CHROMOSOMAL ASSIGNMENT OF THE HL-A COMMON ANTIGENIC DETERMINANTS IN MAN-MOUSE SOMATIC CELL HYBRIDS

SUMMARY In the study presented here, man-mouse somatic cell hybrid clones were examined by means of radioimmunoassays for the presence of both microglo globulin (Pzm) and the HL-A xenoantigenic determinant. In addition, the clones were examined for their karyotype and the expression of enzymes with known chromosomal assignments. The results obtained indicate that the gene coding for the HLA xenoantigenic determinant is carred on chromosome 6. The data obtained provides a direct demonstration that the gene coding for &m segregates independently of that coding for the alloantigenic polypeptide chain of the HL-A molecule, and that the gene coding for P2m is carried on chromosome 15. radioimmunoassay the presence of the /12m determinant for the presence of the HL-A xenoantigenic determinant. The cell clones were also analysed for both chromosomal composition and a variety of enzyme markers, and the results were correlated with those ofthe antigenic analyses. The data obtained indicate that the gene coding for the HL-A xenoantigenic determinant is syntenic with the genes for soluble malic enzyme (MEl) and tetra-meric iodophenol Confirmation coding


INTRODUCTION
Recent studies have demonstrated that the HL-A molecule obtained after papain digestion of cell membranes consists of two polypeptide chains (Cresswell et al., 1973 ;Tanigaki, et d., 1973a, Tanigaki, et al., 1974. The 11,000-dalton fragment constitutes one polypeptide chain while the remaining portion of the HL-A molecule, with a molecular size of approximately 33,000 daltons, constitutes the second polypeptide chain. This 33,000-dalton polypeptide chain carries the HL-A alloantigenic determinant (Nakamuro et al., 1975;Tanigaki er ~l . , 1975). The 1 1,000-dalton portion of HL-A molecule has been demonstrated in free form in both serum and urine and has been identified as beta-2-microglobulin (p2m) (Grey et d., Nakamuro et al., 1973;Peterson et al., 1974). The film polypeptide chain is structurally identical in all HL-A molecules, regardless of the alloantigenic specificity. HL-A molecules carrying different HL-A alloantigenic specificities share certain common antigenic structures that stimulate antibody production in heterologous animals. These common antigenic structures which have been designated HL-A common antigenic determinants, have been detected by use of rabbit antisera directed against HL-A molecules on both the 11,000-dalton and 33,000-dalton polypeptide chains, and they have been shown to be distinct from the HL-A alloantigenic determinants (Miyakawa er al., 1973). In this study, the common antigenic determinant on the 11,000-dalton portion will be termed the P2m determinant, while that on the 33,000-dalton portion will be termed the HL-A xenoantigenic determinant to distinguish it from the HL-A alloantigenic determinant. The xenoantigenic determinant, or a very similar one, has been shown to be present on the Rhesus monkey histocompatibility antigens, RhL-A, and has previously been referred to as the HL-A primate cross-reacting determinant (Katagiri er al., 1974). Since this determinant has been located on the 33,000-dalton polypeptide chain carrying high HL-A alloantigenic activity (Nakamuro et al., 1975), the HL-A common antigenic determinant apparently represents a part of the invariant portion of the product of the genes which code for HL-A alloantigenic chains. Further insight into the relationship between HL-A alloantigenic determinants, the HL-A xenoantigenic determinant and the /12m determinant, may be obtained through genetic studies on the segregation of genetic markers in man-mouse somatic cell hybrids. Through such studies it is possible to establish the number of genes responsible for the production of the structural components of HL-A molecule, and more importantly, to clarify the interrelations of these genes in the antigenic expression of the HL-A molecule.
In the present study, man-mouse somatic cell hybrid clones were examined by means of radioimmunoassay for the presence of the /12m determinant and for the presence of the HL-A xenoantigenic determinant. The cell clones were also analysed for both chromosomal composition and a variety of enzyme markers, and the results were correlated with those ofthe antigenic analyses. The data obtained indicate that the gene coding for the HL-A xenoantigenic determinant is syntenic with the genes for soluble malic enzyme (MEl) and tetrameric iodophenol oxidase (IPO B) on chromosome 6, the chromosome that also carries the genes coding for the alloantigenic determinants of the HL-A molecules (van Someren et al., 1974;Jongsma el al., 1973). Confirmation was obtained that the gene coding for the P2m determinant is carried by chromosome 15 (Goodfellow et al., 1975;Smith et af., 1975). Moreover, it was shown that the production of HL-A, as determined by the presence of HL-A xenoantigenic determinant, occurs in hybrids in which no detectable human P2m is produced.

