THE JOURNAL OF BIOLOGICAL CHEMISTRY Vol. 264, No. 1, Issue of January 5, pp. 251-258,1989 Printed in U. S. A.

Analysis of the Methylation State of the T Cell Receptor /3 Chain Gene in T Cells and Large Granular Lymphocytes*

(Received for publication, July 18,1988)

Shigeru SakamotoS, John R. Ortaldop, and Howard A. YoungSV From the $Laboratory of Molecular Immunoregulation and the §Laboratory of Experimental Immunology, Biological Response Modifiers Program, National Cancer Institute, Frederick Cancer Research Facility, Frederick, Maryland 21 701-1013

We have evaluated the methylation state of the T cell receptor B chain gene (TcR B) in freshly isolated human large granular lymphocyte (LGL) and T cell DNA in order to investigate the relationship between LGL and T cells with regard to methylation of this region of genomic DNA. In addition, we wished to determine whether hypomethylation of specific regions of the chain gene DNA might account for the production of only a nonfunctional 1.0-kilobase (kb) TcR B mRNA transcript in LGL. Our analysis indicates that the het- erogeneous pattern of methylation seen in LGL DNA lies predominantly in the J B 1 region of the TcR /3 DNA structure. The CCGG sequences located at the begin- ning of the HTF island (CG-rich region) in the JB2 region are nonmethylated in both LGL and at least half of the T cell DNA, suggesting that the HTF island is nonmethylated in both LGL and T cell DNA. In addi- tion, specific methylation differences between T cell and LGL DNA can be detected 0.7 kb 3’ to the last exon of Cs1, just 5‘ to the first exon and within the second exon of the CB2 region. The heterogeneous methylation state of the TcR @ J B 1 region in LGL DNA may be due to and a result of the differential use of the JB1 segment for generation of the nonfunctional 1.0- kb mRNA in LGL. These results and our previous studies (Sakamoto, S., Ortaldo, J. R., and Young, H. A. (1988) J. Immunol. 140, 654-660 and Sakamoto, S., and Young, H. A. (1988) Nucleic Acid Res. 16, 2149- 2163) indicate that DNA methylation may be one method by which functional gene expression is con- trolled in specific lymphoid cell populations.

In 1948, Hotchkiss (3) reported that bases in DNA are methylated. A few years later, Wyatt (4) reported that methyl C is the only modified base in higher eukaryotic cells. In 1962, J. Doscocil and F. Strom (5) demonstrated that most of the methyl C nucleotides in higher eukaryotic DNA reside in CpG sequences. After the discovery of a restriction enzyme which recognizes methylated CpG (6), there appeared many reports which suggested a possible biological role for this specific modification of DNA. A gene which is actively transcribed in a tissue is often hypomethylated in that tissue, but the same gene is highly methylated in the tissue where the gene is not transcribed (7-16). Treatment of cells with inhibitors of DNA methylation, such as 5-azacytidine, often results in expression of genes which are not usually expressed in those cells (17- 22). More recent studies of the DNA structure have shown

* The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked “aduertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

!I To whom correspondence should be addressed.

that there exists in genomic DNA CpG rich islands, which are nonmethylated in all cells (23). The occurrence of such islands appears to be associated with genes required for basic cell functions (i.e. housekeeping genes) (23) and may be responsible for constitutive expression of these genes. Meth- ylation of DNA also appears to play a role in development as the maternal and paternal genetic contributions to a mam- malian zygote nucleus do not function equivalently during subsequent development (24-28), and differences in methyl- ation of the genome during male and female gametogenesis may be responsible for the different parental genetic contri- butions to the zygote nucleus (ie. allele-specific methylation) (29,30). Furthermore, methylation of DNA is reported to play an important role in X chromosome dosage compensation (31). Additional unique features of gene methylation is that the pattern of methylation of a gene is maintained from cell cycle to cell cycle except when cell differentiation is associated with an expression of the gene (somatic heritability) (32-34).

LGL’ are a population of lymphoid cells that mediate natural killer activity (35-39). There have been, however, conflicting reports about the cell classification of LGL. Al- though LGL have unique functional and phenotypic charac- teristics (36, 37, 40), some reports have shown that they also have many of the phenotypic and genotypic properties of T cells. For example, LGL can be grown in culture with inter- leukin 2 (41), produce interleukin 2, and interferon-y (42,43), and express CD3, T10, and CD8 antigens (36,38,40). Analysis of T cell receptor gene rearrangement in cloned cells or leukemic cells having LGL-like properties indicated that most of these cell lines and leukemic populations have rearranged T cell receptor genes ( i e . a marker of T cell lineage) (44-48). However, Lanier reported that CD16-positive, low density lymphocyte of human peripheral blood (enriched for natural killer cells) did not contain rearranged TcR /3 genes (49), and there have been no reports of TcR (3 chain protein expression in LGL. Furthermore, this laboratory reported previously that both human and rat LGL do not express a full length 1.3-kb TcR /3 mRNA but do express a nonfunctional 1.0-kb TcR /3 mRNA (50).

