Tissue Amyloid P Component in Normal Human Dermis is Non-covalently Associated with Elastic Fiber Microfibrils

Stephen M. Breathnach, M.A., M.D., Ph.D., M.R.C.P., Mark B. Pepys, M.A., M.D., Ph.D., F.R.C.P., M .R.C.Path ., and Helmut Hintner, M .D. Department of Medicine (Dermatology) (SMB). Charing Cross 2nd Wesnninster Medical School, London; MRC Acute Phase Protein Research Group. Immunological Medicine Unit, Royal Postgn.du:ne Medical School (MBP), London; and Department of Dermatology (HH). University of Innsbruck, Innsbruck. Austria

Tissue amyloid P component (TAP), a protein that crossreacts immunohistochemically with the normal plasma glycoprotein serum amyloid P component (SAP), is invari­ably associated with elastic fiber microfibrils in adult humans. We have investigated the nature of this association. Aliquots of minced, homogenized dermis, obtained following ethy­lenediamine tetraacetic acid (EOTA) separation of whole adult human skin, were extracted with different tea gents, and the presence or absence of TAP in the pellet and in the supernatant following centrifugation was determined by SOS-PAGE and immunoblotting using anti-SAP antibodies. TAP was extractable from dermis using reagents which disru£t non-covalent bonds, including sodium dodecyl sul­fate (SOS) and guanidine hydrochloride. TAP was not ex­tracted by high molarity salt solutions, non-ionic detergents, or the reducing agents dithiothreitol and 2-mercaptoethanol.

Amyloid P compenent (AP) is a glycoprotein found in tissue deposits of all forms of amyloid in which it has been sought, including the various rypes of cutaneous amyloidosis, with the single exception of the intracor­tical plaques in Alzheimer's disease [1-4]. AP is indis-

Manuscript received March 24. 1988; accepted for publiCltion June I, 1988.

This work was supported in part by grants from the Skin Disease Research Fund and the University of London Central Resellirch Fund to 5MB, lIind MRC Program Grant G979/ 51 to MBP. HH was in receipt of a European Program Award of the Royal Society while a Visiting Professor:u the Char­ing Cross and Westminster Medical School.

Reprint requem to; Stephen M. Breathnach. M.A., M.D., PhD .. M.R.C.P., Department of Medicine (Dermatology), Ch:ning Cross Hospi­tal, Fulh:un Palace ROllid, London W6 8RF. U.K.

Abbreviations: AP: amyloid P component OTT: dithiothreitol EDT A: ethylenediamine tetra2cetic lIicid HRP-protein A; horscr.dish pcroxid<lSc-conju~;Hcd Sflliphylococcal

protein A MAGP: microfibril-associated glycoprotein MFP: microfibril protein 2-ME: 2-mercaptoethanol PAGE: polY2cryiamide gel electrophoresis PBS: phosphate buffered saline SAP: serum amyloid P component SDS: sodium dodecyl sulfate TAP; tissue amyloid P componem

EOTA solution was similarly unsuccessful at eluting TAP from the dermal preparation, indicating that the association of TAP with elastic fiber microfibrils is not simply the result of Ca++-dependent binding. Collagenase solubilized some TAP, bue this does not prove covalent linkage to elastic tissue of part of the TAP, because the afparent M, ofT AP extracted was identical to that of norma SAP subunits. We cannot completely exclude the possibiliey that a few subunits in each multimeric TAP molecule are covalenrly attached to the mi­crofibrils. However, our findings that denaturing agents alone extracted most of the TAP from normal human dermis strongly suggest that the great majority of the dermal TAP is non-covalently bound to elastic fiber microfibrils. Thus TAP is not an imegral constitucem of elastic fiber microfibrils. J ltlllest Derrnaro/92:53-58, 1989

tinguishable from, and derived from. the normal r.1asma glycopro­tein serum amyloid P compenent (SAP) 11,5,6 . SAP, which is synthesized by hepatocytes [71, is compesed of 1 0 identical glycosyl­ated polypeptide subunits of approximate molecular weight of 25,000 daltons, non-covalently associated together in the form of two pentameric discs interacting face-to-face. SAP closely resem­bles C-reactive protein (CRP), the classical acute phase reactant, in terms of molecular configuration. ultrastructural appearance, and amino acid sequence. SAP and CRP together form a distinct family of proteins of common evolutionary origin termed pentraxins, and are encoded for by homologous genes that have been mapped to the same region of the long arm of human chromosome 1 [8 - 10]. SAP may be extracted from normal human serum by virtue of its Ca++­dependent affinity for agarose, the result of binding to the pyruvate acetal of galactose [11-13]. SAP exhibits specific Ca++-dependent binding to a variery of other ligands, including fibronectin. C4 binding protein, some glycosaminoglycans, keratin filament aggre­gates. and DNA and chromatin, in addition to isolated amyloid fibrils in vitro [14 - 191.

