MSU

 

 

 

RETURNING MATERIALS:
P1ace in book drop to

 

 

 

LJBRARJES remove this checkout from
w your record. FINES will
be charged if book is
returned after the date
stamped be1ow.
‘.:: x b

 

 

 

MOLECULAR CLONING OF HUMAN HLA-DS AND MURINE THY-1 ANTIGENS:
MEMBRANE PROTEINS WITH HOMOLOGY TO IMUNOGLOBULINS

BY

Hsiu-Ching Chang

A DISSERTATION

Submitted to
Michigan State University
in partial fulfillment of the requirments
for the degree of

DOCTOR OF PHILOSOPHY

Department of’Microbiology & Public Health

198”

TO MY HUSBAND, YEN-MING HSU AND OUR PARENTS

ACKNOWLEDGEMENTS

I would like to express my deeply appreciation to Dr. Jack Silver for
his advice, encouragement and most imortant, the financial support during
my entire graduate study. I am also grateful to Drs. Jerry Dodgson, Harold
L. Sadoff, John L. Wang, Halter Esselman and Susan Conrad for serving as
my research Guidance Committee. Specially, I thank Jerry for teaching me
the molecular cloning technology, and Susan for giving me the space to
finish up my thesis and being my advisor after Jack left M.S.U.

I would like to express my gratitude to Dr. Tetsuya Moriuchi for his
collaborative efforts that aided in the progress of this work, and also I
would like to thank Drs. Tetsunori Seki and Roger Denome for their help

and allowing me to use their unpublished data here.

111

LIST OF TABLES .
LIST OF FIGURES .
ABBREVIATIONS .
INTRODUCTION . .

CHAPTER I LITERATURE REVIEW . .

TABLE OF

CONTENTS

I. The major histocompatibility complex

General introduction of the NBC

The 3-2 complex and the human leuckocyte antigens (HLA)

in mouse and man

Structure and function of the MHC antigens . . . . . . .

II. Class II antigens of the NBC

The molecular structure of class II antigens

The isolation of genes coding for class II antigens
Polypeptides of class II antigens . . . . . . . . .

Exon and intron organization of class II genes
Genetic map of the class II region of the NBC .

Polymorphism of class II antigens . . . . . . .

Biosynthesis of class II antigens . . . . . . .
Gene expression and functional studies of class

II genes

Page

vi

vii

III. The Thy-1 antigen--a membrane protein with homology to immune-
810bu11n O O O O O O C O O O O O O O O O O O O O O O O O O O 0

Tissue distribution of the Thy-1 determinant . . . . . . . .
Molecular properties of Thy-1
PNEGIOSDPmtureOfThy-1.o...o........o..
Homology to lg and speculation on the function of the Thy-1

antigens . . . . . . . . . . . . . . . . . . . . . . . . . .

References

CHAPTER II THE HEAVY CHAIN OF THE HUMAN B CELL ALLOANTIGEN HLA-DS HAS A

VARIABLE N-TERMINAL REGION AND A CONSTANT IMUNOGLOBULIN-
LIKE REGION

References

iv

ix

‘OOQ 00 030) U) m

21

21
22
23

25
29

36
A9

CHAPTER III THY-1 cDNA SEQUENCE SUGGESTS A NOVEL REGULATORY MECHANISM

References

CHAPTER IV THE NUCLEOTIDE SEQUENCE OF THE MOUSE THY-1 GENE . . . .

sumary I O O O O O O O O O O O O O O O O O O O O O O O O O O 0

Introduction

ReSUItS O O O O O I I O O O O O O O O O O O O O O O O O O O O 0

Isolation and characterization of mouse Thy-1 genomic clones

StruCture Of the mouse Thy-1 gene 0 o o o o o o e e o o o o o
The gene organization of mouse, rat and human Thy-1 . . . . .

Northern blot analysis of RNA from mouse and rat tissues

Comparison of 5' untranslated region. . . . . . . . . . . . .
the Thy-1

Comparison of the predicted amino
86188 I O O O O O O O O O O I O 0

Discussion

Experimental procedures . . . . . .

DNA cloning . . . . . .
DNA preparation . . . .
DNA sequencing analysis
RNA analysis

Acknowledgements

References

CLOSING STATEMENT

acid sequences of

51
62
6::
65
66
69
69
724
77
77
so
as
90
9!:
9n
9::
9s
96
97
93

103

LIST OF TABLES

Table Page
Chapter I

1. Amino acid analyses of the Thy-1 glycoproteins . . . . . . . . . . 2h
Chapter IV

1. Degree of predicted Thy-1 protein sequence homology . . . . . . . . 88

2. Comparison of amino acid compositions of the Thy-1 antigens . . . . 92

vi

LIST OF FIGURES

Figure
Chapter I

1.

Genetic map of the M30 in mouse and man . . . . . . . . . . . . .

2. Structure of class I and class II molecules . . . . . . . . . . .
3. Comparison of published amino acid sequences of class II anigens
A. Exon-intron organization of class II and related genes . . . . .
Chapter II
1. cDNA and predicted amino acid sequence of the DSe subunit clone,
pDSo~12 . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
2. Comparison of the protein sequences of DR and D80 subunits
predictedfromcDNAclones .
3. Comparison of the DNA sequences of D87 and D8", 6a subunits . .
A. Comparison of the protein sequences of DSM, 6 and DS7e subunits
Chapter III
1. Partial restriction map and strategy for sequencing Thy-1 cDNA
clone, pT6A . . . . . . . . . . . . . . . . . . . . . . . . . . .
2. Primary structure and predicted amino acid sequence for pT6u . .

3.

Comparison of the DNA sequences of mouse VA1 (MOPC1OHE) and rat

Thy-1 o e o o o e o e o o o e s m o e o e o o o o e o e o e o o 0

Chapter IV

1.

2.

Scheme for the use of cZRB vector in the construction of a mouse
CS7BL/6 cosmd library I O O I O O O 0 O O O O O O O 0 O O O O 0

Restriction enzyme maps of Thy-1 genomic clones . . . . . . . . .

sequence or the mouse Thy-102 gene 0 e o s e e o o o o o o o o 0

Comparison of the gene organization of mouse, rat and human Thy-1

Northern blot analyses of RNA from mouse rat tissues

vii

12
15

H0

H2
H5
”7

55
57

61

71
73
76
79
82

Comparison of the 5' and nucleotide sequences of rat cDNA and mouse
and mt smomc DNA 0 O O O O O O O O O O O I O I O O O O C I O O 0 8”

Comparison of predicted amino acid sequences of mouse, rat and
human Tm-1 O O O O O O O O O O O I I O I O O O O O O O O O O O O O 87

viii

MHC
HLA
Ia
Ia
BZ-M

2-D

CP

CY

3‘UT

SDS

V domain

C domain

VL

HV

Ampr

EDTA
Tris

a-LA

ABBREVATIONS

major histocompatibility complex
human leukocyte antigen
immunoglobulin

I region associated
32-microglobulin

two dimensional

leader

connecting peptide

transmembrane peptide

cytoplasmic domain

3' untranslated region

sodium dodecyle sulfate

variable domain

constant domain

variable domain of the light chain
variable domain of the heavy chain
hypervariable

ampicillin resistance

kanamycin resistance
(ethylenedinitrilo)-tetraacetic acid
Tris (hydroxymethyl)aminoethane

a-lactoalbumin

ix

INTRODUCTION

This thesis decribes the molecular cloning of two cell surface
alloantigens, HLA-DS and Thy-1, both of“which are molecules with homology
to immunoglobulins.

The HLA-DS alloantigen is a class II alloantigen of the human MHC
(Major Ristocompatibility Complex). The NRC is a group of closely linked
loci coding for molecules that play a critical role in rejection of organ
transplants and the control of the immune response. The loci fall into
three classes, class I, class II, and class III. The class I and class II
molecules, typified by transplantation antigens and the Ia (I-region
associated) antigens, respectively, are integral membrane proteins
involved in the recognition that permits the immune system to distinguish
between self and nonself. The class III family encodes several components
involved in the activation pathway of complement. These genes and
molecules have been studied intensively over the last five decades by
geneticists, biochemists and immunologists, but only recently has the
isolation of the genes by molecular biologists facilitated their precise
characterization. Many surprising findings have been made concerning their
structure, multipilicity, organization, function and evolution. Since my
studies relate to the molecular cloning of RLA-DS, I will concentrate my

review on the structure, organization and evolution of the MHC class II

molecules.
Thy-1 was originally defined in mice as a cell surface alloantigen of
thymus and brain with two allelic forms, Thy-1.1 and Thy-1.2. It is

1

2
present on the surface of T lymphocytes, but not B lymphocytes, of the
mouse, so it has been used as a mouse T cell marker. Subsequently, the
Thy-1.1 alloantigenic determinant was identified in rats. In contrast to
mice, the rat Thy-1 determinant is present on bone marrow cells, but not T
lymphocytes. Thy-1-like molecules have also been found in many species,
and in a variety of cells. The tissue distribution of the Thy-1
determinant reveals a very unusual pattern within species, and between
species. Recently, the proteins expressing the Thy-1 determinant have been
isolated, and their protein sequences have been determined. The homology
between Thy-1, Ig, and NBC molecules has been noted. Therefbre, I take the
liberty to discuss the Thy-1 molecule with the MHC molecules, and will
generally discuss the isolation, tissue distribution, relateness of the
Thy-1, Ig and MHC molecules and the speculative functions of the Thy-1

molecule .

Chapter I
LITERATURE REVIEW

1. The Major Histocompatibility Complex in mouse and man

General introduction of the MHC. The HBO was discovered by P.A.
Gorer in the 19303 (1). It was found that tumors of a particular mouse
would grow only in mice of the same inbred strain and that the alloantigen
of red blood cells, antigen II, is important in determining the fate of
these transplants (2). Subsequently, the genes postulated in the genetic
theory of tumor transplantation were referred to as histocompatibility
genes, H-2, by G.D. Snell (3). By grafting tumors or skin among such mice
and fellowing the rejection or acceptance of the graft, the 3-2 was
determined to be a large, complex genetic region, and was denoted the
Major Histocompatibility Complex, M30. The cell surface structures
responsible for graft rejection were initially characterized using allo-
antibodies produced by crossimmunization of mice differing only at the
MHC. The specificities of such alloantisera for gene products of the
murine MHC have permitted the identification of three classes of molecules
(u). Class I molecules, typified by transplantation antigens, serve as
restricting elements for the response of cytotoxic T cells to virally
infected cells. Class II molecules, also called I-region associated, or Ia

antigens, regulate the recognition of foreign antigens on the surface of

macrophages and B cells by T cells (5). Class III molecules include
several components in the activation stage of complement cascade (6), and

will not be discussed further.

u

The 3-2 Complex and the Human Leukocyte AntigensA(HLA). The genetic
maps for the MRC of’mouse and man have been constructed by serological
analysis of recombinant MHC chromosomes (fig. 1) (ll, 5, 7, 8). The H-Z
complex is located on chromosome 17, and spans about 2 centimorgans of
DNA, which corresponds to approximately NOOO Kb of DNA. The complex is
divided into six subregions called K, I, S, D, Ga, and Tla (",5). Two
categories of class I antigens are encoded at four loci of the 8-2
complex-K, D, Ga and TLa. The class I loci, K and D, encode cell surface
molecules termed transplantation antigens, denoted R-ZK, H-ZD and H-ZL
antigens, which mediate the graft rejection initially used to define the
MHC. The class I molecules belonging to this category represent the
classical 3-2 antigens. They are expressed on all tissue types, and are
highly polymorphic, with more than 50 different alleles (",5). The second
category of class I genes, Qa and Tla, encode the Qa-1, Qa-Z and Tla
antigens, which are structurally closely related to the class I
transplantation antigens. However, they differ from classic 3-2 antigens
in several aspects. The Qa antigens are preferentially expressed on B and
T cells, and the Tla antigens on thymocyte and certain leukemia cells.
More striking, the Ga and Tla antigens exhibit a much lower level of
polymorphism among mice (9). Class II genes lie between the K and D loci,
in the region called I of 3-2, and encode the cell surface Ia antigens,
which are primarily expressed on the surface of B cells, macrophage, and
activated T cells, and also exhibit extensive serological polymorphism (h,
5). TwOIdistinct types of'murine Ia molecules, I-A and 1-8, have been
identified (10).

The human MNC called human leukocyte antigens (HLA) complex is

located on chromosome 6, and encompasses about 3 centimorgans of DNA,

 

 

CHROMOSOK I7 WWW/WW.
Locn K ~55, E. [04.516] [b.L.n] 00'2,3 m 00-!
cuss ‘lnln 11 mm 111 l l I
COMPLEX , 11.2 It Tla g
REGION 11‘; 1 II S n D l 00 11 T10 ,
Recouammou _.. _.._ _.._____ ____.
FREQUENC|ES(CM) ”Dy—OJ 0-“ 026 LS
cuaouosoua s WWW/”fiflm
€2.81
“’3' GL0 [SB-.589] [0699.094981] cancas 3 C ‘
cuss n n n n n n g g x n x
COMPLEX A HLA g

 

nacousmnnou A u
Fneousucnas (can ‘ ”3 v~ 0.7---0.3--an-—o.7—o

Figure 1 Genetic map of the MHC in mouse and man.

(Taken from reference 8)

Class I ' Class II

 

Figure 2 Structure of class I and class II molecules. S-S indicates
disulfide bridges in the extracellular domains of class I and class II

molecules and BZ-microglobulin (B 2-11). (It is taken from reference 7).

6
extending perhaps over 6000 Kb of DNA (fig. 1). It also consists of three
types of loci (A, 5, 7). Class I loci RLA-A, -B, -C (analogous to K and D
of’mouse 3-2), encode the transplantation antigens which are present on
virtually all cells and elicit strong allograft reactions. The gene
organization of human class I loci is different from their mouse
counterparts, since the RLA-A, -B, -C genes are contiguous to one another,
whereas in the mouse R-2 K gene is separated from the other class I genes
by 3-2 I and 3-2 8 loci. The human counterparts to mouse 0a and Tla
antigens have been charaterized serologically and biochemically, but their
genes have not yet been mapped (11). Class II loci RLA-D (analogous to I
of‘mouse 3-2) encodes three distinct types of human class II molecules, DR
(12), DC(DS) (13), and SB (111) antigens which stimulate response of T
cells in the mixed lymphocyte reaction and probably modify the response to

antigen on macrophage.

Structure and function of the MHC antigen_. Both class I and class
II antigens are heterodimeric cell surface glycoproteins (fig. 2) (7). The
class I loci encode for a polypeptide chain with a molecular weight of ”5
RD, which contains three extracellular domains (c1, a2 and<13 ), each
about 90 amino acids long, a transmembrane region, and a cytoplasmic
domain (A, 5, 7, 8). The third extracellular domain is noncovalently
associated with 82-microgloblulin (32-M) on the cell membrane. The BZ-M
is a 12 KD polypeptide (15) which shows homology to the constant domain of
Ig and is encoded on the murine chromosome 2 and human chromosome 5 (16).
Thus the cell surface molecules of class I appear to be heterodimers, but
only the class I molecules are anchored on the membrane. Class II antigens

consist of two non-covalently associated polypeptide chains with

7
molecular weight of 33,000 and 28,000, designated heavy chains (a) and
light chains (8), respectively. Both of these polypepetides are encoded in
the MHC, each contains two extracellular domains, a transmembrane region,
and a small cytoplasmic domain (A, 5, 7, 8, 16).

All the cell surface antigens of the MHC have a cruical role in the
regulation of the immune response. The class I antigens, originally
recognized as transplantation antigens have been shown to be involved in
the recognition and killing by cytotoxic T cells, that permit T cells to
detect foreign antigens in the context of self (17). In the process of
imunosurveillance, the T cell receptor on the cytotoxic T lymphocyte
must recognize both foreign antigen and self transplantation antigen,

This phenomenon has been termed major histocompatibility restriction, or
H-2 restriction in mice. The function of non polymorphic class I antigens,
0a and Tla, is unknown. Class II antigens also function as restricting
elements, but for regulatory T cells (18). The proliferative response of T
cells to antigen presented on the macrophage surface required histo-
compatibility at the H-2 I region of the two cell types. Similarly, the
cooperation between helper T cells and B cells exhibits 8-2 restriction.
Thus, the Ia antigens determine several immunologic reactions including
control of the level of immune response (proliferation and suppression),
delayed—type hypersensitivity (DTH), disease-susceptibility, and primary
and secondary allogeneic T cell reactions.

Both the importance of T cell recognition in the positive and

negative regulation of the immune system and a number of unusual

characteristices of class II molecules, make the studies of the class II
molecules important. For instance, the class II molecules possess unusual

features of biosynthesis, conformational change, and interaction. Their

8
structures suggest evolutionary relationships with both class I, and Ig
molecules. Like class I antigens, they form a highly polymorphic multigene
family with a complex pattern of antigenic determinants. Recently, the
rapid progress of molecular biology has dramatically advanced our
understanding of the MHC. The isolation of genes has facilitated the
precise characterization of the MRC class II molecules. Therefore, I will
concentrate the following discussion on the topic of the molecular
structure, gene organization, and functional studies of class II antigens

derived from molecular cloning.
II. Class II antigens of the MHC

The molecular structure of class II antigens. The initial efforts to
purify class II molecules on the basis of reactivity with specific
antisera led to the identification of a number of integral membrane
glycoproteins composed of two chains, a heavy chain (a) and a light (8)
chain (7, 8, 16). In mice, two families of Ia molecules encoded by the I-A
and I-E subregions have been identified and characterized. Both I-A and
I-E molecules consist of two noncovalently associated polypeptide
subunits, I-Aa and I-AB, I-Ea and I-EB, respectively. However, the I-A and
I-E molecules are structurally distinct based on partial N-terminal
sequence and peptide map analysis (10, 19). Futhermore, I-A and I-E
molecules display different patterns of structural variability. Hhen‘
allotypic I-E molecules are structurally compared, their small (8)
subunits are found to be very different, whereas their large (a) subunits
are generally invariant. I-A molecules, in contrast, display structural

variability in both subunits when allotypes are compared.

