Differential Tissue Expression of the Lewis Blood Group Antigens

From www.bloodjournal.org by guest on February 6, 2015. For personal use only.
Differential Tissue Expression of the Lewis Blood Group Antigens: Enzymatic,
Immunohistologic, and Immunochemical Evidence for Lewis a and b Antigen
Expression in Le(a-b-) Individuals
By Torben F. Orntoft, Eric H. Holmes, Philip Johnson, Sen-itiroh Hakomori, and Henrik Clausen
The Lewis blood group system comprises two main carbohydrate antigens, Le’ and Leb. Lewis typing has traditionally
been based on serologic determinations using erythrocytes
and saliva. Several recent studies have demonstrated that
erythrocyte Lewis phenotype may change during pregnancy
or disease, and inappropriate Lewis antigens have been
found in both normal and neoplastic tissue. To evaluate
whether these observations are in conflict with the presently
proposed genetic and biosynthetic basis of the Lewis blood
group system, we performed a combined enzymatic, immunohistologic, and immunochemical study of Lewis antigen
expression in normal and neoplastic tissues, as well as
erythrocytes, plasma, and saliva of Le(a- b-)-typed individu-
als. Of six cancer-bearing patients typed Le(a-b-), three
were identified as nongenuine owing to the presence of a1 +
4fucosyltransferaseactivity (a1 + 4FT) and Lewis antigens in
saliva and three were identified as genuine (lacking a1 -+ 4FT
and Lewis antigens in saliva). These genuine Le(a-b-)
individuals were shown to express significant a1 + 4FT in
tissues, and Lewis antigens were detected in tissues by
immunohistology as well as immunochemistry. We conclude
that the Lewis phenotype obtained by serologic determination of erythrocytes and saliva does not apply to all tissues.
We discuss biosynthetic and genetic consequences of this
finding.
@ 1991 by TheAmerican Society of Hematology.
T
In this study, enzymatic, immunohistologic, and immunochemical methods were used to gain insight into expression
of Lewis a and b antigens and corresponding a1 + 4FT
activities in individuals typed as Le(a-b-) by standard
hemagglutination. Owing to the rare occurrence of
Le(a-b-) individuals ( 6% of the white population), only
six individuals were examined. Two groups of Le(a-b-)
RBC-typed individuals were identified. One group (consisting of three individuals) was classified as nongenuine
Le(a-b-) (a1 -+ 4FT activity detected in saliva); the other
(also consisting of three individuals) was classified as
genuine Le(a-b-) (no a1 + 4FT activity detected in
saliva). The nongenuine Le(a-b-) individuals had chemical amounts of Lewis active neutral glycolipidsin their RBC
and serum similar to Lewis’ individuals, regardless of lack
of hemagglutination. In the genuine Le(a-b-) group,
none or very small quantities of Lewis a and b antigens were
HE STRUCTURES (Table 1) of Le” and Lebdeterminants were first deduced from serologic inhibition
tests” and subsequently confirmed by isolation of oligosaccharide fragments from Le”- and HLeb-active glycoproteins
isolated from ovarian cyst f l ~ i d s . ’ ~ Le”
, ’ ~ antigen
(GalPl+ 3[Fucal-+ 41GlcNAc + R) is formed by the action of the Lewis gene-encoded a1 -+ 4-~-fucosyltransferase (a1 + 4FT) on type 1chain (Galpl+ 3-GlcNAc + R)
endings in glycoconjugates.’ In tissues from individuals who
encode an active a1 + 2-~-fucosyltransferase, both the
galactosyl and the N-acetyl-glucosaminyl residues in the
type 1 chain are substituted with L-fucose and an Leb
(Fucal -+ 2Galp1 + 3-[Fucal + 41GlcNAc + R) structure is formed.I5Lewis antigens on erythrocytes differ from
ABH antigens in that they are not synthesized by the
erythrocyte but rather are secondarily acquired from
plasma.16 The individual Lewis phenotype has traditionally
been based on serologic determinations for erythrocytes and saliva. The Lewis’ phenotypes, Le(a+b-) and
Le(a-b+), have been believed to result from action of a
transferase encoded by an active allele Le at the Lewis
locus, whereas the Lewis- phenotype, Le(a-b-), which
comprises ~ 6 of
% the white and 35% of the black population, has been believed to result from the homozygous
presence of the silent allele le.”
Exceptions to the present concept of regulation of Lewis
antigens, however, have been reported during recent years
as either Lea, Leb,or both antigens have been detected in
of individuals
small intestine,” saliva,18 or ~rothelium’~
typed as Le(a-b-) in hemagglutination assays. Similarly,
Lewis a and b antigens have been detected in various cancer
tissues of RBC Le(a-b-)-typed individual^.^^^** Whether
these observations are in conflict with the proposed genetic
and biosynthetic basis of the Lewis blood group system is
unclear because of reports demonstrating that the Lewis
phenotype on RBC may change with various conditions,
including carcinomas?o,2z-aWith regard to breast cancer,
the current lack of knowledge regarding regulation of Lewis
blood group antigen expression has led to recent controversies regarding the frequency of this disease in Le(a-b-)
individuals.2628
Blood, Vol77, No6 (March 151,1991: pp 1389-1396
-
From The Biomembrane Institute and Department of Pathobwlogy,
University of Washington, Seattle; the Departments of Experimental
Clinical Oncology, Danish Cancer Society, and Clinical Chemistry,
Aarhus Kommune Hospital, Aarhus, Denmark; the Pacific Northwest
Research Foundation, Seattle, WA; the Division of Immunochemical
Genetics, Medical Research Council, Harrow, England; and the Royal
Dental College, Copenhagen, Denmark.
