American Journal of Pathology, Vol. 140, No. 3, March 1992

Copyright © American Association of Pathologists

Induction in Mice of HumanLight-chain-associated Amyloidosis

Alan Solomon,* Deborah T. Weiss,* andMark B. PepystFrom the Department ofMedicine,' Human Immunologyand Cancer Program, University of Tennessee MedicalCenter/Graduate School ofMedicine, Knoxville, Tennessee;and the Department ofMedicine, t Immunological MedicineUnit, Royal Postgraduate Medical School, HammersmithHospital, London, United Kingdom

Primary (idiopathic) or multiple myeloma-associated amyloidosis is characterized by the depo-sition in tissue of monoclonal light chains or light-chain fragments (AL amyloidosis). In contrast toother types of amyloidosis information regardingthe pathogenesis of light-chain-related amyloid hasheretofore been limited due to the lack ofa suitablein vivo model. The authors report the successful ex-perimental induction ofhuman AL amyloid deposits.The repeated injection into mice ofBenceJones pro-teins obtained from two patients with AL amyloid-osis produced the histopathologic lesions character-istic of this disease. Partial dehydration of animalsbefore protein injection resulted in the accelerationofamyloidformation The human proteins were de-posited as amyloid within the mouse renal bloodvessel walls and parenchymal tissue, as well as inother organs. The deposits were Congo red-positive,exhibited green birefringence, and had afibrillar ul-trastructure. As evidenced immunohistochemically,the experimentally induced amyloid deposits con-sisted of the injected human light chains, and in ad-dition, contained mouse amyloidP component (AP);mouse immunoglobulin (Ig) or inflammatory-associated amyloid A protein was not detected Ex-traction and characterization of the amyloid depos-itsfound within the mouse kidney revealed the pres-ence of a predominately intact human lightpolypeptide chain Mice injected in identical mannerwith a non-amyloid-associated BenceJones proteinhad no or only rare amyloid deposits. The experi-mental mouse model provides a means to ascertainthe amyloidogenic potential ofhuman monoclonallight chains and to study further the pathogenesis ofAL amyloidosis. (AmJ Pathol 1992, 140:629-63 7)

In 1964, Osserman and colleagues" predicted thatmonoclonal immunoglobulins (Igs) found in patients withplasma cell dyscrasias could be implicated in amy-loidogenesis. This hypothesis was subsequently con-firmed by Glenner et al,4 who demonstrated the seminalrole of the light polypeptide chain in the pathogenesis ofprimary or multiple myeloma-associated amyloidosis, i.e.,AL amyloidosis. This disease is distinguished by deposi-tion in tissue of monoclonal light chains or light-chain-related fragments, and like in all types of amyloidosis,these deposits have characteristic tinctorial and ultra-structural features, including green birefringence afterCongo red staining and 8- to 1 0-nm nonbranching fibrilsevidenced by polarizing and electron microscopy, re-spectively.5 6 Additionally, a minor but universal constitu-ent of all amyloid deposits, both in human and animals, isthe amyloid P component (AP), a nonfibrillar glycoproteinof the pentraxin family7 identical to a normal circulatingprotein designated serum amyloid P component(SAP).89

The fact that the mere presence of a serum or urinarymonoclonal Ig does not invariably result in amyloid for-mation implies that either certain types of light chains areinherently amyloidogenic or that other pathophysiologicfactors are responsible for the development of amyloid.The availability of an in vivo model for AL amyloidosiswould thus provide a means to study further the patho-genic role of protein, as well as host factors, in this dis-ease. Although it is possible to induce another form ofamyloidosis experimentally, e.g., through injection ofmice with casein or endotoxin,5 the type of amyloid pro-tein deposited is not Ig-related. Rather, this protein, des-ignated amyloid A, represents a degradation product ofan acute-phase serum protein (serum amyloid A or SM)and is the principal constituent of the amyloid associatedwith chronic infectious or inflammatory disease, i.e., AAamyloidosis.610

Supported in part by U.S. Public Health Research Grant CA 10056 fromthe National Cancer Institute, the Ina M. Barger Memorial Grant for CancerResearch (IM-430) from the American Cancer Society, the Stein CancerResearch Fund, and the United Kingdom Medical Research Council.