9
Recent studies have identified three distinct types of human class

II molecules. The HLA-DR molecules are structurally homologous to the
murine I-E molecules (10, 12), whereas the DC(DS) molecules are the human
equivalent to the murine I-A molecules (13). These molecules display the
same patterns of structural variability as their murine counterparts. The
third human class II molecule, SB, has not been found in the mouse,
although the SB gene appears to crosshybridize most strongly with the

mouse 282 gene (7, 32).

The isolation of genes codigggfor‘glass II antigggg. Due to the
availability of specific antisera and recent advances in molecular cloning
techniques, the genes coding for class II molecules have recently been
isolated. Two procedures were used to isolate human class II cDNA clones.
First, cDNA clones were selected for their ability to bind class II mRNA,
which could be identified by in 1&222 translation and immunoprecipitation
of the synthesized polypeptide. Such an approach was used in the cDNA
cloning of DRe:(20), DCa(21), DRB(22) and DCB(23). In an alternative
approach, short oligonucleotides were synthesized corresponding to the
nucleotide sequences predicted from amino acid sequences of class II
polypeptides, and were used as hybridization probes to screen for the
class II cDNA clones (2H). The genomic clones containing human class II
genes were isolated by hybridization with the cDNA clones (25).
Subsequently, the cDNA and genomic clones encoding mouse class II antigens

were isolated by cross species hybridization with human cDNA clones

(26-32).

Polypeptides of class II antigens. Amino acid sequences deduced from

10
DNA sequences of cloned genes (19-A2) have complemented the N-terminal
sequence data and other structural studies in providing a clear picture of
the molecular structure of class II antigens. These sequences, together
with studies demonstrating selective proteolytic cleavage of the dimeric
complex, show that each molecule is composed of four extracellular
domains, two N-terminal domains with no discernible homology to lg ( a1
and B1), and two immunoglobulin-like domains (closest to the membrane)
with sequence similarities to the constant Ig domain (a2 and 82). A short
hydrophilic peptide connects the extracullar domains to a short membrane-
binding hydrophobic region, and continues with a short hydrophilic
carboxyl—terminal domain located in cytoplasm (fig. 3).

The first domain of class II heavy chains are variable in length
(from 85-88 amino acids), and contain one glycosylation site at homologous
residues (residues 82 for A11). The sequences of DRG 1 from different
alleles are invariant, and homologous to murine B domain. The DC(DS)c1
sequences are variable between different alleles, and show sequence
homology to the murine Au domain. The first domain of the class II light
chain is 95 amino acids long and has a glycosylation site at residue 19.
Two cysteines at positions 15 and 79 form a disulphide loop of 6“ amino
acids. The human class II B1 domains have strong homology with A81 and 881
domains, and low homology with human class I u1 domain, and c2 domain (35,
38).

The econd domains of both chains ( a2 and 82) are 95 amino acids
long, bear a 56 amino acid disulphide loop (from residues 111 to 167 in
e2 for Au , and from 118 to 17A in B»2), with one glycosylation site in
the e2 domains, at residue 122 for Au. It has been shown to have strong

sequence homology to lg constant regions CY3 and Cu”: the 03 domain of

11

Figure 3. Comparison of published amino acid sequences of class II
antigens. The sequences are divided into three parts accroding to the Au
chain exon structure: (a) the first extracellular domain, (b) the second
extracellular domain and (c) connecting peptide, transmembrane and
cytoplasmic regions. Known exon boundaries are indicated by arrows. The
identical residues are indicated by dashes. The gaps are inserted to
improve homology. The glycosylation site triplets are boxed, and the
disulfides are indicated by arrows.

Aa : Alleles k, d, b, f, u, q (30). Ea: allele K (27); d (39). D03 : DRA,6
(21). Dsa : DR? (33). DRa : DRAJI (38); DR maja (110); DR untypied, DR2,2,
("1). AB : alleles b, d, k, (“2). EB : allele d (28). DCB : DRn,6 (35);

DR2,2 ( Taken from ref. 7.). DRB : DRu,6 (3H); DR2,2 (Taken from ref. 7.).

um
Ag“)
‘9“)
lg“)

Rpm.”
“Hm.“
”Hm.“

”(39.6)
”m.”

(b)

‘00.)
M“)
“(5)
An“)
Am(u)
uh)

”(I)
It“)

“duh”
050(1)!”
his

mum.”
Mamas)
Reflux“)
HAHN.”

A80)
“(6)
AI“)

”(4)

aqua“)
begun.»

mun-2.1)

(e)
um
Add)
um
um
Ash)
am

In“)
In“)
”duh”
MM?)
m

a.

mum.»
Idle”)
makeup“)
Cum.”

(5)
a“)

“(5)
h“)
smut.»
stuns.»

mum.»
Imam)

12

 

 

 

 

 

 

 

 

 

 

”I; [10 :20 ‘30 ‘60 ISO ‘60 n :70 mam

 

I :hmdmil§vmwm4mmnm ' mam

l
I WWI

 

BW‘UHIWEITGWMWIM 6141mm

 

I. i»
n—m
_
I‘-
III

 

i

 

 

 

 

1 v

n

 

hwmficxmgwwm

I

hm“.

1
mm

 

 

”IMAM! 50K"""'°fl.l-I
' r — :

Miranda's»-

Figure 3

13
class I antigens, B2-microglobulin and Thy-1 antigens (25, 35. 33)-
Following the second domain, there is a hydrophilic peptide called

the connecting peptide (CP) that connects the extracellular domain with
the transmembrane region. The CP of the heavy chains are 13 amino acids
long and rich in glutamic acid and proline, whereas the CP of the light
chains are 11 amino acids long and rich in serine. There is no sequence
homology between the C? of the heavy chain and light chain.

As expected, there is no detectable homology between the heavy chain
and light chain in the transmembrane region. However, there are striking
homologies between different heavy chain transmembrane hydrophobic
regions, and also between light chain hydrophobic sequences. Strong
homologies in the hydrophobic regions of other molecules have been
reported, including membrane IgG1, IgG2 and 13M ("3), Thy-1 antigens of
mouse, rat and human, and glycophorins of man, ox, and pig (nu). This
indicates that the structure of the hydrophobic domain is conserved to
serve an important, specific, but unknown biological function for each
molecule. In class II molecules, it has been suggested that interaction
between the two chains in the hydrophobic region provides a selective
pressure to conserve these homologies. The hydrophobic region ends, as
with every transmembrane protein sequenced, with a cluster of positively
charged residues that are thought to interact with the negatively charged
phospholipid headgroups of the inner leaflet of the membrane. These
residues are followed by short cytoplasmic tails of variable length

containing many charged and hydrophilic residues. Mouse and human class I
and class II genes show similar variations in the length of their

cytoplasmic domain. The cytoplasmic tails of membrane bound polypeptides

have been proposed to mediate the effects of external stimuli to the cell

1“
interior. It is not clear whether the variations in length of the
cytoplasmic domains are important for different effector functions of the
molecules and for distinct interactions with cytoskeletal proteins, or

whether they simply reflect evolutionary variations of a non functional

carboxyl terminus.

Exon and intron organization of class II_gggg§. Genomic cloning of
03a (25), DCB(8), Eu (2?), 38(28) and AB (32) genes have determined the
intron-exon organization of these genes. There is a striking correlation
between the organization of exons in the genes and the structural domains
of the class II molecules-ie., each structural domain is encoded by a
separate exon (fig. A). In the DRa and Ba , the leader peptide (L) and
first domain (a1) is separated by a large intron; the transmembrane
region, cytoplasmic exon and part of the 3‘ untranslated region is encoded
in a single exon, followed by an intron dividing the 3' untranslated
region into two exons (an unusual feature for an eucaryotic gene). A
similar gene organization can be found in the gene for BZ-M (H5). The gene
organization of class II light chains, A8 , EB, and 003 are different from
that of the heavy chains. The B1 and 82 domains are separated by a large
intron, 2.“ kb in AB and D03 and 3.9 Kb in 38‘ The transmembrane region is
encoded by a single domain, whereas the cytoplasmic domain is split into
two exons and the 3' untranslated region is encoded in a single exon. The
gene organization of the class II light chain genes are similar to class I
heavy chain genes (8). Thus, the structure of class II a and 8 genes
differ in several regards. The DOE cytoplasmic region lacks eight amino
acids present in the other light chains of I-A and I-E, which correspond

to a single exon in the A3 and EB (28, 32). Whether this reflects

Lanaz aaTM CY“ 3'U'1‘

 

 

 

 

 

 

 

 

 

 

L 81 82 mo} 3'UT
b .
E8 4 i
d L Bl 82 TM all,R 3'UT
AB ' I I I I a

L 81 82 TM CY, 3'UT
—+ I M

L 0.1 (12 mpCY,3'UT
d '
EU I I I 'a

L a1 a2 TM,CY,3'UT
DR 4 M

L

a 3'UT 3'U

T

 

 

Figure 1!. Exon-intron organization of class II and related genes. L, "
leader peptide; a1, c2, 8 1, 82, extracellular domains; TM, transmembrane
domain; CY, cytoplasmic domain; 3'UT, 3' untranslated region. Data taken

from references in Figure 3.

16
deletion of the exon, loss of a splice site, or alternative splicing
events is not clear. One of two DR light chain cDNA clones reveals a lack
of all of the transmembrane and cytoplasmic region (16). Whether this

represents a secreted form of class II molecules needs to be clarified.

Genetic map of the class II mion of the MIC. The genetic map of
the genes coding for class II antigens has been divided into five
subregions: I-A, I-B, I-J, I-E and I-C, by recombinational analysis (A,
5). Recently, the organization of the class II genes in the I region of
the BALB/c mouse has been studied by chromosome walking procedures (26). A
human DRa cDNA probe was used to screen a cosmid library. This was
followed by using left-ended or right-ended single-copy probes for the
sequential screenings, and over 200 Kb of DNA representing most of the I
region of BALB/c mice has been isolated. Several class II genes, Aa, AB,
Ec, EB, and two unknown genes or pseudogenes called A82 and BB 2, have
been mapped in this region (26). The Ea, Au, and A3 genes were identified
by direct DNA sequence analysis, the 83 gene has been identified by
hybridization with an oligonucleotide probe specific for the E:
polypeptide, and the BB 2 gene has been identified by cross hybridization
to the mouse EB and human DC 8 cDNA probes (26). The A3 and EB , together
with A32 have the same 5' to 3' orientation, whereas the Eu, Au genes
show'an opposite orientation. The location of the Au gene between the A3
and 83 genes is different from the gene order of Aa-AB-EB suggested by the

peptide map analysis of A polypeptides from intragenic recombination
studies (“6). However, the gene cloning data unequivocally shows that the

class II gene order in the BALB/c mice is 5'-A32-AB-Au-EB-Ea-EBZ-3' (16).

The organization of human class II genes is not known to the same

17
extent as in the mouse, although an intensive effort is being made to
understand the organization of the D region, which appears to be more
complex than the I region. Nhen DRB and DCB cDNA probes are used in the
Southern blot analyses of human DR homologous cell lines, ”-13 bands
(depending on the restriction enzyme used) can be detected, suggesting the
existence of at least three categories of light chain genes in human MHC
(u7-u9). Isolation of cosmid clones, together with mapping using
Y—ray induced deletion mutants, show that five or six heavy chain genes
are located in the D region (50, 51). It is currently believed that there
are at least seven human light chains (three DRB , two DCB and two SBB)
and at least five or six human heavy chains (one DRa, three or four DCa-

related, and one SBa).

Polymorphism of class II antigens. Serological studies have defined
the several alloantigenic systems of the MIC, LA and I-E antigens in
mouse and DR, DC(DS), and SB antigens in human. Each of these allo-
antigenic systems can be distinguished from the others on the basis of
imunodepletion and two-dimensional (2-D) gel analyses. These systems are
all highly polymorphic. By using alloantisera and mixed lymphocyte
reactions, at least 13 DR alleles can be defined (52). Biochemical
studies, including 2-D polyacrylamide gel electrophoresis, tryptic peptide
analysis, and N-terminal protein sequence determination, have suggested
that the polymorphism of class II antigens results from their structural
differences. In most biochemical studies, the e chains have been cited as
being divergent, while very limited polymorphism has been attributed to
the enchains (10). The amino acid sequences derived from cDNA clones of

class II antigens have led to the conclusion that alleles of polymorphic

18
class II loci are very different from each other (Figure 3). The
differences are found not only on the light chains but also on the heavy
chains of A and DC.

The comparison of DR achains has revealed that the DRc chain is
invariant, or nearly so, since only one amino acid difference (Val/Leu in
the cytoplasmic tail) has been found in several DRa clones. The Ba chain
sequences from alleles d and k differ only by two amino acids (one in the
a 2 domain, the other one in the transmembrane region). However, the heavy
chains of the I-A and DC antigens show more variation between different
alleles. Comparison of Au sequences from several alleles shows 7-18 amino
acid differences, whereas 11 amino acid differences can be found between
the DCa chains of DR? and DRu,w6 cell lines (33). Most of these
differences are concentrated in the a1 domain and are clustered within a
few hypervariable regions (31, 33). when the amino acid sequences of class
II light chains are compared between different alleles, they reveal 13-15
amino acid differences between different A5 alleles, 17-27 amino acid
differences between DCB chains, and 29 amino acid differences between DRB
chains of’DRu, 6 and DR2, 2 cell lines (fig. 3). The overall sequence of
the a chain is less ploymorphic than that of the 3 chain. However, the 81
domains are not substantially more polymorphic than a1 domain of the I-A
and DC(DS).

The molecular mechanisms involved in the generation of the extensive
allelic polymorphism of class II antigens have not been established.
Several hypotheses have been proposed. An analysis of nine intra-I region
recombinants has demonstrated that all nine recombination events have
occured within a distance of 8 Kb or less at the EB locus-ie., a hot spot

of recombination (26). Together with the data from studies of determinant

19
shuffling, this suggests that, as for class I antigens (53), recombination
and gene conversion between members of the class II multigene family are
important mechanisms for generating diversity (26). However, another
hypothesis has been suggested to explain the generation of Au allelic

polymorphism (32).

Biosynthesis of class II antigen . As mentioned above, the class II
molecules have a very unusual feature of biosynthesis. It has been
reported that the biosynthetically immature class II antigens are
associated with an invariant third chain, provisionally called In (Ii,Y )
chain (SH-56). More specifically, newly synthesized c and B chains form a
complex with the In chain in the endoplasmic reticulum. After transport of
the protein complex to the Golgi apparatus, and concomitant with terminal
glycosylation, the In chain dissociates from the c and B chains. At
least a fraction of the In chains subsequently become integrated into the
plasma membrane independently from class II antigens. The In chain is a
basic, methionine rich, transmembrane glycoprotein, and is not encoded
within the MHC (55). The biological function of the In chain is not yet
known. It has been suggested that it may regulate the intracellular
transport of class II antigens, and it may prevent the formation of class
II hybrid molecules-- ie., molecules composed of a and 8 subunits
encoded by different loci. The cDNA clone of a human In chain has been
isolated (56). The amino acid sequence derived from.DNA analysis suggests
that the N-terminus of the In chain resides on the cytoplasmic side of
membrane, and the In chain may be devoid of an N-terminal signal. DNA
mediated gene transfer experiments of DR or I-A genes to mouse L cells,

which contain very little In chain, indicate that the In chain may not be

20

essential for the surface expression of class II molecules. The precise

role of the In chain still remains to be clarifed.

Gene expression and functional studies of 912§§,IIIE§9°3- The class
II molecules have long been proposed to serve as restricting elements that
permit the regulatory T cell (helper, suppressor, and amplifier) to view
antigen in the context of self on the surface of other T cells,
macrophages, or B cells. Class II genes appear to control the
proliferation of the regulatory T cells as well as the effector reactions
carried out by these T cells, such as the amplification of other T-cell
subsets and the promotion of B-cell differentiation. The isolation of
genes, gene transfer and in vitro site-directed mutagenesis techniques
offer complementary approaches to dissect the function of each molecule,
and to clarify the structure-function relationships existing among Ia
molecules, foreign antigens and T cell receptor molecules. Malissen et al.
have transferred the mouse Acx and AB genes into mouse L-cell fibroblasts
and hamster B cells (57). The I-A molecules expressed on these two types
of cells appear normal by serological assays and 2-D gel electrophoresis,
and can present certain antigens to T cell helper hydridomas. Futhermore,
the I-A molecules on the L cell can act as targets for allogenic cytotxic
T cells. Rabourdin-Combe and Mach also have successfully transferred the
DRa , DRB and the DR associated In chains into mouse L cells (58). The
molecules expressed have been identified by monoclonal antibodies. Whether

the molecules also possess the function of DR antigens is still unknown.

21

III. The Thy-1 antigen--a membrane protein with homology to immunoglobulin

Tissue distribution ofgtheiThy-1 determinant. The Thy-1 antigen was
first identified in the mouse by A. E. Reif and J. M. Allen (59) as a cell
surface alloantigen of thymus and brain, with two allotypic forms called
e-AKR (Thy-1.1) and e-C3H (Thy-1.2). Both antigenic determinants are
coded by the Thy-1 locus on chromosome 9 in mice (60). By using mouse
alloantisera, the Thy-1.1 but not the Thy-1.2 alloantigenic determinant
was recognized on rat thymocytes and brain cells (61). The mouse and rat
Thy-1.1 determinants are similar but not identical, since the affinity of
cross-reacting antibody differs by 10-fold between two species. Other
antigenic determinants on the Thy-1 molecule can be recognized by
xenogeneic antisera, and these include species-specific and cross-reacting
determinants (62). The Thy-1 antigens defined either by allo- or xeno-
antisera have been shown to be located on the same molecular complex,
which is referred to as the Thy-1 molecule. The Thy-1 antigens have also
been identified in dog and man by cross-reaction with rabbit antibodies to
rat Thy-1 (63).

