Adenosine and Adenine Nucleotides Inhibit the
Autonomous and Epidermal Growth
Factor-Mediated Proliferation of Cultured
Human Keratinocytes
Paul W. Cook, Nina M. Ashton, and Mark R. Pittelkow *
Scio s Nova Inc. , Mo untajn Vi ew, Ca li fornja; and " Departments of Dcrmatology, B iochemistry, and Mo lecular Biology, Mayo C lini c/
Foundation, Rochester, Minncsot:t, U .S.A.
Previous investigations have shown disparate effects
of adenine nucleotides on epidermal cell proliferation. Our present study demonstrates that adenosine
and its related nucleotides (ATP, ADP, AMP) are
antiproliferative for nOrlllal human epidermal keratinocytes cultured in the absence or presence of
exogenous epidermal growth factor. Furthermore,
the inhibitory effects of these compounds occur at
concentrations less than 100 p.M, are reversible, and
do not affect the viability of the keratinocyte cultures. Our current investigation also demonstrates
that both selective and nonselective adenosine receptor agonists ' are themselves approximately as potent
as keratinocyte proliferation inhibitors, but are all
less potent inhibitors than adenosine. These observations are consistent with the theory that adenosine
mediates its antiproliferative response Ilia a novel or
more poorly characterized adenosine purinoreceptor
subclass . Moreover, our present study demonstrates
that ATP and ATP-,},-S are significantly more potent
antiproliferative agents than either a,/3-methylene
ATP or /3,'}'-methylene ATP. Based on previous studies that have demonstrated that P zy purinoreceptors
possess this type of ligand specificity and that the P Zy
purinoreceptor may be expressed by keratinocyte
cultures, we propose that ATP may mediate its antiproliferative effects via this purinoreceptor. Collectively, our results indicate that adenosine and adenine nucleotides abrogate exogenous epiderll1.al
growth factor-dependent and -independent keratinocyte proliferation at submillimolar concentrations
and may be ill1.portant physiologic regulators of keratinocyte growth ill "i"o. Further, th'e se results suggest that these or related compounds may have application as treatments for epidermal proliferative
pathologies in which the epidermal growth factor
receptor-signaling pathway has been activated. Key
words: purine/lU.cieoside/epiderrnal/gl'Owtlr. ] Invest Dermatol 104:976-981, 1995
A
and in some cases was performed on cells treated under culture
conditions (elevated calcium and high cell density) that normaliy
pr.omote qui escen ce and cellular d ifferentiation . Still other investigatio ns ha ve suggested that adenosine, deoxyadenosine, and adeni.J1e nucleotides are antiproliferative for both human and porcine
keratinocyte explants, albeit at high concentrations of the adenin e
llucleosides/nucleotides [16-19]. The earliest investigations [16,17]
exa mining the effects of adenine nucleosid es/ nucleotides on human
epj dermal cell outgrowth are also diltjcult to interpret because the
epjdermal exp lants were cu ltured using ill-defined cell culture
conditions (elevated calcium and bovi.ne serum) that promote
keratinocyte differentiation and quiescen ce as w ell as significant
fi broblast contam in ation. In addition, these latter studies [16,17]
did not assess dire ctly tb e efFects of adenosine and its related
nucleotides o n changes in cell proliferation over ex te nded periods
Or de termin e the direct e ffects on thymidine incorporation .
Our present investigation explored the effects of both adenosine
and adenine nucleotides and their related receptor-specific analogs
on the amphiregulin-dependent autonomous [20-24] and epiderrna l growth f.,ctor (EGF)-d e pendent proliferation of seco ndary
kel'atinocytes ill lIi/ro . In co ntrast to a previolls report [8] , w e found
that all of these compound s can reversibly inhibit keratinocyte
number of investigations have shown that adenosine
and adenine nucleotid es ca n modulate physiologic
responses through purinoreceptor-dependent signali.ng pathways [1-7], and examples of both stimulation and inhibition of ce ll prolife ration by these
compounds have bee n doclUnented [8-17]. Previous studies have
reported opposing activities of purin e nucleotides on epid e rmal cell
proliferation . Adenos in e triphosphate (ATP) concentrations of
1-100 fLM have been demonstrated to stimulate thymid in e incorporation in cultures of normal human epiderma l keratinocytes [8].
However, the reported ATP-depe ndellt stimulation of thymidine
in corporation in these keratinocyte c ulture s was low in magnitud e
Manuscript rcceivcd Ju ly 29, 1994; fina l revision received February II ,
1995; acceptcd for publica tion February 20, 1995.
Rcprint requests to: Dr. Paul W . Cook, V A Mcdi ca l Center Dermato logy Service (11 C2- P), 3710 S.W. U.S. Veterans HospitaJ Road, Portland,
OR 97207.
Paul W . Cook's present address: VA Medi ca l Center, Dermato logy
Service, Portland, OR 97207.
Abbreviations: All, amphircgulin; C HA, N" cycloh cxyl adenosine; 11'3'
inositol trisphosphate; NECA , 5' -N-ethylcarboxal11ide adcnos in e; R-PIA.
(-)-N"-(2-phcnyljsopropyl) adenosinc.
0022-202X/95/S09.S0 • SSDIO022-202X(95)00072-S • Copyright © 1995 by The Society for Inv estigative Dermato logy. Inc.
976
VOL.
