Serum antibodies against porphyric hepatocytes in patients with porphyria cutanea tarda and liver disease

GASTROENTEROLOGY

1983;84:1483-91

Serum Antibodies Against Porphyric Hepatocytes in Patients With Porphyria Cutanea Tarda and Liver Disease EDUARDO Department Spain

BARAVALLE

of Medicine,

Clinica

and J. PRIETO Universitaria

The existence of serum antibodies against porphyric or normal rat hepatocytes was investigated in patients with porphyria cutanea tarda and in other forms of chronic liver disease. Ten patients with porphyria cutanea tarda, 7 of them with chronic active hepatitis or cirrhosis (group la) and 3 without significant liver damage (group lb], 8 patients with nonporphyric chronic active hepatitis (group 21, and 8 alcoholic cirrhotics, 3 of them with superimposed severe alcoholic hepatitis (group 31, were studied. In an antibody-dependent cell-mediated cytotoxicity test using isolated hepatocytes from normal and hexachlorobenzene-treated (porphyric) rats as targets, it was found that sera from group la produced high cytotoxicity against porphyric hepatocytes and low or zero cytotoxicity against normal hepatocytes (p < 0.001). The opposite cytotoxic pattern was observed when sera from group 2 was tested. Sera from groups lb and 3 failed to produce cytotoxicity against both targets. The cytotoxic effect on porphyric hepatocytes was significantly reduced by preincubation of serum with free uroporphyrin or by serum absorption with a sepharose-uroporphyrin immunosorbent. Immunofluorescence studies confirmed the existence of antiporphyric hepatocyte antibodies in group la. In conclusion, our results show that antiporphyric hepatocyte antibodies are present in some patients with porphyria cutanea Received January 12, 1982. Accepted December 31, 1982. Address requests for reprints to: Dr. J. Prieto, Departamento de Medicina Interna, Clinica Universitaria, Pamplona, Spain. This work was supported by the Spanish Comision Asesora de Investigacibn Cientifica y TBcnica. This paper was presented in part at the spring meeting of the British Society of Gastroenterology, Bristol, April 9-10, 1981. The authors thank Dr. E. Elias and Dr. B. Potter for revision of the manuscript; Dr. R. Enriquez de Salamanca for the chromatographic study of the commercial uroporphyrin mixture and quantitation and characterization of the liver porphyrins, and Celia Asensio and Carmen Miqueo for their technical assistance. ‘D 1983 by the American Gastroenterological Association 0016-5085/83/061483-09$03.00

Medical

School,

University

of Navarra,

Pamplona,

tarda and indicate that hepatocellular porphyrin might be partially responsible for the antigenicity of the liver cells. The role of these antibodies in the pathogenesis of the liver lesion remains to be elucidated. In porphyria cutanea tarda (PCT), skin lesions are associated with a characteristic disorder of porphyrin metabolism. There is now evidence for the existence of a uroporphyrinogen decarboxylase defect in PCT (l-3), and it has been shown that this defect is genetically transmissible (1,4). The low activity of the uroporphyrinogen decarboxylase and the induction of the hepatic deltaaminolevulinic acid synthase activity by extrinsic factors are considered to be responsible for the accumulation of uroporphyrin and seven-carboxyl porphyrin in the liver, and for the increased urinary excretion of these same products (5,6). Although liver disease is frequent in PCT, the severity of the histopathologic damage is extremely variable (7-13): some patients have normal livers, minimal changes, or slight fibrosis, while others show chronic active hepatitis or cirrhosis. The pathogenesis of the liver lesion in PCT remains obscure. The wide spectrum of liver damage observed in patients with the same metabolic abnormality suggests that a direct hepatotoxic action of accumulated porphyrins is not the only causative factor (9). Chronic alcohol ingestion increases the likelihood of hepatocellular injury, but liver disease is also found in the absence of ethanol consumption (9,14,15). Iron is unlikely to be a major factor since iron stores are only moderately increased in this disease (16)and liver damage can be observed without accompanying siderosis (7). Antibodies against heptocyte antigens have been found in a variety of chronic liver diseases (17). The presence of active chronic inflammation in liver biopsy specimens from patients with PCT led us to

1484

GASTROENTEROLOGY Vol. 84, No. 6

BARAVALLE AND PRIETO

investigate the existence of serum antibodies directed against liver cell antigens. We postulated that porphyrins might act as haptens in the induction of an immune response against hepatocellular constituents. Accordingly, we have looked for the existence of antibodies reacting with porphyric or with normal hepatocytes in the serum of PCT patients. With this aim, we have used an antibody-dependent cellmediated cytotoxicity and immunofluorescence techniques using hepatocytes from normal and hexachlorobenzene-treated rats as targets. Hexaclorobenzene ingestion produces an enzymatic blockade in rats and a metabolic disturbance identical to that seen in PCT patients (18,19). This study shows that serum antibodies reacting with porphyric hepatocytes, and not with normal rat hepatocytes, are found in association with PCT liver disease. Some of these antibodies appear to be directed against uroporphyrin determinants. The relevance of this humoral immune response in the pathogenesis of the PCT liver lesion remains to be clarified. Materials

and Methods

Patients Patients were divided into three groups. We studied 10 patients with PCT (group 1). Their clinical and biochemical findings are listed in Table 1. The diagnosis of PCT was made on the basis of characteristic skin changes, increased urinary uroporphyrin excretion, and liver biopsy specimens with red autofluorescence. This group was subdivided according to the presence (la) or absence (lb) of significant liver disease assessed histologically. Group

Table

1. Clinical

Findings

in Patients

Urinary uroporphyrin”

Age/ sex

Alcohol excessb

Drugs

P.E.

51/m

_

_

7.2. M.S.

43/m 52/m

_

E.R. A.F. J.F.

62/m 40/m 53/m

F.M. D.F. F.P. P.P.