Accepted for publication September 20, 1991.Address reprint requests to Dr. Alan Solomon, University of Tennes-

see Medical Center at Knoxville, 1924 Alcoa Highway, Knoxville, TN37920.

629

630 Solomon, Weiss, and PepysAJP March 1992, Vol 140, No. 3

In an effort to elucidate the protein and/or host factorsessential to the development of AL-type amyloid, wehave used the experimental model first described in 1976by Koss, Pirani, and Osserman.11 These investigatorsfound that a single intraperitoneal injection into mice of ahuman Bence Jones protein produced the renal lesionscharacteristic of myeloma (cast) nephropathy. We haveconfirmed this finding in a study of 40 different BenceJones proteins whereby we determined that the injectionof certain Bence Jones proteins into mice resulted notonly in the renal deposition of the human light chains ascasts, but also as crystals or basement-membrane pre-cipitates.12 In addition, the form and location of BenceJones protein deposition noted experimentally was com-parable to that found clinically. In contrast to these exper-iments, the induction of amyloid in mice required multipleinjections of Bence Jones proteins obtained from patientswith AL amyloidosis.2113 We report in detail the condi-tions that resulted in the experimental in vivo deposition ofhuman light chains as amyloid, the characterization of thelight chains extracted from the induced amyloid, and thedemonstration that such deposits also contained mouseamyloid P component.

Methods and Materials

Proteins

Monoclonal light chains, i.e., Bence Jones proteins, wereisolated by preparative zone (block) electrophoresis11from urine specimens obtained from patients with AL am-yloidosis or multiple myeloma. The purity of the proteinpreparations was established by electrophoresis in aga-rose gels and by SDS/PAGE. The proteins were analyzedimmunochemically using our rabbit anti-human K and Alight-chain isotype-specific antisera as well as anti-VK- oranti-V,-subgroup-specific antisera.14 The VL-subgroupnature of the Bence Jones proteins determined serolog-ically was confirmed by amino acid sequence analyses.

Experimental Protocol

Six-week-old C3H/HEJ mice (weighing 15-20 g) wereinjected intraperitoneally with reconstituted, sterile filteredsolutions of Bence Jones protein using a 3-ml luer-lockMonoject syringe (Sherwood Medical, St. Louis, MO) and25-gauge needle. The protein was dissolved in 1 ml of asterile physiologic buffer solution, pH 7.1, prepared byadding 0.5 ml of 4% sodium bicarbonate (Neut®, AbbottLabs., North Chicago, IL) to 10 ml of preservative-free0.9% sodium chloride (Abbott Labs, North Chicago, IL)and centrifuged 3580 X g for 15 minutes. To remove in-soluble protein and to sterilize the solution, the superna-tant was passed through wetted 0.45-fim and 0.22-,um

filters (Millipore Corp., Bedford, MA). Maximum recoveryof protein was ensured by washing the filters with an ad-ditional 0.5 ml of buffer. A sufficient amount of BenceJones protein was dissolved initially to yield, after centrif-ugation and filtration, the requisite final concentration ofprotein, as determined by a modification of the Folin-Ciocalteau method.14 The desired volume of fluid in-jected was -1.5 ml (maximum volume, 3 ml) and con-tained up to 300 mg of protein. After the designated pe-riod of protein administration, the mice were sacrificed bycervical dislocation and the organs were removed. Tis-sue was processed in two ways: for light microscopy andimmunohistochemistry, sections were either placed in10% buffered formalin and embedded in paraffin or wereembedded in optimum cutting temperature (OCT) me-dium (Miles Inc., Elkhart, IN) and snapfrozen in anisopentane bath chilled with liquid nitrogen; for electronmicroscopy, sections were placed in 2.5% glutaralde-hyde-0.05 M sodium cacodylate buffer, pH 7.2, and after1 to 2 hours, were transferred to the cacodylate bufferand embedded in Epon®.

Histopathology

For light microscopy, 4- to 6-pum tissue sections were cutand stained routinely with hematoxylin-eosin; for detec-tion of amyloid, the sections were treated with a freshlyprepared alkaline Congo-red solution and examined un-der polarized light using a Leitz filter polarizer with a gyp-sum plate and a filter analyzer. For electron microscopy,Epon®-embedded sections were examined with a Zeiss9S transmission electron microscope and photographed.