104, NO.6
JUN E 1995
ADEN I NE NUC LEOSIDE/ TIDE I N HIBITI ON OF KERATI NOCYTE GR OWTH
p ro life ration at con centrations less t h an 1 00 fLM . T h e res ul ts of our
c urre nt investigation , in conjun c tion with t h e r es ul ts of othel'
studi es d e m o n strating aden osin e -indu ced cycli c AMP (cAMP) form a tion in e pide rmis and ATP-indu ced IP 3 fo rm ation in hum an
e pidermal ce lls [8, 18,19,25- 27], are con sistent with purinore cepto r-med iate d g rowth inhibiti on in keratin ocytes. A lth o u g h t h e
res ults of o ur c urre nt inv estigatio n are pre liminary and warrant
furth e r in vestiga ti o n , these findings su ggest that adenin e nucl eosides a nd nucl eotid es m ay b e importan t physiologi c re gulators of
k eratin ocyte g rowth ;/1 1) ;110 .
977
ri.
(;
u
.~
-ci
.c
I-
'0
~
MATEIUALS AND METH ODS
0
bovine in sulin ,
hUI11 011l
tra nsfe rrin , and bovine pituitary extract as a
:r:
0
g
Matcrials M ed ium 154 (modified MCD B 151) was purchased fr0111
Cascade Bi ologics, In c. (Po rtland, OR) and was used to cul tivate selectivel y
no rmal fores kin kcratin ocytcs , as described previously [24,28 - 30J. Mediu.J11
154 was supplemented w ith human recombin an t EGF, hydrocortisone,
<1>
'C
~
:J'
[
CO l'l1 -
bined supp lem ent to yield "complete medium. " M C DB 153 [28J was u sed
as the basal m cdium for all human keratinocyte clo nal-density assays.
Defi ned medium co ntaining hydrocortisonc, insulin , and EGF wa s used foc
clonal-den sity experim ents . "Sta nd ard " medium refers to basa l MCDB 1 S3
o r mcdium 154 supplcmented with hydrocortisone o nl y. I-Iuman recomb.inan t EGF was purchased fro m UBI (Lake Placid, NY). Hepa rin (porc in e
intestinal mu cosa) was purchased fi'om Celsus Labs (lot PM-1 9286),
disso lvcd in solu tion A (HEPES-buffered normal sa li ne and glucose , pH 7.4)
at 100 mg/ ml , and stored at -20°C. Adenosine, ATP, AD P, AMP, inosine,
inosin e m onop hos phate, in osine tripho sphate, guanosine, guan osine mOI1Ophosphate, guan os ine triphosphate, N ECA (5' -N-ethylcarboxa mide aden os in c), N ° cyclohexyl adenosin e (C HA), (-)-N "-(2-phenyli so propyl) adenos in e (R-PIA) , a,J3-methylene ATP, a.nd J3;y-methylene ATP were a ll
purchased fi'om Sigma C hemi ca l Co . (St. Lo uis, MO) . ATP-y-S WaS
purcha sed from Boehringer Mannheim. [' H) Methyl thymidin c (6.7 Ci/
mmol) was purchased from New England Nucl ear (l3 osl'O n , MA) .
Cclls and Ccll Culturc Primary cultures of no rmal fo reskin keratinocytes were esta blished liom newbo rn fo reskin tissue as described previ ousl y
[24,28 - 30]. Kcratin ocytes were ro utinel y cultured in com ple te m edium 1 S4
or complete M C DB 153 as described above.
Thymidinc Incorporation Assays First-passage keratinocytes werc
plated into six -well dishcs at 10,000 cell s/cm 2 in 2 ml compl ete mediut11
154. Keratinocytes were cultured for 24 h and subsequentl y was hed twice
w ith 3 ml of soluti on A. T he cell s were thcn cultured for an additi onal 24
h in 2 ml of sta nda rd medium 154, after whi ch expcrim ental f:lcto rs were
added to each well as indi cated in the fig ure legends. After 30 h of cul ture,
2 fLC i [' H)methyl thymidine was added to each well , and the cell s were
subseq uentl y cultured fo r an additiona l 24 h. T he ceUs were then washed/
fix ed three times ill cold 10% trichloroaceti c ac id and once wi til water. Si"
hundred Inicro litcrs of 0.2 N Na O H containin g 40 fLg/ ml herring spernl
DNA was added to each well and agitated on an o rbital shaker for 45 nl in.
Three hundrcd micro liters of each well 's con tents was then counted ill a
liquid scintiU ation co unter ill 6 ml of sciJl tillallt.
Ccll Prolifcration Assay First-passage keratin ocytes were plated in to
si.<-well dishcs at 25 ,000 ce ll s/cn/ in 2 ml co mpl ete med ium 154.
Keratin ocytes were then cultured for 24 h and subsequentl y washed twi e
w ith 3 ml o f solutio n A. Two millili ters of sta nd ard medium 154 was add e d
to each well , and experimental f.,cto rs were lh en added as indica ted in the
figu re legend s. After 5 d of culture, the cell s were tI)'psinized a l\d
res uspcnded after cen trifugation, and the tota l number of cell s in each we ll
was determincd by counting in a hemocytometer.
Clonal Growth A.says To test the co ncclltrati on dependence of adenosine and its rela ted nucleotides on the clo nal growth of kerati nocytcs, 'we
perform ed a clonal- density growth assay as dcscribed previonsly [23) . To
test the reversibility of adenosin e ;\IId adenine nu cleotide- mcdiated inhibition on clonal keratinocytes , w e perform ed a modifi cation of a previ ollsly
described protoco l [20). BrieAy, keratin ocytes werc pla ted at 1000 cells per
well (six-we ll cu lture plates) in complete med ium 154 . After 24 h ot
culture , adenosin e, AMP, ADP, and ATP wcre ad dcd to some of the w ell s .
T he ce ll s we re cultured fo r an additional 24 h and subsequ cntly Wet'e
wa shed tw ice with ph osphate-buffe red normal saline so lu tio n. T he m edium
in each well WaS th cn rep laced with complete m edium 154 containing the
origina l concentra tion of adenin e nucleoside/nucl eotid e o r w ith compl ete
medium 1 54 with out these compounds, and the cells were then cultured for
an additi ona l 8 d. Co lo ni cs were visualized by fixati on and sta ining as
described previ o usly [201- As positi ve contl'ols, some ce ll cu ltu res weI'"
120
~
6
rC
b
100
80
Qj
.0
E
::::l
Z
Qj
0
0
0
0
~
9..