M A T E R I A L S A N D M E T H O D S Cell types used
In this study the two human cell types used were peripheral blood lymphocytes isolated from two healthy donors and the established lymphoid culture line cells ODY and EB4. Each of thesecell types were fused with RAG cells, a hypoxanthine-phosphoribosyl-transferase deficient mutant mouse cell line (Ruddle el al., 1970).

Man-mouse somatic cell hybrids 107 Enzynie analyses
Lysates for enzyme studies were prepared by washing cells in Dulbecco's phosphate buffered saline (pH 7-2) and then subjecting the cells to one or two cycles of freezing and thawing. Lysates derived from different hybrid clones were examined for the presence of thirteen different murine, human and hybrid enzyme phenotypes by means of starch gel electrophoresis. The enzymes studied are listed in Tables 2 and 3. The enzyme detection systems and the chromosomal assignment for these enzymes are referenced in Table 3.

1975).
The methods used for chromosomal analyses have been previously described (Smith et al.,

Radioimmunoassaj for HL-A and P2n1
The radioimmunoassays for the HL-A xenoantigenic determinant (Katagiri et al., 1974;Nakamuro et af., 1975), the HL-A2 alloantigenic determinant (Tanigaki et af., 1973) and film (Smith et af., 1975) were performed as previously described. The P2m activity and the HL-A activities were assayed in cell lysates. In solubilizing cells for use in the P2m and HL-A assays, 3 -5 x lo7 cells/ml were treated with a 1 % solution of the non-ionic detergent, IGEPAL CA-630 (GAF Corporation, New York). The lysates were centrifuged for 10 min at 1000 RCF and the supernatants were assayed. Activity was defined as the capacity of a specific quality (i.e. 50 pl) of cell lysate to inhibit the binding of a defined amount of antibody to a defined amount of radioiodinated reference antigen.
In the HL-A assay systems, samples were considered positive if the inhibition of binding of labelled antigen to antibody caused by the prior addition of the test sample was over 10%.
Control lysates from human cell clones showed 83-95 7; inhibition. In the P2m assay system, samples were considered positive if the inhibition of binding of labelled antigen to antibody caused by the prior addition of the test sample was over 5 %. Control samples from human cell clones showed 95-98 % inhibition.