In addition to the universal presence of AP in dcpo:siu of amyloid, and the occurrence of SAP as a normal plasma protein, a protein that cross-reacts immunochemically with SAP has been shown to be a normal matrix glycoprotein of glomerular basement membrane, in which it is covalently linked to collagen and/ or other matrix pro­teins [20]. We have shown in immunohistochemical studies that this normal tissue fonn of amyloid P component (TAP) is also invariably associated with elastic fiber microflbrils in normal adult human skin and other tissues [2,21,22], and that alterations in the

0022-202X/ 89/ S03.50 Copyright © 1989 by The Society for Investigative Dermatology. Inc.

53

54 BREATHNACH ET AL

staining parcern with anti-SAP correlate with abnormalities of the microfibrillar component of elastic fibers in disease states [23]. In man, the presence of TAP in normal skin is age-related. because TAP is first readily detected in association with elastic fibers from age 4 onwards [24].

The significance of the association between TAP and e1ascic fiber microfibrils is unknown. One group speculated that molecules of TAP may form the tubular core of the microfibrils, on the basis of their finding that TAP is associated with microfibrils in the footpad and zonulae of the eye of the mouse [25]; we have never been able to detcct any SAP, or SAP-cross reactive antigens, in normal tissues in this species. SAP has been reported to act as an inhibitor of elastase, and it has therefore been proposed that TAP may protect elastin from degradation [26]. We have not, however, been able co confirm that SAP has elastase inhibitory properties [Pepys, unpublished datal. We have suggested that TAP associated with elastic fiber microfibrils may play an important role in the maintenance of denno-epidermal adhesion and connective tissue architecture [4.22J. via an interaction with tissue fibronectin [14,15J and glycos­aminoglycans (16J. The close association between elastic fibers and deposits of amyloid in various types of cutaneous amyloidosis has been noted by many authors [3.27J. and. because SAP binds amyloid fibrils [19]. we have proposed that TAP on elastic fiber microfibrils may provide a nidus for deposition of amyloid material [3.4J. In view of the potential importance of TAP in health and disease. we have studied the biochemical nature of the interaction between TAP and elastic tissue in normal human dermis. We report here that the majority, if not all, of TAP in human skin is associated non-co­valently with elastic fibre microfibrils, and that this association in­volves morc than simple Ca++ -dependent binding.

MATERIALS AND METHODS

Prepara tion of SAP SAP was isolated from normal human serulll as previously described {1I,12J.

Antiserum Anti-human SAP antibodies were raised by immuni­zation of rabbits with isolated purc SAP as described [11,121. and the IgG fraction of rabbit anti-human SAP was obtained by staphylo­coccal protein A absorption for use in immunoblotting studies.

Preparation of Dermal Extracts Dermal material was obtained free of epidermis by incubating approximately 1.0 g of normal human skin obtained at surgical operations with 15 ml of 0.02M ethylenediamine rerraaceric acid (EOTA) buffer for 90 min at 4S0C as previously described [28]. Approximately 800 mg wet weight of dermis was minced with scissors, homogenized using a ground glass homogenizer. and suspended in 15 ml of Dulbecco's phosphate buffered saline (PBS; Gibeo, Paisley. Scotland). After two washes with PBS, I-ml aliquors of the dermal suspension were centrifuged for 15 min at 450 g, and the pellets were incubated for 4 h at room temperature with constant agitation with one of a battery of extrac­tion agents listed in Table I. Following centrifugation at 450 g for t 5 min, extracted dermal material was divided into residual pellet samples and extraction supernatant samples, which were subse­quently processed separately in parallel. In other experiments, 15 ml of EDTA solution, which was incubated with whole skin (in order [0 separate dermis from epidermis), was concentrated to 400 ,ul, and 45 ml, of PBS used to wash approximately 800 mg of dermis (after mincing and homogenization) was concentrated to 650,ul (Minicon B 125 concentrator; Amicon Corp., Lexington, MA). The concen­trates were then processed as below in the same way as the other extraction supernatant samples. Pellet samples were washed in PBS and then boiled for 20 min in 1,200,uL of buffer containing 0.0625M Tris-HCL pH 6.8. 2% sodium dodecyl sulfate (SDS). 10% glycerol. 5% 2-mercaptocthanol (2-ME). 0.01% Bromo­phenol Dlue. and 1% Triton X-IOO ("extraction buffer"), and the remaining debris was removed by centrifugation. Extraction super­natant samples were recentrifuged for 30 min at 10,000 g. and the supernatants were mixed with two times their volume of Laernmli sample buffer 129] containing 5% 2-M.E, and boiled for 10 min. Supernatants from dermal material extracted with 2M NaCI,

THE JOURNAL OF INVESTIGATIVE DERMATOLOGY

T able I. Extraction of AP from Normal Human Dermis: Agents Used

AP Present In-

Extr.lcOon Agent(s) Pellet Supematult

IP~ ~ 2. D.DIM EOTA. D.DIM Tris. D.14M N~CI.

pH 8.0 ~ 3. 2M N,CI ~

4. 2M N.CI. O.OIM EOTA ~ 5. I % Triton X- tOO in PBS +" 6. 1 % Nonidet P-4D in PBS +lo 7. O.O IM EOTA. O.OIM Tris. 0.14M N.CI.