Submitted March 29,1990; accepted November 19, 1990.
Supported by Outstanding Investigator Grant No. CA 42505 from
the National Cancer Institute (NCI), Bethesda, MD (to SH) and by
fundsfrom The Biomembrane Institute. H.C. was supported in part by
Fru Jenny Vising, Lundbeck Fonden, and Sundhedsvidenskabelige
Forshingrad, Denmark. T.F.Q. was supported by the Danish Cancer
Society; he was a visiting Scientist at The Biomembrane Institute.
E.H.H. was supported by NCI Research Center Development Award
No. K04CA01343 and NCI Grant No. CA41521.
Address reprint requests to Dr Torben Qmtoft, Dept. Clinical
Chemistry, Aarhus Kommunehospital, DK-8000 Aarhus C, Denmark.
The publication costs of this article were defrayed in part by page
charge payment. This article must therefore be hereby marked
“advertisement” in accordance with 18 U.S.C. section 1734 solely to
indicate this fact.
0 1991 by The American Society of Hematology.
0006-4971/91/7706-OO21$3.OalO
1389
From www.bloodjournal.org by guest on February 6, 2015. For personal use only.
ORNTOFT ET AL
1390
Table 1. Biosynthesis of Lewis Antigens and Related Isomeric Structures Based on Type-2 Chain Poly-N-Acetyllactosamine Structure
Antigen
Lewis Blood Group
Genotype
Lea+b-
Le, se
Type 1
V
Le”
O
.
Fucal
1
Fucal --+
Le
4
2Galpl+3GlcNAcpl-+R
SelHl
V
Leb
Le“*
Led
Le”
Fucal -+ 2Galpl+4GlcNAcpl+R
HISel
3
T
Fucal
nonsec
Le’
Les-b+ sec
0
.
O
.
0.0
Le, Se
Lea-b-
nonsec
le. se
Lea-b-
sec
le, Se
O
.
A
0
.
0
A
X(Le)
O
.
H
0
.
0
Two distinct a1-2 transferases,encoded by the secretor and H genes, respectively, are presently believed to exist.’”,3The secretor gene-encoded
FT has preference for the type-1 chain substrate, whereas the H gene-encoded transferase has preference for the type-2 chain.‘ The Lewis
gene-encoded FT may use both the type-1 and type-2 chain substrates in vitro, thus participating in formation of both Le”mand Le*?.6 In vivo,
however, it may use only the type-1 chain structure.’ Independent al-3FT activities have been recognized and are believed to be encoded by the X
At present, no information is available regarding the existence of an a1+4FT activity independent of that believed to be encoded by the
Lewis gene.
Abbreviations: Le, Lewis gene-encoded a1-4FT; Se(H), secretor or H gene-encoded al-2FT; 0,Gal; 0, GlcNAc; 0, Fucal+2: V, Fucal-4; A,
Fucal+3.
‘The LeEantigen has been reported to be an elongated variant of the basic Gal pl+3GlcNAcp disaccharide with the type-2 chain Le” hapten.”
f o u n d on RBC or in serum, whereas Lewis antigens and low
but readily detectable a1 -+ 4FT activity were found in
colon and bladder tissues.
MATERIALS AND METHODS
Samples. Fifty-milliliter samples of human whole blood were
obtained from six individuals by venipuncture, partly used within
24 hours for serology, and partly centrifuged and stored at -80°C
as serum and cells. Saliva was obtained simultaneously with the
blood samples, and 1 mL was stored at -80°C for enzyme analysis
and 1 mL was boiled and stored at -20°C for hemagglutination
inhibition studies. Biopsies were taken from colon tumors as well as
morphologically and microscopically normal colonic mucosa far
from the tumor. From two individuals, biopsies were obtained from
transitional cell carcinomas of the urinary bladder. All biopsies
were immediately frozen at -80°C.
Serology. RBC were grouped by use of polyclonal human,
rabbit, and goat antisera (Ortho Diagnostic Systems, Raritan, NJ)
as well as Dolichos biflom and Ulex europaeus agglutinin (Sigma
Chemical, St Louis, MO) according to routine blood bank procedure. Identical sera and lectins were used to detect Lewis substance in saliva by low-ion-strength hemagglutination inhibition. In
all assays, appropriate known controls were included.
FT assays on saliva. Saliva was thawed and frozen twice and
centrifuged at 800g, and the supernatant was used for assays. For
assays of a1 + 31T and a1 + 4 l T , 20 pL enzyme source was
added to GDP-L-[*~C]-FUC
(0.28 nmol, 70,000 cpm), MnCl, (1
pmol), acceptor (0.5 pmol), ATP (0.5 pmol), Triton X-100 (500
pg), and 0.1 mol/L Tris-HCI (pH 7.2) in a total volume of 100 pL.