The tissue distribution of the Thy-1 antigen follows unusual patterns
within a species, and shows surprising differences between species (63).
In all species studied the molecule is a major constituent of the brain
cell membrane, where there are about 600,000 molecules per cell (6A). It
is predominantly found on neuronal cells. In lymphoid tissues the Thy-1
antigen is likely to be the most abundant cell surface molecule of’mouse
and rat thymocytes, but is found in reduced amounts in dog thymocytes and
not at all in human thymocytes. In the mouse, Thy-1 is absent from stem

cells, and has been widely used as a marker for T lymphocytes. In contrast

22
to this, most rat T lymphocytes lack Thy-1, but the molecule is found on a
sub-set of bone marrow cells (BO-“51), including immature B cell and
haemopoietic stem cells (65). In rodents some fibroblasts (66), epidermal
cells (67), breast cells (68), and muscle cells (69) also display the
Thy-1 antigen. In many of these tissues the levels of Thy-1 expression

undergoes dramatic changes during differentiation (68, 69).

Molecular properties of Thy-1. All of the Thy-1 antigenic
determinants from the thymus and brain of rats and mice are expressed in
glycoproteins with an apparent molecular weight on SDS-polyacrylamide gels
of about 25 KD. The molecules have been purified from both tissues in both
species (70, 71). The molecular weights, determined by sedimentation
equilibrium measurements, are 17500 for brain Thy-1 and 18700 for
thymocyte Thy-1, and in each case the molecular weight of the polypeptide
is 12500. The amino acid compositions of the Thy-1 glycoprotein from brain
and thymus are very similar (72), but the carbohydrate composition of
brain Thy-1 glycoprotein differs from that of thymocytes, accounting for
the difference in molecular weight (72).

Both thymocyte and brain Thy-1 glycoproteins have the properties of
molecules that bind directly to the plasma membrane. Rat Thy-1
glycoprotein can be solubilized as a monomer in deoxycholate (DOC) and
detergent Brij 96, and binds one micelle of DOC/molecule. If the detergent
is removed, the molecule self-associates to form an oligomer. Mouse
thymocyte Thy-1 glycoprotein is labelled in membranes by a lipophilic,
photoactivatable reagent which labels only the hydrophobic section of
membrane proteins (73). These properties suggest that the molecule has a

hydrophobic portion in its structure and this is presumably responsible

23
for integration into the membrane since there has never been any
indication of Thy-1 glycoprotein binding to another molecule in

immunoprecipitation studies (7h).

Protein structures of Thy-1. The molecules expressing the Thy-1
determinants have been isolated from the brain of mouse (81), rat (70),
dog (75), human (76, 77), chicken (78), frog (79) and squid (80). These
glycoproteins all have a structure related to the mouse and rat Thy-1
molecules. The amino acid compositions of the Thy-1 molecules from brains
of chicken, mice, rats, men and squids, and from human foreskin have been
determined (76, 78, 80, 81) (Table 1). The protein sequences of the Thy-1
membrane glycoprotein of mouse and rat brain have been reported.

The rat brain Thy-1 sequence was determined from tryptic and V-8
proteinase peptides, and consists of 111 amino acids. The molecule
contains two disulphide bonds, Cys-9 to Cys-111 and Cys-19 to Cys-85, and
three N-linked amino acid sugars, located at Asn-23, -7u, and -98 (81).
The C-terminal peptides were unusual, in that they were either obtained in
a highly aggregated form, or could only be purified after binding to
detergent Brij 96 micelles. They thus appeared to have hydrophobic

Properties, yet did not contain any extended sequence of hydrophobic amino
acids. The C-terminal peptides also contained some unidentified ninhydrin-

postive material, glucosamine and galactosamine. This suggested that the
hydrophobic properties of the C-terminal peptides may be due to the
linkage of lipid.

Mouse Thy-1 glycoproteins were purified from the brain of AKR strain
(Thy-1.1) and CBA strain (Thy-1.2) mice, and the protein sequences were

determined (80). The mouse sequences of mouse Thy-1 are very similar to

2::

Amino acid analyses of the Thy-1 glycoproteins

Human Rat Mouse Chicken Squid

Brain Brain Brain Brain Brain

Thy-1 Thy-1 Thy-1.2 Glycoprotein Glycoprotein
CYS 5 A A A 8
ASX 9 15 14 1A 15
THR 11 10 10 10 7
SER 12 8 11 12 7
GLx 11 10 11 10 12
PRO 5 3 3 3 3
CL! 5 u A 6 8
ALA u 2 3 5 5
VAL 8 9 8 7 8
MET 1 1 1 2 2
ILE 2 5 A 6 5
LED 11 13 12 12 6
TYR 5 2 A 5 u
PHE u 2 u 3 u
HIS 5 A u 3 u
LYS 8 8 9 8 9
ARC 6 9 7 6 6

Data were taken from references: human Thy-1 (76); mouse Thy-1.2 (80); rat

Thy-1 (81); chicken Thy-1-like glycoprotein (78); squid brain glycoprotein
(80).

25
that of rat Thy-1. Both molecules consist of 112 amino acids, with the
two allotypic forms differing by only one amino acid at residue 89, where
Thy-1.1 has arginine and Thy-1.2 has glutamine. This fits well with the
amino acid composition, which is similar except that Thy-1.1 has one more
Arg and one less Gln than that of Thy-1.2 (76). It also explains the

finding that Thy-1.1 is more basic than Thy-1.2 in isoelectrOphoresis
(82).

Homology to lg andyspeculation on the function of the Thy-1 antigens.
Homology between Thy-1, which has a domain-like structure including a
disulphide loop of the appropriate size, and the C domain of Ig has been
noted at the protein sequence level by R. Cohen et al-(83). Nith the mouse
protein sequences and further analysis, it is now clear that the Thy-1
sequence fits better along the whole sequence of the Ig V domain. This was
especially suggested by a comparison of the protein sequence of mouse Ig
VA) and rat Thy-1 (81), and analysis of identities between the protein
sequences of Thy-1 and VL and Va (80). There are 25 and 23 sequence
identities between mouse Thy-1 and VL and VB domain, respectively, and 9
of these in each case are the residues that are conserved in more than 90
t of all VL and VB sequences. There are a number of long stretch of
identical sequence among the mouse, rat and human Thy-1 molecules. These
are interrupted in three places by blocks of residues that differ among
the species, residues 23 to 29, 60 to 68, and 99 to 103. The position of
these blocks of divergent sequences correspond to the locations of
hypervariable regions of the Ig V domain (8“). These data suggest that the
Thy-1 molecule has homology to both the C and V domains of Ig.

The comparison of the protein sequences of Thy-1 andfaz- M suggests

26
that the homologies are at least as good as those between 82-M and C
domains of lg (85). DNA and amino acid sequence analyses have revealed
homology among the MHC antigens, B 2-M and Ig domains, and suggest that all
three have evolved from a common ancestral gene encoding a primitive
domain (85-87). On the basis of sequence homologies, it was suggested that
Thy-1 could be the primordial domain of the Ig superfamily. One homolog of

the Thy-1 molecule with the similar amino acid composition to the other
Thy-1 molecules was found in the squid brain (80). A partial amino acid

sequence showed homology with rodent Thy-1 and IgAV domain sequences,
suggesting that this molecule may be the first invertebrate member of the
Ig superfamily. This finding greatly strengthened the idea that the Thy-1
molecule is primitive, and is likely to be the primordial domain of Ig
family.

Despite the fact that Thy-1 is so abundant on the surface of thymus
and brain cells, and that its molecular structure is so well
characterized, the function of Thy-1 is still an open question. Because of
sequence homology to Ig, it has been suggested that molecules displaying
the Thy-1 antigen play a role in cellular recognition and morphogenesis in
the nervous system (81). The expression of the Thy-1 antigen accompanying
tissue differentiation has been noted since Thy-1 was first discovered.
Reif and Allen noted that the expression of Thy-1 in the brain
accompanies the development of brain, is. neonatal brain possesed only
1.5-2.8! of the content of antigen found in adult brain. In contrast,
the antigen was almost fully developed in the neonatal thymus (59). In
addition, J. E. Lesley and V. A. Lennon also observed that the expression
of the Thy-1 antigen on skeletal muscle was transitory, being expressed on

myoblasts and newly developed myotubes, but disappearing as myotubes

27
differentiated. This finding suggests that Thy-1 may be functional in
normal muscle development (69).

Recently, many studies have tried to use monoclonal antibodies
against the Thy-1 antigen to define the function of Thy-1. Several lines
of data have been reported. R. Dulbecco et al have used monoclonal
anti-Thy-1 sera to prevent the differentiation of'mammary cells (68).
Certain monoclonal antibodies to the Thy-1.1 antigen prevent formation of
new domes and cause disappearance of preexisting one. This suggests that
the specific interaction of these antisera with the Thy-1 antigen
redirects the differentiation programs of these cells. A role for Thy-1 in
cell interactions leading to differentiation within the thymus, and in
enhancing the regeneration of processes by rat retinal ganglion cells,
have been suggested (88, 89). The function of Thy-1 in the immune system
has also been studied by using either allogeneic antisera or monoclonal
antibodies. By using rabbit anti-mouse brain antisera, the Thy-1 antigens

were suggested to be involved in T cell activation, since antibodies

against Thy-1 stimulate T cell proliferation and induce T cell growth
factor production in normal T cells (90). One such monoclonal antiserum
which is able to induce interleukin-2 in T cells, was suggested to
recognize the Thy-1 molecule (91). Therefore, the Thy-1 antigen has been
proposed to be equivalent to human T3 antigen, which has been shown to
play a critical role in the T cell antigen recognition process. However,
another line of data suggests that the mouse antigen equivalent to human
T3 antigen is not the Thy-1 molecule, although it is a 25 KD glycoprotein
with the properties very similar to the Thy-1 molecule (92). Nhether the

Thy-1 antigen is involved in T cell receptor recognition remains to be

clarified.

28
Since the amino acid sequence of Thy-1 has been reported, it is
possible to use recombinant DNA methods to isolate Thy-1 gene, and study
its genomic structure and its relatedness to the immunoglobulins.
Furthermore, by using the isolated gene, we are able to study the
expression and function of Thy-1 in various cell types by gene transfer

and in vitro site-directed mutagenesis techniques.

10.

11.

12.

13.
1”.

REFERENCES

Gorer, P. A. (1937) J. Pathol. Bacterial. HA, 691-697.

Gorer, P. A. (1938) J. Pathol. Bacteriol. 97, 231-252.
Snell, G. D. (19u8) J. Genetics 99, 87-103.

(Klein, J. (1975) Biology of the mouse histocompatibility-2 complex.

pp. 192-230. (New York-Springer-Varlag)

Snell, G. D., Dausset, J. and Nathenson, S. (1976) Histocompatibility
New York Academic Press.

Bianco, C. and Nussenzweig, V. (1977) Contemp. Top. Mol. Imunol.

6, 1N5-176.

Steinmetz, M. and Hood, L. (1983) Science 222, 727-733.

Hood, L., Steinmetz, M. and Malissen, B. (1983) Ann. Rev. Immunol. 1,
529-568.

Yokoyama, K., Stockert, E., Old, L. J. and Nathenson, S. (1981) Proc.
Natl. Acad. Sci. U.S.A. 78, 7078-7082.

Cullen, S. E., Freed, J. H. and Nathenson, S. G. (1976)
Transplantation Rev. 30, 236-270.

Cotner, T.. Mashimo, H.. Kung, P. C., Goldstein, G. and Strominger,
J. L. (1981) Proc. Natl. Acad. Sci. U.S.A. 78, 3858-3862.

Ferrone, S. Pellegrino, M. A. and Reisfeld, R. A. (1977) J. Immun.
118, 1036-10fl1.

Goyert, S. M. and Silver, J. (1981) Nature 299, 266-286.

Hurley, C. K., Shaw, 8., Nadler, L., Schlossman, S. and Carpa, J. D.
(1982) J. Exp. Med. 156, 1557-1562.

29

15.

16.

17.
18.
19.

21.

23.

211.

25.

26.

30
Cunningham, B. A., Hang, J. L., Berggard, I. and Peterson, P. A.
(1973) Biochemistry 12, “811-9821.
Kaufhan, J. F., Auffray, C., Korman, A. J., Shackelford, D. A. and
Strominger, J. L. (1984) Cell 36, 1-13.
Zinkernagel, R. M. and Doherty, P. C. (1980) Adv. Immunol. 27, 51-177.
Klein, J. and Figueroa, F. (1981) Immunol. Rev. 60, 23-57.
Cook, R. G., Siegelman, M. H., Capra, J. D., Uhr, J. N. and Vietta,
E. S. (1979) J. Immunol. 122, 2232-2230.
Lee, J. S. Trowsdale, J. and Bodmer, N. F. (1982) Proc. Natl. Acad.
Sci. U.S.A. 79. 595-599.
Auffray, C., Korman, A. J., Roux-Dosseto, M., Bono, R. and
Strominger, J. L. (1982) Proc. Natl. Acad. Sci. U.S.A. 79. 6337-63H1.
Long, E. 0., Hake, C. T., Strubin, M., Cross, N., Accolla, R. S.,
Carrel, S. and Mach, B. (1982) Proc. Natl. Acad. Sci. U.S.A. 79,
7H65-7u69.
Niman, K., Larhammar, D., Claesson, K., Gustafsson, K., Schenning, L.
Bill, P., Bohme, J., Denaro, M., Dobberstein, B., Hamerling, 11.,
Kvist, 8., Servenius, B., Sundelin, J.. Peterson, P. A. and Rask, L.
(1982) Proc. Natl. Acad. Sci. U.S.A. 79. 1703-1707.
Stetler, D., Das, H., Nunberg, J. H., Saiki, R., Sheng-Dong, R.,
Mullis, K. B., Neissman, S. M. and Erlich H. (1982) Proc. Natl. Acad.
Sci. U.S.A. 79, 5966-5970.
Korman, A. J., Auffray, C., Schamboeck, A. and Strominger, J. L.
(1982) Proc. Natl. Acad. Sci. U.S.A. 79, 6013-6017.
Steinmetz, M., Minard, K., Horvath, 3., McNicholas, J., Srelinger,
J., Hake, C., Long, 3., Mach, B. and Hood, L. (1982) Nature 300,

35-132 0

27.

28.

29.

31.

32.

33.
311.

35.

36.

370

38.

31
Mathis, D. J., Benoist, C. 0., Williams, V. B., Kanter, M. R. and
McDevitt, H. O. (1983) Cell 32, 795-75".
Saito, H., Maki, R. A., Clayton, L. K. and Tonegawa, S. (1983) Proc.
Natl. Acad. Sci. U.S.A. 80, 5520-5529.
Mengle-Gaw, L. and McDevitt, H. 0. (1983) Proc. Natl. Acad. Sci.
U.S.A. 80, 7621-7625.
Benoist, C. 0., Mathis, D. J., Kanter, M. R., Hilliam, V. E. and
McDevitt, H. 0. (1983) Proc. Natl. Acad. Sci. U.S.A. 80, 539-538.
Benoist, C. 0., Mathis, D. J., Kanter, M. R., William, V. B., and
McDevitt, H. 0. (1983) Cell 39, 169-177.
Larhammar, D., Hamerling, U., Denaro, M., Lund. T., Flavell, R. A.,
Rask, L. and Peterson, P. A. (1983) Cell 39, 179-188.
Chang, H-C., Moriuchi, T. and Silver, J. (1983) Nature 305, 813-815.
Long, 0. B., Wake, C. T., Gorski, J. and Mach, B. (1983) EMBO J. 2,
389-39“.
Larhammar, D., Schenning, L., Gustafsson, K., Wiman, K., Claesson,
L., Rask, L. and Peterson, P. A. (1982) Proc. Natl. Acad. Sci. U.S.A.
79. 3687-3691.
Gustafsson, K., E-oth, B., Widmark, B., Bohme, J., Peterson, P. A.
and Rask, L. (198") Nature 309, 76-78.
Roux-Dosseto, M., Auffray, C., Lillie, J. N., Boss, J. M., Cohen, D.
DeMars, R., Mawas, C., Seidman, J. G. and Strominger, J. L. (1983)
Proc. Natl. Acad. Sci. U.S.A. 80, 6036-60H0.
Larhammar, D., Gustafsson, K., Claesson, L., Bill, P., Niman, K.,
Schenning, L., Sundelin, J., Nidmark, 8., Peterson, P. A. and Rask,

L. (1982) Cell 30, 153-161.

39.

90.

91.

92.

93.

99.

us .
96.
:17.

98.

99.

50.

51.

52.

53.

59.