50/m 31/f 36/m

Case

With Porphyria

W24

h)

Cutanea

la consisted of 7 patients with chronic active hepatitis or active liver cirrhosis. Moderately increased stainable liver iron was found in ail of them, fatty changes in 2 patients (PE and JF), central vein sclerosis in 1 patient (JF), and alcoholic hialine in none. Liver cirrhosis, when found, was of the micronodular type. Four patients from this group had circulating non-organ-specific autoantibodies.

In all cases, however, the serum titer was not higher than 1: 10, except for case JZ, where antinuclear antibodies were positive at 1:320 serum dilution. None of the patients of group la showed clinical or histologic signs of lupoid hepatitis, nor received steroids as treatment. No history consistent with a viral etiology of the liver disease was obtained in all the other patients except for patient AF. Group lb was formed by the 3 other porphyric patients. Two of these patients had minimal portal fibrosis and the other had normal liver histology but marked red autofluorescence. Group 2 was comprised of 8 patients with nonporphyric chronic active hepatitis (CAH): 3 patients were HBsAg positive, 1 patient exhibited typical lupoid hepatitis, 1 patient was an intravenous drug abuser, and the others belonged to the group of cryptogenic chronic active hepatitis. Only two of these patients ingested excessive amounts of alcohol. Group 3 was formed by 8 HBsAg-negative patients with alcoholic liver cirrhosis. Three patients of this group had superimposed histologic signs of severe alcoholic hepatitis; the others had slighty raised serum glutamic oxaloacetic transaminase (SGOT) levels. In groups 2 and 3 the urinary uroporphyrin was within normal limits. A liver biopsy was performed in all patients.

Target

Cells

Male Wistar rats were made porphyric by daily oral administration of 200 mg/kg body wt of hexachloroben-

Tarda”

Urinary coproporphyrind

SGOT”

(c~ti.24 hl

(mu/L)

ANA

ASMA

AMA

_

_

Autoantibodiesf

9.244

475

48

_

-

2.136 8.020

221 121

30 29

+ _

_

_ -

+

_

4.392

-

_

+

_

1.445 11.228

45 287 51

38 42 19

+ + +

+ +

_ _ _

+

_

986

204 179 113 108

27 5 10 28

+ _

_

+ _

+

-

+ +

_

estrogens _ _

417 3.459 1.292

not tested

+

+ _

Liver biopsys

chronic active hepatitis active cirrhosis chronic active hepatitis active cirrhosis active cirrhosis chronic active hepatitis active cirrhosis normal histology minimal fibrosis nonspecific changes

a All patients showed characteristic skin lesions. b It was considered with a cross when it was in excess of 80 g/day. c Normal s 60 &24 antibody: AMA: h. d Normal 5 150 /q/24 h. “Normal 5 12 mu/L. f ANA: antinuclear antibody; ASMA: anti-smooth-muscle antimitochondrial antibody. g All patients showed tissue autofluorescence when liver biopsy specimen was illuminated with ultraviolet light except in case FP, in which case this test was not done.

]une 1983

ANTIBODIES AGAINST PORPHYRIC HEPATOCYTES

zene for 60 days. Hexachlorobenzene was given in a suspension with drinking water. Urinary uroporphyrin was increased (34.6 +- 3.35 pg/24 h, mean ? SE) and urinary coproporphyrins remained low (4 5 2 /q/24 h). Liver biopsy specimens showed diffuse and irregular red autofluorescence. Liver porphyrin concentration was determined (20) in one porphyric animal, which was in excess of normal values (39.52 @g/g wet liver wt as compared with 0.395 + 0.043 pg/g wet liver wt in normal rats). The porphyrin chromatogram (20) of the porphyric liver showed 63.69% of uroporphyrin, 31.83% of 7-carboxyl porphyrin, 2.45% of 6-carboxyl porphyrin, and 2.03% of 4-carboxyl porphyrin. This pattern closely resembles the one found in PCT. Porphyric (PH) and normal (NH) hepatocytes were isolated from normal and porphyric rat livers by the method of Berry and Friend (21). The livers were perfused with collagenase type IV (Sigma Chemical Co., St. Louis, MO.) after initial perfusion with Hanks’ buffer solution (Microbiological). The suspension was shaken for 5 min at 37°C and filtered through sterile gauze. The hepatocytes were washed twice, and the viability was then determined with trypan blue exclusion. The viabilities obtained were between 80% and 96%.

Effector

Cells

Mononuclear cells were isolated from heparinized venous blood of normal subjects by centrifugation in Ficoll-Urovison gradients, by the method of Boyum (22), washed twice with Tc 199 (Microbiological) and resuspended in Tc 199 with fetal calf serum (FCS) (Difco Laboratories, Detroit, Mich.).

Antibody-Dependent

Cytotoxicity

Assay

In this assay the release of the cytoplasmatic marker enzyme lactate dehydrogenase (LDH) was used as an indicator of target damage (23). Classic antibody-dependent cytotoxicity assay (ADCC) methodology (24) has been modified by using a greater number of target cells to obtain higher enzyme concentrations and make measurements easy and reproducible. Briefly, the procedure was as follows: in each trial, 2 x lo5 hepatocytes, from porphyric or control animals, were incubated, for 30 min at 37°C in an atmosphere of 95% O2 and 5% COZ, in a total volume of 2 ml with heat-treated (56°C for 30 min) sera, diluted in RPM1 1640 (Microbiological), from patients or obtained from a normal donor. Serum dilutions 1:10 and 1:50 were used in the assays. After incubation, the hepatocytes were washed twice and later resuspended in 1 ml of RPM1 1640 with 40% FCS. Peripheral blood mononuclear cells from healthy human donors in 1 ml of Tc 199 with 10% FCS were added in a target:effector ratio (T: E) of 1:100. The mixture was incubated for 4 h in an atmosphere of 95% O2 and 5% CO2 at 37°C. All assays were performed in duplicate. Lactate dehydrogenase was measured in the supernatant by determining the rate of oxidation of NADH at 340 nm (Boehring Manheim, monotest opt.). Measurements were performed immediately after completion of the ADCC tests. The overall intraassay coefficient of variation for duplicate LDH determinations was 10.84% 2 0.74

(mean k SE). The cytotoxicity from the following expression:

index

1485

(CI) was calculated

CI = [(LDH released by patient serum - LDH released by normal serum) + Total LDH content of hepatocytes]

X 100.