Immunohistochemistry

Six-p.rm paraffin-embedded tissue sections were cut on amicrotome, mounted on poly-L-lysine-coated slides,placed in a 370C incubator overnight, and deparaffinized.Four-p.m snapfrozen tissue sections were cut in a cryo-stat, placed on poly-L-lysine-coated slides, and air-driedovernight. Immunostaining was performed as describedpreviously. 15

For detection of mouse AP, air-dried cryostat sections(5 p.m) of snapfrozen tissue were wetted and blocked byincubation at room temperature for 10 minutes with 10%normal rabbit serum in 10 mM Tris-buffered saline (TBS).The sections were exposed for 60 minutes to a 1:200dilution of a specific sheep anti-mouse SAP antiserum16(this dilution of the anti-SAP antiserum had been previ-ously demonstrated to give optimal specific staining ofamyloid deposits, without any nonspecific background,in titration experiments on spleen from mice with casein-induced amyloidosis). As controls for specificity of stain-

Experimental Production of Human AL Amyloidosis 631AJP March 1992, Vol. 140, No. 3

ing, other sections were incubated with 1:200 dilutions ofnormal sheep serum and the sheep anti-mouse SAP an-tiserum, which had been absorbed with pure mouse SAPto remove all anti-SAP activity. The sections were washedtwice for 5 minutes with TBS containing 0.005% TritonX-1 00 (TBS-T), followed by TBS alone for a further 5 min-utes, and finally stained with a 1:80 dilution of fluoresceinisothiocyanate (FITC)-conjugated rabbit anti-sheep IgG.After a 30-minute incubation with the FITC conjugate, thesections were washed twice with TBS-T and once withTBS before being mounted in Citifluor (Citifluor Ltd., Lon-don, United Kingdom).

Amyloid Extraction

The methods employed for extraction of amyloid fromtissue were as described by Pras et al.17 Mouse kidneywas homogenized with 1.5 ml of cold saline in an ice bathusing an Omni-Mixer (Omni International, Inc., Water-bury, CT). The extract was centrifuged at 10,000 rpm for30 minutes at 40C and the pellet re-extracted twice morewith cold saline, once with 0.1 M Tris-sodium citrate-buffered saline, pH 8.0, and then again with saline untilthe A280 Of the supernatant was < 0.10. The resultant10,000-rpm pellet was homogenized with cold distilledwater and the extract centrifuged at 35,000 rpm for 3hours at 40C. The pellet obtained from the water extractwas then lyophilized.

Immunoblotting

The immunodetection of light chains and related compo-nents (Bence Jones protein and extracted amyloid) wasperformed by Western blotting; SDS/PAGE was done un-der reducing conditions using 20% homogeneous gels(Phastgels, LKB Pharmacia, Piscataway, NJ). Proteinswere transferred by electroblotting onto PVDF mem-branes (Millipore Corp, Bedford, MA). After blocking non-specific protein-binding sites with a 1% nonfat dry milksolution, the membranes were immunostained withmouse anti-human light-chain antibodies. Bound anti-body was detected using an avidin-biotin complex (ABC,Vector Laboratories, Burlingame, CA) and alkaline phos-phatase as the enzyme marker.

Results

Based on our finding that human A chains of a particularVA subgroup-V,vi-are preferentially associated withAL-type amyloid,1 20 we initially selected for study in themouse model one such Bence Jones protein. Attemptsto produce amyloid lesions in mice by single 300-mg

intraperitoneal injections of this XVI light chain were un-successful, i.e., no Congo-red positive deposits werefound in the kidneys of mice sacrificed 48 hours postin-jection. Similarly, no Congophilic deposits were foundwhen mice received -100 mg of protein each of 5 con-secutive days for periods ranging from 2 to 5 weeks (900to 2200 mg of administered protein). However, when theperiod of injection was extended to 12 weeks (5300 mgof protein), Congo-red-stained sections of the mousekidneys revealed occasional green birefringent perivas-cular deposits.We had previously found in other short-term experi-

ments that dehydration of mice before injection of certainhuman Bence Jones proteins accelerated the depositionof these proteins as renal tubular casts.21 We thus inves-tigated if water deprivation would also enhance in micethe formation of human light-chain-related amyloid de-posits. In the first series of experiments, mice were dehy-drated by temporarily removing the water bottle from thecage for a 48-hour period before protein injection. Underthis condition, the administration of a 200-mg dose of theXVI Bence Jones protein given three times over an8-week period (600 mg of protein) resulted in the forma-tion of birefringent Congo-red positive renovascular de-posits comparable to those found experimentally in thenondehydrated mouse that received -5 gm of the hu-man light chain.

