Studies of the linkage relationship of beta‐2‐microglobulin in man‐mouse somatic cell hybrids

Beta-2-microglobulin (beta2mu) production has been studied in 33 primary man-mouse hybrid clones and in 26 secondary man-mouse hybrid clones. These clones have also been examined for the presence of 15 human enzyme phenotypes. Karyotypic analyses have been carried out on clones. From these studies the following conclusions can be drawn: (1) Gene(s) determining human beta2mu production in humans are apparently syntenic with the MPI gene on chromosome 15. (2) Long-term fibroblast lines may be of limited use in mapping studies as chromosomal rearrangements frequently occur in these lines. (3) The gene(s) determining beta2mu production in humans segregate independently of chromosome 6. If the assignment of the genes determining HL-A alloantigens to chromosome 6 is correct, our results imply that beta2mu and HL-A alloantigens are determined by genes carried on different chromosomes, despite the fact that beta2mu forms an integral part of the HL-A molecule.

ratio of RAG cells to Chang cells was 100 t o 1. I n the RAG lymphocyte fusions, equal numbers of the two cell types were used. Equal numbers of cells were used in lymphoid fusions.
After fusion, cells were grown in the supplemented McCoy's 5a medium described above, but containing 10-4 M hypoxanthine, 10-6 M aminopterin, 4 x M glycine; this is the HAT selective medium described by Littlefield (1966). In the HAT selective medium, RAG cells and RAG-RAG fused cells were progressively eliminated. Normal peripheral lymphocytes and the lymphoid cell lines were non-adherent cells and were completely discarded during sequential changes of the selective culture medium. Thus, in the RAG cell-lymphocyte fusions and in RAG cell-lymphoid line fusions only hybrid cells grew in the selective HAT medium and adhered to the vessel surface. However, in the RAG-Chang fusions, hybrid cells had to be separated from parental Chang cells by means of cloning. All hybrid cells were cloned using the method described by Ham & Puck (1962).
Radioimmunoassay for 82p. The concentrations of human p2p in spent culture media and in cell lysates were determined by means of a radioimmunoassay employing 1261-/32p and a double antibody technique for the separation of antibody-bound from free l26I-B2p (Hunter, 1973).
Purified human B2p and rabbit anti-human-82p antiserum were kindly provided by Dr M. D.

Poulik.
Cell lysates for ,82p assay were prepared by washing cultured cells in four changes of Dulbecco's phosphate-buffered saline (pH 7.2) and then subjecting them to three cycles of freezing and thawing. The lysates were centrifuged at 200 g and the supernatants were assayed for p2p.
Studies of cell-surface-associated 82p. The presence of 82p on the surfaces of viable cells was sought by means of an indirect immunofluorescence technique (Thomson & Alexander, 1973). Monolayers of cell clones were trypsinized, and the resulting single cell suspensions were incubated a t 37" C. in unsupplemented McCoy's 5a medium for 2-4 hr. The cells were then incubated with a 1: 10 dilution of rabbit anti-human p2p for 30 min., washed, and finally exposed to fluorescein-conjugated goat anti-rabbit IgG. I n control experiments, cells were incubated with normal rabbit serum, washed and then exposed to fluorescein-conjugated goat anti-rabbit IgG. After washing, the cells were examined for surface fluorescence employing a Reichert fluorescence microscope with an incident light source as described by Thomson & Alexander (1973).
Enzyme analyses. The presence of 15 murine, human, and hybrid enzyme phenotypes was sought in cell lysates from individual clones by means of starch-gel electrophoresis. The enzyme markers in question are distributed on 12 of the human chromosomes. I n preparing lysates for enzyme studies, cells were washed in Dulbecco's phosphate-buffered saline buffer (pH 7.2) and were then subjected to 1 or 2 cycles of freezing and thawing. The enzymes studied are listed, and the methods employed in their detection are referenced in Tables 1-3. For the enzymes selected for study, the human enzyme phenotypes differed clearly in electrophoretic mobility from those of the mouse Karyotype analyses For karyotype analyses, cell cultures were treated with colchicine, a t a h a 1 concentration of 2.5 pg./ml. for 90 min. The cell monolayers were then trypsinized and subjected to hypotonic treatment for 15 min. in 0.075 M-KC~ a t 37" C. The resulting preparations were fixed in a solution consisting of 3 parts of methyl alcohol and 1 part of acetic acid for a t least 2 hr. Air-dried drop preparations were made, and these chromosomes were then banded using a modification of the trypsin-Giemsa banding technique of Seabright (1971). Slides were placed in a 60' C. oven overnight. They were then cooled and immersed in a solution of 0.1 yo M thymidine and Table 1. were then rinsed twice in HBSS at 4' C. Only the first rinse contained 1 yo foetal calf serum. Such preparations were finally stained in 4 % Harleco Giemsa stain in GUIT'S buffer, a t pH 6.8 for 4 mjn., and then rinsed in the same buffer.