O.OIM OTT. pH 8.0 ~ 8. 2MN.CI.O.OIMEDTA.O.IMOTT ~ 9. 2M N.cl. O.OIM EOTA. O.IM OTT. 1%

Triton X-loo 10.5% 2-ME t 1. 2M Hydroxylammonium chloride. pH 10.5 12. 0.1% SOS in PBS 13. 2% 50S in PBS 14. 0.1% 50S. O.IM OTT in PBS 15. SM Guanidine Hydrochloride pH 7.0 16. 6M Guanidme Hydrochloride. O.5M Tris.

0.00 I M EOTA. pH 8.5 17. 6M Guanidine Hydrochloride, O.SM Tris,

O.ooIM EOTA. O.IM OTT. pH 8.5 18. I % Collagenase t 9. 'Extraction buffer';

D.0625M Tris. 2% SOS, 10% glycerol. 5% 2-ME. 0.01% bromophenol blue, 1% Triton X-IOO

+/-"

+/-" +'"

NAf

• As delctITuned by SOS-PAGE ::.nd IInmunoblomng With alld-SAP. 10 +: fe::.Ji ly detecfllble ~ - : completely llbKnl d +/-: ::.bscm on SOS-PAGE bUllfllCC detC'cublc on LTnTnunoblomng; ~ qU::'llrity of AP III reSldu:li ~lIel rC'duCC'd. f NA: nOlllpplicablc

~ ~ ~

~

~ ~

+b

2M hydroxylammonium chloride, or 6M guanidine hydrochloride were dialyzed overnight against PBS before addition of rhe Laemmli sample buffer containing 5% 2-ME. Fifty microliters of each pellet sample and 100 JiL of each extraction supernatant sample were then analyzed by SOS-polyacrylamide gel electrophoresis (SDS-PAGE) on 10.5% continuous gcls followed by immunoblot­ting to identify TAP as below.

In addition, 150 mg wet weight of dermis was incubated with 1.5 mg of collagenase type IA (from Clostridium histolyricum, 380 U/mg; Sigma Chemical Co .. St. Louis. MO) for 18 h at 37°C. Oermal material was also extracted essentially according to the pro­cedure described by Ross and Bornstein [30J for the isolarion of elastic fibers from fetal bovine ligamentum nuchae. In brief, 1.500 mg of human dennis was minced with scissors. homogenized with a ground glass homogenizer, suspended in 50 ml of 5M guani­dine hydrochloride buffer pH 7.0, and incubated for 24 h at room temperature. The buffer was changed twice during the incubation. and aliquors were analyzed by SOS-PAGE and immunoblotting. After three washes with 0.2M Tris pH 7.4, and three further washes with distilled water, the pellet following centrifugation was incu­bated with 15 mg collagenase (1 % of che wet weight of the original dermal sample) in 0.2M Tris pH 7.4 containing O.S% penicillin­streptomycin (Gibco) for 18 h at 37°C. washed twice with 0.2M Tris, and once with distilled water. The residual pellet was then processed for SDS-PAGE and immunoblotting.

Immunoblotting Immunoblorcing was performed essentially as previously described [28,3 1,32]. Proteins transferred to nitrocellu­lose paper were first stained with 0. 1 % Fast Green FCF (Poly­sciences Inc .. Warrington. PA). and their position on the paper was marked by punching with a needle r31]. TAP was then detected using rabbit anti-human SAP (dilution 1: 500) followed by horse­radish peroxidase-conjugated staphylococcal protein A (HRP-pro-

VOL. 92, NO. I JANUARY 1989

tein A) (Kirkegaard and Perry Labs, Gaithersburg, MO) at a dilution of 1 : 200; the peroxidase reaction product was visualized using the diaminobenzidine reaction [28]. The specificity of the rabbit anti­human SAP antibody was confirmed by immunoblotting studies using pure SAP (1 ,ug in 10,uL Laemmli sample buffer), or normal human serum as a source of SAP (20 ,uLof a 1 : 15 dilurion of normal human serum in Laemmli sample buffer containing 5% 2-ME).