In a previous study, we reported that the TcR /3 gene is more methylated in LGL than in T cells and LGL are heter- ogeneous in the methylation state of TcR (3, suggesting that LGL are a heterogeneous lymphoid population distinct from T cells (1). However, we could not precisely localize which region of the TcR /3 gene in LGL DNA differed in its meth- ylation state from T cell, B cell, and monocyte DNA. We could also not determine which portion of the LGL TcR /3 gene was associated with the heterogeneous methylation pat- tern, due to the limitation of the methods which we used in

The abbreviations used are: LGL, large granular lymphocytes; TcR p, T cell receptor beta chain gene; kb, kilobase.

251

252 T Cell and LGL T Cell Receptor Gene Methylation

the previous study (i.e. comparison of the restriction fragment pattern after only HpaII or MspI cleavage). In this study, we utilized the combination of HpaII and other restriction en- zymes to obtain more precise information concerning the methylation pattern of TcR p in LGL and T cell DNA. The predicted restriction enzyme digestion patterns of TcR p, based on the location of the cleavage sites for the restriction enzymes utilized in this study, were derived from published DNA sequence data (51) and our own restriction enzyme analysis and have enabled us to localize methylated and nonmethylated regions of DNA within the genomic DNA structure. Based on these results, we will offer a hypothesis on the origin and relationship of CD3-LGL to other lympho- cyte populations.

MATERIALS AND METHODS

Cell Preparation-LGL and T cells were obtained from the buffy coats after leukophoresis from eight normal healthy volunteers as described previously (52). Mononuclear cells were separated by cen- trifugation on a Ficoll-Hypaque gradient and depleted of monocytes by adherence on a plastic surface. Nonadherent cells were applied onto a nylon wool column, and eluted cells were then fractionated on a seven-step discontinuous gradient of Percoll (Pharmacia LKB Biotechnology Inc.) at concentrations ranging from 40 to 60% (52). LGL were collected from the low density fractions (2 and 3 counting from the top), whereas resting T cells were collected from the high density bottom fraction. Contaminating T lymphocytes were further removed from the LGL fractions by rosetting with sheep red blood cells at 29 "C for 1 h. The LGL, which are nonrosetting at this temperature, were recovered from the interface of a Ficoll-Hypaque gradient. Cell preparations contained an average of 90-95% LGL, as determined by morphological analysis of Giemsa-stained cytopreps and were about 90% CD16-positive and NKH1-positive (35, 36). Contaminating cells in LGL preparations include CD3-, CD4-, and sIg-positive cells ( 4 % each) and nonspecific esterase staining cells (<3%). T lymphocytes contained less than 1% LGL and were >95% CD3-positive (36).

Restriction Enzyme Analysis-Genomic DNA was prepared from cell pellets from individual donors with proteinase K digestion, phenol-chloroform extraction, ethanol precipitation, and RNase A digestion. DNA (10 pg) was first digested with 50 units of either EcoRI, BglII, XbaI, or HindIII (Bethesda Research Labs), followed by phenol extraction and ethanol precipitation. The DNA was then (i) digested overnight with 50 units of HpaII (New England Biolabs, Beverly, MA), followed by readdition of 50 units of HpaII for another overnight digestion (2 x HpaII digestion) or (ii) 2 x HpaII-digested, phenol-extracted, and digested overnight with 80 units of HpaII (3 X HpaII digestion). Digested DNA was subjected to electrophoresis on an 0.8% agarose gel, transferred to a nylon membrane (Nytran, Schleicher and Schuell), and hybridized according to established procedures (53). The location of the restriction enzyme sites in the genomic TcR DNA was derived from the published sequence data (51).

DNA Probes-The TcR p probe (HBJUR 2) was a generous gift from Dr. Tak Mak (Ontario Cancer Research Center) (54). A 5' portion of the Cp region (sequence 115-505 of Ref. 54) was isolated in low melting point agarose from this cDNA following digestion with BglII (Bethesda Research Labs) for use as a Cp probe. The DNA fragments were 32P-labeled by random oligonucleotide priming with a commercially available kit (Pharmacia LKB Biotechnology).