RESULTS
Classijication of the hybrid cells by radioimmunoassay for P2p in spent culture media. The cells of hybrid clones were classified as P2p-positive or 82p-negative on the basis of whether or not P2p was detectable, by radioimmunoassay, in their spent culture media. Depending upon the age of the 1261-P2p, the radiolabelled product used in the assay, the lower limit of sensitivity of the radioimmunoassay for P2p ranged from 1 ng./ml. to 6 ng./ml. Fresh McCoy's 5a medium supplemented with 10 % foetal calf serum, identical medium which had been in contact with 106 RAG cells for 48 hr., and medium from RAG cells that had been exposed to the full fusion procedure in the absence of human cells, always gave assay values below the level of sensitivity of the procedure. Spent culture media which had been exposed to lo6 Chang cells for 48 hr. gave P2p values of 100-200 ng./ml.
Man-mouse hybrid clones were considered to be P2p-positive if medium exposed to l o 6 cells for 48 hr. gave assay values for human P2p that were greater than 10 ng./ml. Man-mouse hybrid clones were classified as p2p-negative if the assay values for P2p were the same as, or less than, those obtained with medium exposed to RAG cells alone. On the basis of these criteria, 27 of the 33 primary hybrid clones studied were classified as P2p-positive, while 6 of the 33 primary clones examined were classified as P2p-negative (see Table 1 and Appendix). Of the 26 secondary RAG-lymphocyte hybrid clones, 17 were P2p-positive while 9 were P2p-negative ( Table 2).
The absence of P2p from the spent culture media of P2p-negative clones may have been due to any of the following causes: ( a ) cessation of P2p synthesis by the cells in question; (b) continued synthesis of p2p by the cells in question with incorporation into the cell membrane, but failure of the material to be secreted into the culture medium; or ( c ) continued synthesis of ,192p but with failure of the material to be either secreted into the culture medium or incorporated into the cell membrane. These possibilities were evaluated in the studies of cell lysates and in the immunofluorescence investigations of the hybrid cell clones under consideration.
Radioimmunoassay for ,8211 in cell lysates. The lysates of 2.5-5.0 x lo6 RAG cells showed no evidence of P2p content. Similar results were obtained with the lysates derived from hybrid cell clones classified as 82p-negative. On the other hand, lysates derived from the same numbers of cells from hybrid clones which had been classified as /32p-positive gave assay values for p2p of 9-17.3 ng./ml.
Immunojluorescence studies. Immunofluorescence studies of viable cells derived from primary hybrid clones classified as P2p-positive demonstrated specific surface staining for B2p. Cells of P2p-negative primary hybrid clones showed no such evidence for 1 2 p on their surfaces.
Studies of enzyme phenotypes in hybrid clones. The 33 primary clones and the 26 secondary clones were examined for the expression of 15 different enzyme phenotypes. The presence or absence of the human enzyme phenotypes in each clone are shown in Tables 1 and 2; the segregation analyses of human P2p and the 15 human enzyme markers examined in the 33 primary clones are summarized in Table 3.     Table 4