RESULTS

The results of attempted extraction of TAP from normal human dermis using a battery of different agents are summarised in Table I. The presence or absence of TAP in the residual pellet and the ex­traction supernatant preparations was determined by SOS-PAGE and immunoblotting. Results of representative experiments are il­lustrated in Figs 1 and 2, respectively. Following the final treatmellt of the residual dermal pellets with boiling "extraction buffer," nu­merous proteins with a wide range of molecular weights were solu­bilized, as analyzed by SOS-PAGE (Table I, extraction agent 19; Fig 1). TAP was identified on SDS-PAGE as a protein with an approxi­mate molecular weight of 30 kO that co-migrated with isolated purified SAP (Fig 1, latl' J). The high apparent molecular weight of TAP/ SAP seen on SOS-PAGE is due to the anomalous behavior of SAP in this system as previously discussed [8] . In all instances in which a protein band, co-migrating with purified SAP on SOS­PAGE, was detected in either the residual pellct or the extraction supernatant. it was identified as TAP on immunoblotting. as a pro­tein that co-mjgrated with SAP in normal human serum (Fig 2, lalle 1) and with isolated purified SAP (Fig 2, la", 2). The specificity of the rabbit anti-human SAP antibody used in the immunoblotting study was confirmed by the absence of any reactivity with other proteins in normal human se rum, or with any o f the numerous other proteins released from dermis by the "extraction buffer" (Fig 2).

1 2

~ -3

p

4

s p

5 6

AMYLOID P COMPONENT TN NORMAL HUMAN DERMIS SS

SOS-PAGE and immunoblot analysis of concentrates of EOTA solution incubated with normal human skin in order to induce splitting, and of PBS used to wash the dermis after mincing and homogenization, did show trace quantities of SAP present. How­ever, both ptcparations were visibly conram..inated by blood before concentration, and , moreover, the pattern on SOS-PAGE analysis closely resembled that obtajned on analysis of nonnal human serum (data not shown). TAP was not so lubilized from preparations of homogenized and extensively washed normal human dermis by extraction with PBS, further EDTA, high molar salt solution, de­tergents, the reducing agents OTT or 2-ME., or hydroxylammon­ium ch loride (Table I, extraction agents 1 to 11 ; Fig I ,ltmes 4-7; Fig 2, falles 3 - 6), because TAP remained detectable only in the residual pellet and not in the ex tfdction supernatant. The above extraction agents did, however, ex tract a number of unrelated proteins which appeared as bands on SOS-PAGE analysis of the extraction superna­tants (Fig I, Jall es 5 and 7). Dy contrast, TAP was detected in extrac­tion supcrnatants following ex traction of dermis with SOS alone or in combination with OTT (Table I, extraction agents 12 to 14) , and with guanidine hydrochloride (Table I, extractio n agents 15 to 17; Fig 1, lllt/es 8-12; Fig 2, laPles 7- 11). Extraction of dermis with guanidine hydrochloride resulted in absence of many of the lower molecular weight protein bands on 50S-PAGE of the residual pel­let, with a corresponding increase in low molecular weight proteins in the extractio n supernatant (Fig 1, lants 8-11). The finding of TAP in the extraction supernatants was associated with a corre­sponding decrease in the amount of TAP in the residual pellets. 6M guanidine hydrochloride extracted virtually all the TAP from the dennis, such that TAP in the residual pellet was undetectable by SOS-PAGE (Fig I , latl" 8 and 10), and detectable only in trace amount by immullobiotting (Fig 2, la~les 7 and 9).

Normal human dermis was also ex tracted according to the

s p s p s p s

7 8 9 10 11 12 13 Figure 1. SDS~PAGE analysis of residual pellet material and extraction supernatants following extraction of nonnal human dermis usingsdected extraction OIge llts from among those listed in TOIbie I; proteins stained with Coomassie Blue.LAtlt 1: molecular weight markers; lane 2: normal human serum control; lant J: i50l:ued pure SAP control; lane 4: pellet. and lane 5: supernatant, following extncrion with 2M N2CI, O.DtM EDT A, 0.1 M OTT, 1 % TritonX-IOO; lant 6: pellet, and latl! 7: supernatant . following ex traction with 2M hydro xylammonium chloride pH 10.5; Jane 8 pellet. and la llt' 9: supernatant. following extraction with 6M guanidine hyd rochloride, O.SM Tris, 0.00 1 M EDT A pH 8.5; laltt 10: pellet. and lane 11 : supernat.ant . following ex rr.lccion with 6M guanidine hydrochloride. O.SM Tris. O.OOtM EDTA, 0.1 M DTT pH 8.5; latH! 12: pellet following sequential extraction with SM guanidine hydrochloride and collagenase; and lant 1 J : supernatant following extraction with co llagenase alone. Arrowheads denote b2nds identified as SAP by co-migration with pure SAP as shown in ftltle J.