Total LDH content of hepatocytes was calculated after freeze thawing of the rat hepatocytes. Different normal sera did not produce different LDH release by target cells. The serum of one of us was elected to be used in all the assays of this study. Lactate dehydrogenase release in the presence of normal serum and effector mononuclear cells was similar for normal and porphyric hepatocytes (26% 5 3.2% and 31.5% f 3.2% of total LDH content, respectively) and these values were not different from those obtained when medium alone was used instead of serum (31% t 2.2% for normal hepatocytes and 27% ? 2% for porphyric cells). Control experiments were performed using identical systems from which mononuclear cells were omitted. In these cases LDH release (22% + 5.9% and 21% + 1.2% of total LDH content for normal and porphyric hepatocytes respectively) was similar to the LDH measured when hepatocytes were incubated in medium alone (16% ? 2% and 23% + 2.6% for normal and porphyric cells). When mononuclear cells were incubated alone in medium, LDH release was negligible. The paired observation t-test and the Student’s t-test were used to determine the statistical

significance

of the results.

Blockade of Cytotoxicity Immunoglobulins

With Aggregated

Four grams of Cohn’s fraction dissolved in 100 ml of 0.154 M saline

II (Sigma)

were

solution at room temperature and then incubated for 12 min at 63°C. Afterwards, 26.8 g of NaZS04 were added. The solution was incubated for 1 h and then centrifuged. The precipitate was resuspended in phosphate-buffered saline (PBS) 0.0066 M pH 7.3 and dialyzed overnight against PBS. After centrifugation, the protein concentration was measured in the supernatant. Mononuclear cells were incubated with 50 mg of aggregated immunoglobulins per milliliter, and then the ADCC was carried out as usual.

Blockade of Cytotoxicity by Serum Absorption With Porphyric Hepatocytes Porphyric hepatocytes were isolated by liver perfusion using collagenase, and then 150 ~1 of heat-treated serum was incubated with an equal volume of porphyric hepatocytes of protein content 100 mg/ml for 30 min at 4°C. After centrifugation, the supernatant was absorbed at room temperature with an equal volume of porphyric hepatocytes for 30 min. These absorbed sera diluted 1: 50 with PBS were used in the ADCC experiments.

Blockade

of Cytotoxicity

A mixture of commercial uroporphyrin (type Waldenstrom,

With

Uroporphyrin

urinary and fecal human Sigma] was used in an

.sa!pnls

anua:~sa.rorqjounruur~

u! pasn uaql se,m tumas paqJosqe s!qJ ‘u!rn oc ~03 salA:, -oledaq ie~ IPUIJ~I~I30 auIn{oA {Enba ue ql!Az aJn$eJaduIal IUOOJ le paqJosc{t? SP.M InelerIJadns ‘3,fi

1P uy

-olcdaq

()E Jo3 u!aloJd

IR.I {wuJou

schl tunaas

30 amn{oA

paleaJl-iear{ .sie{~

-yew

puu

aseuaZt?~{o:)

aJam sa{du~es

-yaq

salholedaq

aJaM

aql ‘uogeZn3!Jluao

30 ~uIkhII Ienba

001

Zuy!e~uo3

ue ql!nz

.~,:)d 30 sJal!{oJyuI uotsn3Jad ‘u!uI

epJel

[)c

paleqn:)u! paJ{1un’{

s 01 n Jo3 saJnl{nn ql!;~

1e.1 [WIJON wnJas

anss!l J~A!{

{J;3[J.IO:JilJ SLM

aI{.L ‘Ir{Z!{

‘yJUzc M;lN “03

~rI,vwJori{f

‘~,NI{ ss!ay,

‘sarn!$

{~III.I~J~ u! ‘01:

t{1!.\2 {~~~eZn!uo:)

UO!~l?~{~K,II! aJO3’)‘{

Jal3v salh uaql

pale{os! pa]eaJl

lu

ey,TqdJod

e Japun paqse.44

1 pa)n{!p

‘%,01 pUe

~lue$euJadns aq] u! paloaiap aq p{no:’ srr!;rAqdlodom ou pm S8d ql!M paqsem SBM a{dures {a8 12 ‘asn aJoJaa ‘{a23 pa!.tp 30 cum8 Jad punoq SBM pueZ~{ 30 rue.tZ!{{!ur au0 ‘palt?ln?{ez sem asoJeqdas aql 01 punoq pueZ!{ 30 lunoure aql pUe AJlauroJ -onUoJ13ads kq suogn{os Zu!qsFM aq1 u! pamseaur aJaM su!Jhqd.rodoln ‘pun03 aJam sdnoJZ pal:)eamn ou pm ‘(92) yes pue e{{!ed 01 Zu!pJome p!3e :Jyo3{nsauazuaqo.r~!uy qJ!m &no payn?o a.raM sdnoJ8 z&y 30 sno!$euy~cs$a(l ‘q{nq uroo.nfJep paJ e 30 $qZ:!{ aql lapun ]no pa!J.re:) alam su+iqdlodo.m %I!A{OAU! suo!yz?Jado {{v ‘pasn {gun 3,~ le palots pue ‘[oe,~ {Id :u!urnqIe ({o”/{oA) E~\I ~0.01

(‘A’N

paAJasq0 aJaM

iiaql

Ue II! 3,LF:

%q’O

tumas ‘“NeN

Ja33nq

e

au!Aoq ({OAW) ql!,w

{etuJou

JO I:)a!qns ‘p{ay

tuoJ3 (Jelr!{ tie

ql!,~

uo!~i?)!3!tIZf?uI

P [“It?