RESULTS

Analysis of DNA Digested with BglII and HpaIII-The results of restriction enzyme digestion with either BglII alone or BglII followed by HpaII of T cell DNAs and LGL DNAs are shown in Fig. lA. DNA from different donors was utilized in these studies to eliminate any possible difference due to restriction enzyme polymorphism. In addition, the majority of T cells can have two different TcR ,O gene structures (51). BglII digestion of T cell DNA (lanes 1 and 4 ) and LGL DNA (lanes 7 and 10) and hybridization with the Cp probe yields a 9.0- and 1.0-kb band. As shown in Fig. lB, the 9.0-kb band is

the B2-B3 fragment, and the 1.0-kb band is the B4-B5 frag- ment. Further digestion of BglII-digested T cell DNA with HpaII revealed a 5.0-, 2.3-, LO-, and 0.7-kb band (lanes 2, 3, 5, and 6). As seen in Fig. lB, the 5.0-kb band is the DNA fragment from B2 to the MspI site near M11. As there are many MspI sites clustered closely around M11, it is not possible to determine the precise HpaII-sensitive (nonmeth- ylated) CCGG site around M11. The 2.3-kb band is the B2- M8 fragment, the 1.0-kb band is the B4-B5 fragment, and the 0.7-kb band is the B4-M30 fragment. As the intensity of the 5.0-kb band is equal to or slightly weaker than that of the 2.3-kb band, about half of T cell DNA without a deletion of the Cpl region has a nonmethylated M8 site, and at least half of T cell DNA without any allelic deletion of TcR @ has a nonmethylated MspI site near M11. It is also possible that multiple sites near M11 are hypomethylated. We cannot evaluate the methylation state of the MspI site near M11 in T cell DNA when the T cell DNA contains a nonmethylated M8 site. Since the intensity of the 0.7-kb fragment is much greater than that of the 1.0-kb fragment, most of the M30 sites are nonmethylated in T cell DNA. Further digestion of BglII-digested LGL DNA with HpaII revealed a 5.0-kb band and a 1.0-kb band (lanes 8,9, 11, and 12), indicating that the M8 site is methylated and a MspI site or multiple MspI sites near M11 are nonmethylated (Fig. 1B). As extensive HpaII digestion (3 X HpaII) did not change the hybridization pattern in any of the experiments reported here, incomplete digestion of the DNA sample by HpaII can be ruled out in all of these restriction enzyme studies.

Analysis of DNA Digested with HindIII and HpaII-The results of restriction enzyme digestion with either HindIII alone or HindIII followed by HpaII of T cell DNAs and LGL DNAs are shown in Fig. 2 A . HindIII digestion of T cell DNA (lanes 1 and 4 ) and LGL DNA (lanes 7 and 10) revealed 7.5- and 3.8-kb bands. As shown in Fig. 2B, the 7.5-kb band is the H5-H6 fragment, and the 3.8-kb band is the H4-H5 fragment. The H6-H7 fragment cannot be detected by our present experimental conditions. After further digestion of HindIII- digested T cell DNA with HpaII, the 7.5-kb band disappeared, the intensity of the 3.8-kb band decreased, and new 3.2-, 2.1-, and 0.6-kb bands appeared (lanes 2, 3, 5, and 6). The 3.8-kb band is the H4-H5 fragment, the 3.2-kb band is the H4-M8 fragment, the 2.1-kb band is the M28-H6 fragment, and the 0.6-kb fragment is the M29-H6 fragment. As the M28 or M29 sites are almost completely nonmethylated in T cell DNA (as shown in the XbaI + HpaII digestion below), very little of the 3.5-kb M27-H6 fragment can be detected. As the intensity of the 3.8-kb band is identical to or slightly weaker than the intensity of 3.2-kb band, about half of the M8 sites are nonmethylated in T cell DNA. In addition, since the intensity of the 0.6-kb band is much stronger than that of the 2.1-kb band, the M29 site is nonmethylated in more than half of T cell DNA. Much of the T cell DNA which contains a methylated M29 site has a nonmethylated M28 site as the intensity of the 2.1-kb band is relatively strong.

After the further digestion of HindIII-digested LGL DNA with HpaII, the 7.5-kb band disappeared, the 3.8-kb band remained relatively constant, and a new 2.1-kb and a very faint 0.6-kb band (not visible in Fig. 2 A ) appeared (lanes 8, 9,11, and 12). As shown in Fig. 2B, the region of hybridization at approximately 3.8 kb represents both the 3.8-kb H4-H5 and the 3.5-kb M27-H6 fragment, whereas the 2.1-kb band is the M28-H6 fragment, and the very faint 0.6-kb band is the M29-H6 fragment. These results indicate that the M8 site is methylated, and most of the LGL DNA contains a methylated M29 site since only a very faint 0.6-kb band appeared. As the

1

FIG. 1. Digestion of T cell and LGL DNA with BgnI and HpaII. A, T cell or LGL DNA was either digested with BgnI (PI) alone (lanes I, 4, 7, and 10) or first digested with BgnI and then twice (2 X) (lanes 2,5,8, and I I ) or three times (3 X) (lanes 3, 6, 9, and 12) with HpaII (Hp). 10 pg of digested DNA was subjected to electrophoresis on a 0.8% agarose gel, transferred to a nylon mem- brane, and hybridized with a radiola- beled T cell receptor @ chain cDNA probe. Roman numerals represent differ- ent donors, and numbers on the side of the figure refer to size of the DNA in kilobases. B, schematic representation of the BglII + HpaII restriction fragment pattern in T cell and LGL DNA. The cDNA probe utilized hybridizes with the first exon of both Cj3l and Cj32 genomic DNA.