S6 BREATHNACH ET AL THE JOURNAL OF INVESTIGATIVE DERMATOLOGY

Se SAP P S P S P S P S P S

•

1 2 3 4 5 6 7 8 9 10 11 12 Figure 2. Immunoblot analysis with anti-SAP of residual pellet material and extraction supernatants following extraction of normal human dermis using agents as shown in Fig. 1. Lwt 1: normal human serum control; lane 2: isol:ued pure SAP control: lant 3: pellet. and lant 4: supernatant, following extr.action with 2M NaCI. O.OIM EDTA. O.1M OTT. 1% Tricon X-tOO; lant 5: pellet. and lane 6: supernatant. following extraction with 2M hydroxylammonium chloride pH to .S; lane 7: pellet, and lant8: supernatant, following exttacnon with 6M guanidine hydrochloride, O.SM Tris. O.OOtM EOTA pH 8.S; lane 9: pellet, and lant 10: supernatant. following extraction with 6M guanidine hydrochloride, O.sM Tris. O.OOlM EOTA, O.IM OTT pH 8.5; Itmt 11: pellet following sequential extraction with sM guanidine hydrochloride and collagenase; and lane 12: supernaunt following extraction with collagenase alone.

method of Ross and Bornstein [30], with SM guanidine hydrochlo­ride followed by collagenase. TAP was detectable in the extraction supernatant in the first and second changes of the extraction buffer, bur was absent from the third change of buffer I despite the fact that some TAP remained detectable in the residual dennal pellet even after 24 h incubation with guanidine hydrochloride. This residual TAP was entitely extracted after further incubation with collagen­ase, because TAP subsequently became undetectable in the residual pellet by SDS-PAGE and immutloblotting (Fig I, fa .. , 12; Fig 2, fa .. , 11). In the absence of guanidine hydrochloride, collagenase treat­ment alone of the dermis partially solubilized TAP, because TAP became detectable by SDS-PAGE in the extraction supernatant as well as in the residual pellet (Table I, extracrionagent 18; Fig 1, Jarle 13; Fig 2, fa .. , 12).

DISCUSSION

Elastic tissue within the dennis is composed of three sets of inter­connecting fibers. termed oxytalan. elaunin, and elastic fibers, re­spectively [33]. Mature elastic fibers in the reticular dermis connect with a more superficial plexus of thinner e1aunin fibers running praUel to the dermo-epidermal junction, from which in turn thin oxytalan fibers run upwards in the papillary dermis, perpendicular to the dermo-epidennal junction, to be inserted directly into the basal lamina. Under the electron microscope. elastic fibers are com­posed of a central amorphous elastin core surrounded by a peripheral

mantle of tubular microfibrils, of approximate diameter 10 to 14 nm, termed elastic fiber microfibrils. Ultrastructurally, oxytalan fibers are seen to consist of elastic fiber microfibrils only, al though an immunohistochemical study has reported that they also contain non-amorphous elastin {30,34- 37]. Elaunin fibers possess an amor­phous central core, and are intermediate in character between elastic and oxytalan fibers. We have previously shown in immunoultra­structural studies that TAP is invariably associated with elastic fiber microfibrils in normal adult human tissue throughout the body [2,21-23].

The relationship of TAP to other constituents of elastic fiber microfibrils is uncertain. Biochemical characterization of elastic fiber microfibrils has been hampered by the difficulty experienced in extracting the microfibrils in pure form, and in consequence our knowledge of their composition is restricted. In contrast to elastin. elastic fiber microfibrils are susceptible to digestion by proteolytic enzymes such as trypsin, chymotrypsin, and pepsin, but are resistant to elastase and collagenase l30.34]. Extraction of fetal bovine liga­mentum nuchae with dithiothreitol in SM guanidine hydrochlo­ride, which nonenzymatically solubilizes elastic fiber microfibrils, yields a glycoprotein preparation rich in cystine and/olar amino acids, but lacking hydroxyproline, hydroxylysine. an the desmo­sines [30,34]. Ultrastructural histochemical investigation of bovine elastic tissue has confirmed the association of glycoproteins with elastic fiber microfibrils [38]. Fibroblast cultures derived from fetal