[
LJ I
DayO
0
::>
»
»
s::
'1J
Cl.
<1>
::>
0
»
0
'1J
~
'1J
:r:
-lB
~
:J'
(J)
:J'
<1>
Day 5
F igurc 1. Effcct of adcnosinc and adcninc nuclcotidc. on thc
autonomous prolifcration of kcrat ino cytcs. a) Keratinocytes were
cul tured and assayed as described for tllytTlidin e in corporation assays in
Mllteria ls nlld Metlwd.'. Experim en tal f., ctors were control (no addition),
heparin (300 fLg / ml), aden osine (1 mM), and AMP (1 mM) . Each point
represents the average [' H) thymidine in corporation (Thd . in corp .; cpm) of
three determinations:!: SEM. iI) Keratinocytes were culrured and assayed as
described for cell proli feration assays in I\ll aterials nlld MetllOds. Experim ental
f.,ctors were co ntro l (no additio n), adenosin e (1 mM) , AMP (l mM), ADP
(1 mM ), ATP (1 mM), and heparin (300 fLg/ m l). Each point represents ti,e
average ce ll number/ well of three determinatio ns:!: SEM.
carri ed throu gh the assay but were not treated with any experim en tal
agents .
R ESULTS
Adenosine and Adenine Nucleotides Inhibit the Autonomous Proliferation of Human Epidermal Keratinocytes In
t h e fi r st e xpel;me nt, keratinocytes were c ultured unde r conditions
t hat promote amphireguli n-dependent auton o mous proliferation
[20-24,3 1]. T h e effects of h e parin, aden osine, and A MP on DNA
synth esis of k e ratinocytes were assessed b y [3H]m e th yl thymidine
inco rp onltion , and t h e r es ults of t h ese ini tial exp e rim e nts are sh own
in Fig la . T h e results demon str ate t h at, in a m anner simil ar to
h epa rin , both aden osiJ1 e and AMP suppress thymidine in corpo r at io n in a u ton omo u sly proliferatin g k e ratinocytes. To ru le out
pote ntial confoundin g e ffects of adenosin e and AM P on ["H]methyl
th y midin e up take and in corporation b y ke ratinocytes, we performed a simil ar exp e l;m e n t a nd assessed the effects of h eparin a nd
adenosin e and a d e nin e nucl eotid es (AMP, ADP, ATP) o n the actual
cell numb er per well after 5 d of autonomous g r owth in culture .
T h e r es ults of t hi s exp erim e n t (Fig lb) d e m o n strate that adenosin e,
AMP, AD P , and ATP exerte d sig nifica nt antiproliferative effects o n
a uto n o m o u sly pro liferating k e ratin ocyte c ul t ures. Moreover, this
latter exp e rim e n t cl ea rl y indi cates t h at t h e inhibitory effects of
978
THE JOURNAL OF INVESTIGATI VE DER.MATOLOGY
COOK ET AI-
250
•
ATP
0
•
ADP
X
--0-
AMP
C')
C')
b 200
x 175
E
Cl. 150
oS
ci.. 125
E
a. 150
~
ci.. 125
0
0
0
u
.~
c
-0
~
0 .3 1 3
0 .3 1 3
0.31 3
L--.J
L--.J
L--.J L--.J
»
a.
Q
~
"U
»
s:
"U
(l)
:::J
o
S·
U>
0 .33
~
Q.
(l)
F igure 2. Effect of adenosine and adenine nucleotides on the
proliferation of n eon atal hum an epidermal keratinocytes in the
presence or absence ofEGF. Kerati nocytes were c ultured a nd assayed as
described for thymidine inco rp o ratio n assays in F ig la. ATP, AMP, and
adenosine were tested at 1 mM concentratio n . Contro l desig nates no
additio n of 11llcleosidc o r nucl eotid es. All experimental conditi o ns we re
performed with the indi cated concentrations ofEGF. Eac h po in t represents
the average eJ-f]th ymidin e incorpo rati on (Thd. in corp.; cpm) of three
dctcnllinations :!: SEM.
cdenosin e and the aden.ine nucl eo tid es o n thym idin e in co rp o ration
(Fig 1a) co rrelate with the expected response of these compound s
on cell prolife ration. We also used the propidium iodid e exclusio n
assay [32] and determined that inhibi tory co ncentrati o ns of ade nosine and its re late d nucl eotide. (l mM) did not pe nn ea bilize or
othe rwise indu ce cytotoxicity in o ur keratinocyte c ultu res, as the
ce ll viabili ty of the treated cells co nsistentl y exceeded 99% (data not
shown) .
We also investigated til\! efFects of g uano sin e and inos ine, their
related nucleotid es, uridin e triphosphate, and cy tosine triphosphate
o n the autonomous pro li feration of keratinocytes. Despite the
stru ctural similatity of some of the se co mp o unds , the res ults
de monstrated that hi g h co ncentratio ns (1 mM) of t hese agents did
not significa ntly inhibit thymidine in co rpo ra tio n in auto nomou s
keratinocytes (data not shown) wh en co mp ared with the inlubitory
activity of adenosine and adenin e nucl eotides.