The mixture was incubated for 2 hours and chromatographed on
Whatman no. 40 paper in ethyl acetate/pyridine/water ( 1 0 4 3
vol/vol/vol) for 48 hours. The papers were dried, scanned for
radioactivity (Packard radiochromatogram scanner, Downersgrove, IL), and the mobility of peaks was measured relative to
known compounds; these areas were then cut out and counted by a
liquid scintillation counter. N-Acetyllactosamine (Galpl +
4GlcNAc) was used as acceptor for the a1 431T determination,
and lacto-N-biose I (Galpl 4 3GlcNAc) was used as acceptor for
the a1 + 41T.29
FT assay on tissues. Tissue samples ( = 1 g) were homogenized
in 2 vol50 mmol/L HEPES buffer (pH 7.2), 0.5 m o w sucrose, and
1mmol/L EDTA by two strokes of a Potter-Elvehjem homogenizer
and used for characterization of enzyme activities present in each
fraction. The FT activity was determined as previously described”’
in reaction mixtures containing 2.5 pmol HEPES buffer (pH 7.2),
40 pg lactotetraocyclceramide (Galpl-3GlcNAc~1-3Gal~l4Glcp1-1Cer) (Lc4) or lactoneotetraosylceramide ( G a l p l 4GlcNAcpl-3Galpl-4G1cpl-1Cer)
(nLc,), 100 pg taurodeoxycholate, 1pmol MnCI,, 0.5 pmol CDP-choline, 15 nmol GDP-[’4C]-Fuc
(15,000 cpndnmol), and 100 to 300 pg protein in a total volume
of 0.1 mL. The reaction mixture was incubated for 2 hours at 37°C
and terminated by addition of 6 pmol EDTA and 0.1 mL
chloroform-methanol (CM) 21. The entire reaction mixture was
streaked onto a 4-cm-wide strip of Whatman no. 3 paper and
chromatographed with water overnight. The glycolipid remaining
at the origin was extracted with 2- to 5-mL washes of chloroformmethanol-water (CMW) 10:5:1. The solvent was removed by
nitrogen stream, and the glycolipid was dissolved in 20 pL CM 2 1 .
A 10-pL aliquot was removed, spotted onto an high-performance
thin-layer chromatography (HPTLC) plate (Merck, Darmstadt,
FRG), and developed in CMW 605:l containing 0.02% CaCl, as a
From www.bloodjournal.org by guest on February 6, 2015. For personal use only.
1391
DIFFERENTIAL TISSUE EXPRESSION OF LEWIS ANTIGENS
final concentration. Standard glycolipids were visualized by orcinol
spray. Radioactive glycolipid bands were located by autoradiography, scraped from the plate, and counted by liquid scintillationcounter.
Isolation of glycolipids. RBC ( = 20 mL) and serum ( = 30 mL)
were extracted overnightat 4°C with solvent A (isopropanol-hexanewater [IHW] 65:25:20).Tissues were extracted with 10 vol solvent
A in a Potter-Elvehjem homogenizer, sonicated, and centrifuged at
SOOg for 20 minutes. The insoluble pellet was re-extracted twice in
an identical way, and the combined supernatants were evaporated
under nitrogen stream. The near-dry samples were then dissolved
in CMW 30603 and subjected to chromatography on diethylaminoethanol (DEAE) Sephadex A - 2 9 to separate total neutral
glycolipids from gangliosides. The total neutral glycolipid fraction
was dried and acetylated with 1 mL pyridine and 0.5 mL acetic
anhydride, followed by chromatography on a Florisil column and
deacetylation with s-vol 0.5% sodium methoxide in CM 21. The
samples were dried and resuspended in CM 21, and 5 JLLof this
total neutral glycolipid fraction was spotted on HPTLC plates,
developed in CMW 5040:10, and stained with orcinol. The amount
of glycolipid in each lane was standardized according to the extent
of orcinol staining.
TLC immunostaining. Immunostaining of glycolipids separated
by HPTLC was performed according to the procedure of Magnani
et allz as modified by Kannagi et
The monoclonal antibodies
(MoAbs) used were anti-le” clone CF4-C, (IgG,)34and anti-Leb
(IgM) donated by Dr Donald A. Baker (Chembiomed, Alberta,
Canada). TLC immunostaining with the anti-LebMoAb produced
staining in the hexaosylceramide region. In glycolipid extracts of
tissues, staining of the H type 1 pentaosylceramide region was also
observed, indicating cross-reactivitywith H type 1 structures. Only
one anti-Leb MoAb was used for TLC immunostaining studies
owing to the limited amount of lipid extract available.
Immunohistochemistry. Immunohistochemicalstudieswere performedwith two additional IgM anti-LebMoAbs: one from BiotestSerum Institut, GMBH, Frankfurt, FRG, the second donated by L.
Messetter, Malm@Hospital, Sweden. The specificity of the latter
MoAb has been studied in detail?5An IgG, MoAb, FH7:6 directed
to the sialosyl derivative of Le” (sialosyl-Le;,), was also used.
A modification of previously described methods3’.” was used.
Formalin-fixed, paraffin-embedded 3-km sections were deparaffinized and rehydrated. Endogenous peroxidase was blocked by
0.8% H202in absolute methanol for 30 minutes. Sections were
washed in TrisPhosphate-buffered saline (PBS), incubated with
10% normal rabbit serum in PBS for 10 minutes, and then
incubated overnight at 4°C with MoAbs diluted 1:40 to 1:80 in PBS.
After repeated washings, MoAb binding was visualized by incubation for 60 minutes with biotinylated rabbit anti-mouse immunoglobulins (diluted 1:loO in PBS), followed by avidin-biotin-peroxidase
complex method according to the manufacturer’s instructions
(Dakopatts, Copenhagen, Denmark) and 0.04% 3-amino-9ethylcarbazole. The three anti-LebMoAbs showed similar staining
reactions. MoAb FW7 (antisialosyl-Le;,) stained more cells than
the anti-le” MoAb in colonic sections.
RESULTS
Genuine &(a - b - ) individuals. Three individuals were
identified as genuine Lewis-negative individuals because
their RBC typed Le(a-b-) and their saliva contained no
a1 -+ 4FT activity (Table 2). Normal colon and colonic
cancer tissue was available from only two of three and in
one only bladder cancer tissue was available. Normal colon
tissue and bladder tumor tissue from Lewis-positive individuals served as control.
