Immunogenetics 24: 184—190, 1986
`
`11221211111 ,
`gene as
`© Springer-Verlag 1986
`
`HLA-DQ Polymorphism Analyzed by Sequential Restriction
`Endonuclease DNA Digestion
`
`Massimo ’Iruccol, Sharon Rosenshinel, Isabella Cascinol, and René J. Duquesnoy2
`
`l The Wistar Institute of Anatomy and Biology, 36th Street at Spruce, Philadelphia, Pennsylvania 19104
`2 Division of Clinical Immunopathology, Department of Pathology, University of Pittsburgh School of Medicine, Pittsburgh, Pennsylvania 15261
`
`Abstract. A direct, physical correspondence between cer—
`tain Pst I-generated genomic DNA fragments and Taq 1-
`generated fragments, revealed with HLA—DQO, or —DQ3
`gene probes, has been demonstrated. As an immediate con—
`sequence, the nature of the DQ and DX hybridized genes
`contained in the fragments was established. Taq I—
`generated DQ allelic forms which associate with serologi-
`cally defined DRl, DR2, and DRW6 specificities were also
`proven to be sensu stricto “splits” of the Pst I-generated al-
`lelic form associated with all three DR specificities.
`
`Introduction
`
`The class II HLA molecules are heterodimers encoded by
`oz and 6 genes located in three HLA-D subregions, DR, DQ,
`and DP. While the polymorphisms of the DR subregion are
`primarily restricted to the [3 genes, it has been observed that
`both or and 6 genes in the DQ subregion exhibit extensive
`polymorphisms (Korman et al. 1985, Trowsdale et al.
`1985).
`Our investigations have focused on the relation between
`restriction fragment length polymorphism (RFLP) analy-
`sis of the HLA-DQ subregion and the expression of corresp-
`onding cell— surface products detected by specific anti-
`bodies and alloreactive T-cell clones. RFLP analysis of Pst
`I-digested genomic DNA yields three allelic DQO, and four
`allelic DQB bands, each corresponding to a distinct sero-
`logical and/or cellular antigenic determinant (Cascino et
`al. 1986). Our findings suggest that the polymorphisms of
`DQ must be defined at the individual at and 6 chain levels.
`The current antigenic system of DQWl—3, even when re—
`cently recognized subtypes and variants are included, does
`not represent a truly allelic system because Dle cor—
`responds at the molecular level to a DQO, gene allele,
`
`Ofi‘print requests to: Dr. M. Trucco, Pittsburgh Cancer Institute, 230
`Lothrop Street, Pittsburgh, PA 15213—2592
`
`whereas DQWZ and DQw3 correspond to DQ6 gene al-
`leles (Cascino et al. 1986). Recent studies have shown addi-
`tional cellularly defined polymorphisms of DQ which do
`not exactly correspond to Pst I RFLP alone (Rosenshine et
`a1. 1986). We have observed that Taq I RFLP could resolve
`this problem in that it could be demonstrated that the
`DR2—associated DQ specificity of two alloreactive T—cell
`clones (ZeeVi et al. 1983) corresponded to a combination of
`DQO, and DQ5 alleles defined by Pst I and Taq I RFLP.
`Although the Taq I RFLP of DQ was considerably more
`complex than the Pst I RFLP of DQ, we found that certain
`Pst I restriction bands always associated with distinct
`groups of two or three Tan bands. For instance, a 15.2 kb
`Pst I band of DQa, corresponding to Dle, always as-
`sociated with a 2 .6 kb Taq I band from DRl cells, with a 5.8
`kb Taq I band from DR2 cells, and with a 6.2 kb Taq I band
`from DRW6 cells. This suggested that, at the molecular lev-
`el, the Pst l-defined DQO, allele could be subdivided or
`“split” into three alleles which could be defined by Taq I
`RFLP. Since these three Taq I bands were smaller than the
`Pst I band, we postulated that by digesting the isolated Pst
`I band with Taq I we could determine the existence of three
`Taq I splits of the Pst I—defined allele.
`In this paper we present evidence that this concept is
`correct. We also demonstrate that this type of sequential
`DNA digestion with different restriction enzymes im-
`proves RFLP analysis of HLA-DQ subregion genes and
`also permits a better differentiation between allelic bands
`attributed to the cross-hybridizing DQ and DX genes.
`
`Materials and Methods
`
`Cell lines. Epstein—Barr virus—transformed, HLA—DR—homozygous, lym—
`phoblastoid cell lines used in this study have been described (Trucco et al.
`1978) and are listed along with their complete HLA typing in Table l.
`
`Probes. The 2.1—kb Pst I fragment of DQa chain genomic DNA (Trows-
`dale et a1. 1983) and the ~ 600—bp Ava I fragment (covering the entire first
`two exons) of pIIBI (DQfi) cDNA (Larhammar et al. 1982) were isolated
`
`