VOL. 92. NO. t JANUARY )989

bovine ligamentum nuchae accumulate extracellular fibrils mor­phologically identical to elastic fibre microfibril s. and secrete two glycoproteins, a collagenase sensitive protein of molecular weight 150 kD distinct from collagen cype VI (designated MFP I) and a collagenase-resistant protein of molecular weight 300 kD (desig­nated MFP II), which cross-react immunochemicaHy with elastic fiber microfibrillar protein [39 - 41]. MFPI and MFPII are distinct from TAP {21J. However, the antiserum used to monitor the cross­reactivity of MFPI and MFPII with elastic fiber microfibril s also displays affinity for non-elastic tissue microfibrils, and the status of MFPI and MFPII as elastic fiber microfibrillar proteins has therefore been challenged [421. A 31 kD acidic glycoprotein termed microfi­bril-associated glycoprotein (MAGP) has been identified as the major antigen of elastic fiber microfibrils extracted from fetal bo­vine ligamentum nuchae, using immulloblocting and immunohis­tochemical techniques [43}. MAGP was extractable using NaC I. urea, or guanidine hydrochloride only if the reducing agent dithio­threitol was present, indicating that disulfide bonding is important for the association of MAGP with elastic fibers. On the basis of our current findings, TAP would appear to be extractable from elastic tissue under entirely different conditions to those required for the extraction of MAGP, confirming that MAGP is structurally and immunologically distinct from TAP [43J.

The results of this study have shown that TAP is extractable from nonnal adult human dermis using chaotropic reagents, including SDS and guanidine hydrochloride, which disrupt non-covalent bonds. Reagents that disrupt covalent bonds, including the reducing agentS Dl1 (which interruptS disulfide bridges) and 2-ME, or high molarity salt solutions or detergents, were neither necessary nor sufficient for the extraction of TAP from dermis. Alchough trace quantities of SAP were present in concentrates of EDT A solution used to split whole skin and of the PBS used to wash the dermis after homogenization, this was 1110St likely due to contamination with blood. Certainly, EDTA solution was unsuccessful at eluting the great majority of the TAP from the dermal preparation, indicating that the association of TAP with elastic fiber microfibrils is nOt simply the result of Ca++-dependent binding. Interestingly, colla­genase treatment alone of the dermis did result in solubilization of some TAP. Therefore, although our results strongly suggest that the majority of the TAP associated with elastic fiber micro fibrils is bound to the microfibrils by non-covalent linkage. we cannot en­tirely exclude the possibility that a fraction of the TAP is covalently bound.

The fact that the majority of the TAP associated with elastic fiber microfibrils is non-covalently bound to the microfibrils would sug­gest that TAP is not, as has been suggested [25J. an integral constitu­ent of the microfibrils. This is in keeping with the report that TAP is not present in fetal skin or in the skin of children up to 2 years of age [24L despite the unequivocal presence of elastic fiber microfi­brils. It is unknown whether elastic tissue-associated AP is derived from SAP, or whether it is synthesized locally. We have been unable to confirm a report [44} that TAP is detectable by immunofluores­cence in no rmal human cultured fibroblasts, and have failed to de­tect TAP in fibroblast culture supernatant, despite the fact that the cel ls were demonStrably synthesizing Ct by radioimmunoassay [Reid and Pepys, unpublished datal. We have similarly been unsuc­cessful at detecting TAP in ex tracts of fibroblasts using immuno­blotting with anti-SAP IHintner. Pepys, and Breathnach, unpub­lished datal . Although AP wouldappearto be non-covalently bound to elastic fiber microfibrils, this fact does not exclude an important role for elastic tissue-associated AP in the maintenance of normal connective tissue architecture. via its Ca++-dependent binding to tissue fibronectin and glycosaminoglycans, or a role in the deposi­tion of amyloid via its binding affinity for amyloid fibril s l14 -16,191·

WI! thank Miss Jl4fil! Boo/uri B,Se., (HId Miss Ursilia Sta'12/, B.Se., far providing (xedlt ,l! ttchnieal assistatlCt.

AMYLOID P COMPON~NT IN NORMAL HUMAN DERMIS 57

REFERENC ES

I. Pepys MB, Baltz ML, de Beer FC. Oyck RF. Holford S. Breathnach SM. Black MM. Tribe CR, Evans OJ, Feinstein A: Biology of serum amyloid P component. Ann NY Acad Sci 389:286-298, 1982

2. Bre:uhnach SM, Bhogal B, Dyck RF. de Beer Fe. Black MM. Pepys MB: Immunohistochemicalsrudies of amyloid P component in skin of normal subjects and patients with CUtaneoWi amyloidosis. Br J Dermato!t05:115-124,1981

3. Breathnach SM, Mel rose SM, Bhogal B. de Beer FC, Black MM. Pepys MB: Ulcrastrucrurallocalisation of amyloid P component in primary localised cutaneous amyloidos is. Clin Exp Oermatol 8:355-362, 1983

4. Breathnach SM: The cutaneous amyloidoses; pathogenesis and ther­apy. Arch DermatoI121 :470 - 475, 1985

5. Baltz ML, C:aspi D, EV:1.Il5 OJ, Rowe IF, Hind CRK. Pepys MB: Circu lating se rum amyloid P component is the precursor of amyloid P component in tissue amyloid depositS. Clin Exp Immunol 66:691-700, 1986