32
McNicholas, J., Steinmetz, M., Hunkapiller, T., Jones, P. and Hood,
L. (1982) Science 218, 1229-1239.
Lee, J. S., Trowsdale, J., Travers, P. J., Carey, J., Grosveld, P.,
Jenkins, J. and Bodmer, W. F. (1982) Nature 299, 750-752.
Das, K. H., Lawence, S. K. and Weissman, S. M. (1983) Proc. Natl.
Acad. Sci. U.S.A. 80. 3593-3598.
Choi, B., McIntyre, K., Germain, R. N. and Seidman, J. G. (1983)
Sceince, 221, 283-287.
Tyler, B. M., Cowman, A. F., Gerondakis, S. D., Adams, J. M., and
Bernard, 0. (1982) Proc. Natl. Acad. Sci. U.S.A. 79, 2008-2012.
Murayama, J-I., Tomita, M. and Ramada, A. (1982) J. Membrane Biol.
69, 205-215.
Parnes, J. R. and Seidman, J. G. (1982) Cell 29, 661-669.
Rose, S. M. and Cullen, S. E. (1981) J. Immunol. 127, 1972-1977.
Bohme, J., Owerbach, D., Denaro, M., Lernmark, A., Peterson, P. A.
and Rask, L. (1983) Nature 301, 82-89.
Owerbach, D., Lernmark, A., Rask, L., Peterson, P. A., Platz, P.
and Svejgaard, A. (1983) Proc. Natl. Acad. Sci. U.S.A. 80, 3753-3761.
Wake, C. T., Long, E. 0. and Mach, B. (1982) Nature 300, 372-379.
Trowsdale, J., Lee, J., Carey, J., Grosveld, F., Bodmer, J. and
Bodmer, W. (1983) Proc. Natl. Acad. Sci. U.S.A. 80, 1972-1976.
Auffray, C., Kuo, J., DeMars, R., and Strominger, J. L. (1983)
Nature 309, 179-177.
Bodmer, J. G. (1978) Br» Med. Bull. 39, 233-290.
Weiss, B. H., Mellor, A., Golden, L., Fahrner, K., Simpson, 3.,
Hurst, J. and Flavell, R. A. (1983) Nature 301, 671-679.

Day, C. B. and Jones, P. P. (1983) Nature 302, 157-159.

55-

56.

57.

59.
60.
61.
62.

63.

6”.

65.

66.
67.

69.

70.

71.

33
Kvist, S., Wiman, K., Claesson, L., Peterson, P. A. and Dobberstein,
B. (1982) Cell 29, 61-69.
Claesson, L., Larhammar, D., Rask, L. and Peterson, P. A. (1983)
Proc. Natl. Acad. Sci. U.S.A. 80, 7395-7399.
Malissen, B., Price, M. P., Goverman, J. M., McMillan, M., White, J.,
Kappler, J., Marrack, P., Pierres, A., Pierres, M. and Hood, L.
(1989) Cell 36, 319-327.
Rabourdin-Combe, C. and Mach, B. (1983) Nature 303, 670-679.
Reif, A. E. and Allen, J. M. (1969) J. Exp. Med. 120, 913-933.
Blankenhorn, E. P. and Douglas, T. C. (1971) J. Hered. 230, 259-263.
Douglas, T. C. (1972) J. Exp. Med. 136, 1059-1062.
Williams, A. F. (1976) Eur. J. Immunol. 6, 526-528.
Dalchau, R. and Fabre, J. W. (1979) J. Exp. Med. 199. 576-582.
Acton, R. T., Morris, R. J. and Williams, A. F. (1979) Eur. J.
Immunol. 9, 598-602.
Hunt, S.V., Mason, D. W. and Williams, A. F. (1977) Eur. J. Immunol.
7, 817-823.
Stern, P. L. (1973) Nature (London) New Biol. 296, 76-78.
Scheid, M., Boyse, E. A., Carswell, E. A. and 01d, L. J. (1972) J.
Exp. Med. 135, 938-955.
Dulbecco, R., Bologna, M. and Unger M. (1979) Proc. Natl. Acad. Sci.
U.S.A. 76, 1898-1852.
Lesley, J. F. and Lennon, V. A. (1977) Nature 268, 163-165.
Kuchel, P. W., Campbell, D. G., Barclay, A. N. and Williams, A. F.
(1978) Biochem. J. 169, 911-917.
Trowbridge, I. S., Weissman, I. L. and Bevan, M. J. (1975) Nature
256, 652-659.

72.

73.

79.

75-
76.

78.

79.

80.

81.

82.

83.

89.

85.

86.

39
Barclay, A. N., Letarte-Muirhead, M., Williams, A. F. and Faulkes, R.
(1976) Nature 263, 563-567.
Owen, M., Knott, J. C. A. and Crumpton, M. J. (1980) Biochemistry 19,
3092-3099.
Ledbetter, J. A. and Herzenbrg, L. A. (1979) Immunol. Rev. 97, 63-90.
McKenzie, J. L. and Fabre, J. W. (1981) Transplantation 31, 275-282.
Cotmore, S. F., Crowhurst, S. A. and Waterfield, M. D. (1981) Eur. J.
Immunol. 11, 597-603.
Ades, E. W., Zwerner, R. K., Acton, R. T. and Balch C. M. (1980) J.
Exp. Med. 151, 900-906.
Rostas, J. A. P., Shevenan, T. A., Sinclair C. M. and Jeffrey, P. L.
(1983) Biochem. J. 213, 193-152.
Mansour, M. H. and Cooper, E. L. (1989) J. Immunol. 132, 2515-2523.
Williams, A. F. and Gagnon, J. (1982) Science 216, 696-703.
Campbell, D. G., Gagnon, J., Reid, K. B. M. and Williams, A. F.
(1981) Biochem. J. 195, 15-30.
Hoessli, D., Bron, C. and Pink, J. R. L. (1980) Nature 283, 576-
578.
Cohen, F. B., Novotny, J., Sternberg, J. 8., Campbell, D. G. and
Williams, A. F. (1981) Biochem. J. 195. 31-90.
Bernard, 0., Hozumi, N. and Tonegawa, S. (1978) Cell 15, 1133-1199.
Peterson, P. A., Cunningham, B. A., Berggard, I. and Edelman, G. M.
(1972) Proc. Natl. Acad. Sci. U.S.A. 69, 1697-1701.
Steinmetz, M., Frelinger, J. 0., Fisher, D., Hunkapiller, T.,
Pereira, D., Weissman, S. M., Dehara, H., Nathenson, S. and Hood, L.
(1981) Cell 29, 125-139.

87.

88.

89.

90.

91.

92.

35
Larhamar, D., Schenning, L., Gustafsson, K., Wiman, K., Claesson, L.,
Rask, L. and Peterson, P. A. (1982) Proc. Natl. Acad. Sci. U.S.A. 79,
3687-3691.
Raelder, A., Arndt, R., Raedler, E., Jablonski, D. and Thiele, H. G.
(1978) Eur. J. Immunol. 8, 728-730.
Leifer, D., Lipton, S. A., Barnstable, C. J. and Masland R. H. (1989)
Nature 229, 303-306.
Norcross, M. A., Smith, R. T. and Shimizu, S. (1989) J. Immunol. 132,
833-838.
Gunter, K. C., Malek, T. R. and Shevach, E. M. (1989) J. Exp. Med.
159. 716-730.
Allison, J. P., McIntyre, B. W. Ridge, L. L., Gross-Pelose, J. and
Lainier, L. L. (1989) Fred. Proc. 93, 1593.

Chapter II

The heavy chain of the human B cell alloantigen HLA-DS has a variable

N-terminal region and a constant immunoglobulin-like region

by

Hsiu-Ching Chang
Tetsuya Moriuchi

Jack Silver

Department of Microbiology and Public Health
Michigan State University

E. Lansing, MI 98829

Reprinted by permission from Nature 305, 813-815.

Copyright (c) 19-- Macmillan Journals Limited.

36

37

The HLA-D region of the major histocompatibility complex (MHC) of man
encodes polymorphic glycoproteins found predominantly on the cell surfaces
of B cells and macrophages. These proteins mediate interactions, required
for the induction of immune responses, among cells of the immune system
and consequently are referred to as Ia (immune-response associated). Two
families of Ia molecules, DR and DS (also known as DC), have been defined,
the former analogous to the I-E (ref. 1) and the latter to the I-A
molecules of murine we (2-9). Both DR and DS molecules consist of two
non-covalently associated polypeptide chains with molecular weight 33,000
and 28,000 designated a.and B , respectively. The polymorphism of DR
molecules is due to structural variation in the small subunit, DRB, with
the large subunit, DRa, being constant in structure (5-7). In contrast,
both subunits, D80, and DS 3, are structurally variable when 08 allotypes
are compared (3). We have now isolated a cDNA clone from a DR7 cell line
that contains the entire coding sequence for the DSc subunit and have
compared its predicted amino acid sequence with that previously deduced
from a D80 cDNA clone isolated from a DR9,w6 cell line (8). This
comparison reveals that 10 of 11 amino acid differences are located within
the 011 (N-terminal) domain and that the 9:2 or immunoglobulin-like domains
are identical.

Poly(A)-containing RNA was isolated from a consanguineous homozygous
DR7 cell line (LG-10) and cDNA was synthesized by reverse transcription
using an oligo(dT) Primer followed by second-strand synthesis using DNA
polymerase. A cloned cDNA libray was constructed by insertion of the cDNA
into the Pet I site of pBR322 using the poly(dG).poly(dC) homopolymeric
extension method (9). Colonies containing cDNA were screened with a nick-

translated cDNA probe, pDCH1 (provided by C. Auffray), which encodes the

38
entire DS(DC)<1 subunit except the leader sequence and 17 amino acids of
the N-terminus (8). From an initial screening of 10,000 colonies, four
strongly hydridizing clones were detected, all of which had similar
restriction patterns (data not shown); and the one with the largest
insert, pDSa-12, was subjected to DNA sequence analysis (Fig. 1). The cDNA
insert of pDSa-12 consists of 921 nucleotides. Starting with the
methionine initiation codon, ATG, at nucleotide positions 20-22, there is
an open reading frame of 762 nucleotides encoding 259 amino acids and
terminating with the TGA codon at nucleotide positions 782-789. This is
followed by a 3' untranslated region of 123 nucleotides and a poly(A) tail
of 19 nucleotides. Nineteen bases upstream of the poly(A) tail is the
canonical polyadenylation signal AATAAA. The first 23 codons, amino acids
-23 to -1, encode the hydrophobic leader sequence. The amino acids
predicted from the next 25 codons are in nearly perfect agreement with the
previously determined N-terminal protein sequence of a 087 alpha subunit
(3). By convention we have given the DS allotype the same numerical
designation as the DR allotype of the cell line from which it was
isolated. Accordingly, DS molecules isolated from a DR7 cell line are
designated D87. The two discrepancies are probably due to protein
sequencing errors although the existence of a second D87 alpha-like
subunit cannot be ruled out. Thus, pDSa -12 contains the entire coding
sequence for the 087a subunit.

As has been shown previously (8), the DS(DC)<nsubunit displays strong
amino acid and DNA sequence homology to the DRc subunit (Fig. 2). However,
despite the overall high degree of homology between DRa and DSa (59$)
there are several regions of very low homology, notably the first 51 amino

acids of the coding region, (including the leader sequence) where only 12

39

Fig. 1. cDNA and predicted amino acid sequence of the D83 subunit
clone, pDSa-12. The nucleotide sequence of the coding strand and the
predicted protein sequence (single letter amino acid code) are shown. The
nucleotide sequence immediately follows the 15 guanine residues used to
insert the cDNA into the Pet I site of pBR322. The 23-amino acid signal
sequence is numbered -23 to -1 and the termination codon at nucleotide
positions 782-789 is indicated by an asterisk. The canonical
polyadenylation signal AATAAA located in the 3' untranslated region is
underlined. The restriction enzyme sites that were 5' end labelled and
used for DNA sequencing are shown. The nucleotide sequence of an extensive

series of overlapping fragment was determined by the procedure of Maxam

and Gilbert (18).

40

-23
M I L D K
G SASPCTGCCTTGGGAAGAAG ATG ATC CTA AAC AAA
m
-1 1
T V M S P C G G E D I
ACC GTG ATG AGC CCT TGT GGA GGT GAA GAG ATT
20
NLYQSYCPSGQ
AAC TTG TAG GAG TCT TAG GGT CCC GGC GAG
40
F Y V D L E R K E T V
TTC TAT GTG GAG GTG GAG AGG AAG GAG ACT GTC
60
F D P Q P A L T N I A
TTT GAG CCG GAA TTT GGA GTG AGA AAC ATC GCT
80
K R S N S T A A T N E
AAA CGC TCC AAC TCT ACC GCT GCT AC AAT GAG
,-_£
100
P V T L G Q P N T L I
CCC GTG AGA CTG GGT GAG CCC AAC ACC CTC ATC
120
V N I T W L S N G H S
GTC AAC ATC ACC TGG CTG AGC AAT GGG GAG TGA
EEEEB
140 150
L S K S D H S F F K I
CTC TCC AA ACT GAT CAT TCC TTC TTC AAC ATC
170
E I Y D G K V E H W G
GAG ATT TAT GAG TGC AAC GTG GAG GAG TGG GGC
190
P E I P A P M S E L T
CCT GAG ATT CGA GGA CCT ATG TGA GAG CTC A '
210
V G L V G I V V G T V
GTG GGC CTC GTG GGC ATT GTG GTG GGG ACC GTC
lflfll
230
A S R H Q G P L *

A
GCT

L
CTC

an:

TGG

TGT

130

ACT

CTC

E T
GAG ACT

L I
TTG ATC

M
ATG

L G
CTC GGG

CCT CTC

T F
ACC TTC

P L
CCT CTT

C

V
CTC TGT

R G
CGA GGC

A
GCC

L A
CTC CCC

10

TCT TAC

GGA GAC

H
CAC AGA

N I
AAC ATC

P S
TTT TCC

F P
TTT CCT

E T
GAG ACC

P S
CCT TCT

K H
AAA GAC

200
L G
CTC 666

220
R 8
COT TGA

L
CTG

180

TGG

T
ACC

TCT

160

GAT

GCT TCC AGA CAC CAA GGG CCC TTG TGA ATCCCAT CCTGAAAAAG AAGGTGTTAC CTACTAAGAG ATGCCT

GGGG TAAGCCGCCC AGCTACCTAA TTCCTCAGTA ACATCGATCT AAAATCTCCA TGGAAGGAAT AAATTCCCTT T

AAGAGAAAAA AAAAAAAAAC

21 IEEE:iI

Figure l

127

190

253

316

379

642

505

568

631

699

757

827

902

91

Fig. 2. Comparison of the protein sequences of DR and 08a subunits
predicted from cDNA clones. The DRa protein sequence was predicted from
the nucleotide sequence of a DRa cDNA clone previously isolated (19), and
the protein sequence of the DSa.subunit was predicted from the nucleotide
sequence of the pDSa-12 cDNA clone described here. The DRa and DSa
subunits may be organized into a leader (L) region, two extracellular
domains (a1 and a2), a connecting peptide (CP), a transmembrane (TM)
region and an intracytoplasmic (CY) region based on the studies by Korman
et al. (10), Auffray et al. (8) and Lee et al. (11). Homologous regions of
DR and DSc subunits are boxed. The homology between these two molecules is
21.71 in the leader sequence, 98.2% in the c 1 domain, 65.21 in the a2
domain and 72.51 in the combined CP, TM and CY regions. The numbering

system is based on the D8 sequence.

DR
DS

DR
DS

DR
DS

DR
DS

DR
DS

DR
DS

DR
DS

42

 

 

 

 

 

.. L --9‘ no

I L ‘—? 1‘ ~ cl
AISCVPVEEFFIIAVL AQESWA KEE VII-QAE
ILDKALMLGALALTTV PCGGED van VASYGVN

 

 

 

 

 

30 40

c1 '
IPEII.N'P D o‘s c M F p r - 'tnéarnfilEinnA K x E T naginnnnii
L Q s Y 6 P's c Q T a E 2 Y L E R K a r V‘W'K L P L

 

 

 

60 7o 80
_ H *1
c F A s E A G[A L A N I A v D A N L E I M r Y P I

 

'

V P P E V T V L
E

100 110 120
c2
T N S P V E L R E V L I C F I K F T ' ' ' ' ‘ V

 

32

130 140 150 160
L R K P V T T G V S E T V F LlP R.E D HIL R F H Y L P F L P S T E
L S H S V T E G V S E T S F L‘S K.S D H S F I S Y L T F.L P S A D

 

 

 

 

 

 

 

 

 

 

170 180 190

92 v-P'1'$---- cm'-———-——-o I
D v . R c F DnniisnfilL P T o '
JE?!E!E]E!KLV’E H.W o L D a P L L x unvnan>£:1 A. M s 1.&lEl

200 210 ' 220 ‘ 230
[a m . :+__.. .l

N V‘V(3AnL o L'T1IC:L v cnt1:1767innr'glglnc;n.n.x snsg'A.E RIR' I'L
T v v c A L c L s v c L v c I v v,g_ng L I R16 L R s v c A s a R Q o P L

 

 

 

Figure 2

93

of 51 (29$) of the residues are identical and amino acid positions 97-59.
where only 9 of 13 (31$) of the residues are identical. In addition, the
3' untranslated regions show no significant homology above random (30$).