3 4 5 6 7 8 9

T (1 1) T ( I I 1 ) LQL (VI I) n n n n

X X n X I I

X X X 0 N 0 cy0

x x

+ m

+ m m m m

+ + + m m

I 1 I

9.0- I" 9 . 0 - l - 9.0-

1.0-

-5.0- i

-2.3- c #

-1.0- -0.7-

-5.0-

-2.3-

-1.0- ' -0.7-

L

-5.0

10 11 12

LQL (V)

X I n n

(YP) x x

+ + m m m

. .

-5.0

- 1.0

i-ti 0.7 kb

5.0 kb w LG L

I I 1.0 kb u Bpnl+ H p l l

intensity of the 2.1-kb band is relatively intense, the M28 site in most of the DNA is nonmethylated as it was cleaved by HpaII. However, since the M27-H6 fragment overlaps with the H4-H5 fragment, the methylation state of M27 cannot be analyzed by double digestion with HpaII and HindIII as the two bands cannot be resolved under these electrophoresis conditions.

Analysis of DNA Digested with Xbal and Hpall-The re- sults of restriction enzyme digestion with either XbaI alone or XbaI followed by HpaII of T cell DNA and LGL DNA is shown in Fig. 3A. XbaI digestion of T cell DNA (lanes 1 and 4 ) and LGL DNA (lanes 7 and 10) revealed 10- and 2.5-kb DNA fragments. The 10.0-kb band of T cell DNA is less distinct and broader than that observed in LGL DNA due to the polyclonal rearrangement of TcR @ in T cells which includes the DP2 and JP2 segments. As seen in Fig. 3B, the 2.5-kb band is the X2-X3 fragment and the 10-kb band is the X4-X5 fragment. After the further digestion of XbaI-digested T cell DNA with HpaII, the 10-kb band disappeared, the intensity and width of the 2.5-kb band increased, and a new

1.0-kb band and a very faint 5.0-kb band appeared (lanes 2, 3, 5, and 6). The 2.5-kb band is a summation of fragments, X2-M8, X2-X3, M28-M30, the 1.0-kb band is the M29-M30 fragment, and the 5.0-kb band is the M28-X5 fragment. These results indicate the majority of T cell DNA contains M28, M29, and M30 sites which are not methylated.

After the further digestion of XbaI-digested LGL DNA with HpaII, the 10-kb fragment disappeared and new 6.2- and 5.0- kb bands appeared, whereas the 2.5-kb band showed no change (lanes 8,9, 11, and 12). The 6.2-kb band is the M27- X5 fragment, and the 5.0-kb band is the M28-X5 fragment. As the intensities of the 6.2- and 5.0-kb bands are almost identical, the majority of LGL DNA contains methylated M29 and M30 sites. These data are in agreement with the results seen with double digestion with HpaII and HindIII or HpaII and EcoRI. In addition, about half of the M27 and M28 sites are nonmethylated in LGL DNA. As the intensity of the 2.5- kb band did not change following HpaII digestion, there is no M28-M30 fragment in the 2.5-kb band.

Analysis of DNA Digested with EcoRI and Hpall-The

254 T Cell and LGL T Cell Receptor Gene Methylation

hybridizes with the first 'exon Cgl and CB2 genomic DNA.

A 1 2 3 4 5 6 7 8 9 10 11 12 T (11) T(1II) LQL (IV) LQL (VI) a n X I x x ( Y m

n n r x n a I x X I x x x x

+ + + + ( Y m ( Y m

I X I r x x + + + + r x x 1 x 1

n a x x w ( 3

7.5 - 3.8 -

7.5d

-3.8- '3.2

"21 -21 -21

FIG. 2. Digestion of T cell and LGL DNA with HindIII and H'II. -0.6 4 . 6 A , T cell or LGL DNA was either di- gested with HindIII (H) alone (lanes 1, 4, 7, and 10) or first digested with HindIII and then twice (2 X) (lanes 2,5, 8, and 11) or three times (3 X) (lanes 3, T Cell M22 "M'B no 6, 9, and 12) with HpaII (Hp). 10 pg of digested DNA was subjected to electro- \ / MZ7 M28 M29

phoresis on a 0.8% agarose gel, trans- 1 'I!'