Based on these results, we increased in subsequentexperiments the total amount of protein injected to -2gm. Mice were dehydrated for 24 hours before each 200-mg injection of the XVI Bence Jones protein given 10times over an 8-week period. Extensive renovascularamyloid was found, as evidenced in Congo-red-stainedsections in which green birefringent deposits werepresent within blood vessel walls (Figure 1, left), as wellas interstitially. These Congophilic deposits contained theinjected human light chain, as demonstrated immunohis-tochemically using a monospecific anti-human X-light-chain antiserum (Figure 1, middle). The experimentallyinduced amyloid deposits did not react with anti-mouseIg or anti-mouse amyloid A antisera. Under the sameexperimental conditions, no renal amyloid deposits werefound in mice injected with a Bence Jones protein ob-tained from a patient who did not have amyloidosis. Noanti-human light-chain antibodies were detected in thesera of mice injected with either Bence Jones protein.

To determine if the human light-chain amyloid depos-its induced in the mouse contained AP, kidney sectionswere examined immunohistochemically for its presence.Using a specific anti-mouse SAP antiserum, bright fluo-rescence was noted in vascular structures correspond-ing precisely in location to those that were Congo-redpositive and displayed green birefringence (Figure 1,right). Sections stained either with anti-mouse SAP an-

632 Solomon, Weiss, and PepysAJP March 1992, Vol. 140, No. 3

ts

]_._r

_._

_

*s_

1 b

I

kDa A

20- I4

tiserum, which had been absorbed with pure SAP, or withnormal sheep serum, as well as with anti-human SAPantiserum, were negative, i.e., no fluorescence wasnoted. Furthermore, AP was not present in the kidneys ofmice injected with the non-amyloid-associated BenceJones protein.

Definitive evidence for the amyloid nature of the ex-perimentally induced light-chain deposits was found byelectron microscopy. Ultrastructural analysis revealed thepresence of 8- to 1 0-nm nonbranching fibrils comparablein location to the green birefringent deposits depicted inFigure 1. The fibrillar nature of the experimentally inducedamyloid was comparable to that found in the renovascu-

Figure 1. Experimental production of human AL (A) amyloid deposits: Renaldeposition. (Left) Birefringent Congo-red positive deposits in mouse renal bloodvessel wall (polarizing microscopy; x400). (Middle) Human X-light-chain depositsin mouse renal blood vessel wall (immunoperoxidase technique; primary antise-rum: anti-human X chain; x 400). (Right) Mouse AP-component deposits in mouserenal blood vessel wall (immunofluorescence technique;primary antiserum: anti-mouse SAP; X400).Figure 3. Experimental production of renal intratubular non-birefringentCongo-redpositive deposits containing human X light chains and mouse amyloidP component. (Left) Congo red stain (x400). (Middle) Immunoperoxidase stain-ing with anti-human X-light-chain antiserum (x 400). (Right) Immunofluores-cence staining with anti-mouse SAP antiserum (x400).Figure 4. Characterization ofthe human light chainsfound in the experimentallyinduced amyloid renal deposits. (Left) The green birefringent Congo-red positivenature of the sediment from the water-soluble extract is shown (X400). (Right)Immunoblot analysis of the human XVI Bence Jones protein injected into mice(lane A) and of the amyloid extract (lane B) (SDS/PAGEfollowed by immuno-blotting with a specfic anti-human XVI antiserum; the position of the 20-kDamolecular weight marker is as indicated).

lar amyloid deposits of a patient with XVI light-chain-associated amyloidosis (Figure 2).

In addition to the presence of vascular and interstitialamyloid in the kidneys of mice injected with the human Achain, occasional renal tubular casts, located mainly inthe loop of Henle, were Congophilic and contained boththe injected light chain and mouse AP but did not exhibitgreen birefringence (Figure 3). Ultrastructural examina-tion of the renal tubular casts did not reveal fibrils.

In an effort to characterize the molecular properties ofthe human X-chain protein contained within the experi-mentally induced amyloid deposits, we extracted thismaterial from the mouse kidney. The protein sediment

il.