When the nucleotide sequence of the pDSa-12 cDNA insert, which codes
for the D87a subunit, is compared with that of the pDCH1 cDNA insert,
which codes for a nearly complete DS9,6a subsunit, 11 nucleotide substitu-
tions as well as deletion of a codon at nucleotide positions 259-256 are
observed (Fig. 3). A comparison of the predicted protein sequences for the
DSa and D89,6a subunits (Fig. 9) reveals 11 amino acid sequence
differences, 10 of which are located whithin the N-terminal portion of the
molecule which by analogy to the DRo subunit has been designated the a1
domain (10,11). The u2 domain, which represents an immunoglobulin-like
region conserved among class I and class II histocompatibility antigens
(12-15), is identical for both the D89, 6 and D87 a subunits. The 11th
amino acid sequence difference is found in a region of the molecule
referred to as the connecting peptide and bridges the a2 domain and the
transmembrane segment. It is intrguing that 9 of the amino acid
differences are clustered at position 97-56, one of several regions, as
pointed out above, where DR and D8 subunits differ substantially. Thus,
there is even a suggestion of a hypervariable region. Indeed, recent
studies by M. Davis et al. (personal comunication) reveal that the murine
Aa subunits (homologues of D8) show a high degree of variability at these
same positions. Note also that of nine nucleotide differences within the
01 domain (the deletion of the codon at positions 259-256 is considered as
a single substitution event), seven result in amino acid differences. This
high precentage of non-silent substitutions suggests that there are fewer

structural and functional constraints on the c1 domain than for other

99

Fig. 3. Comparison of the DNA sequences of DS7 and D89, 6 a
subunits. The upper line represents the DS7c DNA sequence and lower line
the D89, 6 a DNA sequence (from ref. 8). The sequences are identical
unless otherwise indicated. Positions 259-256 repesent a deletion in D87 0
with respect to D89, 6c . The first N-terminal amino acid is indicated by

an arrow and the termination codon by an asterisk.

45

615AGGCTGCCTTGGGAAGAAGAIGATCCTAAACAAAGCTCTGATGCTGGGGGCCCTCGCCCTGACCACCGTGATGAG

 

CCCTTGTGGAGGILAAGACATTGTGGCTGACCACGTTGCCTCTTACGGTGTAAACTTGTACCAGTCTTAEGGTCC
I

CTCTGGCCAGTTCACCCATGAATTTGATGGAGACGAGGAGTTCTATGTGGACCTGGAGAGGAAGGAGACTGTCTG

 

U H“

GAAGTTGCCTCTGTTCCACAGACTTAGA---TTTGACCCGCAATTTGCACTGACAAACATCGCTGTGCTAAAACA
'1: “ ' ‘”‘

 

U I ”I!“

TAACTTGAACATCCTGATTAAACGCTCCAACTCTACCGCTGCTACCAATGAGGTTCCTGAGGTCACAGTGTTTTC
1T

 

CAAGTCTCCCGTGACACTGGGTCAGCCCAACACCCTCATCTGTCTTGTGGACAACATCTTTCCTCCTGTGGTCAA

 

CATCACCTGGCTGAGCAATGGGCACTCAGTCACAGAAGGTGTTTCTGAGACCAGCTTCCTCTCCAAGAGTGATCA
TTCCTTCTTCAAGATCAGTTACCTCACCTTCCTCCCTTCTGCTGATGAGATTTATGACTGCAAGGTGGAGCACTG
GGGCCTGGATGAGCCTCTTCTGAAACACTGGGAGCCTGAGATTCCAGCACCTATGTCAGAGCTCACAGAGACTGT
GGTCTGIGCCCTGGGGTTGTCTGTGGGCCTCGTGGGCATTGTGGTGGGGACCGTCTTGATCATCCGAGGCCTGCG
\o
e
TTCAGTTGGTGCTTCCAGACACCAAGGGCCCTTGTGAATCCCATCCTGAAAAAGAAGGTGTTACCTACTAAGAGA
\I

TGCCTGGGGTAAGCCGCCCAGCTACCTAATTCCTCAGTAACATCGATCTAAAATCTCCATGGAAGCAATAAATTC

CCTTTAAGAGA14615

Figure 3

75

150

225

300

375

450

525

600

675

750

825

900

mm

“.377: f”

96

Fig. 9. Comparison of protein sequences of D89, 6 and D87cxsubunits.
The predicted protein sequence of the D37a subunit (Fig. 1) is compared
with the predicted protein sequence of the D89, 6a subunit taken from
ref. 8. As the D89, 6 a cDNA clone did not contain the first 17 amino
acids, residues 1-17 are based on protein sequence analysis (9). Domain
organization is as in Fig. 2. There are 11 amino acid differences between

these two molecules, 10 in the a1 domain, the 11th in the connecting

peptide region.

 

Idr

97

“
MM
E
on
“
MM
all
on
(E

057
059.6

|¢m-Hv-—-

 

cg

 

”In...

"
ILL

mm

MM
I.I.
1.1.
mar.
MM
nuns
w:::
LL
nunu
I.I.
I.v.
r.r.
3.3.
r.:.
away
so.
NnIu
t.t.
r.r.
MM

T5
'6'-

on
‘6
SS
"
m::.
'V
SS
00
V'-
ILL

057
054.6

 

——-w~—4>I

”(:3

057
054.6

”A
“

 

MM
1.9.
MI
mull
mva.
l.v.
I...
0..
cr.r
MM
III"
on
V.v.
Isl
super»
ME
Ian;
1".
I.I"
l...
awnu
pl...
.L.L
v.1.
mus.
I...
as.)
r...
as.)
r.:.
w.V.
vu-n
V.V.
t.r.

weir

MM
[E

057
054.6

m1

 

 

057
054.6

rope
mul.
I.l.
vuvl
3.x.
even
nan.
rats
”cf:
"M
0.9.

"
II
[are
PP
tel.
MM
”NM
NM
Isl.
Imus
PP
Car.
00
nuns
65
MM

If.
UV
MM
CC
00
UV

09
DSLG

 

Figure 4

 

057
054.6

98
proteins. We believe that pDSa-12 and pDCH1 represent products of alleles
at the same locus because 126 of 127 nucleotides at the 3' untranslated
(UT) regions are identical. Although, at this time we cannot entirely rule
out the possibility that these clones are products of different loci that
have recently diverged (16, 17), it is difficult to see how they could
have accumulated so many differences in the a1 domain while maintaining a
relatively constant 3' UT region. Additional studies should resolve this
issue.

Thus, the 08a,subunit consists of an N-terminal domain with a high
degree of variability and a highly conserved immunoglobulin-like domain.
The constancy of the a2 domain may reflect a functional requirment such as
the ability to interact with the immunoglobulin-like domain of the D8 8
subunit, interactions analogous to those that occur between immunoglobulin
heavy and light chains. The role of variability in the a 1 domain,

however, remains to be determined.

1.

2.

3.

10.

11.

12.

130

REFERENCES

Allison J.P., L.E. Walker, W.A. Russell, M.A. Pellegrino, S. Ferrone,
R.A. Reisfeld, J.A. Frelinger, and J. Silver. Proc. Natl. Acad. Sci.
USA 75. 3953-9356 (1978).
Goyert, S.M. and J. Silver. Nature (London) 299, 266-268 (1981).
Goyert, 8.M., J.E. Shively and J. Silver. J. Exp. Med. 156, 550-566
(1982).
Bono, R. and J.L. Strominger. Nature (London) 299, 836-838 (1982).
Silver, J. and S. Ferrone. Nature (London) 279. 936-937 (1979)-
Silver, J. and 8. Ferrone. Immunogenetics 10, 295-298 (1980).
Goyert, 8.M., J.J. Hubert, R.A. Curry and J. Silver. J. Hum. Immun.
1, 161-175 (1980).
Auffray, C., A.J. Korman, M. Roux-Dosseto, R. Bono and J.L.
Strominger. Proc. Natl. Acad. Sci. USA 79. 6337-6391 (1982).
Moriuchi, T., H.C. Chang, R. Denome and J. Silver. Nature (London)
301, 80-82 (1983).
Korman, A.J., C. Auffray, A. Schamboeck and J .L. Strominger. Proc.
Natl. Acad. Sci. USA 79, 6013-6017 (1982).
Lee, J.8., J. Trowsdale, P.L. Travers, J. Carey, F. Grosveld,
J. Jenkins and W.F. Bodmer. Nature (London) 99. 750-753 (1982).
Orr, H.T., J.A. Lopez de Castro, D. Lancet and J.L. Strominger.
Biochemistry 18, 5711-5720 (1979)
Malissen, M., B. Malissen and B.R. Jordan. Proc. Natl. Acad. Sci.
USA 79. 893-897 (1982)

99

19.

15.

16.

17.
18.
19.

50
Larhammar, D., K. Gustafsson, L. Claesson, P. Bill, K. Wiman, L.
Schenning, J. Sundelin, E. Widmark, P. Peterson and L. Rask. Cell 30,
153-161 (1982).
Larhamner, D., L. Schenning, K. Gustafsson, K. Wiman, L. Claesson, L.
Rask and P.A. Peterson. Proc. Nalt. Acad. Sci. USA 79. 3687-3691
(1982).
Trowsdale, J., J. Lee, F. Grosveld, J. Bodmer and W. Bodmer. Proc.
Natl. Acad. Sci. USA 80, 1977-1976 (1982)
Auffray, C. EMBO J. 2, 121-129 (1983)
Maxam, A.M. and W. Gilbert. Meth. Enzym. 65, 999-560 (1980)
Kajimura, Y., H. Toyoda, J.E. Shively, S. Goyert, J. Silver and

K. Itakura. DNA (submitted).

Chapter III

THY-1 cDNA sequence suggests a novel regulatory mechanism

By

Tetsuya Moriuchi, Hsiu-Ching Chang,

Roger Denome and Jack Silver

Department of Microbiology and Public Health
Michigan State University

E. Lansing, MI 98829

Reprinted by permission from Nature, 301, 80-82

Copyright (c) 19-- Macmillan Journals Limited.

51

 

I]. A

52

Thy-1 was originally defined in mice as a cell-surface alloantigen of
thymus and brain with two allelic forms, Thy-1.1 and Thy-1.2 (ref. 1).
Subsequently, the Thy-1.1 alloantigenic determinant was identified in rats
(2). In both species, Thy-1 is present in large amounts on thymus and
brain cells (3) and in smaller quantities on the fibroblasts (9),
epidermal cells (5). mammary glands (6) and immature skeletal muscle (7).
In many of these tissues the level of Thy-1 expression changes
dramatically during cell differentiation. The molecules expressing the
Thy-1 antigenic determinant have been isolated from rat and mouse brain
cells and have been shown to have a molecular weight of 17.500 (ref. 8).
One-third of the Thy-1 molecule is carbohydrate and the remainder is a
polypeptide of 111 amino acids whose sequence has been fully determined
(9). We report here the isolation and characterization of a cDNA clone
encoding the rat thymus Thy-1 antigen but find that the DNA sequence ends
prematurely at a position corresponding to amino acid 103. It appears to
be a complete transcipt, however, as the last codon is followed directly
by a poly(A) tract.

Poly(A) containing RNA was isolated from W/Fu rat thymocytes, and
cDNA was synthesized by reverse trancription using oligo(dT)12-18 primer.
Double-stranded cDNA was synthesized as described previously (10). A
cloned cDNA library was constructed by inserting the total cDNA population
into the Pst I site of pBR322 using poly(dG).poly(dC) homopolymeric
extensions (11). Colonies containing cDNA were screened (12) with a
synthetic 32P-labelled oligodeoxynucleotide (17-mer) mixture composed of
all 32 possible sequence permutations corresponding to amino acids 82-87
of Thy-1. Two hydridizaton-positive clones were isolated from 10,000

colonies. One clone (pT69) contained cDNA encoding the Thy-1 antigen and

53
its entire sequence was determined by procedure of'Maxam and Gilbert (13).
The overall structure of the cDNA clone pT69 and the sequencing strategy
used are shown in Fig. 1.

The DNA sequence contains a 95-base 5'-untranslated region followed
by 57 bases coding for a presumptive leader peptide of 19 amino acids
starting at the first methionine codon (numbered -19 in Fig. 2). The
leader sequence exhibits several features characteristic of leader peptide
present at the amino terminus of membrane bound proteins (19); it contains
many hydrophobic amino acids (11 non-polar residues) and terminates in a
residue with a small neutral side chain (glycine). The remainder of the
predicted amino acid sequence is in agreement with the published amino
acid sequence for rat brain Thy-1 (9). There are, however, several unusual
features at the 3' end of the Thy-1 cDNA clone. The DNA sequence ends
prematurely at a position corresponding to amino acid 103, 8 amino acids
earlier than predicted from the protein sequence, but followed directly by
a poly(A) tract. There is, however, no termination codon. Furthermore, a
presumptive polyadenylation signal, 5'-AATAAA-3'(ref. 15), which is part
of the coding sequence, is found 12 nucleotides upstream from the poly(A)
tract. Two alternative explanations for these unusual features at the 3'
end can be proposed: (1) a deletion of the sequence encoding the
C-terminal peptide occurred during the cloning procedure while retaining
the sequence corrsponding to the poly(A) tract, or (2) the AATAAA sequence
at positions 293-298 was recognized as a polyadenylation signal in yiyg
and position 310, which is 12 nucleotides downstream from the AATAAA
sequence, served as a polyadenylation site.

The first explanation cannot be excluded although such a cloning

artefact has not been described before. The second explanation is a very

59

Fig. 1. Partial restriction map and strategy for sequencing Thy-1
cDNA clone, pT69. The entire cDNA was cleaved at selected sites with
restriction endonuclease and fragments corresponding to the insert were
purified by acrylamide gel electrophoresis and electroelution. Fragments
were treated with calf intestinal or bacterial alkaline phosphatse and
32P-labelled at both 5' ends with T9 polynucleotide kinase as described
(13). Alternatively, some fragment were 32P-labelled at both 3' ends with
DNA polymerase (26). Double-end-labelled fragments were separated by
acrylamide gel electrophoresis, electro-eluted and subjected to partial
chemical degradation sequence analysis as described by Maxam and Gilbert
(13). Partially cleaved fragments were separated on 8, 15 and 20$
acrylamide gels. The extent of sequence determined from each fragment is
indicated by the length of the arrow. The closed circles and vertical
lines at one end of each arrow indicate labelling at the 3' end or 5' end,

respectively. The sequencing strategy used here allowed us to sequence

both DNA strands completely.

55

TH :_( 4.

a”?

400
l

 

1H 05... L

300
l

 

 

a... t 1

200
n

in new A

 

 

Ounu awn. L

 

 

 

rumor o A.

.32 A. .r

Figure 1

56

Fig. 2. Primary structure and predicted amino acid sequence for
pT69. Amino acids are numbered relative to the amino-terminus of the
protein sequencec as detertmined by Campbell et a1 (9). The 19-amino acid
leader sequence 5' to the codon for the first amino acid is numbered -19
to -1. The hexanucleotide sequence, AATAAA, at the 3' end is underlined. A

29-base poly(A) tail directly follows codon 103.

Primary structure and predicted amino acid sequence for pT69

ile
ATC

arg
AGG

asp
GAC

ser
ACC
50

val
GTT

phe
TTT

asp
GAC

asn

ser
ACC

val
GTG

cys
TGC

1eu
CTG

pro
CCC

ile
ATC

tyr
TAC

lys

57

ile thr 1eu 1eu 1eu ser val
TCA CTC

ATC

ile
ATC

20
erg
CCT

thr
ACC

glu
GAG

lys
AAC

met
ATG

100
thr

AAT AAA ACT

ACT

ser
ACC

his
CAT

arg
CGA

his
CAC

val
GTC

cys
TGT

ile
ATC

CTC

1eu
CTC

glu
GAG

glu
GAG

thr
ACT

70
Ian
CTT

glu
GAA

asn
AAT

CTG

thr
ACA

asn
AAT

lys
AAC

tyr
TAC

thr
ACT

1eu
CTT

103
val
CCT

CTT

ala
GCC

asn
AAC
40

lys
AAC

arg
CGC

1eu
CTA

arg
CGA

(A)

cys
TGC

thr
ACC

lys
AAC

ser
TCC

ala
CCC

val
GTC

29

10
Ian
CTG

asn
AAC

his
CAC

srg
CGC

asn
AAC

90
ser

TCG

Figure 2

1eu
TTG

val
CTC

1eu
TTC

val
CTC

val
GTC

phe
TTC

sly
soc

-19
met asn pro val
CTGCAAGCTACGGGAGCCCAGACCCAGGACGGAGCTATTGGCACC ATC AAC CCA CTC

gln met ser

CAC

asn
AAC

pro
CCC

1eu
CTG

60
asn
AAC

thr
ACC

gln
CAG

ATC

gln
CAG

ile
ATC

ser
TCG

leu
CTT

thr
ACC

asn
AAT

TCG

asn
AAC

30
gln
CAG

sly
ccc

phe
TTC

lys
AAC

pro
CCC

arg
CGA

1eu
CTT

his
CAT

thr
ACC

ser
ACT

asp
CAT

thr
ACA

-1

sly
CGA

arg
CGA

glu
GAG

1eu
CTG

.asp

GAC

80
glu
GAG

ser
ACC

l
gln
CAG

1eu
CTG

phe
TTC

sly
coo

arg

sly
ccc

ser
TCC

58
attractive model for a mechanism regulating rapid changes of Thy-1 gene
expression during cell differentiation and is similar to that observed in
other systems. For example, during differentiation, B lymphocytes undergo
a shift from expression of membrane-bound IgM to secreted IgM. It has been
shown that the transcription unit for secreted and membrane-bound IgM
heavy chains contains two separate poly(A) addition sites. The basis for
selection between these two gene products apparently resides in the choice
of polyadenlyation site of the precursor RNA followed by differential
”splicing out" of introns (16, 17). Expression of the late adenovirus
transcription unit is alo controlled through differential poly(A) site
selection (18). A similar mechanism may be responsible for regulating
Thy-1 gene expression. Differential poly(A) site selection may control the
expression of the C-terminal coding segment, which is apparently the
region of Thy-1 responsible for membrane integration (9) and, in the
presence of differentiation-inducing factors such as thymic hormones,
complete expression of cell-surface Thy-1 may be induced. Komuro and Boyse
demonstrated that Thy-1 negative precursor cells in the bone marrow and
spleens of mice can be induced to differentiate in yitgg into Thy-1
positive T lymphocytes within 2 h of incubation with crude thymus extract
(16). Our hypothesis might explain the mechanism responsible for this
rapid induction of Thy-1 on the cell surface.