FT activity of tissue from Le(a-b-) individuals. In
normal colonic mucosa, colon carcinoma tissue, and bladder cancer tissue, a1 +. 4FT activity was readily detectable
with the type-1 chain glycolipid acceptor Lc4 (Table 2). The
specific activity of the a1 + 4FT was similar in all five
specimens examined and considerably lower (5% to 16%)
than the activity in Lewis-positive controls (Table 2).
In contrast, significant a1 +. 3Fuc transfer into the type-2
chain glycolipid acceptor nLc, occurred in all samples
regardless of Lewis status of donor. TLC analysis of
reaction products from transfer of I4C-Fucfrom GDP-[I4C]Fuc to the type-1 chain acceptor Lc4 and the type-2 chain
Table 2. Serologic, Enzymic, Immunohistologic, and Immunochemical Findings in Genuine Le(a- b-) Individuals
Serology’
Material
Normal
Colon Co,
Colon Co,
Cancer
Colon Co,
Colon Co,
Bladder B,
Lewis-positive control
Normal colon (n = 3)
Cancer, bladder (n = 5)
Enzyme Activity (pmollhtmg Protein)
Saliva
RBCs
Le’
ABH
Lewis
A,
A,
Le(a-b-)
Le(a-b-)
2
2
A,
A,
A,
Le(a-b-)
Le(a-b-)
Le(a-b-)
Leb
Tissue*
Salivat
ABH
a1+3
al-4
8
8
>256
>256
90
340
0
0
1,990
1,320
2
2
4
8
8
8
>256
>256
4
90
340
308
0
0
0
16
16
>256
>256
>256
>256
581
293
Immunohistology§
HPTLC
Le‘
Leb
Le’
Leb
60
70
+
-
+
+
(+)
i
2,000
1,460
608
47
177
98
+
+
+
2,480
472
1,190
632
a1-3
u1-4
-
+
+
+
+
+
+
+
+
+
+
*Hemagglutination with polyclonal sera, Dolichos biflorus, and Ulex europaeus lectins defined RBC phenotype; hemagglutination inhibition was
used to determine titer in saliva.
tSaliva was incubated with GDP-L-[“CI-FUC (0.28 nmol, 70,000 cpm), MnCI, (1 pmol), acceptor (0.5 kmol) lacto-N-biose I (al-4FT) or
N-acetyllactosamine (al-3FT). ATP (0.5 pmol), Triton X-100 (500 kg), and 0.1 mol/LTris-HCI (pH 7.2) in a total volume of 100 pL.
SHEPES buffer (pH 7.2) 2.5 pmol, Lc, or nLc,40 pg. taurodeoxycholate 100 pg, MnCI, 1 pmol, CDP-choline 0.5 pmol, GDP-[14C]-Fuc15 nmol(l5,OOO
cpdnmol), and 100 to 300 pg protein in a total volume of 0.1 ml. Material for assay was homogenized in 50 mmol/L HEPES buffer (pH 7.2). 0.5 mol/L
sucrose, and 1 mmoVL EDTA.
§Anti-lea and three different anti-LebMoAbs were used to demonstrate Lewis structures on tissue sections and TLC. Similar staining was observed
with the three anti-LebMoAbs.
From www.bloodjournal.org by guest on February 6, 2015. For personal use only.
0RNTOFT ET AL
1392
active glycolipids when total neutral glycolipids wcre extracted from RBC and serum and subjected to immunochemical characterization (Fig 2A). Normal and malignant
tissuc spccimcns examined by this mcthod all Contained
pcntasaccharidc Le' and hcxasaccharidc Lehantigens however, (Fig 2A). Thcsc findings wcre furthcr supported by
immunohistochemicaldctcction of Lewis antigcns on tissue
scctions (Fig 3) from all fivc spccimcns. Thc prcscncc of
small quantities of Le' and Lehactivc glycolipids in RBC
extracts of thc bladder carcinoma patient (Fig 2A, B,
samplcs) is obscure. The migration of thc glycolipids, as
well as the immunorcactivity, indicates that thcsc arc
authentic Le' and Lehglycolipids.This patient had no W o r
Lehactive glycolipids in scrum extract.
Nongenuine I / ( a 4-) individuals. Three individuals,
two with colon canccr and onc with a bladder tumor wcrc
serologically dcfincd as Le(a-b-) individuals. Detailed
study of thcir saliva showed a1 4FT activity and secretion
of Lewis antigcns, howcvcr; therefore, thcy are classified as
nongcnuinc Le(a-b-) (Tablc 3). Normal colon tissue was
obtained from the two colon carcinoma paticnts; from the
third patient. bladder carcinoma tissue was availablc.
FT activity in tissue from non-genuine Le(a4-)individuah. In thc two spccimcns of normal colon mucosa, an
a1
4FT activity similar to that in the Lewis positive
control group (Table 2) was detcctcd (Table 3). In thc
bladder carcinoma, the al + 4FT activity was in the same
rangc as that of gcnuinc Lewis-negativc individuals (Tablc
3). The a1 3FT was activc to an cxtcnt similar to that of
the other individualsexamined (Tablc 3).
Lewis antigens in tissues jivm nongenuine L.e(a-b-) individuals. lmmunostainingof extracted total neutral glycolipids showed pcntasaccharide Le' and hexasaccharidc Leh
antigens in tissues, and interestingly, also in RBC and
scrum although RBC could not bc agglutinated by antiLewis sera (Fig 3B). As cxpcctcd, immunohistologyshowcd
Lewis antigcns in these individuals (Tablc 3).