Adenosine and Adenine Nucleotides Inhibit the Proliferation of Human Epidermal Keratinocytes in the Presence or
Absence ofEGF Our previo us inv estigatio ns h ave de mon strated
that EGF or tra nsform ing gro wth factor-a lph a is capable of C0 11l1teracting the inhibitory effect of he parin o n auto no mou s ke ratinocyte c ultures, presumab ly by circumventing the keratin ocyte's
requirement fo r e nd oge no usly produced amp luregulin (AR) or
other he pal;n-se nsitive growth factors [20,21]. W e therefore exami ned w he ther EGF co uld also override inh.ibi tion of keratinocyte
autonomou s grow th mediated by aden os in e or adel1in e nucl eotid es .
Autonomous ly pro li ferati ng kerati nocytes were ex posed to ATP,
AMP, or ade nosine in the presence or absen ce of EGF at con centrations known to reverse the inhibitory effects of he parin. T he
results of thi s expe riment are sho w n in Fig 2 and demonstrate that
adenosine and its related nucl eotides inhibited keratinocyte proliferation in the presence and absence ofEGF. G reater conce ntratio ns
of EGF were also tested in thi s assay and fa iled to reverse the
in hibitory e ffe cts of adenosine and the adenine nucl eotides (data not
shown) . T he res ul ts described above were o bta in ed with purin e
nucl eos ide / nucl eo tide concentrations 10- to 50-fold hi gh er than
that req uired to obtain ma ximal growth inh.ibi tio n. T hus, it re mains
to be determin ed w he th e r EGF ca n co unte ract the antipro li fer ative
effects of lower concentratio ns of these inhibi tors.
100
75
.c
50
I
25
f-
o~~~~~~~~--~~-L~~~
EGF (ng/ml)
225
C')
0
0
.01
.1
mM
Figure 3. Conccntration-dcpcndcnt cffccts of adcninc nuclcotidcs
on thc a utonomous prolifcration of keratinocytcs . Keratinocytcs
wel,e c ultured and assayed as described in Fig la. Expe rime ntal factors
(ATP, ADP, AMP) were tested at the indicated conce n tra tio ns. Each po int
re p resents the average [3J-I]thymidin e in corpo ra tion (Thd . in corp .; c pm) of
three d eterminations:!: SEM.
Concentration-Dependent Inhibition of Autonomous Hum.an Epidermal Keratinocyte Proliferation by Adenosine
and Adenine Nucleotides and Their ·Related Agonists It
has been esta blished previously that high concentrations of ATP
(greater th an 1 mM) ma y m ediate responses via perm eabilization in
certain target cell s, whe reas lo w e r concentrations o f ade nin e
nu cleosid es and nucleotides are thought to m ediate their effects
thro ug h purinoreceptor-mediated activation of intracellul ar sig naltraJlsd uc ti o n pathways. Indeed, ATP-induced pro du ction ofTP) and
e le vatio n of in trace ll ular cal cium have bee n reported in human
kera tino cyte cu ltures [8], suggestin g that P 2y recepto rs may be
expressed by th ese cells. With referen ce to th ese prior o bserv ations,
we pe rform ed experim ents to es tab li sh the range of co ncentrations
over whi ch the ade nin e nucl eotides exe rted their antiproliferative
actions on auto nomou sly prolife ra ting keratinocyte cultures. F igUl"C 3 demonstrates the do se-de pend ent inhibitio n ind uced by ATP,
AD P, and AMP in keratinocyte cnltures . T he results show tha t each
of the aden in e nucleotide s m ediated grow th inhibi tio n in autonom o usl y proliferating keratin ocytes with apparent 50°;', inhibi tion
ra n ges of 8 - 20 MM.
A vari e ty of adenos in e purinorece ptor agoni sts have been synthesized , and some hav e been d emon strated to exert selective
ac tiv ity o n partic ular adenosin e purin oreceptor subtypes (Ai, A2).
B ecause no study to date has examin ed the relative pote ncies of
th ese compounds on keratinocyte pro life ration, we exa min ed th e
e£!:ects of the Ai purinoreceptor-selective agonists (R-PIA and
CHA) and of a nonselective agonist (NECA) of the All A2
purinoreceptors. Figure 4a de m o nstrates th e co ncentratiol1depe nd e nt effects of th ese agents on the auto nomo ns proliferation
of keratin ocytes. T he resu lts ind ica te that all of the nucl eos id e
ana logs tested showed dose-depe nd e nt growth inhibitio n . However, no ne of the selectiv e adenosin e purinorecep tor ana logs
appeared to be as pote nt as ade nosine , thu s i.nd icating that a nove l
Ot more poorly chara cterized class of aden osine purinoreceptors
l11 ~y mediate gro wth inhibition in the autonomous keratinocyte
cultures. T lus latte r conclu sion is suppo rted by our inability to
demonstrate specific high-affinity (11M) bil~din g of either radiolabeled CHA or N ECA to the keratino cyte c ultures (data not shown).
It has been re ported that ATP and ATP-,),-S possess greater
affi nity for th e P 2y purinoreceptor subtype [3,6] than do e ither
O',p-me th yle ne ATP o r {3, ,),-methylene AT P, wh ereas these latter
two compo und s are known to be potent P 2 ., re ceptor agoni sts
[3-6] . With regard to the se o bse rvations, w e examined th e effects
of ATP and these ad e nin e nu cleotide analogs on the autonomous
pro life ration of keratinocytes ill IIil ro. Figure 4b dem o nstrates the
cOJl centrati o n-d epe nd ent effects of these reagents o n th e proliferation ofke ratin ocytes. T he res ults indi cate th at both O',{3-me th yle n e
VOL. 104, N O.6 J UNE 1995
250
'?
0
x
225
ADE N INE NU CLEOSIDE/ TIDE INHIBITION OF KERATINOCYTE GROWTH
a
200
____
Adenosine
---0-
R-PIA
C HA
E
c.
