretors regardles of blood group. The specific human "anti-0 sera, however, are only inhibited by the rare soluble substance

which appeared to be the 0 gene product. The previously

designated "anti-0 sera should, therefore, be termed "anti-H", and the designation "anti-0 reserved for re-

agents like "Mrs. G serum. It was suggested that Hirsz-

feld's theory could be modified to accommodate this- information by considering H as the primary gene which gives rise to basic H substance. H substance is then an evolutionary precursor of 0 , A and B substances, and H

gene is in the process of mutation to pure, or complete, 0, A

or B. Thus, most red cells come from transitional forms, and have variable amounts of H substance on red cells. In 1949, Grubb obtained antisera from chickens by immu-

nization with "0 substance" from group 0 secretors, or with

Shiga bacilli, which reacted preferentially with group 0 red

cells. These had all the characteristics of anti-H reagents, and Grubb said that these should be considered as anti-H (Grubb, 1949 and 1950). Jonsson (1944) had discovered an eel serum with a high titer for 0 cells, and this was one of the best anti-H reagents available, according to Grubb's results.

Eel anti-H was very useful in discriminating group 0 secre-

tors from nonsecretors.

There are a number of plant seed extracts which have

anti-H activity, such as those of certain Ulex, Cystisus,

Lotus and Laburnum species (Renkonen, 1948; Cazal and Lalurie, 1952). There are usually called "lectins". The dis- cussion of group substances will be concluded in sec- tion 19.9, dealing with the nature and biosynthesis of these materials. Our present understanding of the biochemical genetics of the ABO system (section 19.9) makes it quite

clear that the 0 gene is, in fact, silent, in the sense that it

does not make a product that is responsible for the synthesis of any ABO (H) substance. However, the gene may make a product, which is immunologically related to the products of the A and B genes (see in section 19.9.3).

Blood Groups-ABO and Secretor Systems

The material in section 19.4.2 is excellently reviewed and enlarged upon by Watkins and Morgan (1955).

19.4.3 bagglutinins in body fluids other than serum

In 1919, Bond carried out experiments in which he looked for isoagglutinins in a number of body fluids, but, based on some observations he had made on blood, he was operating under the assumption that drying samples down, pul- verizing them and then reconstituting them, or else simply exposing them to mechanical friction had some enhance- ment effect on the isoagglutinin content. The results with body fluids were not very clear-cut. Kirihara (1924) reported finding isoagglutinins is pleural and pericardial fluids, but not in cerebrospinal fluid. In 1928, Yosida conducted an extensive investigation of isoagglutinins in body fluids and secretions, and found them to be present in many cases. In most examples of tears, saliva, seminal plasma, and in pleural and pericardial fluids, isoagglutinins were found. They could be found in urine too, if it were first concentrated

to '/5 to % or its original volume (Yosida, 1928a and 1928b).

Schwartmann ( 1928) found isoagglutinins in vaginal secre- tions, and in ovarian and vaginal cyst fluids. In milk and colostrum were found isoagglutinins corresponding to those in the serum. Happ (1920) said that isoagglutinins occurred in human milk and corresponded to those in the serum. Heim (1926) agreed with this finding. Hara and Wakao (1926) reported that milk and colostrum can contain a non- specific general hemagglutinin in some cases, and they found isoagglutinins corresponding to the serum in about half the samples examined. Hirszfeld (1928) found that isoagglu-

tinins were present in the milk of group 0 mothers, but

lacking in about half group A or group B mothers.

Yosida's observations on the presence of isoantibodies in the saliva (1928a and 1928b) have been confirmed, but with some modifications. Some of the Japanese investigators have tended toward the view that the isoagglutinins were inher- ited, or that the ability to secrete them in saliva was inherited. Prokop ( 1961) indicated, however, that "non- secreting" parents could have "secreting" children, and that the presence of isoantibodies in saliva is much more frequent

in group 0 people than in A or B people. In addition, the

antibodies from 0 subjects can be of the cross-reacting type

in some cases (Prokop, 1963). Putkonen (1930) had ob-

served that isoagglutinins were most frequent in group 0

salivas as well. These observations which have been confirmed by others do not support a genetic basis for the phenomenon. Boettcher (1967~)found agglutinins in a

significantly higher percentage of 0 salivas than in A or B

ones, and a higher incidence of anti-B in A2 than in A, saliva. Jakobowicz et al. ( 1966) got similar results. Schlesinger and Osihka (1964) found the very same thing, and in addition, they reported that a lower percentage of pregnant women showed aggutinins in saliva. Bell and Fortwengler (1971) found anti-A and anti-B in the whole salivas of most group

0 males. The titer of the agglutinins increased upon immu-

nization with A and B substances. They thought that the agglutinin activity was due primarily to secretory IgA (see