M30

ferred to a nylon membrane, and hybrid-

,'3 chain cDNA probe. Roman numerals ized with a radiolabeled T cell receptor

on the side of the figure refer to size of represent different donors, and numbers

the DNA in kilobases. B, schematic rep-

striction fragment pattern in T cell and resentation of the HindIII + HpaII re- n4

LGL DNA. The cDNA Drobe utilized

I b n7

M3l

Rearranged ''

M9- M'B n7 no

n5 \,!pll,, , , p 7

MJO M31

of both 1 5 1 I 1 1 1 ' ' ~

-3.8

c2.1

results of restriction enzyme digestion with either EcoRI alone or EcoRI followed by HpaII of T cell DNA and LGL DNA are shown in Fig. 4A. EcoRI digestion of T cell DNA revealed a 4.0-kb band and a faint broad band of hybridization from 7- to 12 kb (lanes 1 and 4 ) . As shown in Fig. 4B, the 4.0-kb band is the E3-E4 fragment, and the faint, broad hybridization is due to a polyclonal rearrangement which includes the D@1 and J@1 region. EcoRI digestion of LGL DNA revealed a 12- and 4.0-kb band (lanes 7 and 10). As shown in Fig. 4B, the 12-kb band is the El-E2 fragment, and the 4.0-kb band is the E3-E4 fragment. After the double digestion of T cell DNA with EcoRI and HpaII, the broad band of hybridization dis- appeared, the intensity of the 4.0-kb band decreased and new 1.5- and 1.0-kb bands appeared (lanes 2,3,5, and 6). Most of

the 4.0-kb band consisted largely of the M6 or M7-E2 frag- ment, since the broad hybridization disappeared following HpaII digestion. The 1.5-kb band is the E3-M30 fragment and the 1.0-kb band is the M29-M30 fragment. This result indicates that most T cell DNA contain a nonmethylated M30 site. However, the intensity of the 1.0-kb fragment is two or three times stronger than that of the 1.5-kb band, indicating that T cell DNA contains a low level of nonmethylated M29 sites.

As shown in Fig. 4A, after double digestion of LGL DNA with EcoRI followed by HpaII, the intensity of the 12-kb band decreased considerably, the 4.0-kb band showed almost no change, and the new 7.5-, 5.5-, and 1.5-kb bands appeared (lanes 8, 9, 11, and 12). Another faint 1.0-kb band appeared

T Cell and LGL T Cell Receptor Gene Methylation 255

FIG. 3. Digestion of T cell and LGL DNA with XbaI and HpaII. A, T cell or LGL DNA was either digested with XbaI ( X ) alone (lanes 1, 4, 7, and IO) or first digested with XbaI and then twice (2 X) (lanes 2,5,8, and 11) or three times (3 X) (lanes 3, 6, 9, and 12) with HpaII ( H p ) . 10 pg of digested DNA was subjected to electrophoresis on a 0.8% agarose gel, transferred to a nylon mem- brane, and hybridized with a radiola- beled T cell receptor @ chain cDNA probe. Roman numerals represent differ- ent donors, and numbers on the side of the figure refer to size of the DNA in kilobases. B, schematic representation of the XbaI + HpaII restriction fragment pattern in T cell and LGL DNA. The cDNA probe utilized hybridizes with the first exon of both Cpl and CD2 genomic DNA.

A 1 2 3 T ( I I 1 )

X I a n

01- x x

+ + x x x

10.0-

2.5-

B T

10.0- -5.0

-2.5-

-1 .o

x 1 C U P ) x x

+ + x x x

n a

10.0-

"5 .0

-2.5-

"1.0

7 8 9 10 11 12 LOL (Iv) LOL (VI)

n m x 1

n n x x 01- + +

x x x

X I x x 01- + +

x x x

Xbal + ia kb M

in LGL DNA from one donor (lanes 11 and 12). The 7.5-kb band is fragment M1 or M2-E2, the 5.5-kb fragment is M4 or M5-E2, and the 4.0-kb band consists largely of the E3-E4 and M6 or M7-E2 fragments. This indicates that LGL DNA contains a heterogeneous methylation pattern in the JB1 region. The 1.5-kb band is the E3-M30 fragment, and the 1.0- kb band is the M29-M30 fragment. These results indicate that a small fraction of LGL DNA contains a nonmethylated M30 site and that in one donor a t least some of the M29 sites are nonmethylated. Thus, the methylation state of the M29 site can differ among different donors.

DISCUSSION

Based on the results of these studies, a comparison of the methylation pattern of the TcR p chain gene among the

different leukocyte populations is shown in Fig. 5 and can be summarized as follows.