979
Adenosine
NECA
B
e-
/.IM--
8
o
0.5
5
.!:
'ci
.t::
~
'"
.1
mM
____
b
400
o
0.5
o
10
50
100
o
500
1000
2000
5
AT P
--0-
ATP-y-S
_
a,P-Molhyleno ATP
p,y-Molhylone ATP
--0-
b
AMP
ADP
~
x
E
c.
B
ATP
.01
.1
mM
Figure 4 . Concentration-dependent effects of adenosine, ATP,
and their related agonists on the autononl0US proliferation of
neonatal h u man epidermal keratinocytes. a) Keratinocytcs were
cultured and assayed as described in Fig la . Adenosine ,md its related
agonists (R- PIA, CHA , NECA) were tested at the indicated concentrations.
Each point represents the average [' HJthymidine incorporation (Thd.
in corp.; cpm) of three determinations ::!: SEM . b) Keratinocytes were
cultured and assayed as described for thym idine incorporation assays ij,
Maler;"ls "lid Met/tads. ATP and its related agonists (ATP-y-S, a,/3-metllylene ATP, /3, y-methylene ATP) were tested at the indi cated concentrations.
Each point represents the 'lVerage [' HJthymidill e incorporation (cpm) of
three determinations ::!: SEM.
ATP and {3, y-methylene ATP e li cited a less potent dos e-d ep end e nt
growth inhibition than that observed for ATP and ATP-y-S. T hu s,
these resu lts are co nsi stent with a mode l wherein the growth
inhibition induced by both ATP and ATP-y-S is mediate d throug h
ce \1-surface P Z Y purinoreceptors, a receptor that ha s been suggested
to be expressed by keratillocyte cultures [8].
Concentration- Dependent Inhibition of EGF-Mediated
Human Epidermal Keratinocyte Colony Formation by
Treatment With Adenosine, Adenine Nucleotides, and
ATP-,),-S Prev ious inv estig ations hav e demonstrated that hi gherdensity ke ratino cyte cultures are ca pab le of AR-dependent autono mous growth, whe reas at clonal d en sity, ke ratinocytes require
exogenous growth factors (e.g., EGF, transforming growth factoralpha, AR or fibrobl ast growth facto rs, and insulin or insulin-like
grow th factor) to initiate co lony formation [20 - 24]. To examine
the potentia l d ensity-dep ende nt diffe rences in the antipro Liferative
effects of these compound s, we investigated the do se-dep ende nt
e ffects of ade nosin e and several adenine nucl eotid es on EGFd e pe nde nt colony formation of clonal- de nsity keratinocyte cu ltures. The resu lts (Fig 5) show that EGF- treated keratinocytes
c ulturec at clonal d e nsity were g rowth in hibite d by adenosine a.ud
by al\ the adenin e nucl eo tides tested. It is interesting that both
adenosine an d AMP appeared to be noticeab ly more potent 011
clonal-density keratinocytes than on high-d ensity cell cu ltut:es
(compare with Figs 3,4). These findings suggest that high e rden sity kera tinocyte cultmes more rapidly modify or transport these
agents (e.g., adenosine deamillase or cellular nucl eosid e transporte rs). [n contrast, both ADP and ATP were shown to be less pote nt
ATP-y-S
Figure 5. Conccntration-dcpcndcnt cffects of adcnosine, adcninc
nuclcotides, and ATP- y-S on thc EGF-lllcdiated colony formation
of kcratinocytcs . Keratinocytes wcre cultured and assayed as dcscribed
und er clonal growtll assays in lIifatelials alld !lifelhods. Thc adcnine nllclcosides and nllclcotides were tcstcd at thc indi cated conccntrations. Each dish
is represcntative of thrcc colony growth detcrminations.
cell-proliferation inhibitors on clonal-density keratinocytes than o n
the high-d en sity keratinocyte cultures (compare with Figs 3,4).
A TP-y-S also was tested on clonal-den~ity keratinocyte cu ltures
and elicited a more potent inhibition of keratino cyte co lony
formation than ATP, even though a similar potency was displayed
b y ATP and ATP-y-S on high-de n sity keratinocyte cultures (see
Fig 4b). The pronowlced d ecreases in app are nt potencies of th ese
agents in the c1ona.! assay may be due to the longer duration of
g rowth exa mine d in the clonal response assay (10 d) comp ared with
the thY111id.ine incorporation assay (30 h) and the inherent instability
of ATP and ADP (but not of ATP-y-S) in soluti on at the cellculture incubation temperature of 37°C. In addition, exposure of
keratinocyte cultures to adeno sine and adenine nucleotides is
reve rsibl e, as keratinocyte c1ona.! cultures treated .with inhibitory
concentrations of these compounds for 24 h can recover and foml
larger colonies (see Fig 6).
DISCUSSION
Previous studies have reported disparate and opposing activities of
purine nucl eotid es on keratlllocyte proliferation. For example, ATP
h as been d e m onstrated to inhibit termlllal differentiation and to
stimulate thymidine incorporation in keratiJl0cyte cu ltures [8],
w hereas other earlier investigations using ill-defin ed cell culture
conditions have suggested that adenosine, deoJ..'y adenosine, and
adenine nucleotid es can inhibit the epidermal outgrowth from both
human and porcin e keratinocyte expl ants [1 6 -1 9]. T h e results o f
ou.r present lllVestigation are in agreem ent with the earlier epider-
980
T H E JOURNAL OF INV EST IGATIVE DERMATOLOGY
COOK ET AL
-1Control
+1+
+1-
Adenosine
AMP
ADP
ATP
Figure 6. Reversibility of adenosine- and adenine llucleotidemediated keratinocyte growth arrest. C lona l growth of human keratinocytes wa s perfonlled as described in Materials alld Metllods. Keratinocytes
were plated at 1000 cell s per well (six-well culture plates) in comp lete
medium 154. After 24 h of culture, adenosine, AMP, ADP, and ATP were
added to some of the wells. The adenine nucl eosides and nucl eoti des were
tested at the following concentrations : adenosine (0.1 mM) , AMP (0.1
mM), ADP (1.0 mM), and ATP (1.0 mM). Keracinocytes were cultured
with the indicated adenjne nucleoside/nucleotide for 24 h , washed, and
subsequently cu ltured in complete medium 154 containin g the corresponding adenine nucleoside/nucleotide (+ / + ), or cultured in compl ete m ed iulll
without these compounds (+ / - ). Con trol keratinocyte cultures (- / - ) not
treated with any experimental agents were also performed. Each dish is
representative of three colony growth determinati ons .