-
1
2
3
4
nc
Flg 1.
aM-h of FTproductrfrom le, a d nLc, utr)ynd
f" hom Lmb-pOrhh 8 d gmUim w + n e g & e
donon. Autoradiographs of product. from Lc, with Lewicpcattive
donor W e 1). nLc. with Lewis-podtive donor (lane 2). Le, with
genuina Lewh-negative donor (lane 3). and nLc, with genuine Lewisnegative donor (lane 4). Reaction conditions am deaccribed in the
Materiah and Methods section. The solvent syrtem was C M W
60:35:8. TLC mobility of standard glycolipids ia shown.
dUU0
acceptor nLe4 was conducted with enzyme fractions from
Lewis-positive and genuine Lewis-negative donors. These
results arc shown in Fig 1. Strong bands corresponding to
formation of a l 3Fuc derivativcsof nLc, wcrc found with
eithcr fraction (lanes 2 and 4). as was a product corrcsponding to a1 2 fucosylation of the terminal Gal of Lc, (lanes
1 and 3). most probably associated with thc sccrctor gcnc
status of the donors. Fucose transfer in a1 4 linkagc to
Lc, rcsultcd in a strong band with the Lewis-positivc donor
(Fig 1, lane 1). which was greatly reduced but still dctcctablc with the genuinc Lewis-ncgativc donor (Fig 1, lanc 3).
Further analysis of thc isolated band corrcsponding to
III'FucLc, from the Lewis-negative donor was conducted
by TLC analysis of the acetylated derivative." Thc acctylated product comigrated with standard III'FucLc, but not
III'FucnLc,, indicating that the product Contained an a1
4 linked Fuc on Lc, (data not shown). The rclativc intensity
of thc bands reflects the competition betwccn the a1
4FT and a1 21T for their mutual acceptor Lc4and that of
the a1
3FT and a1
2FT for their mutual acceptor
nLc,. In the former case, the a1
4FT dominatcs in
Lewis-positive individuals (Fig 1, lane 1). whereas the a1
2FTdominates in Lewis-negativeindividuals(Fig 1, lane 3).
In the latter case, the a1 -* 3FT is so strongly expressed in
most individuals as comparcd with a1 -* 2FT that only
insignificant amounts of a1 + 2-fucosylated products are
formed (Fig I, lanes 2 and 4).
Lewis antigens in tissues of genuine L.e(a-b-) individuals.
The genuine Le(a-b-) status of the examined individuals
was confirmed by the nearly complete absence of Lewis
-
-
-- - -
-
-
-
-
-
-
DISCUSSION
Lewis blood group status is traditionally determined by
RBC and saliva phenotyping. Based on such phenotyping,
wc idcntificd thrcc pcrsons as genuine Lewis-ncgativc
individuals having Lc(a-b-) RBC and no a1
4FT
activity in saliva. Thrcc other individuals were identified as
nongcnuinc Le (a-b-) bccausc their RBC wcre typed
Le(a-b-) by serology but their saliva contained at 4FT
activity and Lewis a and b antigens.
In the group of gcnuinc Le(a-b-) individuals, we were
ablc to idcntify Lewis a and b antigens and detect low a1 -*
4FT activity in colon and bladder tissue, indicating in vivo
activity of the a1
4FT. In the group of nongenuine
Lc(a-b-) individuals, an al 4FT activity corrcsponding
to that detected in Lewis-positivc individuals, as well as
Lewis antigens, were dctcctcd in colon and bladder. Owing
to grcat individual variation in enzyme activity, homozygous
and heterozygous individuals cannot be scparatcd, but the
relatively high frequency of heterozygous (Le, le; IC, Le)
individuals in the population makes it unlikely that the
rclativcly rare nongcnuinc Le(a-b-) individuals should
simply be the heterozygous individuals.
-
-
- -
From www.bloodjournal.org by guest on February 6, 2015. For personal use only.
DIFFERENTIAL
nssw EXPRESSION OF LEWIS
ANTIGENS
1393
NORMAL TISSUE
A
Standord
CANCER TISSUE
Cor
~8Leb IJo?Leb
Leo
b
Le
orc
R S R S RSTRST T T T T T T R S T R S T
NORMAL TISSUE
B
1
Stondord
Leb
CANCER TISSUE
I
1
1
TrCo4
Leo
Leb
@Leb
--
Leo
Leb
orc
-
c
=e
-
R S R S
R S T R S T
T T
R S T
R S T
no 2
o(w.ndw.ahng)yedipidrbync kmn-ining.iotli
m
i gmpidr-~erbnt.dfrom
RBC, ..~m,cdon,
and Moddar tkrcm...pmhdby HPTLC, and lmmunmtaimd with anti-Le' and anti-Lsb M
o
m
.Pam1A: Snnpk. from p.tknta d . u M as
gonuim L d a - b - ) indbidualr based on saliva M a phenotype and a1 4FT nttvfty. Staining of bands migrating ar Le' and Lo' glydlpldr
lmmunortaimd by the reapectlve MoAbr In RBC of E,. but not in M N ~wm
. weak. The nature of the fast-migrating band in aome individualr is
obacum but could muk from variation. in the rphingblipid portlon of th.molecuiea or from ataining of the H t y p l pentaoaylcarrmlde region
( d d b e d in the Materiair and Method. d o n ) . Panel B: Sampln from patient. clauifled ar nongenuine Le(a-b-) individual.. Standard
glycolipids were extracted from 0 Le(a+b-) and 0 Le(a-b+) RBC. Co,, CO,, TrCo, and TrCo,: colon tirrue donora. E, and TrB,: bladder tirrue
-
donor. R, RBC; S, aerum; T, tirsue; On,orcinol rtaining of cancer tiaaue from Co, and TrCo,. b n e r labeled Le* and Lebare autoradlogramr of
plat- developed In CMW 0:35:8and immunortalned with anti-Le' (fromthis laboratory) and anti-Leb(fromChemblomed) YoAba; deacribed in
the Materiala and Method. &ion.