R E S U L T S
From the results of the radioimmunoassays for HL-A xenoantigenic activity and for P2m activity, shown in Table 1, it will be seen that clones could be classified into four categories as follows : HL-A-positive and &m-positive; HL-A-positive and j,m-negative; HL-Anegative and P,m-positive; and HL-A-negative and B,m-negative. In Tables 2 and 3, the segregation of human B2m, HL-A and thirteen human enzyme markers are shown. A high degree of discordance was observed between the expression of human P2m and HL-A. A similarly high degree of discordance was observed in the expression of HL-A and mannose phosphate isomerase (MPI) which maps on chromosome 15 (van Heynigen et al., 1975).
A high degree of concordance was observed between the expression of HL-A and the expression of enzymes ME1 and IPO B which have been assigned to human chromosome 6 (van Someren et af., 1974;Jongsma et af., 1973). In the case of segregation of HL-A and ME1 two discordant clones were observed. Both of these clones were, however, positive for IPO B. From Table 4 it will be seen that in the case of segregation of IPO B and HL-A, the t Activity is expressed as yo inhibition.
$ The first three are non-lymphoid cell clones and the others are lymphoid cell clones.
5 This is the parental mouse cell clone that was used for cell fusion.
concordant clones were almost equally distributed in the HL-A-positive and IPO B-positive clones, and in the HL-A-negative and IPO B-negative clones. Thus, the observed concordant segregation cannot be attributed to insufficient chromosomal loss from hybrid clones. In the case of segregation of ME1 and HL-A, a similar distribution of concordant clones was observed. A high degree of discordance was seen in the segregation of HL-A and the remaining enzyme markers examined. These markers have a known distribution of nine chromosomes ( Table 3). Table 4 also shows a high degree of concordance in the expression of P2m and the expression of the enzyme MPI.
Chromosomal analyses were carried out on twenty-four of the thirty-five clones assayed for HL-A xenoantigenic activity. The results of the chromosomal analyses are shown in Table 5. It can be seen that HL-A xenoantigenic activity was expressed in all of the clones in which chromosome 6 was present. Similarly HL-A xenoantigenic activity was absent from clones in which chromosome 6 was absent. Furthermore, no other chromosome was found to be consistently present in HL-A-positive clones and consistently absent from HL-Anegative clones.   Several clones of RLyMS, derived from the fusion of RAG cells with the normal peripheral lymphocytes from an individual (MS) whose HL-A phenotype is known to be 2 , 5 , 7 and 12, were tested for HL-A2 alloantigenic activity by the radioimmunoassay. Two clones were positive for HL-A2 alloantigenic activity. The two clones, RLyMSAe and RLyMSWb, were also positive for HL-A xenoantigenicactivity and had only chromosomes 6 and 10 in common.  , Shows, 1970 Nabholz et al., I969 Ruddle, 1974Ruddle, 1974 ? ~~~ ~~ ~ * HL-A as determined by the HL-A xenoantigenic activity. The results of this study indicate that the HL-A xenoantigenic determinant can be assayed in man-mouse hybrid cells by means of radioimmunoassay and that the expression of this determinant in such hybrids is apparently dependent only upon the presence of human chromosome 6. The results also provide direct evidence that the human Pzm gene segregates independently of the gene coding for both the HL-A alloantigenic and the HL-A xenoantigenic portions of the HL-A molecule. These results are, therefore, in agreement with previous reports on the chromosomal linkage relationships of HL-A antigens (Goodfellow el Smith et al., 1975). Studies on the inheritance of HL-A alloantigens in families have revealed that the HLalloantigens are determined by three closely linked structural gene loci (Thorsby, 1974), and that the genes determining HL-A alloantigens are linked to the gene for the enzyme phosphoglucomutase (PGM 3) (Lamm et al., 1971). Studies on man-Chinese hamster somatic cell hybrids have demonstrated that the PGM 3 gene and the genes determining ME1 and IPO B are located on chromosome 6 (van Someren et al., 1974;Jongsma et af., 1974). van Someren et al., have also demonstrated a good correlation between the occurrence of HL-A alloantigens and the enzymes MEI, IPO B, and PGM 3 in man-Chinese hamster somatic cell hybrids. However, in their study, a number of discordant clones were observed. In six out of thirty-five clones examined, IPO B and HL-A alloantigenic activity segregated discordantly, while in five out of thirty-one clones, ME1 and HL-A allogenic activity segregated discordantly. In attempting to account for the observed discordance, they drew attention to the diffi- The use of rabbit antisera directed against HL-A molecules greatly facilitated the demonstration of HL-A expression in hybrid cells having different HL-A phenotypes. Since the rabbit antisera contain antibodies directed to the antigenic determinant common to the alloantigenic polypeptide chain of HL-A molecule, it was possible to detect the presence of HL-A alloantigenic chains in clones derived from hybrid cells produced by the fusion of RAG cells with lymphocytes of different individuals, regardless of their HL-A phenotypes. However, even with this radioimmunoassay method for HL-A, the HL-A activity detected was extremely low in the hybrid clones tested, and the amount detected on hybrid cells was usually 1/100, or less, than that produced by human lymphoid cell lines (unpublished data). In order to overcome this difficulty. it was necessary to use cell lysates in high concentration.
The present data indicates that the HL-A alloantigenic chain, as detected by the presence of the HL-A xenoantigenic determinant, is expressed in cell hybrids that do not carry human chromosome 15 and, thus, do not produce any human P2m. Similarly, hybrid clones were obtained that expressed B2m in the absence of demonstrable HL-A xenoantigenic activity. However, in both types of clones, we have obtained evidence that hybrid molecules may be formed between human /12m and mouse H-2 alloantigenic chains, and between mouse P2m and human HL-A alloantigenic chains (unpublished data).
( 1 975) reported that a Burkitt lymphoma cell line, Daudi, does not produce P2m and yet produces a polypeptide chain that is devoid of HL-A allospecificity but is very similar in the xenoantigenic and physicochemical characteristics to the HL-A alloantigenic polypeptide chains isolated from other human cell lines and suggested that P2m and the HL-A alloantigenic chains are under separate genetic regulation. They also mentioned that Daudi cells have a deletion of chromosome 15.