mal exp lant culture studies [16-1 9] and have rigorou sly demonstrated that both the EGF-dependent and - independent (autonomous, AR-dependent) proliferation of secondary keratinocyte
cultures can be inhibited by exoge nous adenosine, AMP, ADP,
ATP, and ATP-'Y-S at submmimolar co ncentrations. These observations, in conju n ction with our findings that the inhibitory effects
of adenosine and its related nucleotides are reversible and do not
penneabilize or induce cytotoxicity of keratinocyte cul tures, suggest that the effects of these age nts on keratinocyte proliferatio n are
appa rently specific an d of potentia.! physiologic relevance. Moreover, as other investigators [33] have proposed for psoriasis, we also
specu late that adenosine, adenine nucl eoti des, or related compounds might be tested as small-mol ec ule-based therapeutics for
epiderma l proliferative pathologies (psoriasis) in which the EGF
receptor pathway is activated [34-37].
Previous investigations hav e shown that adenosine and AMP ca n
act as purinoreceptor liga nd s and subs eq uently elevate intracellular
cAMP [2], and som e investigations hav e demonstrated that these
agents can elevate intrace llular cA MP in epidermal cells [19,25 ,
27,38]. Our current investigation has shown that purinoreceptor
agonists that are reported to b e e ither selective for adenosine type
1 receptor (CHA, R-PLA) or nonsel ective for both the type 1 and
2 receptor (NECA) subclasses are significantly less potent iJlhibitors
of ker.. tinocyte proliferation than is adenosine. Moreover, we have
b ee n unabl e to demonstrate specific high-aftlnity binding sites for
either CHA or NECA in the keratinocyte cultures . Collectively,
these res ults suggest that the inhjbitory effects of adenosine (and
possibly AMP) on keratinocyte proliferation are purinoreceptormediated, possibl y Ilia a novel or more poorly characterized
adenosine re ceptor subclass that e levates intracellular cAMP 311d
does not d emonstrate hjgh-affinjty interaction with CHA or NECA .
Our iJ1Vestigation has demon strated that both ATP and ATP-'Y-S
are antiproLiferative for keratinocyte cultures. Because ATP-'Y-S is
both poorly hydrolyzed and i.neftl ciently transported into the cell,
Qur results suggest that the site of action of these co mpounds on
keratinocytes is the membrane surface. Moreover, we demonstrate
that both ATP and ATP-'Y-S are significantly more potent antiproliferative agents for these cells than is either a,{3-methylene ATP or
{3, y-methylene ATP. Because previous studies have demonstrated
that P2 y purinoreceptors possess this type of comparative ligand
potency/specificity [3,6] and have also suggested that ATP elevates
intcacellular IP 3 and Ca ++ Ilia P 2 y purinorece ptors in keratinocyte
c:u ltures [8], we propose that ATP might m ediate its antiproliferative effects in keratinocytes Ilia this purinorece ptor subtype.
Recent investigations have demonstrated that elevation of intracellular cAMP has the ability to attenuate EGF-initiated signa] transductiOl1 by inhibitiJlg Ras stimulation of Raf-l kinase and subsequently
decrease activation of downstream kin<lses such as mitogen-activated
j?rotein kinase [39,40] . Thus, the results of om current investigation
<\.nd previous obselvations that adenosine and AMP can elevate
intracellular cAMP in intact epidemljs suggest t11at exogenous aden~sine and AMP mjght inhibit keratinocyte proliferation via a cAMPdependent attenuation of the EGF-activated signal-transduction pathway. In addition, it has been demonstrated that treatment of renal
~ pjth e Lial cells or human A431 epidermoid cell line with extracellular
ATP decreases the binding ofEGF to its receptor [13 ,41 ]. Thus, it is
ternpting to speculate that the antiproLiferative action of ATP on
kecatinocyte cu.!tures might be mediated by decreasing the ability of
13GF or of endogenously produced growth factors (transformjng
grow th facto r-a]pha, AR, heparin- binding EGF-like growth fa ctor) to
iJ1teract with EGF receptors.
Our preliminary investigation rai ses new questions co ncerning
the mechanisms by which adenosine and adenine nucl eotides
inbjbit keratinocyte proliferation. Because our cell proliferation
assays use approximately 1.5- to 10-d incubation periods b efore the
effect of the antiproliferative agents are assessed, agents that block
ceU-d epe ndent metabolism of these co mpounds (nucleoside transport, adenosine deaminase, and nucl eotidase inhibitors) will be
req uired to de linea te further their precise actions on keratinocyte
proliferation. Moreover, subsequent studies examllling the express ion of purinoreceptors in keratinocyte cultures, the effects of
purinoreceptor agonjsts and antagonists, and the signal-transduct ion pathways used by ad eno sine and adenine nucleotides wiJI also
be required to determin e the specific celJular and intrace llular
mechanjsms by which these co mpounds attenuate keratinocyte
proliferation.
Tllis ",ork lIIas sllpported ill part by grallts (M.R.P.) ji-O//l t"e Mayo m,d J. I~?