Dpl, Jpl, and Cpl Regions-1) LGL DNA contains a heterogeneous methylation pattern in the JB1 region; 2) T cell DNA which still contains the Cp1 region has almost all nonmethylated M6 or M7 sites in the Jp l region; 3) B cell and monocyte DNA have a highly methylated 581 region (2); 4) LGL, B cell (2), and monocyte DNA (2) contain a meth- ylated M8 site, but about half of T cell DNA in which TcR p has not been deleted contains a nonmethylated M8 site.

Dp2, JB2, and Cp2 Regions-1) specific MspI sites of the 562 HTF island are nonmethylated in T cell, LGL, B cell (2), and monocyte DNA (2), and it is hypothesized that the entire HTF island in J82 is nonmethylated; 2) about half of LGL DNA, half of monocyte DNA, and all of B cell DNA have a

256 T Cell and LGL T Cell Receptor Gene Methylation

A 1 2 3 4 5 8 7 8 9 10 11 12

T (1) T (111) LGL (Iv) LGL (VI I I)

a n n n n n I S S I S I

n n X I

x x x x N O N O N O

x x

+ + + 4- w w w w w w

x x N O

C + + c + + 0 w w w ,o w w w - uz . 2: .

a N e g [ r I I

r

4.0- -4.0-

1

FIG. 4. Digestion of T cell and LGL DNA with EcoRI and HpaII. A, T cell or LGL DNA was either digested I

with EcoRI ( E ) alone (lanes 1, 4, 7, and 10) or first digested with EcoRI and then twice (2 X) (lanes 2,5,8, and 11 ) or three B times (3 X) (lanes 3, 6, 9, and 12) with Cell HpaII ( H p ) . 10 pg of digested DNA was subjected to electrophoresis on a 0.8% agarose gel, transferred to a nylon mem- brane, and hybridized with a radiola- beled T cell receptor 0 chain cDNA probe. Roman numerals represent differ- ent donors, and numbers on the side of the figure refer to size of the DNA in kilobases. B, schematic representation of the EcoRI + HpaII restriction fragment pattern in T cell and LGL DNA. The cDNA probe utilized hybridizes with the first exon of both Cpl and Cp2 genomic DNA.

o r -1.5

-1 .0

-1.5 j -1.0 I

I

- 1.5 "1.5

"1.0

I cEy I

EcoRl I Broad - Ej k b l

Hybridizelion (12 kb - 7.0 kb)

1.5 kb

LGL

nonmethylated M27 site (3' to the Jj32 HTF island) (2); 3) approximately half of LGL DNA and all of B cell and mono- cyte DNA has a methylated M28 site (3' to the Jp2 HTF island) (2), whereas at least half of T cell DNA has a non- methylated M28 site; 4) a small fraction of LGL DNA has a nonmethylated M29 site (immediately 5' to the Cj32 region), this site is completely methylated in B cell and monocyte DNA (2), and most of T cell DNA has a nonmethylated M29 site. 5) A small fraction of LGL DNA has a nonmethylated M30 site (in the Cp2 region), B cell and monocyte DNA band contain a methylated M30 site (2), and most (but not all) T cell DNA contains a nonmethylated M30 site.

The results of this study indicates that LGL are a distinct cell population from other leukocyte cell types. However, the

observed pattern of methylation places LGL as being more closely related to T cells than to B cells and monocytes with regard to cell differentiation pathways. Our preliminary data (not shown) that T cell and LGL but not B cells or monocytes have an equally hypomethylated T cell receptor y chain gene also supports this hypothesis. The main cause of the hetero- geneity in the methylation state of the JPl region of LGL DNA is not precisely known, but it could possibly be due to the fact that LGL may be composed of distinct subpopulations which contain more homogeneous methylation patterns or as discussed below, this heterogeneity could arise due to the selection of various Jj3l segments for production of a 1.0-kb mRNA of TcR j3 in LGL. We hypothesize that T cells and LGL share an initial stem cell and differentiation pathway to

T Cell and LGL T Cell Receptor Gene Methylation 257

Methylation State of TO Gene in these cell populations, appear to be important areas of further investigation.

HTF Irland (CpG Rich Segment)

FIG. 5. Diagrammatic representation of the methylation status in leukocyte DNA of the MspI (HpaII) sites in the human T cell receptor @ chain genomic DNA DJC-1 and DJC- 2 regions. M refers to the MspI sites. Black circles are methylated sites, white circles are nonmethylated sites. Combinations of black and white in a circle indicate that the DNA at this site was not completely methylated or nonmethylated in that leukocyte popula- tion.

a branch point from which a cell becomes committed to either the T cell pathway with further demethylation of TcR p (perhaps under the influence of the thymus) or the LGL pathway without further demethylation of TcR p. The meth- ylation state of TcR @ in LGL DNA might reflect the sepa- ration point between the T cell and LGL differentiation pathway.