Experimental Production of Human AL Amyloidosis 633AJP March 1992, Vol. 140, No. 3

HUMAN

Figure 2. Fibrillar nature of mouse and human renal blood vessel wall light-chain deposits. Electronphotomicrograps X27,500, insetX84,000;bar =500 nm,

obtained after high-speed centrifugation of the water-soluble extract exhibited green birefringence after Congored staining (Figure 4, left). Examination of this material bySDS/PAGE revealed two protein components: one rep-resented -80% of the extracted protein and consisted ofan -22-kDa constituent identical in mass to that of thelight chain injected; the second was of lower molecularmass (-18- to 20-kDa). Both proteins reacted in immu-noblotting experiments with an antiserum specific for hu-man XVI light chains (Figure 4, right), as well as with ageneral anti-human X-chain antiserum (not illustrated).

Based on the results obtained with the human amy-loid-associated X light chain, we injected mice with a KBence Jones protein obtained from a patient with exten-sive AL amyloidosis; mice that received 2300 mg of thisprotein over an 8-week period developed renovascularamyloid deposits that contained the human K light chain.Of the total dose of protein administered, 700 mg weregiven over the first 2 weeks in the form of 1 00-mg injec-tions into nondehydrated mice; the remainder of the pro-tein (1600 mg) was given as 200-mg doses given twiceweekly to mice that were dehydrated for 24 hours prein-jection. When the amount of protein injected was in-creased to 4900 mg given over a 12-week period, renaldeposition of the human K chain as amyloid was moreextensive and involved not only blood vessel walls butinterstitium as well. Furthermore, examination of Congo-red-stained sections of mouse liver, heart, lungs, andspleen revealed perivascular green birefringent deposits

(Figure 5). The presence of the human light chain withinthe Congophilic areas was demonstrated using a spe-cific anti-human K-chain antiserum (not illustrated). Elec-tron microscopic analysis of the spleen (Figure 6) andkidneys showed the presence of fibrils having compara-ble ultrastructural features to those found in the mouseinjected with the A Bence Jones protein. Under the sameexperimental conditions, no renovascular amyloid de-posits were found in mice injected over an 8-week periodwith a non-amyloid-associated Bence Jones protein. Af-ter 12 weeks of injection, the only evidence of amyloidformation was the presence of a single renovascularCongophilic deposit exhibiting green birefringence.

Discussion

We have demonstrated that monoclonal light chains ob-tained from patients with AL amyloidosis, when injectedinto mice, were deposited as amyloid. The induced de-posits had the characteristic tinctorial and ultrastructuralfeatures of amyloid, namely, the deposits were Congo-philic and exhibited green birefringence by polarizing mi-croscopy and, by electronmicroscopy, were fibrillar in na-ture. Using specific anti-human K-or X-chain antiserum,the deposits were shown to contain human, not mouse,Ig. The vascular and interstitial locations of the amyloiddeposits found experimentally were comparable to thatfound clinically. In contrast to the rapid development

MOUSE

634 Solomon, Weiss, and PepysAJP Maich 1992, Vol. 140, No. 3

KIDNEY

LIVER

HEART

LUNG

SPLEEN

Figure 5. Experimental production of human AL (K) amyloid deposits: Systemic deposition: Birefringent Congo-red positive, interstitialdeposits in mouse kidney, liver, heart, lungs, and spleen (polarizing microscopy; x 400).

(within 48 hours) of Bence Jones protein-containing renaltubular casts, crystals, or basement-membrane depositsthat resulted from the single injection into mice of 200 to300 mg of human monoclonal light chains,12 we found

that under the experimental conditions employed, con-siderably more protein (-2 gm) given over a muchlonger period (-2 months) was necessary for amyloidformation.

Experimental Production of Human AL Amyloidosis 635AJP March 1992, Vol. 140, No. 3

;-k Se.. by'.t... < ' //'v. \vs .'g - .sCif aNe ,),%f.iX* ; *1 t t 8 % s 4tt' \\ _3 ti'

k^''sS PX , ,E'§t].'.] _ w.

..:+ .^(.7_

...

Figure 6. Fibrillar nature ofexperimentally induced splenic light-chain deposits. Electronphotomicrographs: x 26,000, insetx 75,000; bar = 100 nm.