(t;#]

saliioolvdat{

aJahz

a{dum

-sari

{hue

pappe

IIAIdx

IIInJas

alait\

~a({ s[{a3

B q$tM

put? ‘D,L~

5.~

gb

aso.wqdas

({oA/~M)

94,~ pue sew

I/\T80‘0

pale{os!

i~{{e:J!ueqzaur

sod

(/0~1 {lue

SeM ‘@ID

SJayJOMO3

paJedaJd

sem

eJas aq$ ‘ua4.L

aqL

aJaM

‘S8d

~su!eZt?

slueieuJadns

.palead

-JodoJn

{euy

ayL

[)p’~

‘paq!J~ap

{III/W

f;t’()

SFI pauIJo3Jad

0~ ~03 uo!~n{os

C{QM paleqnzu! LUnJaS

3”

Hd

aJaM

SJa$!{OXJ!Ul

pa$sn!pe

ol

UInJas

r{:)ez

‘05’~

Hd

{:)eN

~,y ~‘0

30 {ti 006 qi!M

‘~:)~i

aql

q”!qM

ql!M

paqJosqe

JO SJ~~~{OJ~~I

tIaaMJ,

aqL

SaZeJaAe

aJam sisal

‘Ja]eT

u!JliqdJodoJn

OM,L

q”!q”

gsav

‘aJn$eJadUIa$ s!ql

S{OJ~UOO pul? sluaged pa.IpUnq

30 [r’ 00~

ruoJ3 sa{dures

‘HOeN

N

L’L

sem $1 ‘NW 8~z.o 30 uogeJ)uanuon

sem uogeJedaJd

01 hpnls

s!ql

30 ssazxa

,‘apnJz,,

UT apwu

SBM

UB u!eluo:)

aql

ldrualle

pue

Zuy:)o{q

‘aJnlxy

ql!M

e 01

aql

aqi

‘su@qdJod

ou

saiholedaq

u!

Lpnd

luaJajJ!p

z!JhqdJod

asne3aa

lqZ!uJaAo

ZU!qSEM

Jaffnq

aJaM eJas

h13y pa.rpunq

au0

aql

3o u!Z!~o

Japu!euraJ

wnJas

payoed

‘~1~3

+OdHZeN

put? seswaJlen:)

SCM luaqJosounmuI!

‘(SZ)

‘tumas J,>d u!e{d UTpun03 uogeJ~uanuo3 u!JkqdJodoJn aql s!qA .{~/Zrl oz se,w ‘palPqnzu! aJam 30 ssa:)xa II! so an{eil s)aZ.tel aq) aJaqhz ‘elas paln{!p aql u! uogel$ua:)uo:) u!JLqd

t pue 01: I) Suognpp

pun? (01;:

UIOOJ ie u!m

()bgt

30 .IaAg 1BJ

F-Z

sir?J

[ 04 paln{!p

(1s:

ala{duror,

SI?h\ ‘{‘“{‘{~I?

01 ‘a{dures

asoJeqdas

{~UI.IOU pue

‘0~

luaqJosounuuu!

SBM asoJeqdas

HV

‘(Lz’gz)

Zu!p~o~e

ql!M

paqJosqe

s{{a:)

ol saihoiedaq

palu{ost le

paze{d

I! U!

“6s.O ‘{:F’N

SBM

e ol paJJa3sueJl

SBM

e 3” {‘u 001 pue ‘S'P Hd {DH 30 {UT 0001 qI!‘N parISeM Irar{l seiu aso.uzqdas aqL ‘{DeN w ~‘0 30 {tu 0001, 30 uo!$!ppe aql hq ua{{oMsaJ pm c 3 .ral{g sse{Z e uo Jalem qj!rn {)aqsm% aJaM (uapams ‘e{esddn ‘s[eoyraq~ auy e!zmu 30 2 s :SMO{{OJ se SRM aJnpanoJd 30 poqlaur -Jeqd)

“41 ‘kua!J8

aql 0)

le

‘(ayJaMZu!Jqag) uerunqgue

O+‘JI

{~assn:~!{~)

paleqnw!

qZ!q

a[paau aq1

L{S!{' !J&l

v ~atr!mr?~n{;i Zu!tI!eluo3 of,gl 1I/1Idx qitm pasn3 {o.~~uo:) pur? :)!Jn^qdJod 30 sJa)\!{ aq,L

alan!

paqsehz

aql ‘uaqA .uogn{os Zu!qsem aql II! palnalap aq p{noo su!JAL{dlodo.m ou {gun (I: E) auexo!p Ja$eM 30 uogn{os UIOOJ IF?IqHyJaAo I? qgM paqsem sem {a8 aqL .a.mleJatlmal uogn{os u!&iqdJodorn aq$ pue aqJ, ‘{UI 01 01 dn s’p

paxyr

Jayeaq

Hd {:)H uI paI”{!p uaql pul? ‘{3&I ql!lu S‘P Hd 01 palsn!pe auexo!p 30 {tu 5.i: UT pai\{oss!p aJaw (OIL 107 ‘qe$cl *ueZo~ u@qdJodoJn) [ u!.thqd~odoJn 30 SUIr?J#!{{!UI UaL ~‘5 Hd (eUI%S) ap&unpoqJe~ Iy I,‘() 30 uo~~n{OS ‘Sl?npOJd

‘pappe

Lq

30 UOy

q$!h9 paddyha

Jno3

aJaqdSOUIle

u!{nqo{ZounuIuy

S:).g

OM$

aonaoSR~Onu(JunUlIu!

Iuap!:)u!

{JRD) adoosoJo!ur

a~a,\z .iaI{,L

'::,(;G30

“4,~ IJCII! 70

U!LII 0~ JOJ at~!{es L{]!M

aA!d

u! pauyzl

h{

Jo3 Dog5

eaueln:]

~s.it?ssl?pu!{q U! s.tamasqo /~{)Ua{xIada[~Il!

-:)alap

,i{ale!{~atuuI! ‘Jalcl

u!a:)sa~on~

satu!] aaJt{i par{se~\ rIarll ;IJ~M /iaq~ '70 '%,se 30 alaqdsomle ue ri! :),,ft; 11,u!Itt (I(: .IO~au!{r:s u! ‘p: I paIn{!{) ‘s~uaged LDd {rznba

30 aurn{o;\ aqL

J a{~ ;Irigs~~!pe ‘s~o{3 aleu!tu!{a ol paJaI{y pue ‘:),L{; lr? tI!w I; ~03 uaypqs sem uo!suadsns aqL ,[zz#)

10 uo!y~JlrIa:)uo {ad{e:)s

ql!"