6. Pep)'s MB: Amyloid P-component: structure and properties. In: Mar­rink J. van Rijswijk MH (cds.). Amyloidosis. Martinus NijhoffPub­lishers. Dordrecht. 1986. pp 43 -49

7. Tauura E, Si~ JO, Shinhama T. Skinner M. Cohen AS: Different regulatory mechanisms foe serum amyloid A and strum amyloid P component synthesis by cultured mouse hepatocytes. J Bioi Chern 258:5414 -5418, 1983

8. Pepys MB. Dash AC. FletcherTC. Richardson N. Munn EA. Feinstein A: Analogues in other mammals and in fish of human plasma pro­teins. C-reactive protein and amyloid P component. Nature 273:168-170,1978

9. Ohnishi S, Maeda S, Shimada K. Arao T : Isolation and characterization of the complete complementary and genomic DNA sequences of human serum amyloid P componenl . J Biochem 100:849 - 858, 1986

10. Floyd-Smith G, Whitehead AS. Colren HR, Francke U: The human C-reactive protein gene (CRP) and serum amyloid P component (APeS) are located on the proximal long arm of chromosome 1. Immunogenetics 24: 171 - 176, 1986

11 . Pepys MB, Dash AC, Munn EA, Feinstein A. Skinner M, Cohen AS, Gewurz H, Osmand AP. Painter RH: Isolation of amyloid P compo­nent (protein AP) from normal serum as a calcium-de~ndent bind­ing protein. Lancet i:l029 - 1031, 1977

12. De Beer FC, Pepys MB: Isolation of human C-reactive protein and serum amyloid P component. J Immunol Methods 50: 17 - 31, 1982

13. Hind CRK, Collins PM, Renn D. Cook RB. Gaspi D • .Baltz ML, Pepys MB: Binding s~cificity of serum amyloid P component for the pyruvate acetal of galactose. J Exp Med 159: 1058 - 1069, 1984

14. De Beer FC. Baltz ML, Holford S, Feinstein A. Pepys Mll: Fibronecun and C4 binding protein are selectively bound by aggregated amyloid Pcomponent.J Exp Med 154:1134- 1149, 1981

15. itostagno A. Fnngione B. Peaclskin E, Garcia-Prado A: Fibronectin binds to amyloid P component. Localization of the binding site to the 31.000 dalton C-terminal domain. Biochem Biophys Res Com­mun 140:12-20. 1986

16. Ham:u.aki H: Ca2+ -mediated association of human serum amyloid P component with hepann sulfate and dermatan sulfate. J Bioi Chern 262: 1456- 1460, 1987

17. Hintner H. Booker J. Ashworth J . Aubock J. Pepys MB, Brearhnach SM: Amyloid P component binds to kentin bodies in human skin and to isolated keratin filament aggregates in vitro. J Invest Derma-10191:22-28,1988

18. Pepys MB, Butler PJG: Serum amyloid P component is the major calcium-dependent s~ci fic DNA binding protein of the serum. Biochem Biophys Res Commun 148:308-313. 1987

19. Pepys MB. Dyck RF. de Beer FC. Skinner M, Cohen AS: Binding of serum amyloid P component (SAP) by amyloid fibril s. Cl in .Exp ImmunoI 38:284-293. 1979

20. Dyck RF, Lockwood e M, Kershaw M, McHugh N. Duance VC, Baltz ML. Pepys MB: Amyloid P component is a constituent of normal human basement membrane. J .Exp Med 152: 1162- 1174, 1980

21. Breathnach SM, Melrose SM. Bhogal a, de Beer FC. Dyck RF. Ten­nent G. Black MM. Pepys MB: Amyloid P component is located on

58 UREATIINACH ET AL

clastic fibrc microfibrils in nomlal human rissue. Nature 293:652-654. 1981

22. Breathnach SM. Melrose SM, Bhogal B. de Beer Fe. Black MM.I)epys MB: Immunohistochemic:ll studies of amyloid P componenr disrri­bution in noml:ll hum:ln skin.J Invest OermatoI80:86-90. 1983

23. Breathnach SM. Melrose SM. Bhogal B. de Beer FC. Black MM. Pepys MB: lnllnunohistochemical srudies of amyloid P componem in dis­orders of cutaneous elastic tissue. fir J Dermatoll07:443 - 452. 1982

24. Khan AM. Walker F: Age rd:lted detection of tissue :lmyloid P in the skin.J P:lthoI143:183 - 186. 1984

25. Inoue S. Leblond CPo Gr:lnt OS. Rico P: The microfibrils of connec­tive tissue: II. Immunohistochemical detection of the amyloid P component. Am J Pathol 176:139 - 152, 1986