The demonstration of small quantities of Le' and Leb
antigens on RBC., but not serum, of a bladder carcinoma
patient classificd as genuine Le(a-b-) raises the possibility that the antigen originates from Lewis a and b antigcn-
positive tumors. They may therefore constitute authentic
"tumor markers" similar to thosc observed for sialosyl
derivatives of Le' antigen.am In the case of glycolipids,
these may be more conccntrated on RBC as compared with
From www.bloodjournal.org by guest on February 6, 2015. For personal use only.
_-.
lV 13 U O l N V 8
-
.
.
..
A-
W&l
From www.bloodjournal.org by guest on February 6, 2015. For personal use only.
DIFFERENTIAL TISSUE EXPRESSION OF LEWIS ANTIGENS
patients may have Lewis antigens in tissues or secretions,
because they probably are genetically identical to Lewispositive individuals. Owing to some unknown rearrangement of their RBC membrane, however, their RBC cannot
agglutinate with anti-Lewis sera although chemically they
contain detectable quantities of Lewis active glycolipids.
The three persons we identified as nongenuine Le(a-b-)
individuals all had terminal cancer. This is in accordance
with previous reports of occurrence of a RBC Le(a-b-)
phenotype in individuals previously typed as Lewis positive
in association with changed biologic conditions such as
pregnancy,= alcoholic cirrhosis and pancreatitis? hydatid
cysts,25and various carcinomas.M~”
The finding of a1 + 4FT activity leading to formation of
Lewis antigens in vivo in genuine Le (a-b-) individuals is
of interest because such persons are assumed to be homozygous for an inactive allele le at the Lewis gene locus. The
operational definition for Lewis-positive and Lewisnegative individuals does not specify any mechanism for the
observed differential expression, however. In view of our
results, the distinction between Lewis-positive and Lewisnegative individuals does not appear to be qualitative and
alternate mechanisms for differential expression might be
proposed as follows.
First, the present findings may reflect differing tissue
distribution of the Lewis FT in Lewis a o r b antigen-positive
individuals as compared with Lewis a o r b antigen-negative
individuals, resulting in the absence of antigen activity in
saliva but presence of antigen activity in certain tissues.
This is consistent with results from human colon in which
high a1 + 2FT activity was observed in cecum and low
activity of the same transferase was observed in rectum,4l
with results from bladder tissue showing synthesis of Leb
antigens in urothelium from individualswho have Le(a+b-)
RBC,” and with results from rats exhibiting variable tissue
expression of sialyltransferase ( > 50-fold), with the highest
levels in liver and the lowest in brain.”
1395
Second, the observed pattern of expression may indicate
the existence of a hitherto-unrecognized a1 --* 41T not
associated with the Lewis gene. Such an enzyme may be
detectable only in Lewis-negative individuals, however, in
whom it is not masked by the Lewis enzyme. The a1 + 2
ITSare examples of similar transferases that are coded by
two different structural genes.
Third, a probably more likely alternative involves an
analogy to the blood group A subgroups. Some members of
such subgroups inherit an A gene coding for an A transferase with high acceptor affinity (A, transferase), whereas
others inherit a gene coding for an enzyme with lower
acceptor affinity and more restricted substrate specificity
(A, transferase), yielding different levels of cell surface A
antigen characterizing the phenotype.44345
Although current
knowledge regarding number and specificities of a1 +
%FTs is a m b i g u o u ~ , ~a1
- ~ .+
~ % transfer apparently is
ascribable to the same protein, based on kinetic data? In
addition, in recent studies using gene expression data in
which human DNA was transfected into COS-1 cells, an
enzyme with both a1 + 3-(N-acetyllactosamine as acceptor) and a1 + 4FT activity was identified and expressed
(John B. Lowe, personal communication, March 1990). By
similar techniques, an a1 + 3FT activity with no a1 + 4FT
activity was recently identified.47If the number of FT genes
is limited to the presently identified Lewis (a1 + %FT) and
“ X ’ (a1 + 3 R ) , the Lewis gene probably comprises several variants similar to the blood group A genes, with differing relative reactivities to type-1 and type-2 chain acceptors.
The specific mechanism for the expression of Lewis a and
b antigens in Lewis-negative individuals is complex and still
unresolved. Although a few individuals were studied, however, we conclude that RBC and saliva Lewis blood group
phenotype do not provide complete information regarding
the Lewis phenotype in epithelia in general, most likely
owing to Lewis gene-encoded transferase proteins with
variable activity.
REFERENCES
1. Oriol R, Le Pendu J, Mollicone R: Genetics of ABO, H,
transferase from the blood group Le gene-specified a - % - ~ Lewis, X, and related antigens. Vox Sang 51:161, 1986
fucosyltransferase in human milk. Biochem SOCTrans 10:445,1982
8. Eppenberger-Castori S, Lotscher H, Finne J: Purification of
2. Kumazaki T, Yoshida A Biochemical evidence that secretor
the N-acetylglucosaminide a(1 + %)fucosyltransferase of human
gene, Se, is a structural gene encoding a specific fucosyltransferase.