DNA and amino acid sequence analyses have revealed homology among the
major histocompatibility antigens, BZ-microglobulin and immunoglobulins,
and suggests that all three have evolved from a common ancestral gene
encoding a primitive domain (20-22). Homology between Thy-1, which has a
domain-like structure including a disulphide loop of the appropriate size,

and immunoglobulins has been noted at the amino acid sequence level (23).

59
We have compared the nucleotide sequences of Thy-1 cDNA amd mouse 11 light
chain variable region (29) which was shown to have the highest degree of
amino acid sequence homology to Thy-1 (23). When the sequences are
aligned as shown in Fig. 3, overall homology between the neculeotide
sequence of Thy-1 and the variable region is 36$. This sequence homology
supports the hypothesis that Thy-1 and immunoglobulin have evolved from a
common ancestral gene. Furthermore, our observation concerning the Thy-1
cDNA clone, pT69, suggests that the Thy-1 gene may even have inherited a
remnant of a mechanism involved in regulating immunoglobulin gene
expression, namely differential polyadenylation. Additional experiments
aimed at defining the genetic organization and regulating elements of
Thy-1 expression are in progress. Thus Thy-1 which has long been used
primarily as a marker for T lymhpocyte differentiation (25), may itself

represent an intriguing system for study of gene expression.

60

Fig. 3. Comparison of the DNA sequences of mouse V11 (MOPC109E) and
rat Thy-1. DNA sequences were aligned with the amino acid squence
alignment of Williams et al. (23), except for certain regions, indicated
by parentheses, that were realigned to increase homology. Identical amino
acid residues are underlined and nucleotide sequence homologies are
indicated by vertical lines. Gaps inserted to maximize homology are
indicated by dashes. The hypervariable regions (HV1, HV2 and HV3) are
lined above the amino acid sequence and the Variable-Constant (V-C)
junction (J region) is indicated by a dotted line. It is interesting that
although the overall nucleotide sequence homology between Thy-1 and Vn1 is
36$, comparisons of small segments reveal considerable variation in the
degree of homology. The homology is 91$ in positions 1-20, 19$ in the
position 21-39 (sequences corresponding to the first hypervariable region
1), 91$ in positions 35-85, 38$ in positions 86-92 (HV3) and 29$ in the

positions 93-103 (sequences corresponding to the J region).

61

ccH ca

Hm> mm¢ can ask emu mm< new mom has cum and mHu AHo new Hm> mu< sea :Ho eNm aux uNm am<

uww w<w uw¢ su¢ 9<< ova uw< <u< coo a<< oflw woo cos can <wo _ 44% www
_ _

one au< can <44 uu< <uo <60 you 099 new ooh a<u o<< uu< 049 009 <90 you was may a<a <H<

.H.|no w w am.— ..3 2:. LHHu a3 a8 2: Hg .5. on. 5: uluom as. 8H. 3H :2 .5 o

eoHuooanuu.> H >u

on an ac
ado :Ho mm¢ mag ugh ask one mm¢ eH< seq ugh mmm.ue> mag emu mam mu< mm¢ mum umm.mmm mm< Hm>
cue udu adu ;«¢ ow¢ uu¢ oak u<¢ Duo <90 90¢ 990 can u<¢ ua¢ aha oco u¢o Hu¢ 099 990 o<¢ use

"11:3

m¢u H¢u Q¢u 90¢ u¢u ¢UU 600 40¢ 08¢ 00¢ 080 000 . Boo u¢¢ U<U duo H¢¢ use 008 Due
30

awn: =Hu use oHu «He mHu use oHH use :oH nH< - . HoH< oaH uo< aHu oHH sou tom aHu
on
uu< tom uu< «as use .Hm oHu one Ho> NHu ooH use NHu Lou ooH
uuu wow uwu use uu< u<u uow uww saw uwwauauu . n n n - u - - uu< uuw owe uwu
_
«us one <u< <4u.mmmzmmm.uuu sue you <uu u<< u<< uu< you you «we <su auu Hue use
new «so uu< . - n aH<o m H o> amm_oum oH< sue one oo< umm aHu aHu oHH aoH NHu use use

 

N >=

us on
34 ea ooH now use oHu on oHu oHH one sou soc use oo< one =Hu oH:
HHquw .uc wee wow was uqu a<u ueu um“ uwu use u<< uw< u<< w<d u<u aeu
A .
”ea uuu uua u<< uuu u<u u<< au< Hue <u< emu sou uuu uu< uu< euu
Ho> one oo< oH< umm,oa< . now use use Ho> oH< aHu tum tom mom

 

H >=
oH oH H
uu< one soc sou uu< ooH oo< oHu oo< Ho> :oH oau oH< any gas uo< oHH H<> uu< :Hu
auu uua ueu use euu emu u<< ueu . n . u<< use use uuu uuu uu< uau uu< ua< uau uuc u<u

Z __ I. _ _ _ _ : _ _ __ _ _ __ :.
uuu aua au< uuu <u< uau <u< <<u you you «up <u< uu< uau cuu you see u<u uu< uuu emu you u<u
a: sum use ooH use Has any =Hu aHu one you use one ooH oH< tom oHu oHu one Hos Ho> oH< oHu

«lush Una

. H4> ammo:

H-aou one

H«> ammo:

Hnaau one

H4> ammo:

«15:8 umm

H4> muse:

Hussy um“

H4> omen:

Figure 3

10.

11.

12.

13.
1".

REFERENCES

Reif, A.E. and Allen, J.M.V. J. Exp. Med. 120, l8134433 (1961!).
Douglas T.C. J. Exp. Med. 136, 105'4-1062 (1972).
Reif, A.E. and Allen, J.M.V. Nature 209, 523 (1966).
Stern. P.I. Nature 2146, 76-78 (1973).
Scheid, M., Boyse, R.A., Carswell, R.A. and 01d, L.J. J. Exp. Med.
135. 938-955 (1972).
Lamon, V.A., Unger, M. and Dulbecco, R. Proc. Natl. Acad. Sci. USA
75, 6093-6097 (1978).
Lesley, J.F. and Lennon, V.A. Nature 268, 163-165 (1977).
Kuchel, P.W., Campbell, D.G., Barclay, A.N. and Williams, A.F.
Biochem. J. 169, 1811-1”? (1978).
Campbell, D.G., Gagnon, J., Reid, B.M. and Nilliam, A.F. Biochem. J.
195, 15-30 (1981).
Efstradiadis, A., Kafatos, F.C., Maxam, A.M. and Maniatis, T. Cell 7.
279-288 (1976).
Villa-Komaroff, I., Estratiadis, A., Broome, S., Lomedico, P., Tizard,
R., Naber, S.P., Chick, V1.1... and Gilbert, N. Proc. Natl. Acad. Sci.
USA 75, 3727-3731 (1978).
SL338, S.V., Wallace, R.B., Hirose, T., Kawashima, B.R. and Itakura,
K. Proc. Natl. Acad. Sci. USA 78, 6613-6617 (1981).
Maxam, A.M. and Gilbert, N. Meth. Enzym. 65, 1499-560 (1980).
Inouye, M. and Halegoua, S. CRC Crit. Rev. Biochem. 7, 339-371
(1980).

62

15.
16.

17.

18.

19.

20.

21.

22.

23.

211.

25.

26.

63
Proudfoot, N.J. and Brownless 6.6. Nature 263, 211-21“ (1976).
Rogers, J., Early, P., Cater, C., Calame, K., Bond, M., Rood, L. and
wall, R. Cell 20, 303-312 (1980).
Early, P., Rogers, J., Davis, M., Clame, K., Bond, M., Hall, R. and
Hood, L. Cell 20, 313-319 (1980).
Nevins, J.R. and Hilson, N.C. Nature 290, 113-118 (1981).
Komuro, K. and Boyse, E.A. Lancet 1, 7flO-7h3 (1973),
Peterson, P.A., Cunningham, R.A., Berggard, I. and Edelman, G.H.
Proc. Natl. Acad. Sci. USA 69, 1697-1701 (1972).
Steinmetz, M, Frelinger, J., Fisher, D., Hunkapiller, '1‘. Pereira, D.,
Neissman, 8.M., Uehara, H. Nathenson, S. and Hood, L. Cell 2“,
125-13" (1979).
Larhamer, D.L., Schenning, L., Gustafsson, K., Wiman, K., Claesson,
L., Rask, L. and Peterson, P.A. Proc. Natl. Acad. Sci. USA 79.
3687-3691 (1982).
Nilliams, A.F. and Gagnon, J. Science 216, 696-703 (1982).
Bernard, 0., Hozumi, N. and Tonegawa, 8. Cell 15, 1133-11uu
(1978).
Raff, H.C. Immunology 19, 637-650 (1970).
Maniatis, T., Pritsch, B.R. and Sambrook, R. Molecular Cloning, 115

(Cold Spring Harbor Laboratory, New York, 1982).

Chapter IV

THE NUCLEOTIDE SEQUENCE OF THE HOUSE THY-1 GENE

Hsiu-Ching Chang and Jack Silver

Department of Microbiology and Public Health
Michigan State University

E. Lansing, MI “882M

Running Title: Genomic cloning and DNA sequence analysis

of the mouse Thy-1.2 gene

64

SUMMARY

The mouse Thy-1.2 gene was isolated from a CS7BL/6 cosmid library.
The nucleotide sequence of the mouse Thy-1.2 gene has been determined from
an B-Kb long Eco RI fragment. The predicted amino acid sequence indicates
that mouse Thy-1.2 antigen contains a 19 amino acid leader peptide and the
112 amino acids previously reported by protein sequence analysis, plus 31
extra amino acids at the carboxyl terminus. These 31 amino acids contain a
stretch of 20 amino acids, at positions 12u—1u3, which is highly
hydrophobic. Northern blot analysis of RNA from mouse tissues indicates
that the sequence coding for these 31 amino acids is present on poly(A)
containing RNA of brain and thymus tissues. The entire coding sequence is
distributed among three exons, encoding for amino acid residues -19 to -8,
-7 to 106 and 107 to 1H3, respectively. The predicted amino acid sequences
and gene organizations of mouse, rat and human Thy-1 were compared. The
sequences are more highly conserved between rodents than between rodents
and human. The carboxyl termini are highly conserved among these three
species, with homologies of more than 85 i. The highly conserved nature of
the carboxyl terminus suggests a functional role for this fragment of the

molecule.

65

il

 

 

 

 

INTRODUCTION

The Thy-1 antigen was originally defined in mice as a cell surface
alloantigen of thymus and brain with two allelic form, Thy-1.1 and
Thy-1.2 (Reif and Allen, 19611). Subsequently, The Thy-1.1 determinant was
identified in rat (Douglas, 1972). In both species, Thy-1 is present in
large amounts on the surface of thymus and brain cells, with about 600,000
molecules per cell on rat thymocytes, and equivalent amounts on brain
cells . Thus Thy-1 is probably the most abundant surface glycoprotein of
both cell types (Acton et al., 197"). In rodents, Thy-1 is also expressed
in small amounts on fibroblasts (Stern, 1973). epidermal cells (Scheid et
al., 1972), mamary glands (Lennon, 1978) and immature skeletal muscle
cells (Lesley and Lennon, 1977). In many of these tissues the level of
Thy-1 expression changes dramatically during cell differentiation.

The molecules expressing the Thy-1 antigenic determinant have been
isolated from rat and mouse brain cells (Letarte-Muirhead et al., 1975,
Trowbridge et al., 1975). Thy-1 has been shown to have a molecular weight
of 17,500, of which one third of the molecule is carbohydrate, and the
remainder is a polypeptide of 111 amino acids in rat (Campbell et al.,
1981) and 112 amino acids in mouse (Williams and Gagnon, 1982).
Glycoproteins that are structurally related to rat and mouse brain Thy-1
have been purified from the brain of’human (Cotmore et a1, 1981), dog
(Mbkenzie and Fabre, 1981), chicken (Rostas et al., 1983) and frog
(mnsour and Cooper, 1989), and even from invertebrate squid (Williams and

Gagnon, 1982). The expression of I‘m-1 seems to be conserved in the
66

67
evolution'of neuroral cells and fibroblasts; however, the expression of
Thy-1 on lymphoid cells varies in different species (Dalchou and Fabre,
1979). Thus, although Thy-1 is the most abundant surface molecule on rat
and mouse thmocytes, the human homolog cannot be detected unambiguously on
human thymocytes. Furthermore, Try-1 is present on mouse T lymphocytes but
absent from B cells. In constrast, Thy-1 is absent from rat T cells, but
present on rat bone marrow cells (Hunt et al., 1978).

The Thy-1 molecule was previously thought to have an unusual mode of
integration within the membrane. It can be heavily labelled in the
thymocyte membrane with an affinity label which labels only the
hydrophobic section of the membrane proteins (Owen et al., 1980). Previous
biochendcal studies had failed to reveal any extended sequence of
hydrophobic amino acid that might function as a transmembrane segment.
Furthermore, the C-termdnal peptide had hydrOphobic preperties despite a
lack of a hydrophobic sequence (Kuchel et al., 1978). Therefore, it was
suggested that the hydrOphobic preperties of the C-terminal peptide were
due to the linkage of lipid which anchored Thy-1 on the cell membrane.
However, recent studies of Thy-1 cDNA clones by T. Moriuchi (personal
communication) have shown that the rat Thy—1 molecule is 192 amino acids
long with a hydrophobic segment at the carboxyl terminus. This suggests
that Thy-1 in fact is an integral protein anchored on the membrane via a
hydrophobic segement.

Thy-1 has long been used as a mouse T cell marker. Despite much
speculation, however, there is no clear understanding of its biological
significance. On the basis of amino acid sequence, secondary and tertiary
structures (Cohen et al., 1978) and nucleotide sequence (Moriuchi et al.,

1983), Thy-1 shows striking structural homologies with Ig domains, and

68
with other members of the Ig superfamily (Larhammer et al., 1982, Korman
et al., 1982). This has led to the hypothesis that Thy-1 is likely to be a
‘ primordial domain from which all the members of the Ig superfamily might
have evolved. It would therefore be interesting to analyze the evolution
of the Thy-1 gene in relation to other members of Ig superfamily.

In this study, we report the isolation of mouse Thy-1 genomic clones
from a CS7BL/6 comsid library, and DNA sequence analysis of the Thy-1
gene. The predicted amino acid sequences and gene organization of mouse,
rat and human genes are compared. The protein sequences and gene
organization of the Thy-1 genes are well conserved among the three

species, with homologies in the third exon of greater than 85$.

RESULTS

Isolation and charavterization of mouse Thy-1 genomic clones.
A cosmid library of mouse C57BL/6 DNA was constructed in the c2RB vector
as described (Bates and Swift, 1983) with a small modification. The
C57BL/6 DNA was partially digested with San 3A to an average size of no
Rb, and dephosphorylated before ligation with Bam RI-Sma I digested 0283
DNA. Ligated DNA was then packaged in yitgg, and infected E. 29;; strain
1096 was used to construct a cosmid library (Figure 1). To isolate a mouse
Thy-1 genomic clone, the cosmid library was probed with the P-labelled
insert of the rat thymocyte Thy-1 cDNA clone, pT69 (Moriuchi et al.,
1983). After screening 175,000 clones, three independent clones were
isolated, cT10, cT3u and cTSF. The partial restriction enzyme maps of
these clones are presented in Figure 2A. The cT10 clone shares only 9 and
2 Kb of 3' end of the Thy—1 gene with cT3u and cTSF, respectively. The
total insert of these clones spans about 75 Kb DNA along the mouse DNA.
The clones were first identified as Thy-1 containing clones by
Southern blot analyses, since the mouse Thy-1 gene had been shown to be a
single copy gene, and to be located in an 8 kb Eco RI fragment in
Southern blot analysis of total cellular DNA (data not shown). The
Southern blot analysis of these cosmid clones shows that each contains
only one hybridizing band. The cT1O clone contains an 8-kb Eco RI band of
the exact size as indicated by the cellular DNA analysis. The Eco RI
fragments of clones cT3u and cT10 were subcloned into the Eco RI site of
pBR322. The Thy-1 containing subclones, ch311 and ch108, were subjected

69

70

Figure 1. Scheme for the use of c2RB vector in the construction of a
mouse CS7BL/6 cosmid library. After digestion with Bam BI and Sma I, c2RB
DNA is ligated to phosphatase-treated mouse C57BL/6 DNA insert, which was
partially digested with San 3A to an average size of approximately 110 Kb.

Note that Bam HI and San 3A have the same cohesive GATC ends. A high ATP
concentration is used during the ligation reaction to prevent the Joining
of the Sma I blunt ends. _I_n vitro packaged DNA was used to infect E. coli

10116 in order to construct the mouse cosmid library.

71

to... (In. cinch
‘32:... a... a... .23....o
.33: 3!...
.o.. H....u .. 30:03.;
....... on... .4

.o......! .H........ u...
.3......
N

P 2.2.... ¢....

 

 

0.8 uCId .0. E .I.

al..

to...:.........

 

AAOQINO 33': :0...

H 20 .a .3... a: In. a. so. 3.... 2.0.0.».
on... no...

a;

He: al....

(80 :0...

a clone 2 In.