In our previous study analyzing B cell and monocyte DNA, we suggested that the nonmethylated HTF island of Jp2 in B cells might play a role in the expression of a 1.0-kb mRNA of TcR p in B cells (2), as it had been reported that cDNA clones of the 1.0-kb mRNA were transcribed from the Jp2 promoter (55). In addition, these cDNAs did not code for functional proteins. We considered that the 1.0-kb mRNA of TcR /3 is transcribed from only the Jp2 promoter in B cells, as Jp2 but not Jpl is nonmethylated in B cell DNA, although there is as yet no evidence to rule out transcription from the Jp l pro- moter in B cells. LGL also synthesize a 1.0-kb mRNA, and we have demonstrated here that both the Jp l and J82 regions are hypomethylated in LGL DNA. Thus, LGL may possibly transcribe a 1.0-kb mRNA of TcR /3 from both the Jpl and Jp2 promoters. The heterogeneity of the Jp l region might reflect which segment of Jp l is used in the generation of a 1.0-kb mRNA transcript in LGL. Molecular cloning of the LGL TcR /3 mRNA will be necessary to resolve this issue.

At the present time, the biological significance of the dif- ferent methylation states of the Jp l and Jp2 regions and the heterogeneity of the methylation of the Jp l region in LGL DNA is unknown. As the Jp regions of the TcR p gene are involved in rearrangement of TcR p in T cells and LGL can be considered to be closely related to T cells, the elucidation of the biochemical mechanisms which control methylation and the effect of methylation on site specific recombination

Acknowledgments-We wish to thank Hilda Porter and Anna Mason for preparation of LGL and T cells, Drs. Dan Longo, Joost Oppenheim, and Luigi Varesio for critical evaluation of this manu- script, and Beth Guyton for secretarial assistance.

1.

2.

3. 4. 5.

6.

7. 8. 9.

10. 11. 12. 13. 14.

15.

16.

17.

18.

19.

20.

21.

22.

23. 24. 25.

26. 27. 28.

29.

30.

31.

32.

33.

34.

35.

36.

37. 38. 39.

40.

REFERENCES Sakamoto, S., Ortaldo, J. R., and Young, H. A. (1988) J. Immunol.

Sakamoto, S., and Young, H. A. (1988) Nucleic Acids Res. 16,

Hotchkiss, R. D. (1948) J. Biol. Chem. 175, 315-332 Wyatt, G. R. (19511 Biochem. J. 48, 581-584 DoskoEil, J., and Sorm, F. (1962) Biochim. Biophys. Acta. 55,

Waalwijk, C., and Flavell, R. A. (1978) Nucleic Acids Res. 5,

Doerfler, W. (1983) Annu. Reu. Biochem. 52,93-124 Holliday, R., and Pugh, J. E. (1975) Science 187,226-232 Razin, A., and Riggs, A. D. (1980) Science 210,604-610 Jones, P. A., and Taylor, S. M. (1980) Cell 20,85-93 Ehrlich, M., and Wang, R. Y-H. (1981) Science 212,1350-1357 Compere, S. J., and Palmiter, R. D. (1981) Cell 25,233-240 Felsenfeld, G., and McGhee, J. (1982) Nature 296, 602-603 Mavilio, F., Giampaolo, A., Care, A., Migliaccio, G., Calandrini,

M., Russo, G., Pagliardi, G. L., Mastroberardino, G., Marinucci, M., and Peschle, C. (1983) Proc. Natl. Acad. Sci. U. S. A. 80,

Strobl, J. S., Dannies, P. S., and Thompson, E. B. (1986) Bio- chemistry 25,3640-3648

Fukunaga, R., Matsuyma, M., Okamura, H., Nagata, K., Nagata, S., and Sokawa, Y. (1986) Nucleic Acids Res. 14,4421-4436

Stein, R., Razin, A., and Cedar, H. (1982) Proc. Natl. Acud. Sci.

Farrar, W. L., Ruscetti, F. W., and Young, H. A. (1986) J.

Mohandas, T., Sparkes, R. S., and Shapiro, L. J. (1981) Science

Groudine, M., Eisenman, R., and Weintraub, H. (1981) Nature

Venolia, L., Gartler, S. M., Wassman, F. R., Yen, P., Mohandas, T., and Shapiro, L. J . (1982) Proc. Natl. Acud. Sci. U. S. A. 79,

Jaenisch, R., Schnieke, A., and Harbers, K. (1985) Proc. Natl.