milk. Glycoconjugate J 6:101,1989
Proc Natl Acad Sci USA 81:4193,1984
9. Holmes EH, Ostrander GK, Hakomori S: Enzymatic basis for
3. Betteridge A, Watkins WM: Variant forms of a - 2 - ~ the accumulation of glycolipidswith X and dimeric X determinants
fucosyltransferase in human submaxillary glands from blood group
in human lung cancer cells (NCLH69). J Biol Chem 2607619,1985
ABH “secretor” and “non-secretor” individuals. Glycoconjugate J
10. Holmes EH, Ostrander GK, Hakomori S: Biosynthesis of
2:61,1985
the sialyl-Le“ determinant carried by type 2 chain glycosphin4. Betteridge A, Watkins WM: Acceptor substrate specificities
golipids (IV3NeuAcIII’FucnLc4, VI’NeuAcV’FucnLc,, and
of human a-2-~-fucosyltransferasesfrom different tissues. Biochem
VI’NeuAcIII’V3Fu~nLc6) in human lung carcinoma PC9 cells. J
SocTrans 13:1126,1986
Biol Chem 261:3737,1986
5. Prieels J-P, Monnom D, Dolmans M, Beyer TA, Hill RL:
11. Hanfland P, Kordowicz M, Peter-Katalinic J, Pfannschmidt
Copurification of the Lewis blood group N-acetylglucosaminide a1
G, Crawford RJ, Graham HA, Egge H: Immunochemistry of the
+ 4fucosyltransferase and an N-acetylglucosaminide a1 + 3fucoLewis blood-group system: Isolation and structures of Lewis-c
syltransferase from human milk. J Biol Chem 256:10456,1981
active and related glycosphingolipids from the plasma of blood6. Johnson PH, Yates AD, Watkins WM: Human salivary
group 0 Le(a-b-) nonsecretors. Arch Biochem Biophys 246:655,
fucosyltransferases: Evidence for two distinct a-3-~-fucosyltrans1986
ferase activities, one of which is associated with the Lewis blood
12. Watkins WM, Morgan WTJ: Specific inhibition studies
group Le gene. Biochem Biophys Res Commun 100:1611,1981
relating to the Lewis blood group system. Nature 180:1038,1957
7. Johnson PH, Watkins WM: Separation of an a-3-~-fucosyl13. Rege VP, Painter TJ,Watkins WM, Morgan WTJ: Isolation
From www.bloodjournal.org by guest on February 6, 2015. For personal use only.
1396
of a serologically active fucose-containing trisaccharide from human blood group Le" substance. Nature 204740,1964
14. Marr AMs, Donald ASR, Watkins WM, Morgan WTJ:
Molecular and genetic aspects of human blood group Leb specificity. Nature 215:1345,1967
15. Grollman EF, Kobata A, Ginsburg V An enzymatic basis
for Lewis blood types in man. J Clin Invest 48:1489,1969
16. Watkins WM: Biochemistry and genetics of the ABO, Lewis,
and P blood group systems, in Harris H, Hirschhom K (eds):
Advances in human genetics, vol10, New York, NY,Plenum Press,
1980, p 1
17. Bjork S, Breimer ME, Hansson GC, Karlsson KA, Leffler H
Structures of blood group glycosphingolipids of human small
intestine: A relation between the expression of fucolipids of
epithelial cells and the ABO, Le and Se phenotypes of the donor. J
Biol Chem 2626758,1987
18. Sakamoto J, Yin BWT, Lloyd KO: Analysis of the expression
of H, Lewis, X, Y, and precursor blood group determinants in
saliva and red cells using a panel of mouse monoclonal antibodies.
Mol Immunol21:1093,1984
19. Limas C: Detection of urothelial Lewis antigens with monoclonal antibodies. Am J Pathol125:515,1986
20. Hirano K, Kawa S, Oguchi H, Kobayashi T, Yonekuni H,
Ogata H, Homma T Loss of Lewis antigen expression on erythrocytes in some cancer patients with high serum CA19-9 levels. J Natl
Cancer Inst 79:1261,1987
21. Ernst C, Atkinson B, Wysocka M, Blaszczyk M, Herlyn M,
Sears H, Steplewski Z, Koprowski H: Monoclonal antibody localization of Lewis antigen in fixed tissue. Lab Invest 50:394,1984
22. Yazawa S, Asao T, Izawa H, Miyamoto Y, Matta KL: The
presence of CA19-9 in serum and saliva from Lewis blood-group
negative cancer patients. Jpn J Cancer Res (Gann) 79538,1988
23. Hammer L, Mansson S, Rohr T, Chester T, Ginsburg V,
Lundblad A, Zopf D: Lewis phenotype of erythrocytes and Leb
active glycolipid in serum of pregnant women. Vox Sang 40:27,
1981
24. Stigendal L, Olsson R, Rydberg L, Samuelsson BE: Blood
group Lewis phenotype on erythrocytes and in saliva in alcoholic
pancreatitis and chronic liver disease. J Clin Pathol37:778,1984
25. Makni S, D a h AM, Caillard T, Compagnon B, Le Pendu J,
Ayed K, Oriol R: Discordance between red cell and saliva Lewis
phenotypes in patients with hydatid cysts. Exp Clin Immunogenet
4:136, 1987
26. Phipps RF, Perry PM: Lewis negative genotype and breast
cancer risk. Lancet 1:1198,1989
27. Petrakis NL, King MC, Dupuy ME: Lewis negative genotype
and breast cancer risk. Lancet 2:41,1989
28. Rowe GP, McGregor M, Napier JAF: Lewis negative
genotype and breast cancer risk. Lancet 2:227,1989
29. omtoft TF, Wolf H, Watkins WM: Activity of the human
blood group ABO, Se, H, Le, and X gene-encoded glycosyltransferases in normal and malignant bladder urothelium. Cancer Res
48:4427,1988
30. Holmes EH, Hakomori S, Ostrander G K Synthesis of type 1