.%umbook in FOE& SefVfogy, Immunology, and Biwhemirtty

below). There was not a good overall correlation of saliva with serum titer in the same person, and it was said that the secretory system is probably independent of serum titer. As Prokop (1961 ) showed, there are sometimes correlations be- tween serum and salivary agglutinin titers, so that they do not seem to be completely independent, yet people with high serum titers can lack salivary agglutinins altogether. Mat- suzawa et al. (1972) noted agglutinins in about H of the 0 , A and B subjects they examined, but said this percentage is

much higher if the agglutination test is carried out in collo-

idal media, such as PVP. The first report of isoagglutinins in tears appears to be that of Hegner (19 16). He found the antibodies in 3 of 20 people suffering from relapses of typhoid fever, but there

were no agglutinins in 20 healthy people who had been

immunized against typhoid, although their serum showed normal isoagglutinin activity. Putkonen (1930) found iso-

agglutinins in the tears of 5 out of 21 0, A and B subjects,

in whom he induced the lachrymal response with onion slices or with bromacetone. Prokop et al. (1963) looked for iso- agglutinins in tears by placing strong test cell suspensions into the subjects' eyes, and examining the lower lid area for microscopic agglutination. Isoagglutinins did not always occur, but those which did correswnded to the serum con-

tent except that, as in saliva, 0 persons may sometimes have

either anti-A or .anti-B and not both. Isoagglutinins have been found in cervical mucus as well. Gershowitz et al. (1958) reported that 17 of 77 women showed isoagglutinin activity in their cervical mucus, and 15

of these were group 0. This work was expanded upon (Solish

et al., 1961) in studies on 128 subjects. When multiple samples were collected from the same subject, the per- centage showing agglutinins in at le&t one of the specimens was 63.4. The occurrence of antibodies was much greater in

persons of group 0. The presence of the agglutinins was not

dependent on the phase of the menstrual cycle, nor on the secretor status of the subject with respect to blood group substances. Parish er al. (1967) confirmed the presence of isoantibodies in cervical mucus, and found immune type anti-A hemolysins in some samples as well.

IgA is the most abundant immunoglobulin in saliva and

colostrum, in which it is dimerized and connected to a smaller "secretory" component. The anti-A and anti-B of these fluids is most probably made up of secretory IgA, which has been detected in urine and nasal mucus as well. Secretory IgA differs from serum IgA in containing the "secretory piece" (Mollison, 1972; Zmijewski and Fletcher, 1972).

19.5 Quantltatlva and Physloochamlcal Approachas Much of the work on the physical chemistry of the iso- agglutination reaction was first carried out by the Filitti- Wurmser group in France, and reported in a lengthy series of papers beginning around 1947. It could be established

that the isohemagglutination reaction is a reversible, equi- librium reaction, amenable to the usual kinds of mass action and kinetic treatment. Filitti-Wurmser and Jacquot- Armand (1947) carried out a series of experiments on the agglutination of B cells by anti-B (A serum), in which a known number of cells were incubated with a constant amount of antiserum, and percentage agglutination scored by determining the number of free cells remaining after the reaction had achieved equilibrium. The same yield of agglu- tination was obtained when the reaction was carried out at 37", or in two stages, the first being at 4", and the second at 37". Similarly, the same yield was obtained when dilution was used as a means of dissociating the agglutinates. In studies on the temperature dependence of agglutination, it was observed that the maximum number of cells aggluti- nated at a particular temperature varied directly with the total number of cells present, and that the number of agglu- tinated cells present decreased as a function of temperature between 4" and 37". The maximum number of cells aggluti-

nated at 4" was designated N4, and the same value obtained

at 37", all other conditions being identical, was designated

N3,. It could be shown that sera differ in their ability to bring

about agglutination as measured by the N4/N3, ratio. Anti-

B sera from AIO, Al, A2 and 0 individuals differed in this

property (Filitti-Wurmser et a1 ., 1950 and 1954). In a some-

what more complicated measurement, designed to give a measure of the dissociation constant for anti-B with B cells, differences between the four different sources of anti-B were also apparent (Filitti-Wurmser et al., 1952). The heats of reaction for the various combinations were determined as well, and differences in this parameter were found to be significant. The reaction is exothermic (Filitti-Wurmser et al., 1952 and 1953a). Measurements conducted with the ultracentrifuge indicated that B(00), @(AIAI)and @(AIO)

differed in s 2 ~ (see in section 1.1.3.4) as well, corresponding

to molecular weights of about 177,000, 300,000 and 500,000, respectively (Filitti-Wurmser et al., 1953a). The MW of the B(AIAI)agglutinin was later revised to about 200,000 (Wurmser and Filitti-Wurmser, 1957). Other

physicochemical measurements were canied out as well, all

of which indicated differences between the anti-B anti-

bodies. It could be shown further that then P(A,O) iso-

agglutinin did not behave like an equal mixture of B(AIAI) and B(00) (Filitti-Wurmser et al., 1953b). These studies were reviewed by Wurmser and Filitti-Wurmser in 1957 in a paper written in English. In sum, the experiments indi- cated that the same isoagglutinin, with the same blood group specificity, differed considerably in its properties depending upon the genotype of the person from whom it was derived. Furthermore, the isoantibodies from persons of a given geno- type showed considerable homogeneity in the studies, while

it is known that antibodies obtained by immunizing animals

show a heterogeneous antibody composition in the resulting antiserum. Kabat (1956) levelled a number of criticisms at the studies of the Filitti-Wurmser et al. group, based on some consultation he had had with physical chemists. Rea- sons were given why some of the data might be deceptive