Imdx 30

of791 wud8 II! paw{!p uaw pue 13~ N L’L 30 {d oz u! paA{os -s!p seal U!Jl(qdJOdOJn [e!nJauIruon 30 UIEJZ~{{~UI au0 .sluaur!Jadxa pasn

‘OQ’L q$!m

Hd Sgd

-aJ SBM aJnpa:)oJd pazk{l?!p

v&g

aleqdsoqd ~/t r;()‘() e ql!~ saurg aaJql paqsem pue pailour -aJ sem aso.mr{das aql ‘uogeqnnu! Jal3y .Jax!ur pua-Jan0 -pus Ue I{~!M q 1 lo3 aJnlF?JadUIal ~100~ I!2 pUE!Z!{ Inoyl!M JO qj!m

Y ‘ON ‘P8 ‘I”:1

e

le

pue

su!JhqdJod

pag!Ja$sauou

‘UyhqdJodoJnqns

‘urJiZqdJod

{dxoq.E-L

pue ‘(0;) ure~Zoleuro~q3

30 (7&‘81)

o/“b.r; ‘u!JhqdJodoJn

(y&l

u!e$uon

01

UMoqs SeM

3yamoJon~

aql 30 Zu!uuens

Jakel u!q$ Lq pau!urlalap

AqdeJZoleuroJqD

aql 30 uog!sodtuoo ayL -.hpoqgue aql u! paa{omt

pue

seM amlx!ur

,shesse L~!~!xo+oJ~(o wapuadap sa!poqgue aql y”o{q 01 ldrualle

.C:)( ~‘IOH:~I.l.N~lOH.J~~V~~

ANTIBODIES

minimal changes in the liver (group lb) also failed to induce cytotoxicity against porphyric hepatocytes. These differences were statistically significant using serum dilutions of both 1:10 and 1:50 (legend Figure 1). When normal hepatocytes were used as targets, low or absent cytotoxicity was found with the serum from PCT patients and alcoholic cirrhotics. In contrast, a high cytotoxic index was obtained with serum samples from patients with nonporphyric chronic active hepatitis. Again, these differences were statistically significant with both serum dilutions (legend Figure 1). It was found by comparing the cytotoxic effect of the same serum sample (paired observation test) against normal or porphyric hepatocytes that serum from group la produced a significantly higher cytotoxic index with porphyric than with normal hepatocytes. The opposite cytotoxic pattern was observed in group 2. Groups lb and 3 did not show cytotoxicity either with normal or with porphyric hepatocytes. These results were observed at both serum dilutions (Figure 1). To ascertain whether or not the cytotoxic effect on porphyric hepatocytes was mediated by humoral antibodies, effector cells were preincubated with aggregated immunoglobulins as described in Methods. Comparing the results obtained using normal and preincubated effecters, a significant blockade of cytotoxicity was observed (Figure 2). To the same aim, sera from 3 PCT patients with active liver disease

AGAINST

PORPHYRI(:

HEPATOCYTES

1487

p~o.001

I

WITHOUT AGGRGATED IMMUNOGLCBULINS

Figure

WITH AGGREGATED IMMUNOGLOI~UL~NS

2. The effect of preincubating effector cells with aggregated immunoglobulins on ADCC against porphyric hepatocytes mediated by serum from 4 PCT patients. Lines link values obtained with serum samples from the same patient. Closed circles: serum dilution 1: 10. Open circles: serum dilution 1 : 50.

were absorbed with porphyric hepatocytes (see Methods), and ADCC assays were performed with absorbed and unabsorbed sera against porphyric hepatocytes. A significant drop of the cytotoxicity was found by using absorbed sera (p < 0.05, paired observation test). Actual values of the cytotoxic

Cytotoxic IIldCX

60

pco.005

p402

p

B

48 36

PsO.05

PH

NH PCT

Figurr

PH

NH CAH

PH AL0

NH

PH

NH PCT

PH

NH CAH

PH

(NS)

NH

ALD

I. Cytotoxicity indexes obtained in ADCC tests using serum samples from PCT patients (PCT), nonporphyric chronic active hepatitis (CAH). and alcoholic cirrhotics (ALD) against normal (NH) and porphyric (PH) rat hepatocytes. Sera were diluted 1: 10 (A] and 1: 50 (B) with RPM1 1640. Lines link the results obtained with the same serum sample. * indicates PCT patients without liver damage: these values were not included in the statistical analysis [paired observation t-test). The mean CI (?SEMJ against NH was 1.47 2 1.31, 5.04 2 5.04, 30.4 t 5.44”, and 1.31 +- 0.47 for PCT patients with liver disease, PCT patients without liver damage, CAH, and ALD, respectively, at serum dilution 1: 10 (A) and 1.34 -t 0.87, 3.03 t 1.6. 26.27 r 7.02, and 2.15 t 0.8 at serum dilution 1:50 (B). Using PH as targets. the values were 32.44 IT 3.94”, 0. 1.76 i 0.93, and 2.60 2 1.03 (A), and 29.76 5 5.51”, 2.44 t 1.98. 2.45 -+ 0.78. and 0.95 2 0.78 (B), respectively, for the same groups. “p < 0.001 and I1p I 0.05 when these values were compared with the other values of the same series (Student’s t-test).