Kicckillifer FOlllldaticllls alld tile Na tioll,,1 Illst itlltes of Healt/, (A1U9547). We
dutllk )lIdit" A. Abral,mll fvr critical CO//lII,wts "" tI,is //Iallllscript alld No ,."",
S ", allsoll alld }(are" Hit z ellWIIII for tlleir excellellt tecl",ienl assist.allce.
REFERENCES
1.
Oubyak GR..: Signal tr;lI1sduction by P2- purillcrgic receptors for
ATI'. I ltll) Resl';" Cell Mol 13;01 4:295-300 . 199 1
cxtr~l c clllllar
VOL.
104, NO.6
JUN E 1995
ADEN IN E NU CLEOS IDE/TIDE IN HIBIT ION OF KERATINOCYTE G R OWTH
Stiles GL: Adenosine re ceptors. ) Bi()1 C helll 267: 645 1-6454, 1992
Williams M: Purine receptors in mamma lian ti ss lI es: phartl1aCology a_ud fun. ctio nal signi fican ce. A t",If R CII P/wrm (fCtl / 1'oxicol 27:3 15-345. "1987
4 . Go rd o n JL: Extra ce llular AT!': efrects. so urces and tilte. Bi()c/ICIII) 233 :309-3 1 9.
1986
5. EI-M oatassim C. DOllland J , Mani J C: Extracellular ATP and cell s ign allil~g.
Bi()chilll 13i()l'h),,, Ac,n 11 3 4:31-4 5, 1992
6. O 'Collllor SE: R ecellt d eve lopme nts in rile class ifi cation and functional signifi ....
C:lII CC o f rece ptors for ATP and UTP. evidence for nucleotide recepto rs. Lije Sci
50: 1657-'1664, 1992
7. Linden J: Structure and fu nction of Al adenosine rece ptors. FASEB) 5:26682676, 199 1
8. PiUai S. 13i klc DD: Adenosin e uojphosphatc stimulates phosp hoinositidc mctab o..li sl11 , mobilizes intracc liuJar ca lc iull l, and i_l1h_ibits term inal differentiation of
human epiderm al ke ratin ocytes. ) C lill III lIcsl 90: 42-5 1, 1992
9. Fang W-G. Pirlli :1 F. Bang Y-J. Myers CEo T repcl JB: P2-purinergic receptor
agonists inhibi t the growth of androgen-independe nt prostilte ca rcinoma ceUs.
) C lill JII "CS' 89: 19 1-196. 1992
10. Tc y HE. Khoo HE. Tan C H: Adenosi ne modulates cell gro wth in hunl ~u1
epiderm o id carcino ma (A 43 1) cells. 13iochc/II Woph)'s R es COII IIIIIIII 187: 14861492, 1992
11. Ra thb o ne MI', Middl emiss pJ , Gysbers JW , DeForge S, Coste llo P, Del Moestro
RF: Purine Ilucleosides and llucJeotidcs stimulate pro life ration of a wide rall g~
of cell types . JII Vi,ro Cdl DCII Bioi 28a: 529 - 536, 19 92
12. Jackson JA, Carlson Ee: Inhjbition of bo vin e retina l micro vascular pericyte
prolireratio n in vitro by adenosine . Jl III) Pltysiol 263: H 634 -640, 1992
'13 . Harada 1-1 , T:li H . Moto mura A, Suzuki S, Suke ta Y : Extracellular ATP-indu c~d
reg ulatio n of epidermal g rowth f.1ctor signalling in cultured renal LLC -PI<1
cells. A ll,) Physiol 264: C956-960, 1993
'14. R. apaport E: Treatm ent of human tumo r cells w ith AD t' o r ATP yie lds arrest of
gro wth in S phase of the ce ll cycle. ) C dl Physiol 11 4:279-283 . '19 83
t 5. R,apaport E, Fishman Il....F. Gerccl C : Growth inhibi tion o f human tumo r cells U1
soft-agar cul tures by treatm ent with low levels of 'ldc llosin e 5'-triphospha te_
C nllcel" R es 43 :4 402- 4406 , 1983
1 6. FlaXJnan AB. Harper flA: In vitro analys is of the contro l of kc.ratinoC)'tv
prolife ra tio n in human e pidermis by phys io logic and pharmacologic agents. )
III "cs' De,.,II nlCll 65:52-59, 1975
"17. Harper RA, Flaxma n llA. C hopra Dp: Eflc ct or phannacological agents 0 0
humall ke ratinocytc mitosis ill vitro. I. Inhi bition by adenin e Iluclcotidcs. P"ot:
Soc Exp Bioi Mcd 146:1032- 1036.1974
18. Taylor JR, Halprin KM . Le vinc V: DNA synthesis in surviving pig skin in vitt"o.
) III "est Derlll n,oI 74:"125-130 , 1980
19. liz uka H, Adac hi R., Koizu11li H, Aoyagi T, O hka wara A, Miura Y : Eltccts of
~ld C II Ositl e aiid 2' - deoxyadcllosinc 0 11 cpide rmal kc ratin oc),te pro life rfltio n : itS
rcl:ltio n to cyclic AMP formati o n . ) T" lIcst D CI'II/flfol 82 :608-6'12. 1984
20. C o o k PW, pittelko w MP, Shiple)' GD: Grow th ra ctor-indepe ndent pro lirera t io o
o f normal hurnan ncollatfll kcr;ltinocytcs: produ ction o f :1utocrinc- a.lld pat'a ....
crin c-acting mitogeni c f.,cto rs. ) CclI Ph),siol '146:277-289, 1991
2 1. Cook PW , M attox PA. K eeble WW , Pitte lkow MR, Plo wllIan G O . Sho)'ab 1\1.
Adelm an Jp. Shipley GO: A heparill sulfate-regulated human keratinoCYte
autocrine f.1 ctor is similar or identicaJ to amphircg ulin. Ado! Cel! Bioi 11 :2547 .....