. Analyses of /32p and the enzymes M P I , ME 1, EST D, N P , Pep B and Pep A in 26 Rag-lymphocyte secondary clones
From the results obtained on examination of primary clones it was not possible to exclude linkage of humanP2p to the following chromosomes: 6, 12, 14) and 13. Furthermore, although the discordance rates between /32p and both MPI (determined by chromosome 15) and Pep A (determined by chromosome 18) were high in the RAG-Chang hybrids, the discordance rates for /32p and these enzymes were low in the RAG-lymphocyte hybrids. I n the RAG-lymphoid line hybrids a high degree of concordance was observed in the occurrence of human P2p and MPI. It is possible that the observed discrepancies between human /32p and MPI occurrence in the RAG-Chang hybrids could be due to chromosomal rearrangements in the long-term Chang cell line. For this reason, 26 secondary RAG-lymphocyte clones were examined. Segregation analyses of human P2p and the human phenotypes of the enzymes MPI, ME 1, Est D, NP, Pep A and Pep B in the secondary clones are summarized in Table 4.
From this table it will be seen that a high degree of concordance was observed only between /32p and MPI.
Studies of hybrid clone karyotypes. Results of these studies are summarized in Table 5. Karyotypic analyses in the primary hybrid clones revealed a high degree of discordance between /32p production and the occurrence of chromosomes 1, 3, 4, 5, 7, 8, 9 and 10 for which no enzyme markers were examined. I n 1 of the P2p-negative primary clones and in 4 of the P2p-negative secondary clones, chromosome 6 could be clearly demonstrated. Karyotypic studies have so far been done on 5 of the secondary RAG-lymphocyte clones; results support the enzyme studies in suggesting linkage of the gene determining /32p production to chromosome 15.  (1975), who recently assigned the P2,u gene to chromosome 15. These workers used a cytotoxicity assay to determine the presence of human P2,u in man-mouse somatic cell hybrids.

Results
In RAG-Chang hybrids MPI and /32p apparently segregated independently. It is well known that chromosomal rearrangements occur in long-term fibroblast lines and this probably limits the usefulness of these lines in mapping studies. Chromosomal rearrangements apparently occur less frequently in lymphoid cell lines (Povey et al. 1973).
It is interesting to note that the clone RLy C was initially classified as P2,u-positive, MPI negative. Two of the 11 secondary clones derived from this clone were ,82,u-positive, MPI positive, while 9 were P2,u-negative, MPI negative. This indicates that the level of /32p production in these cells exceeds that of the MPI production, or that the detection system for human P2p is much more sensitive than the MPI detection system. Therefore, clones in which a low percent age of cells are positive for chromosome 15 could be positive for P2p and apparently negative for MPI. The fact that the primary hybrid clones classified, during this study, as P2,u-negative had lost the capacity to synthesize, rather than the ability to transport ,821~ was demonstrated by both the immunofluorescence studies and by radioimmunoassays performed on cell lysates. Cells of the P2,u-negative clones failed to show the P2,u molecule either in their membranes or in the supernatants of their lysates. Thus, the P2,u-negative clones described here appear to lack the structural and/or the regulator gene for P2,u production.
Discordant segregation of P2,u and chromosome 6 was clearly demonstrated. If the assignment of the gene determining HL-A alloantigens to chromosome 6 is correct, our findings indicate that the gene responsible for P2,u production is not syntenic with the gene responsible for HL-A alloantigen production.

SUMMARY
Beta-2-microglobulin (P2,u) production has been studied in 33 primary man-mouse hybrid clones and in 26 secondary man-mouse hybrid clones. These clones have also been examined for the presence of 15 human enzyme phenotypes. Karyotypic analyses have been carried out on clones. From t'hese studies the following conclusions can be drawn: (1) Gene(s) determining human P2,u production in humans are apparently syntenic with the MPI gene on chromosome 15.
(2) Long-term fibroblast lines may be of limited use in mapping studies as chromosomal rearrangements frequently occur in these lines.
(3) The gene(s) determining P2,u production in humans segregate independently of chromosome 6. If the assignment of the genes determining HL-A alloantigens to chromosome 6 is correct, our results imply that P2,u and HL-A alloantigens are determined by genes carried on different chromosomes, despite the fact that P2p forms an integral part of the HL-A molecule.
We wish to acknowledge the excellent technical assistance of Mrs Lorna Katz and Miss Marthe Fauconnet, and the photographic assistance of Mr P. Rippstein. We thank Dr Jan Schulz for assistance with the radioimmunoassay studies, Dr C. M. Steel (MRC Cytogenetics Unit, Edinburgh) and Dr B. Watson for donating the lymphoid cell lines, and Dr Susan Povey for helpful discussions. This research was supported by a grant from the National Cancer Institute of Canada.