1488

BARAVALLE

AND

GASTROENTEROLOGY

PRIETO

index were 36% 49% and 49% with unabsorbed sera, and ll%,, 4% and 17%, respectively, with absorbed sera. To see whether the antiporphyric hepatocyte antibodies observed in group la were directed against porphyrins present on the liver cell, the sera from these patients were preincubated with a uroporphyrin mixture. It was found that this preincubation significantly blocked the cytotoxicity (Figure 3). To ascertain that this effect was not related to a nonspecific action of the uroporphyrin mixture on mononuclear cell function, the same procedure was carried out in the ADCC system of nonporphyric chronic active hepatitis serum and normal hepatocytes. Preincubation of these serum samples with uroporphyrin did not result in any blockade of the cytotoxicity (Figure 3). To further clarify whether or not antiporphyric hepatocyte antibodies carried combining sites against liver cell uroporphyrin, sera from 3 PCT patients with significant liver damage were absorbed with a sepharose-uroporphyrin immunosorbent. After removal of the immunosorbent, they were tested in an ADCC assay against porphyric hepatocytes. Parallel experiments were done with unabsorbed sera and sera treated with sepharose alone instead of the immunosorbent. A significant decrease of the cytotoxicity (p < 0.05) was found with immunosorbent-processed sera as compared with unabsorbed or sepharose-treated sera. These two produced similar results (Figure 4).

PC1 SERW

AGMNST

KPAlVC

UH

FVWHYRIC

AGAINST

NORMAL

HEPATOCYTES

CyMtcwc

55 -

cyto(oxK

h-dcr

50.

k?dcr

45 -

SERUM

YTES

p

:

N. S.

1

40. 35. 30 25 _ 20 15 . 10 5.

Figure

3. The effect of uroporphyrins (uro) on ADCC against normal and porphyric rat hepatocytes mediated by serum from 4 PCT and 4 nonporphyric chronic active hepatitis (CAH) patients. On the left: PCT serum against porphyric rat hepatocytes, on the right CAH serum against normal hepatocytes. (Closed triangles), 1:lO serum dilution (open circles) 1:50 serum dilution. All sera were diluted with RPM1 1640. The uroporphyrin concentration in the diluted serum was 20 FLgiml. Lines link values obtained with serum samples from the same patient.

Vol.

84, No. 6

cI %

4o ---A 30

peo.05

20 :7

I

I

PS Figure

IAS

SAS

PS

SAS

IAS

4. Sera were absorbed with sepharose uroporphyrin immunosorbent (IAS) or sepharose alone (SAS) and ADCC tests were performed as described. PS indicates untreated patient serum. Serum dilutions 1: 50 in PBS were used in the assays. Lines link values obtained with serum samples from the same patient. Sera from 3 PCT patients with liver damage and 3 nonporphyric chronic active hepatitis patients were tested against porphyric and normal hepatocytes, respectively.

Control assays using the ADCC system with nonporphyric chronic active hepatitis serum and normal heptocytes showed that treatment of the serum samples with the sepharose-uroporphyrin immunosorbent did not result in any changes in cytotoxicity (Figure 4). Indirect

Immunojluorescence

The ADCC assays have shown that PCT patients with hepatic damage, and not with other forms of liver diseases, have a serum factor able to induce cytotoxicity by normal mononuclear cells against porphyric hepatocytes. Immunofluorescence techniques were performed in order to confirm that this factor was an immunoglobulin and to determine whether or not it could be detected with porphyric hepatocytes after serum absorption with normal hepatocytes. Along these lines, normal serum and serum samples from 6 PCT patients were tested, 4 of them from group la, and 2 from group lb. Each sample, after being absorbed with cultured normal hepatocytes, was incubated with porphyric or normal hepatocytes. Patients from group la showed positive immunofluorescence with porphyric hepatocytes and negative with normal hepatocytes (Figure 5). Neither normal serum nor samples from group lb showed positive green immunofluorescence with normal or porphyric hepatocytes. Discussion Serum antibodies directed against liver membrane antigens have been found in various forms of chronic liver disease (17~4). In addition, mononu-

June

1983

Figure

ANTIBODIES

AGAINST

PORPHYRIC

HEPATOCYTES

1489

5. Green immunotluorescence patterns ot isolated normal (A) and porphyric (B) rat hepatocytes after incubation in serum from a PCT patient with significant liver damage. The serum sample was previously absorbed with normal rat hepatocytes. The slight red autofluorescence of porphyric hepatocytes was not seen under the conditions of the photograph.

clear cells from patients with chronic active hepatitis have been shown to be directly cytotoxic for isolated rabbit hepatocytes (29) and there is evidence suggesting that this cytotoxic effect is mediated by non-T cells bearing surface antibodies (30). Gonzales et al. (24), using antibody-dependent cell-mediated cytotoxicity techniques, have demonstrated that normal peripheral mononuclear cells become cytotoxic towards rabbit hepatocytes when targets or effecters were preincubated in chronic active hepatitis sera. The free antibody in the sera that confers cytotoxicity was blocked by adding a membrane lipoprotein fraction of human liver (LSP), suggesting that this was the putative antigen against which the antibody was directed. The purpose of this paper was to investigate whether or not porphyric hepatocytes, as distinct from normal hepatocytes, were antigenic for patients with PCT liver disease. Accordingly, we have looked for the existence of serum antibodies against isolated normal or porphyric hepatocytes. Along these lines, an antibody-dependent cell-mediated cytotoxicity was employed: serum from patients or normal donors, and normal peripheral mononuclear cells, were used in the assays against normal or porphyric hepatocytes. No attempt was made to investigate direct cytotoxicity by mononuclear cells from the patients. This point merits further study. By using the antibody-dependent cell-mediated cytotoxicity methodology, we have been able to show that serum from PCT patients with significant liver disease induces cytotoxicity against porphyric, but not against normal, rat hepatocytes. This cytotoxic effect is blocked by either serum absorption with porphyric hepatocytes or by preincubating the effector cells with aggregated immunoglobulins. These ob-