26. Vachin G. Heck LW. Kaplan MM. GeifandJA.llurkeJF. McAdam KPWJ: Elast2sc inhibition by the amyloid P-component.ln: Peeters H (cd.). Protides of the Biological Fluids, Proceedings of the Thirty­Fourth Colloquium. Pergamon Press. Oxford. 1986. pp 395-398

27. Yanagihara M. Kato F. Fukushima N. Mori S: Intimate structural association of amyloid and elastic fibers in systemic and cutaneous amyloidoses. J Cutan Pathol 12: 110 - 116. 1985

28. Hintner H, Steinert PM. Lawley TJ: Human upper epidemlal cytO­plasmic antibodies are directed against keratin intermediate fib.ment proteins. J Clin Invest 72: 1344-1351. 1983

29. Lacl11mli UK: Cleavage of structural protcins during thc assembly of thc head of bacteriophage T 4 • Narurc 277:680-685. 1970

30. Rou R. Bornstein P: The el:lstic fiber. I. Thr separation and partial characterization of its macromolecular components. J Cdl Bioi 40,366-381.1967

31. Himner H. Lawley TJ: Ker:uin intermediate filaments be:lr antigenic determinants for stratum corneum antibodies. J Invcst Dermatol 8H91 - 495.1984

32. Hinter H. Neises GR. Lawlry TJ: Immunologic properties of enzy­matically degnded human ker:nin intermediate filaments. J Invesr Dermatol 84:108- 1 13, 1985

THE JOURNAL OF INVESTIGATIVE DERMATOLOGY

33. Cona-Pcreira G. Guerra Rodrigo F. Binencourt-S:lmp:lioS: Oxytalan. claunin. and elastic fibers in the human sk.in. J Invest Oenm.tol 66:143 - 148,1976

34. Ross R: The elastic fibre. A review. J Histochem Cytochem 21: 199-208. 1973

35. Uino J: Biochemistry of the elastic fibers in nonnal connective tissues :lnd its alteratiOlu in diseases. J Invest Ocrmatol 72: 1-10. 1979

36. Cleary EG. Gibson MA: EI:lsrin-associ:lted microfibrils and microfi­brillar proteins. Int Rev Conn Tissue Rcs 10:97 -209. 1986

37. Schwartz E. Fleischm:lJcr It: Association of elastin with oxytalan fibers of the dermis and with extracellular microfibrils of cultured slcin fibroblasts. J HistochCIl1 Cytochem 34: 1063 - 1 068. 1986

38. Fanning Je. Cleary EG: Identification of glycoproteins associatcd with elastin-associated microfibrils. J Histochem Cyrochem 33:287-294.1985

39. Scar CHJ. Kewley MA.Jones CJP. Grant ME. Jackson OS: The iden­tification of glycoproteins associated with dastic-tissue microfibrils. BiochemJ 170:715 - 718, 1978

40. Scar CHJ. Grant ME, Jackson OS: The nature of the microfibrillar glycopfOteins of elastic fibres. A b,osynthc[ic srudy. Biochel1l J 194:587 - 598. 1981

4 t. Ay:ld S. Chambers CA, Shuttleworth CA, Grant ME: Isolation from bovine dastic tissues of collagen type VI and characterization of its form in vivo. Biochem J 230:465 - 474. 1985

42. Pros~r IW .Gibson MA. Cleu), EG: Microfibrillu protein from d2Sdc tissue: a critical evaluation. Aust J Exp Med Sci 62:485 - 505. 1984

43. Gibson MA. Hughes JL, FanningJC. Cleary EG: The major protein of e12Stin-associ:ltcd microfibnls IS a 31-kOa glycoprotein. J Bioi Chem 261:11429 - 1 1436.1986

44. Spark EC. Shirahama T. Skinner M. Cohen AS: The identific:ltion of amyloid P-component (protein AP) in nonnal cultured human fi­broblasts. Lab Invest 38:556 - 559.1978

ADVANCES IN SKIN CANCER MANAGEMENT : A MULTI-SPECIALTY APPROACH

A symposium on the advances in skin cancer management: a multi-specialty approach will take place on February 17 - 18, 1989, aethe Ramada Renaissance Hotel, 55 Cyril Magnin Street (Marker at Fifrh). San Francisco. California 94102 (Tel. 415/392-8000). This program is being presented by rhe Department of Dcrmacoiogy and the Division of Plastic Surgery of the University of California School of Medicine at San Francisco. It is sponsored by UCSF's office of Extended Programs in Medical Education, which certifies that this activity meets the criteria for 10 credit hours in Category 1 of the Physician'S Recognition Award of the AMA and the Certificate Program of the CMA. For further infomlation or an application for enrollment, write to: EPMC, Registration Office, Room S7S-U. Universicy of California. San Francisco. CA 94143-0766 (Tel. 415/476-4251.