 

‘ Figure 1

72

Figure 2. Restriction enzyme maps of Thy-1 genomic clones.
(A) Partial Eco RI digestion maps of Thy-1 containing clones, cT10, cT3ll
and cTSF are shown. Clone cT10 shares only 11 and 2 kb of the 3' end of the
Thy-1 gene with cT311 and cT5F, respectively, making the total insert of
these clones span 75 Kb of DNA along the mouse genome.
(B) Partial restriction map of the Thy-1 containing genomic clone, ch108,
and the intron-exon organization of the mouse Thy-1 gene. The thicker
line, and the boxed area denote the coding region of Thy-1. The amino
acids were numbered relative to the amino-terminus of the protein sequence
as determined.
(C) The strategy for sequencing the Thy-1 gene. The pBR322 clones
containing the Eco RI insert of Thy-1 were cleaved at selected sites with
restriction endonucleases as indicated. The fragments were isolated and 5'
end-labelled with T11 polynucleotide kinase, or 3' end-labelled by filling
in with DNA polymerase I. Double-end labelled fragments were digested with
restriction enzymes, separated on the agarose gel, eluted, and then
subjected to DNA sequencing by Maxam and Gilbert's method. The extent of
sequence determined from each fragment is indicated by the length of the
arrow. The solid circle and vertical lines at one end of each arrow
indicated labelling at the 3' or 5' end, respectively. The DNA sequence
was completely determined by sequencing either both complementary strands

or the same strand at least twice, starting with different cleavage sites.

Restriction Enzyme Maps oi Mouse Thy-1

 

73

Genomic Clones

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

EcoR 1 E Bean 1 CT10
A. e ,% fi‘r E ~—-*5kb
/ ‘ i f ‘ :r—CT34
/ I \ CTSF
/ \
/' I \
/ I \
’ \
/ I \
/
\
/ I \
/ \
, ’ Xha I Hind m I Xho I B BamBem \
3 g 1 HI. . -1 ll 1 A, ch 108 '-—'
' ' T" '\ 0 5kb
& t \ -
851 I Sat, at 8st I
/I l \ \ \
II I \ \ \
// I \ \ \
// I \ \ \.
/ / i \ \ \
/ / l \ \ \
/ I I \ \ \
/ / \ \ \
/ L _\
f—w. A ,‘f—L g—fl—AZEi—fl—Muu
- 19 -7 -8 '106 107 143
fiF—‘fi fl 9'3
I 107 14a "
Pvu II Pvu II
AV. 1 ""
HIM! PM! Ava] L08 U A”! Aral 1_0ib T ‘
r— i G I ‘ li UETW'NU I '1 lfi * or. 0'
sux Pqu 3m! am I In! PHI :1 ".1 on! .IIHI IonHI AATAAA
; P 4. H 4.; HF fiz—H‘fi jh———u-Q ‘fi :; —*,
¢--———.H 9.: :
(m..___., e : : ; = =% ; 2 - ———-—i'
' L . L.—-——-'
v Hfiffi-‘u fi

Figure 2

7A
to DNA sequence analysis following the strategies illustrated in Figure
20. The sequence data enabled us to definitively identify these clones as

Thy-1 containing clones.

Structure of the mouse Thy-1 gene.

Comparison of the nucleotide sequences of mouse genomic and rat cDNA
clones, and comparison of the predicted amino acid sequence with the
protein sequence previously reported (Williams and Gagnon, 1982)
established the intron-exon organization of the mouse Thy-1 gene (Figure
23). The coding region of Thy-1 is divided into 3 exons by two introns of
sizes 590 bp and 388 bp, respectively. As is often the case, the first
exon codes for part of the 5' untranslated region of the mRNA and the
first 12 amino acids of the signal peptide, amino acids -19 to -8,
followed by a 590 bp intron. The second exon contains most of the amino
acids of the mature protein, amino acids -7 to 106, followed by a 388 bp
intron. Both introns were occurred by RNA spliced between the first and
second base of the Junctional codons according to GT/AG rule (Nevins,
1983). The rest of the predicted amino acid residues, positions 107-193,
followed by the termination codon, TGA, and the 3' untranslated region of
mRNA reside in the third exon. Since the only difference between the two
mouse allelic forms, Thy-1.1 and Thy-1.2, is amino acid residue 89, in
which Thy-1.1 is arginine and Thy-1.2 is glutamine, the gene we have
cloned is the mouse Thy-1.2 gene. This is consistent with the serological

typing of the CS7BL/6 Thy-1 gene. The predicted protein sequence shows

that the mouse Thy-1.2 gene contain a 19 amino acid presumptive leader
peptide, 112 amino acid residues which fit completely with the protein

sequence reported by Campbell et al. 1981, plus 31 additional amino acids

75

Figure 3. Sequence of the mouse Thy-1.2 gene. The nucleotide
sequence and predicted protein sequence (three letter amino acid code) of
the Thy-1 gene are shown. Exons are underlined, with a thin line for the
untranslated region of mRNA, and with a thick line for the protein-coding
sequence. The initiation codon is boxed and the termination codon is

indicated by an asterisk. The AATAAA sequences are boxed.

76

 

101
.htnokthlfinru use ' I.” I.”

'TIII’II\'!III’.‘.'I\'.
*

   

   

’ "va\'1\"

_1--1H\EWi~

     

    
 
 

 

ltfitlflvlitt

 

1 L'! 1255’}! m’.’.‘.‘.‘1 l'! l \

     

 

Figure 3

77
(Figure 3). The last 20 amino acids at the carboxyl terminus (positions

1211 to 1113) are highly midrophobic and include ten leucine residues.

The gene organization of'mouse, rat and human Thy-1.

Since the nucleotide sequences of’rat and human Thy-1 genes have been
determined (Seki et al., 198“), we were able to compare the gene
organization of mouse, rat and human Thy-1 genes (Figure 1l). The
distribution of the coding sequence among the three exons is identical for
all three species, coding for amino acids, -19 to -8, -7 to 106 and 107 to
193. All the introns follow RNA splicing occurs between the first and
second base of the Junctional codons accroding to GT/AG rule, with
variations in length among these three species. This indicates that the
gene organization of Thy-1 is highly conserved among these three species,
with very well conserved exon structure, and slightly less conserved
intron structures. However, there are some features conserved in the rat
and human thy-1 genes, but not in the mouse Thy-1 gene. Namely, there is a
deletion of one amino acid at position 29, and there are two AATAAA
sequences in the 3' untranslated region of both human and rat Thy-1 gene
located around 550 bp and 1100 bp downstream of the termination codon,
TGA. However, there is only one AATAAA sequence in the 3' untranslated
region of mouse Thy-1 gene, located at a position corresponding to the

second AATAAA sequence of the rat and human gene.

Northem blot analysis of RNA from mouse and rat tissues
To determine whether the 31 extra amino acids predicted from the cDNA and

gene structure are present in the mRNA, northern blot analyses were

78

Figure u. Comparison of the gene organization of mouse, rat and
human Thy-1. The intron-exon organization of rat and human Thy-1 genes
were determined by Seki et al., 1989. The sizes of exons are shown by the
amino acid residues, and the sizes of introns and 3' untranslated region
are shown by the nucleotide base pairs. The locations of the AATAAA

sequences are shown in non-proportional scale.

79

L]: Li: Df

.00 v-

0.0 90

use...

03252.2. o. .32...» noses.

IN ‘0. ‘0. H50

HIIIililJ a: co 1I|I|J use as

Id ‘0. .0. H..

_I|||.|II|._ Do. U. 'nIIIIJ Oi. U.

. iH. 3., H3 2..
[ 1
a. a. >>4>L

mwmcnm e

>>¢>>>

H... a...

. .
>>4>>>

, 3: so

H3...-

>>4>>.

>3»:

>>a$>§

80
performed. A northern blot of poly(A) containing RNA from mouse brain,
liver, spleen and thymus, was hybridized with a Pvu II digested fragment,
PV3, which codes for amino acid residues 107 to 193 (Figure 28). Only one
species of mRNA was detected in brain and thymus tissues (Figure 5A). This
demonstrates that the 31 extra amino acids are transcribed. The same blot
was then hybridized with the entire insert of ch108 two months later. The
same bands were detected again (data not shown). The northern blot of
total RNA from rat thymus and form several human B cell lines was also
probed with the insert of pT6h. There is also only one species of Thy-1

mRNA in the rat thymus, with an approximate size 1.7 Kb.

Comparison of 5' untranslated region

To determine where the Thy-1 gene begins, the nucleotide sequences of the
5' ends of the rat cDNA and rat and mouse genomic clones were compared
(Figure 6). The nucleotide sequences of the first coding exon and 28 base
pairs upstream from the initiation codon, ATG, are identical for the rat
cDNA and genomic DNA; upstream from this point, the sequences differ
substantially. The same situation can be found when the nucleotide
sequences of rat cDNA and mouse genomic DNA are compared. However, the
homology between the sequences of the rat and mouse genomic clones extend
far beyond this point. The rat gemonic nucleotide sequence CTGCAG/A at
positions -30 to -29 fits very well with the consensus sequence of the
splicing acceptor site, TNCAG/G (Nevins, 1983).This indicates the presence

of an intron within the 5' untranslated region of both the rat and mouse
Thy-1 genes. In the mouse thy-1 gene, a sequence close to the consensus
sequence of the splicing acceptor site, at positions -30 to -25, can also

be found. When a Southern blot of Eco RI-Hind III digested ch108 was

81

Figure 5. Nothern blot analyses of RNA from.mouse and rat tissues.
(A) Poly(A) containing RNA was isolated from mouse brain (B), liver (L),
spleen (S) and thymus (T), and subjected to northern blot analysis as
described in Experimental Procedures.
(B) Total RNA of rat thymus (5) and several human B cell lines, 3105 (1),
3106(2), DR2-Ma (3), and DR2-Mo (H) was isolated, and analyzed as

described above, except it was hybridized with the insert of pT69.

82

NORTHERN BLOT ANALYSIS OF
RNA FROM VARIOUS TISSUES

 

Figure 5

83

Figure 6. Comparison of the 5' end nucleotide sequences of rat cDNA
and mouse and rat genoudc DNA. The sequence of the rat cDNA (top line),
starting from position -95 upstream from the initiation codon, ATG, is
compared to the sequences of rat, and mouse genomic clones (middle line
and bottom line, respectively). Identical bases are shown by a line
between the sequences. The initiation codon is boxed, and the putative

splicing points are marked.

RAT
MOUSE

cDNA
RAT
MOUSE

cDNA
RAT
MOUSE

84

COMPARISON OF S'END NUCLEOTIDE SEQUENCES OF RAT AND
MOUSE THY-1 GENOMIC DNA AND RAT cDNA

CAAGTGTCCCTCCAAGAGAAGGGGAGGGGCTGAGTTGTCCATTGTGTGATCCTTGGATCT
TCCCTAAGAGAAGGGGATAAAGAGAGGGGCTGATGTATTCATCATGTGCTCCGTGGATCT

-u5
émcucc'rA
CAAGCCCTCCCOGTAAATGGGGACCCCTCCCTCCTATCAOGGGGCTGACATTCAGCTTAC
CAAGCCCTCAAGGTAAATGGGGACCCACCTGTCCTACCAGCTGGCTGACCTGTAGCTTTC

-30 I -10 1 25
GGGGAGCCCAGACCCAGGACGGAGCTATTGGCAC ACCCAGTCATCAGCATCACTC
CCCA-CTGCAGACCCAGGAOGGAGCTATTGGCACC CCCAGTCATCAGQATCACTC
COCACCACAqfiATCCAAGTCGGAACTCTTGGCAC CCCAGCCATCAGCGTOGCTC

    
  
 

Figure 6

85
hybridized with a full length mouse Thy-1 cDNA clone (a gift form M.
Davis), both fragments hybridized (data not shown). Since the only Hind
III site in ch108 is located 1.5 Kb upstream from the initiation codon,
it suggests that there is an intron in the 5' untranslated leader of the
Thy-1 gene with an estimated size of 1.5 kb. The lack of restriction
enzyme sites within this region, suggesting the presence of highly
repetitive sequences, has made this region difficult to sequence. Introns
in the 5' untranslated region have been observed in the genes of rat
insulin I (Cordell et al., 1979) and chicken ovalbumin (Breathenach et
al., 1978) with sizes of approximately 120 bp. The reason for the
existence of such a large intron in the 5' untranslated region of the
Thy-1 gene is unknown. Hhether it plays some role in regulating the

expression of Thy-1 remains to be determined.

Comparison of the predicted amino acid sequences of the Thy-1 genes

The predictd protein sequences of mouse, rat and human Thy-1 are aligned
with a gap inserted at amino acid 29 in both rat and human Thy-1 to
maximize the homology (Figure 7). All three Thy-1 antigens contain a 19
amino acid leader peptide plus the extra 31 amino acids at the carboxyl
end previously undetected by other investigators. Although, there is a
high degree of overall homology, there are three regions, located at
positions 23 to 29, 60 to 68, and 99 to 105, that vary substantially among
the species. It is intriguing to note that these regions correspond
closely to the locations of the hypervariable regions of Ig V domains
(Bernard et al., 1978), suggesting that as in imunoglobulins there is
little structural constraint in these regions of the molecule.

The degree of protein sequence homology between the three species are

86

Figure 7. Comparison of the predicted amino acid sequences of mouse,
rat and human Thy-1. Identical amino acid residues are underlined. The

hydrophobic segment is underlined.

87

Protein sequence comparison of mouse, rat, and human Thy-1

Mouse
Rat
Human

House
Rat
Human

Mouse
Rat
Human

Mouse
Rat

Human

Mouse
Rat

Human

Mouse
Rat
Haman

Mouse
Rat

Human

Mouse
Rat
Human

Mouse
Rat
Human

Mouse
Rat
Human

Mouse
Rat
Human

-19
met asn

-10
pro ala ile ser val ala leu leu leu ser val 1eu gln
val ......... ile-thr- ‘

 

 

val ser

 

 

leu-ala --------- ile-ala thr— lys
-1 1 10

erg gly gln lys val thr ser leu thr ala cys 1eu val

 

met

 

—arg----ile

 

val

asn gln

 

lys----thr

20
asn leu arg 1eu asp cys arg his glu asn asn thr lys
asn
thr-ser-ser

 

asp

asp asn
1eu-pro
ser-asn

lys his

 

38F

30 no
ser ile gln his glu phe ser leu thr arg glu lys arg
lys
thr-lys

 

 

-------- tyr

50
val leu ser gly thr leu gly ile pro glu his thr tyr
val

 

 

 

phe ------- -val----val——

 

arg ser

60 70
erg val thr leu ser asn gln pro tyr ile lys val 1eu

 

 

asn ----- phe-ser-asp-arg—phe

 

 

thr leu

thr-asn-phe-phe-ser—lys-tyr-asn-met

80
ala asn phe thr thr lys asp glu gly asp tyr phe cys

 

tyr----ser-alar -------- ser

glu 1eu

met
thr

 

thr

 

90 100
gln val ser gly ala asn pro met ser ser asn lys ser
arg gln --------- thr thr

 

 

ala-----his-his------—--his-ser----pro-ile ----- ser-gln-asn

ile ser

---asn-
val-thr ----- ile

1eu leu

110
val tyr arg asp lys leu val lys cys gly gly ile-

----ile
glu

 

 

 

 

120 g * 13o
val gantrp met 1eu 1eu 1eu leu leu ser
’ 1eu

 

 

 

 

ala 1eu

 

 

 

1eu ser

------ -leu

mo 4

leu leu gln ala leu asp phe ile ser leu
phe thr
1eu --------- met ........

 

 

 

Figure 7

88

.ncoxc Hmsoa>aoca coospoo nonmaaoo one amasnxonsoa can amasnxume
.onsoaxamp no noocoscon :aouoaa copoaoonn on» coospon hmoaoao: no common
one ”mmoaoaon cocoscon campona Puhna wouoaooaa mo common .P manna

ion to»
to. too
to. tau
:3 . .
032320»: cox. o5

322:0... 8:263

#00 :5 ens-£2.32:

‘.. ’90 B‘s—:- \~.b

I; #2. .2562!
cc: ecu :38 a:

£396 91>... ._. ho cocoon

89
compared exon by exon in Table 1. The overall protein sequence homology
between mouse and rat Thy-1 is 80$, but only 66$ between mouse and human
and rat and human. Within the first exon, amino acid resiudes -19 to -8,
the homology is 75$, 661 and 751; thus, as expected the signal peptide is
not conserved between the species. The homology for the second exon
coding for amino acids -8 to 106, is 801 between mouse and rat, and 661
and 601 between both rodents and man, consistent with the evolutionary
relationship of the species. The homologies within the third exon,
encoding amino acids 107 to 1H3, are 92$, 89$ and 86$. If only the
hydrophobic segments are compared, the homologies are greater than 851
among all three species. Strong homologies in the hydrophobic regions of
other molecules including membrane IgG1, IgGZa, and IgH (Tyler et al.,
1982), and glycophorins of'man, ox, and pig (Hurayama et al., 1982) have
been reported. This suggests that the structure of the hydrophobic segment
is conserved in order to serve an important and specific, biologic

function.

DISCUSSION

Analysis of the mouse Thy-1 genomic clone, and comparison of the
protein sequences derived from the nucleotide sequences of mouse, rat and
human Thy—1 genes have revealed several interesting features: the mouse
Thy-1 antigen contains 1H3 amino acids instead of the 112 originally
reported. The coding region of the Thy—1 gene is distributed into three
exons, with the third exon encoding amino acids 10? to 1H3. The gene
contains several AATAAA sequences through out the entire Thy-1 sequence,
but only one, located 1110 bp downstream of the termination codon, TGA,
serves as a polyadenylation signal. There is an intron in the 5'
untranslated leader with a size of at least 1.5 kb. The overall structure
of the mouse, rat and human Thy-1 genes is well conserved, including the
intron-exon organization, the protein sequence and the 31 additional amino
acids at the carboxyl terminus. There are three regions where protein
sequence differences are concentrated when Thy-1 molecule from different
species are compared, and these regions correspond closely to those of the
hypervariable regions of the Ig V domain.