servations suggest that the cytotoxicity is indeed antibody-dependent. The existence of serum antibodies reacting with porphyric hepatocytes, and not with normal hepatocytes, was further demonstrated by using immunofluorescence techniques. The finding of a serum antibody able to induce cytotoxicity by normal peripheral mononuclear cells against porphyric hepatocytes was confined to PCT patients with significant liver disease. Indeed, in cases of nonporphyric chronic active hepatitis, the cytotoxic pattern against isolated hepatocytes is the opposite to that found in cirrhotic or precirrhotic PCT patients. The observed cell-mediated cytotoxic effect induced by the serum of nonporphyric chronic active hepatitis patients against normal hepatocytes is in agreement with previously published studies (24). The reason why sera from these patients induce low or zero cytotoxicity against porphyric liver cells, however, is not clear. A possible explanation is that accumulating porphyrins may mask the normal hepatocellular antigens against which the humoral immune response is directed. Similarly, there is no obvious explanation for the fact that serum from PCT patients with chronic active hepatitis induces low cell-mediated cytotoxicity against normal hepatocytes. It must be recalled, however, that other studies have shown that not all patients with different forms of chronic active hepatitis produced cytotoxicity toward normal liver cells (24,30). The antibodies against porphyric liver cells that have been found in PCT patients seem to have combining sites for hepatocellular porphyrins since preincubation of the serum with excess of uroporphyrin results in a significant reduction of the cytotoxic effect. This blocking effect has been shown not to be a nonspecific depressor action on mononuclear

1490

BARAVALLE

AND PRIETO

cell function. In fact, in the antibody-dependent cytotoxicity test with nonporphyric chronic active hepatitis serum and normal hepatocytes, addition of uroporphyrin tends to increase the cytotoxic effect. The above finding is consistent with the hypothesis of a porphyrin haptenization of the liver cell, but it is still compatible with other explanations. It is germane to think that hexachlorobenzene can alter the integrity of the liver cell in such a way that hidden antigens would be exposed and made accessible to antibodies. The free uroporphyrin in the medium could react with these antigens, and thus block the antibody binding. To further clarify this point, a sepharose-uroporphyrin immunosorbent was prepared and used to absorb serum from PCT patients. After removal of the immunosorbent, this serum produced significantly less cytotoxicity than the unabsorbed serum or the serum treated with sepharose alone instead of the immunosorbent. This observation suggests that some antiporphyric hepatocyte antibodies are in fact reacting with uroporphyrin present in the liver cell, and it gives support to the view that porphyrin haptenization is partially responsible for the antigenicity of the porphyric hepatocytes. This paper only provides indirect evidence for this mechanism of immunogenicity and direct demonstration of the presence of immunogenic porphyrin at the liver cell membrane surface should be pursued in further studies. Equally, the existence of uroporphyrin-reacting antibodies awaits confirmation in the future by the use of such techniques as solid phase radioOur immunosorbent experiments immunoassay. should promote, we believe, this type of study. Recently, Vergani et al. (31), by using an antibodydependent cell-mediated cytotoxicity assay and immunofluorescence techniques, have shown that in severe halothane-induced hepatitis, the patients develop circulating antibodies with specificity for halothane-treated rabbit hepatocytes. This group considered that halothane-altered membrane antigens have become immunogenic in this situation. Although a haptenization process by halothane has been proposed, no serum absorption experiments with isolated putative antigens have been done in their study. There is, however, a close parallelism between their methodology and findings, and our design and results. With a similar view it would be plausible to hypothesize that in PCT, porphyrin compounds could become bound to hepatocellular macromolecules, and the exposure at the liver cell membrane surface of these hapten-modified macromolecules could make them accesible to the lymphoid system resulting in immunogenicity. In our PCT patients with chronic active hepatitis or cirrhosis, alcohol abuse was present in 4 and

GASTROENTEROLOGY

Vol. 84, No. 6

HBsAg in 1 patient. Only 2 patients had liver damage unexplained by causes other than PCT. Even in our alcoholic PCT patients, however, the liver biopsy specimen was not typical of ethanol-induced hepatic damage (except for case JF, who exhibited central vein sclerosis), suggesting that other factors may be acting. The involvement of the antiporphyric hepatocytic antibodies in the pathogenesis of the liver lesion is a matter of speculation. The association of these antibodies to active PCT chronic liver disease does not necessarily imply a causative role on the liver damage. In fact, it may well be that the same metabolic process leading to hepatotoxicity produces simultaneous antigenicity. The mechanism of liver damage in PCT remains ill defined. Severe distortion of liver architecture is more common in the elderly (9), but significant hepatic lesions can also be encountered in early childhood (15,32). Siderosis is unlikely to cause hepatocellular damage because iron stores only tend be elevated two to three times the normal levels in PCT (7,16).Alcohol is a well-known contributory causative factor of the liver lesion, but cirrhosis can occur in the absence of ethanol consumption (9,14,33,34). Corms et al. (9) reported findings in liver biopsy specimens from PCT patients that were distinct from other forms of liver diseases, including alcohol-induced hepatopathy. These observations prompted authors to conclude that there is a specific liver damage whose pathogenesis is connected with the metabolic abnormalities that occur in PCT. The severity of liver damage in PCT, however, is very variable, and does not seem to correlate with the degree of the metabolic disturbance, suggesting that disordered porphyrin metabolism per se does not alter the integrity or the liver cells directly (35). For all these reasons, an open mind should be kept on the possible implication of immunologic factors in the development of the porphyric hepatopathy. In conclusion, this paper shows that antiporphyric hepatocyte antibodies able to induce cytotoxicity by normal peripheral mononuclear cells against the porphyric target cell in vitro are found in the serum of some patients with PCT, and it is suggested that hepatocellular porphyrin might be responsible for the antigenicity of the liver cell membrane. The involvement of the antiporphyric hepatocyte antibodies in the pathogenesis of the liver lesion remains to be elucidated. References 1. Kushner JP, Barbuto AJ, Lee GR. An inherited enzymatic defect in porphyria cutanea tarda. J Clin Invest 1976;58:108997. 2. Elder GH, Lee GB, Tovey JA. Decreased activity of hepatic

lune

3.

4.

5. 6.

7.

8.

9.

10. 11.

12.

13. 14. 15. 16.

17.

18. 19.