Bird, A. P. (1986) Nature 321, 209-213 McGrath, J., and Solter, D. (1983) Science 220, 1300-1303 Surani, M. A. H., Barton, S. C., and Norris, M. L. (1984) Nature

McGrath, J., and Solter, D. (1984) Cell 37,179-183 Cattanach, B. M., and Kirk, M. (1985) Nature 315,496-498 Surani, M. A. H., Barton, S. C., and Norris, M. L. (1987) Nature

Reik, W., Collick, A., Norris, M. L., Barton, S. C., and Surani, M. A. (1987) Nature 328, 248-251

Spienza, A. C., Peterson, J., Rossant, J., and Balling, R. (1987) Nature 328, 251-254

Wolf, S. F., Jolly, D. J., Lunnen, K. D., Friedmann, T., and Migeon, B. R. (1984) Proc. Natl. Acad. Sci. U. S. A. 8 1 , 2806- 2809

Pollack, Y., Stein, R., Razin, A., and Cedar, H. (1980) Proc. Natl. Acad. Sci. U. S. A. 77,6463-6467

Stein, R., Gruenbaum, Y., Pollack, Y., Razin, A., and Cedar, H. (1982) Proc. Natl. Acad. Sci. U. S. A. 79, 61-65

Vedel, M., Gomez-Garcia, M., Sala, M., and Sala-Trepat, J. M. (1983) Nucleic Acids Res. 11, 4335-4354

Timonen, T., Ortaldo, J . R., and Herberman, R. B. (1981) J. Exp. Med. 153,569-582

Ortaldo, J. R., Sharrow, S. O., Timonen, Y., and Herberman, R. B. (1981) J. Zmmunol. 127,2401-2409

Herberman, R. B., and Ortaldo, J. R. (1981) Science 214, 24-30 Trinchieri, G., and Perussia, B. (1985) Lab. Znuest. 50,489-513 Reynolds, C. W., Timonen, T., and Herberman, R. B. (1981) J.

Reynolds, C. W., Sharrow, S. O., Ortaldo, J. R., and Herberman,

140,654-660

2149-2163

953-959

3231-3238

6907-6911

U. S. A. 79,3418-3421

Zmmunol. 135,1551-1554

2 11,393-396

292,311-317

2352-2354

Acud. Sci. U. S. A. 82,1451-1455

308,548-555

326,395-397

Immunol. 127,282-287

258 T Cell and LGL T Cell Receptor Gene Methylation R. B. (1981) J. Zmmunol. 127,2204-2208 48. Yoshikai, Y., Yanagi, Y., Suciu-Foca, N., and Mak, T. W. (1984)

693 49. Lanier, L. L., Cwirla, S., Federspiel, N., and Phillips, J. H. (1986)

(1983) J. Zmmunol. 131 , 2379-2385 50. Young, H. A., Ortaldo, J. R., Herberman, R. B., and Reynolds, C.

(1983) J. Zmmunol. 130,988-992 51. Toyonaga, B., Yoshikai, Y., Vadasz, V., Chin, B., and Mak, T.

T., Hussey, R. E., and Reinherz, E. L. (1985) Science 2 2 8 , 52. Timonen, T., Reynolds, C. W., Ortaldo, J. R., and Herberman,

45. Rambaldi, A., Pelicci, P-G., Allavena, P., Knowles, D. M., Ros- 53. Maniatis, T., Fritsch, E. F., and Sambrook, J. (1982) in Mokculur sini, S., Bassan, R., Barbui, T., Della-Favera, R., and Manto- Cloning: A Laboratory Manual, pp. 382-389, Cold Spring Har- vani, A. (1985) J. Exp. Med. 162 , 2156-2162 bor Laboratory, Cold Spring Harbor, NY

46. Ikuta, K., Hattori, M., Wake, K., Kano, S., Honjo, T., Yodoi, J., 54. Yoshikai, Y., Antoniou, D., Clark, S. P., Yanagi, Y., Sangster, R., and Minato, N. (1986) J. Exp. Med. 164,428-442 Van den Elsen, R., Terhorst, C., and Mak, T. W. (1984) Nature

47. Kronenberg, M., Governman, J., Haars, R., Malissen, M., Kraig, 312,521-524 E., Phillips, L., Delovitch, T., Suciu-Foca, N., and Hood, L. 55. Calman, A. F., and Peterlin, M. (1986) J. Exp. Med. 164, 1940- (1985) Nature 313,647-653 1957

41. Vose, B. M., and Bonnard, G. D. (1983) J. Zmmunol. 130 , 687- Nature 310,506-508

42. Kasahara, T., Djeu, J. Y., Doughtery, S. F., and Oppenheim, J. J. J. Exp. Med. 163 , 209-214

43. Handa, K., Suzuki, R., Matsui, H., Shimizu, Y., and Kumagai, K. W. (1986) J. Zmmunol. 136,2701-2704

44. Ritz, J., Campen, T. J., Schmidt, R. E., Royer, H. D., Hercend, W. (1985) Proc. NatZ. Acad. Sci. U. S. A. 81,8624-8627

1540-1543 R. B. (1982) J. Zmmunol. Methods 51,269-277