It should be noted that there are several inconsistencies between
the DNA and protein sequences of Thy-1 (Table 2). For instance, the
predicted amino acid compositions of mouse, rat and human Thy-1 (Table 2,
line b.) contain tryptophan, but protein sequences of mouse, rat and human
brain Thy-1 antigens, or even the squid brain glycoprotein (Table 2, line
a.) do not. Also amino acid composition analyses failed to reveal any
hydrophobic amino acids in the C-terminal peptide (Campbell et al., 1981;

90

91
Williams & Gagnon, 1982; Cotmore et al., 1981). These C-terminal
hydrophobic regions are highly homologous (greater than 85$) among all
three species, and are also present on the mRNA of rat and mouse thymocyte
and brain tissues. According to the length and hydrophobicity of this
segment, this region may serve as a transmembrane segment anchoring Thy-1
on the cell surface (Chou and Fasman, 1978). It is likely therefore that
Thy-1 is in fact an integral membrane protein anchored on the membrane via
a hydrophobic C-terminal segment.

In addition to Thy-1, there are two other proteins which are reported
to have discrepancies between the biochemically determined protein
sequences and the protein sequences predicted from the nucleotide
sequence. One of those is a membrane protein, the Variant Surface
Glycoproteins (VSGs) of the parasitic protozan trypanosome (Boothroyd et
al., 1982). The nucleotide sequence from the cDNA clones of VSGs suggests
that the primary translation product of one VSG gene contains a 23 amino
acids hydrophobic tail at the carboxyl terminus which is not found on the
isolated mature glycoproteins. Another one is a secretory protein, rat
a-lactoalbumin (c-LA), which has been found to have 17 extra hydrophobic
amino acids beyond the C-terminus (Dandekar and Qasba, 1981). Two forms of
rat a-LA have been found with different molecular weights; however, only
one species of mRNA can be found.

Several possible explanations have been proposed for these
discrepancies. It is possible that the extra hydrophobic segments are

removed during the processing of protein in a manner similar to the

cleavage of the N-terminal signal peptide (Lingappa et al., 1978). An
alternative explanation is that proteolytic degradation has occured during

the preparation and purification of’molecules and their peptide products,

92

Comparison of amino acid compositions of the Thy-1 antigens

 

 

 

 

 

Human Rat House Squid

Brain Brain Brain Brain

Thy-1 Thy-1 Thy-1.2 Glycoprotein

a b a b a b a

028 5 h u A h A 8
ASX 9 10 15 17 1A 15 15
THR 11 13 1O 12 1O 11 7
SER 12 19 8 13 11 16 7
GLX 11 12 10 12 11 12 12
PRO 5 3 3 3 3 3 3
CL! 5 5 u 6 A 6 8
ALA A 5 2 3 3 u 5
VAL 8 9 9 1O 8 9 8
NET 1 2 1 1 1 2 2
ILE 2 u 5 7 fl 6 5
L30 11 23 13 2h 12 *’2h 6
TR? 0 1 0 1 0 1 0
TYR 5 5 2 2 u n u
PHE u 6 h 6 3 u u
HIS 5 6 A u u u u
LYS 8 6 8 8 9 9 9
ARC 6 7 9 9 7 9 6

 

 

%

Line a. is the amino acid composition obtained from protein analysis

(residue/molar amount).

Line b. is the amino acid composition predicted from nucleotide sequence

(residue/molar amount).

Data were taken from references: human Thy-1 (Cotmore at al., 1981); rat
Thy-1 (Campbell at al., 1981); mouse Thy-1.2 and squid brain glycoprotein
(Williams and Gagnon, 1982); predicted amino acid compositions of human

and rat Thy-1 (Seki et al., 198A).

93
or that the C-terminal peptide was lost simply because of its
hydrophobicity. Further protein sequence studies and searches for possible

precursors should resolve this issue.

EXPERIMENTAL PROCEDURES

DNA cloning

The cosmid library of mouse CS7BL/6 was constructed in a cosmid vector
c2RB as described (Bates and Swift, 1983) with a minor modification. The
mouse CS7BL/6 DNA was partially digested with San 3A to an average size
of about “0 Kb, then dephosphorylated before ligated into the Dam HI and
Sma I digested c2RB DNA. A high ATP concentraction is used during the
ligation reaction to prevent the Joining of the Sma I blunt ends. In yitgg
packaged DNA was used to infect g, 22;; strain 10u6 to construct mouse
libraries. The mouse library was screened using the insert of the Thy-1
cDNA clone as a hybridization probe. The DNA of putative Thy-1 positive
clones was purified by mini preparation. The DNAs were cleaved with Eco
RI, and other appropriate restriction enzymes and run on a 0.61 agarose
gel for Southern blot analyses. Confirmed Thy-1 positive cosmid clones
were then digested with Eco RI and subcloned into the Eco RI site of
pBR322. The Subcloned DNA was used to transfom g. 29;; strain H8101, and
the colonies were screened with the original Thy-1 probe. The general
procedures for the growth, screening and analysis of cosmid libraries,
nick translation, and Southern blot analyses were performed as described

(Maniatis et al., 1982).

DNA preparation.
The preparation of DNAs from putative Thy-1 clones was perfomed using the

alkaline-extract method (Birnboim and Doly, 1979) with some modifications.

9”

95
A bacterial culture (#0 ml) was cultured overnight without any
amplification. The bacteria were pelleted, and treated with the CTC
solution (50 mM glucose, 10 mM EDTA, 25 mM Tris-HCl, pH 8.0) without
lysozyme. A clear lysate was prepared by the addition of 2 volumes of 0.2
M Na0H-11 SDS, and the cellular DNA was precipitated by the addition of
0.5 volume of 3 M potassium acetate (pH 5.2). The plasmid DNA remaining in
the supernant was precipitated with 0.6 volume of isopropanol. The DNA was
suspended in TE buffer (10 mM Tris-HCl, 1 IN EDTA, pH 8.0), treated with
DNase-free RNase, phenol extracted, and then precipitated with 2.5 volume
of ethanol. The DNA isolated from this procedure can be used for
endonuclease restriction enzyme digestion, and ligation for subcloning.
Large scale DNA preparation was perfomed as described above, except that
CsCl gradent centrifugation was used for purification of the plasmid prior

to RNase treatment and phenol extraction.

DNA sequence analysis

The Thy-1 positive subclones, ch108 and ch3H, were subjected to
nucleotide sequence analyses. The entire genomic clone was cleaved at
selected sites with the restriction endonucleases, and fragments
corresponding to the insert were purified by agarose gel electrophoresis
and electroelution, or electrophoresis on low melting point agarose (BRL)
and extraction with 0.3 M NaOAC (pH 5.2) saturated phenol. Fragments were
dephosphorylated with calf intestinal alkaline phosphatase and
end-labelled with Tu polynucleotide kinase. Only one of the fragments was
and labelled by filling in with the DNA polymerase I. Double-end labelled
fragments were digested with restriction enzymes, separated on agarose gel

or low melting point agarose, and eluted from gels, then subjected to

96
partial chemical degradation sequence analysis as described (Maxam and

Gilbert, 1980).

RNA analysis

The preparation of total RNA from mouse and rat tissues was perfomed as
previously described (Chirgwin et al, 1979). The poly(A) containing RNA
from mouse tissuses was isolated as described (Aviv and Leder, 1972). The
northern blot analysis was performed as described (Lehrach et al, 1977).
The mouse poly(A) containing RNA was run on a 1.25 formaldehyde agarose
gel, botted onto a nitrocellulose filter, and hybridized with 32P-labelled,
PV3, in 50$ formamide solution at h2°C. The northern blot analysis of RNA
from rat thymus and several human B cell lines was done in the same way as

that of mouse, except that it was hybridized with the 32P-labelled insert

of pT6u in a solution of 6x SSC, 10x Denhardt's, 0.1$ SDS at 68°C.

ACKNOHLEDGEMENTS

The authors thank Drs. Jerry Dodgson and Paul Bates for their advice
in cloning technology, Drs. Tetsuya Horiuchi, Tetsunori Seki and Roger
Denome for their kindly providing their unpublished data, and Dr. Susan

Corned for her helpful discussion.

97

REFERENCES

Acton, R. T., Norris, R. J. and Williams, A. F. (197“) Estimation of the
amount and tissue distribution of rat Thy-1.1 antigen. Eur. J. Immunol. A,
598-602.

Aviv, H. and Leder, P. (1972) Purification of biologically active globin

messenger RNA by chromotography on oligothymidylic acid-cellulose. Proc.
Natl. Acad. 8°10 U.S.A. 69, “408-11513.

Bates, P. L. and Swift, R. A. (1983). Double cos site vecters: simplified
cosmid cloning. Gene 26, 137-1u6.
Bernard, 0., Hozumi, N. and Tonegawa, S. (1978) Sequences of mouse immuno-
globulin light chain genes before and after somatic changes. Cell 15,
1133-11HN.
Birnboim, H. C., and Doly, J. (1979). A rapid alkaline extraction proce-
dure for screening recombinant plasmid DNA. Nucleic Acids Research 7,
1513-1523.
Boothroyd, J. C., Paynter, C. A., Coleman, S. L. and Cross G. A. N.
(1982). Complete nucleotide sequence of complementary DNA coding for a
variant surface glycoprotein from Trypanosoma brucei. J. Hol. Biol. 157,
597-556.
Breathenach, R., Benoist, C., O'Hare, K., Ganon, F. and Chambon, P. (1978)
0valbumin gene: evidence for a leader sequence in mRNA and DNA sequences
at exon-intron boundaries. Proc. Natl. Acad. Sci. U.S.A. 75, u853-u857.
Campbell, D. C., Gagnon, J., Reid, K. B. M. and Williams, A. F. (1981).
Rat brain Thy-1 glycoprotein: the amino acid sequence, disulphide bonds
98

99
and an unuaual hydrophobic region. Biochem. J. 195, 15-30-

Chirgwin, J. H. Przybyla, A. E., MacDonald, R. J. and Rutter, W. J.
(1979). Isolation of biologically active ribonucleic acid from sources
enriched in ribonuclease. Biochemistry 18, 5299-5299.

Chou, P. Y. and Fasman, G. D. (197”) Prediction of protein conformation.
Biochemistry 13, 222-295.

Cohen, 1". E., Novotny, J., Sternberg, J. 8., Campbell, D. G. and Williams,
A. P. (1981) Analysis of structural similarities between brain Thy-1 anti-
gen and immunoglobulin domains. Biochem. J. 195. 31-110.

Cotmore, S. P., Crowhurst, S. A. and Waterfield, H. D. (1981). Purifica-
tion of Thy-1-related glycoproteins from human brain and fibroblasts:
comparisons between these molecules and murine glycoproteins carrying
Thy-1.1 and Thy-1.2. Eur. J. Immunol. 11, 597-603.

Cordell, B., Bell, G., Tischer, E., DeNoto, F. H., Ullrich, A., Pictet,
R., Rutter, W. J., and Goodman, H. H. (1979). Isolation and characteriza-
tion of a cloned rat insulin gene. Cell 18, 533-5fl3.

Dandekar, A. H. and Qasba, P. K. (1981) Rat a—lactalbumin has a 17-
residue-long CO0H-terminal hydrophobic extension as judged by sequence
analysis of the cDNA clones. Proc. Nalt. Acad. Sci. U.S.A. 78, “BBB-H857.
Dalchau, R. and Fabre, J. W. (1979) Identification and unusual tissue
distribution of the canine and human homologues of Thy-1 (9). J. Exp. Med.
1H9, 576-582.

Douglas, T. C. (1972) Occurrence of a theta-like antigen in rats. J. Exp.
Med. 136, 10511-1062.

Hunt, 8. V., Mason, D. W. and William, A. F. (1977) In rat bone marrow
Thy-1 anigen is present on cells with membrane immunoglobulin and on

precursors of peripheral B lymphocytes. Eur. J. Imunol. 7, 817-823.

100
Korman, A. J., Auffray, C., Schamboeck, A and Strominger, J. L. (1982) The
amino acid sequence and gene organization of the heavy chain of the
HLA-DR: homology to imunoglobulins. Proc. Natl. Acad. Sci. U.S.A. 79,
6013-6017.
Kuchel, P. 11., Campbell, D. G., Barclay, A. u. and Williams, A. F. (1978).
Molecular weights of the Thy-1 glycoproteins from rat thymus and brain in
the presence and absence of'Deoxycholate. Biochem. J. 169, h11-fl17.
Larhamar, D., Schenning, L., Gustafsson, K., Wiman, K., Claesson, L.,
Rask, L. and Peterson, P. A. (1982) Complete amino acid sequence of an
HLA-DR antigen-like 8 chain as predicted from the nucleotide sequence:
similarities with immunoglobuin and HLA-A, -B and -C antigens. Proc. Natl.
Acad. Sci. U.S.A. 79, 3687-3691.
Ledbetter, J. A. and Herzenbrg, L. A. (1979) Xenogeneic monoclonal anti-
bodies to lymphoid differentiation antigens. Immunol. Rev. ”7, 63-90.
Lehrach, H., Diamond, D., Wozney, J. H. and Boedtker, H. (1977) RNA
molecular weight determinations by gel electrophoresis under denaturing
conditions, a critical reexamination. Biochemistry 16, u73u-u7u2.
Lennon V. A., Unger, N. and Dulbecco, R. (1978) Thy-1: a differentiation
marker of potential mamary myoepithelial cells in 143.32. Proc. Natl.
Acad. Sci. U.S.A. 75, 6093-6097.
Letarte-Huirhead, M., Barclay, A. N. and Williams, A. F. (1975) Purifica-
tion of the Thy-1 molecule, a major cell surface glycoprotein of rat
thymocytes. Biochem. J. 151, 685-697.
Lingappa, V. R., Katz, F. N., Lodish, H. F. and Blobel, G. (1978) A signal

sequence for insertion of a transmembrane glycoprotein. J. Biol. Chem.

53 9 8667-8670 e

101
Maxam, A. M. and Gilbert, W. (1977) A new method for sequencing DNA. Proc.
Natl. Acad. Sci. U.S.A. 7n, 560-56“.
Maniatis, T., Fritsch, E. and Sambrook, J. (1982) Molecular cloning. (Cool
Spring Harbor Laboratory, New York).
Mansour, M. H. and Cooper, E. L. (198”) Purification and characterization
of Rana Pipiens brain Thy-1 glycoprotein. J. Immunol. 132, 2515-2523.
McKenzie, J. L. and Fabre, J. W. (1981) Studies with a monoclonal antibody
on the distribution of Thy-1 in the lymphoid and extracellular connctive
tissues of the dog. Transplantation 31, 275-282.
Moriuchi, T., Chang, H-C., Denome, R. and Silver, J. (1983) Thy-1 cDNA
sequence suggests a novel regulatory mechanism. Nature 301, 80-82.
Murayama, J-I., Tomita, M. and Ramada, A. (1982) Primary structure of
horse erythrocyte glycophorin HA: its amino acid sequence has unique
homology with those of human and porcine erythrocyte glycophorins. J.
Membrane Biol. 6“, 205-215.
Nevins, J. R. (1983) The pathway of eukaryotic mRNA formation. Ann. Rev.
Biochem. 52, HH1-u66.
Owen, M., Knott, J. C. A. and Crumpton, M. J. (1980) Labeling of
lymphocyte surface antigens by the lipophilic photoactivatable reagent
Hexanoylidiiodo-N-(h-azido-Z-nitrophenyl)Tyramine. Biochemistry 19,
3092-3099.
Reif, A. E. and Allen, J. M. (196”) The AKR thymic antigen and its
distribution in leukemias and nervous tissues. J. Exp. Med. 120, u13-H33.
Rostas, J. A. P., Shevenan, T. A., Sinclair C. M. and Jeffrey, P. L.
(1983) The purification and charaterization of a Thy-1-like glycoprotein

from chicken brain. Biochem. J. 213, 1H3-152.

102
Scheid, M., Boyse, E. A., Carswell, E. A. and 01d, L. J. (1972) Serologi-
cally demonstrable alloantigens of’mouse epidermal cells. J. Exp. Med.
135, 938-955.
Seki, T., Chang, H-C., Moriuchi, T., Denome, R., Proegh, R. and Silver, J.
Structural organization of the Thy-1 genes. Science (submitted).
Stern, P. L. (1973) Alloantigen on mouse and rat fibroblasts. Nature 2H6,
76-78.
Trowbridge, I. S., Weissman, I. L. and Bevan, M. J. (1975) Mouse T-cell
surface glycoprotein recognized by heterologous anti-thymocyte sera and
its relationship to Thy-1 antigen. Nature 256, 652-65“.
Tyler, B. M., Cowman, A. P., Gerondakis, S. D., Adams, J. M. and Bernard,
0. (1982) mRNA for surface imunoglobulin c chains encodes a highly
conserved transmembrane sequence and a 28-residue intracellular domain.
Proc. Natl. Acad. Sci. U.S.A. 79, 2008-2012.
Williams, A. P. and Gagnon, J. (1982) Neural cell Thy-1 glycoprotein:

Homology with immunoglobulin. Science 216, 696-703.

CLOSING STATEMENT

Although many of the questions concerning the structure,
polymorphism, and gene organization of class II genes have been answered
by recombinant DNA technology and monoclonal antibodies analysis, there
are still a set of questions that need to be answered. These include the
detailed tertiary structure, the mechanism of biosynthesis, the mechanism
of generation of polymorphism, the molecular function and the evolutionary
history of class II antigens.

It is the same situation for Thy-1. Although the detail genomic
structures of'mouse, rat and human Thy-1 have been studied, many
controversies concerning the Thy-1 antigen remain to be resolved. Such as,
why is human Thy-1 antigen not conserved by evolution; whether the Thy-1
antigen contains the hydrophobic segment on its carboxyl end while
spanning on the membrane. Finally, since Thy-1 is expressed in various
amounts in different tissues, we would like to know more about how its

expression is regulated, and what its role is in these different tissues.

103