and 2 lacto series glycolipid antigens in human colonic adenocarcinoma and derived cell lines is due to activation of a normally
unexpressed p l + 3N-acetylglucosaminyl-transferase.J Biol Chem
262:15649,1987
31. Yu RK, Ledeen RW: Gangliosides of human, bovine, and
rabbit plasma. J Lipid Res 13:680,1972
32. Magnani JL, Smith DF, Ginsburg V: Detection of ganglio-
qRNTOFT ET AL
sides that bind cholera toxin: Direct binding of lzI-labeled toxin to
thin-layer chromatograms. Anal Biochem 109:399,1980
33. Kannagi R, Nudelman E, Levery SB, Hakomori S: A series
of human erythrocyte glycosphingolipids reacting to the monoclonal antibody directed to a developmentally regulated antigen,
SSEA-1. J Biol Chem 257:14865,1982
34. Young WW Jr, Johnson HS, Tamura Y, Karlsson K-A,
Larson G, Parker JMR, Khare DP, Spohr U, Baker DA, Hindsgaul
0, Lemiew RU: Characterization of monoclonal antibodies specific for the Lewis A human blood group determinant. J Biol Chem
258:4890,1983
35. Messeter L, Brodin T, Chester MA, Karlsson K-A, Zopf D,
Lundblad A: Immunochemical characterization of a monoclonal
anti-Lebblood grouping reagent. Vox Sang 46:66,1984
36. Nudelman E, Fukushi Y, Levery SB, Higuchi T, Hakomori
S: Novel fucolipids of human adenocarcinoma: Disialosyl Le"
antigen (II14FucII16NeuAc-IV3NeuAcLc,)
of human colonic adenocarcinoma and the monoclonal antibody (FH7) defining this
structure. J Biol Chem 261:5487,1986
37. Hsu S-M, Raine L, Fanger H: Use of avidin-biotinperoxidase complex (ABC)in immunoperoxidase techniques: A
comparison between ABC and unlabeled antibody (PAP) procedures. J Histochem Cytochem 29577,1981
38. omtoft TF, Wolf H, Clausen H, Hakomori S, Dabelsteen E:
Blood group AB0 and Lewis antigens in fetal and normal adult
bladder urothelium: Immunohistochemical study of type 1 chain
structures. J Urol138:171,1987
39. Saito T, Hakomori S: Quantitative isolation of total glycosphingolipids from animal cells. J Lipid Res 12:257, 1971
40. Magnani JL, Steplewski Z, Koprowski H, Ginsburg V
Identification of the gastrointestinal and pancreatic cancerassociated antigen detected by monoclonal antibody 19-9 in the
sera of patients as mucin. Cancer Res 435489,1983
41. omtoft TF, Greenwell P, Clausen H, Watkins WM: Regulation of the oncodevelopmental expression of type 1chain ABH and
Leb blood group antigens in human colon by ( Y - ~ - Lfucosylation.
Gut (in press 1991)
42. omtoft TF, Wolf H: Blood group A B 0 and Lewis antigens
in bladder tumors: Correlation between glycosyltransferase activity
and antigen expression. Acta Pathol Microbiol Immunol Scand
Suppl4:126, 1988
43. Paulson JC, Weinstein J, Schauer A: Tissue-specific expression of sialyltransferases. J Biol Chem 264:10931,1989
44. Schachter H, Michaels MA, Tilley CA, Crookston MC,
Crookston JH: Qualitative differences in the N-acetyl-D-galactosaminyltransferases produced by human A' and A* genes. Proc
Natl Acad Sci USA 70:220,1973
45. Clausen H, Holmes E, Hakomori S: Novel blood group H
glycolipid antigens exclusively expressed in blood group A and AB
erythrocytes (type 3 chain H): 11. Differential conversion of
different H substrates by A, and A2 enzymes, and type 3 chain H
expression in relation to secretor status. J Biol Chem 261:1388,
1986
46. Watkins WM, Greenwell P, Yates AD, Johnson PH: Regulation of expression of carbohydrate blood group antigens. Biochimie
70:1597,1988
47. Stanley P, Kumar R, Potvin B, Howard D: Use of CHO
glycosylation mutants to clone mammalian glycosyltransferases, in
Sharon N, Lis H, Duksin D, Kahane I (etis): Proceedings of the
Xth International Symposium on Glycoconjugates, Jerusalem,
Israel, 1989, (abstr 54)
From www.bloodjournal.org by guest on February 6, 2015. For personal use only.
1991 77: 1389-1396
Differential tissue expression of the Lewis blood group antigens:
enzymatic, immunohistologic, and immunochemical evidence for
Lewis a and b antigen expression in Le(a-b-) individuals
TF Orntoft, EH Holmes, P Johnson, S Hakomori and H Clausen
Updated information and services can be found at:
http://www.bloodjournal.org/content/77/6/1389.full.html
Articles on similar topics can be found in the following Blood collections
Information about reproducing this article in parts or in its entirety may be found online at:
http://www.bloodjournal.org/site/misc/rights.xhtml#repub_requests
Information about ordering reprints may be found online at:
http://www.bloodjournal.org/site/misc/rights.xhtml#reprints
Information about subscriptions and ASH membership may be found online at:
http://www.bloodjournal.org/site/subscriptions/index.xhtml
Blood (print ISSN 0006-4971, online ISSN 1528-0020), is published weekly by the American
Society of Hematology, 2021 L St, NW, Suite 900, Washington DC 20036.
Copyright 2011 by The American Society of Hematology; all rights reserved.