1983

uroporphyrinogen decarboxylase in sporadic porphyria cutanea tarda. N Engl J Med 1978;299:274-8. Felscher BF, Norris ME, Shih JC. Red-cell uroporphyrinogen decarboxylase activity in porphyria cutanea tarda and in other forms of porphyria. N Engl J Med 1978;299:1095-8. Benedetto AV, Kushner JP, Taylor JS. Porphyria cutanea tarda in three generations of a single family. N Engl J Med 1978; 298:358-62. Elder GH. Porphyrin metabolism in porphyria cutanea tarda. Semin Hematol 1977;14:227-42. Felsher BF, Carpio NM, Engleking DW, Nunn A. Decreased hepatic uroporphyrinogen decarboxylase activity in porphyria cutanea tarda. N Engl J Med 1982;306:766-9. Chlumska A, Chlumsky J, Malina L. Liver changes in porphyria cutanea tarda patients treated with chloroquine. Br J Dermatol 1980;102:261-6. Creutzfeldt VW, Beck K, Clotten R, Bianchi L. Die Leber bei den hepatischen Porphyrien, mit besonderer Berucksichtigung der Porphyria cutanea tarda. Acta Hepatosplen 1966; 13:65-83. Cortt% JM, Oliva H, Paradina FJ, Hernandez Guio C. The pathology of the liver in porphyria cutanea tarda. Histopathology 1980;4:471-85. Uys CJ. Eales L. The histopathology of the liver in acquired (symptomatic) porphyria. S Afr J Lab Clin Med 1963;9:190-7. Pimstone NR. The hepatic aspects of the porphyrias. In: Read AE, ed. Modern trends in gastroenterology. Vol 5. London: Butterworths, 1975:373-417. Lundwall 0, Weinfeld A. Studies of the clinical and metabolic effects of phlebotomy treatment in porphyria cutanea tarda. Acta Med Stand 1968;184:191-9. Doss M, Look D, Hennings H. Chronic hepatic porphyria in chronic aggressive hepatitis. Klin Wochenschr 1970;49:52-4. Elder GH, Gray CH, Nicholson DC. The porphyrias: a review. J Clin Path01 1972;25:1013-33. Welland FH, Carlsen RA. Porphyria cutanea tarda in an 8year-old boy. Arch Dermatol 1969;99:451-4. Turnbull A, Baker H, Vernon-Roberts B, Magnus IA. Iron metabolism in porphyria cutanea tarda and in erytropoietic protoporphyria. Q J Med n.s. 1973;166:341-55. Meyer zum Biischenfelde KH, Manns M, Hutteroth TH, Hopf U. Arnold W. LM-Ag and LSP-two different target antigens involved in the immunopathogenesis of chronic active hepatitis? Clin Exp Immunol 1979;37:205-12. Courtney KD. Hexachlorobenzene (HCB): a review. Environ Res 1979;20:225-66. Doss M, Schermuly D, Koss G. Hexachlorobenzene porphyria in rats as a model for human chronic hepatic porphyrias. Ann Clin Res 1976;17(Suppl 8):171-81.

ANTIBODIES

AGAINST

PORPHYRIC

HEPATOCYTES

1491

20. Day RS, de Salamanca RE, Eales L. Quantitation of red cell porphyrins by fluorescence scanning after thin layer chromatography. Clin Chim Acta 1978;89:25-33. 21. Berry MN, Friend DS. High-yield preparation of isolated rat liver parenchymal cells. J Cell Biol 1969;43:506-20. 22. Boyum A. Isolation of leucocytes from human blood. Further observations. Stand J Clin Lab Invest 1968;97(Suppl 21):3176. 23. Segal AW, Levi AJ. Cell damage and dye reduction in the quantitative nitroblue tetrazolium (NBT) test. Clin Exp Immunol 1975;19:309-18. 24. Gonzales C, Cochrane AMG, Eddleston ALWF, Williams R. Mechanisms responsible for antibody-dependent, cell mediated cytotoxicity to isolated hepatocytes in chronic active hepatitis. Gut 1979;20:385-8. 25. Grossman ME, Bickers DR, Poh Fitzpatrick MB, Deleo VA, Harber LC. Porphyria cutanea tarda: clinical features and laboratory findings in 40 patients. Am J Med 1979;67:277-86. 26. Cuatrecasas P, Wilckek N, Afinsen CB. Selective enzyme purification by affinity chromatography. Proc Nat1 Acad Sci USA 1968;61:636-43. 27. Cuatrecasas P, Afinsen CB. Protein purification by affinity chromatography. J Biol Chem 1970;295:3059-65. 28. Failla D, Santi D. A simple method for quantitating ligands covalently. Anal Biochem 1973;52:363-8. 29. Thompson AP, Cochrane AMG, McFarlane IG, Eddleston ALWF, Williams R. Lymphocyte cytotoxicity to isolated hepatocytes in chronic active hepatitis. Nature 1974;252:721-2. 30. Cochrane AMG, Thompson AD, Moussouros A, Eddleston ALWF, Williams R. Antibody dependent cell mediated (K cell] cytotoxicity against isolated hepatocytes in chronic active hepatitis. Lancet 1976;28:441-7. 31. Vergani D, Mielli Vergani G, Alberti A, et al. Antibodies to the surface of halothane-altered rabbit hepatocytes in patients with severe halothane-associated hepatitis. N Engl J Med 1980;303:66-71. 32. Pifiol Aguadb J, Galy Mascar C, Mascar J, Clinical studies of 63 cases of various porphyrias in Barcelona over the Z-yearperiod 1968-1969. S Afr J Lab Clin Med 1971;17:201-4. 33. Topi G, D’Alessandro Gandolfo L. Inheritance of porphyria cutanea tarda. Br J Dermatol 1977,97:617-27. 34. Taylor JS, Roenigk HK. Estrogen induced porphyria cutanea tarda symptomatic. In: Doss M, ed. Porphyrins in human disease. Basel, New York: Karger, 1976:328-35. 35. Henning H, Vogel HM, Luders CJ. Serum enzymes and iron levels in chronic hepatic porphyrias. In: Doss M, ed. Porphyrins in human disease. Basel, New York: Karger, 1976:348-g.