2557, 1991
22. Cook PW, Ma ttox PA , Ke eble WW , Shiple), GO: Inhibi tio n of auto no m o u S
hum an kera tinocyte prolife ration and all1phiregulin mitogenic activity by
sulfa ted polysacclmridcs . II, Vi,l'O C ell DCI' Bioi 28a:218-222, 1992
23. Pittclkow MR, Cook PW. Shipley GO, Derynck R.. Coffe), R Jr: Autonom o u s
g rowth of human kcratinocy tcs rcquires epidcnnal growth (;lctor receptor
occupan cy. Cell C rolll,h Dillel" 4 :5 13-52'1, 1993
2.
3.
24 .
25..
26 .
27.
28.
29.
30.
3 1.
32 .
33.
34.
35.
36.
37.
38.
39.
40 .
41.
981
Shipl ey G D , Keeb le WW, Hendrickson]E, Coffey RJ . Pittelkow MR: G rowth
of no rmal human keratinocytes and fib roblasts in serulll-free mcdium is
stimul ated by acidic and basic fibrob last growth f.1cto r. ) Cell PhJ'sio /1 38:51 1518, 1989
li z uka H , Adachi K. Halprin KM . Levine V: Adenosine and adenine Ilucleotide
sti mulation of skin (epidermal) adellylate cyclase. Bil,chjm Biophys Acta 444:
685-693, 1976
Suter MM , C rameri FM, Slattery Jp, Millard pJ. Go nzalez FA: Extrace llular ATp
:1I1d SOUle o f its ana logues induce transient rises in cytosolic free calcium in
indi vid ual canine keratinocytes. J lu vest D cr11lfltol 97:223- 229 , 1991
Ducl EA: Adenosine-indu ced a.lterations in the adenosine 3 ': 5' -monophosphate
levels in manl111alian e pidermis. 1\1101 Ph anl/ flco l 18:4 9 -52, 1980
Shipley GO, Pittelkow MR: Contro l or g rowth and diffe ren tiation in vitro or
human pro keratin ocytes cul tured in serum-free m edium. Arch Dc,.",nlol 123:
154 1a-1544a.1987
Pittelkow MR. Scott R.E: New techniques for the in vitro culture of hum an skin
ke ratinocytes and perspectives o n thcir lise for g rafting of patients \vith
extensive bums. Mn)'o Ciill Proc 61 :771-777. 1986
.Boyce ST. Ham R.G: N o rm ol humon e pidermal keratinocytes. In: Weber MM,
Seke ly L (cds. ). JII Vi,ro Models fo r C nllcer R esenrell. C R C Press USA , Boca
Rato n , FL. 1986, PI' 245-274
Piepko m M , Lo C, Plow m an G: A mphiregulin-depe ndent proliferation of
cultured human kera tinocytes: autocrine growth. the effects of exogeno us
recombinant cytokinc, and apparcnt requirement for heparin-likc g lycosaminoglycans. ) Cclll'h l'siol1 59 :1 14 -1 20, 1994
Kelly EW. Wang KK , Swanson NN. Hitzemann KA. Pitte lkow MR.. : Assessment
of cutaneous phototoxicity indu ced by bento porphyrin derivative and hematoporp hyrin derivati ve. In: Spinelli P, D al Fan te M . M arch esini R (eds.).
Photody"amic T"erapy alld Biomedical LasCJ's. Elsevier USA. New Yo rk. 1992. PI'
493-502
Kiehl R. lo nescu G: A defective purine nucleotide synthesis pathway in psoriatic
patie n ts . Aclll D er/ll Vell creol 72:253-255 . '1 992
Elder J T, Fisher GJ, Lindquist PB, Bennett GL. Pittclkow MR., Coffey RJ Jr.
Ellingsworth L. Derynck R . Voorhees JJ : Overexpression of transforming
g rowth factor cr in psoriatic epidennis. ScicHce 243:8 11- 814 . 1989
Cook PW . P ittclkow MR, Keeb le W \ XI, G raves-Deo! R , Coffey RJ, Shipley GO:
Amphiregulin mcssenger It.NA is e1cv:lted in psoriatic e pide nnis and gastrointestina.l carcino mas. Ca llccI'Rcs 52:3224-3227 , 1992
IG ng LEJ, G ates RE, Stoscheck C M. Nanney LB: Epiderm al growth f.1ctor/
transforming growth factor alph <l receptors in psoriasis. J 1"lIcst Del'll/ fliol
95: IOS- 12S .1 990
Go ttlie b A, C hang CK, Poshett ON. Fanelli B. Tam JP: Detection of tron,rOmlillg growth fa ctor a in normal , malignant and hypcrproliferative human
ker" ti nocytes. ) Exp Med 167:670-675 , 1988
li z uka H. Adachi K. Halprin KM. Levine V: Cyclic AMP accumulation in
psoriatic ski n: difi"c rc ntiai responses to lti stamine . A MP and epinephrine by the
unin vo lved and invo lved epidermis. ) III"es , Denlla,ol 70:250-253, 1978
W II J, Dent P.J eline k T . Wolfman A. Weber MJ . Sturgil TW: Inhibitio n of the
EG F-a cti v<ltcd MAP kin ase signalin g pathway by adcnosine 3'.5 '-monophosphate. Sricllce 262:1055-10 69 . 1993
Cook SJ , M cCormi ck F: Inhibitio n b)' cAMP ofras-dependen t activation of raf.
Sciellce 262: 1069- 1072. 1993
H osoi K, Fujishi ta M , Sugita K, K urih ara K , Atsumi T, Murai T, U cha T : 1'2
purinerg ic receptors and cellular calcium m etabolism in A43 1 human epiderm oid carcino m a cells. A lii) I'h)'sio l 262 :C635-643, 1992