In addition to the human light-chain nature of the ex-

perimentally induced amyloid deposits, we demon-strated that these deposits also contained mouse AP.The AP was detected in vascular and interstitial areas

that contained birefringent Congophilic fibrillar light-chaindeposits; in addition, this component was found in theoccasional nonbirefringent Congophilic renal tubularcasts that contained the injected human light chain. Thehomogeneous nature of the deposited light chain mayhave obscured the detection of fibrils within the protein-aceous casts and, in any case, the sample viewed byelectron microscopy was necessarily limited comparedwith that seen by light microscopy; thus, sparse fibrilsmay have been missed. Such sparsity could also ac-

count for the lack of green birefringence despite Congo-philia since it is well recognized that birefringence is de-pendent on the presence of a sufficient density of amy-loid. In this respect, the immunohistochemicalidentification of AP in amyloid deposits has been shownto be more sensitive than green birefringence for the de-tection of amyloid.22

The contributory role of host AP in amyloid depositionhas been evidenced by the demonstration of human APin porcine insulin-containing amyloid induced at the siteof insulin injections in a patient with diabetes mellitus.23The presence of mouse AP in the amyloid deposits de-scribed here is the first demonstration that mouse SAPbinds to human amyloid fibrils and extends to this uniquexenogenic model of amyloidosis the invariable associa-tion of AP with amyloid deposits. It has long been recog-

nized that a number of precursor proteins can be in-duced to form amyloid fibril-like structures in vitro,'- butwhether SAP is required or involved in amyloid fibrillogen-esis and deposition in vivo remains unresolved, even

though there is experimental and clinical evidence thatSAP levels and turnover are increased during amy-

loidogenesis.2126

The finding that under similar experimental conditionsBence Jones proteins obtained from patients with amy-loidosis were more readily deposited as amyloid in themouse (as opposed to a non-amyloid-associated lightchain) indicates the inherent amyloidogenicity of certainlight chains. A structural basis for light-chain amyloidoge-nicity has been construed from the prevalence of X ver-SUS K chains in amyloid deposits, the presence of un-usual (hydrophobic) amino acid substitutions in the light-chain variable domain (VL), and the electronegativenature of amyloid versus non-amyloid-associated pro-tein.5'27 28 Even more striking is that A light-chain mem-bers of a rare V-region subgroup - VVl - are prefer-entially associated with the amyloid process.1820 Al-though the V region of XVI chains contains a subgroupdistinctive, two-residue insertion in the third frameworkregion, unusual insertions, deletions, or substitutions arenot readily apparent among amyloid-associated A chainsof other V. subgroups, as well as among amyloid-associated K chains. To discriminate amyloid versus non-amyloid light chains on the basis of primary structure, i.e.,to identify amyloid-specific residues, is confounded bythe inherent V-region amino-acid sequence variabilitythat results from the large number of human VL genes,recombination of the two VL-encoding genes, V and J,and especially, somatic mutation.29 Whether amyloid-associated light chains have common tertiary structuralfeatures that permit the formation of the characteristicp-pleated structure6 remains to be determined. Conceiv-ably, most or all monoclonal light chains are potentiallyamyloidogenic, as evidenced by the generation of Con-gophilic, fibrillar VL fragments3'' and VL-related pep-tides36 through enzymatic digestion of Bence Jonesproteins. However, the lack of correlation between the invitro production of amyloid and the amyloidogenicity ofthe same protein in vivo33 suggests that additional (host)factors are required for amyloid formation. Our experi-mental findings indicate that one factor of pathophysio-logic importance - dehydration - which potentiatesmyeloma cast nephropathy,21 may also promote light-chain amyloid formation.

In most instances, the major portion of the light-chainprotein extracted from the amyloid deposits found in pa-tients with AL amyloidosis consists only of the VL domainor the VL plus up to approximately one-half of the light-chain constant domain (CL); only rarely is the intact pro-tein present.5-6 Whether these fragments represent syn-thetic37 or catabolic products or if a partially degradedlight chain is necessary for amyloid formation has not yetbeen conclusively established. Our analyses of the pro-tein extracted from the experimentally induced XVI-associated amyloid demonstrated that the depositedprotein consisted predominately of intact light chainsand, to a lesser extent, of a light-chain-related fragment.

636 Solomon, Weiss, and PepysAJP March 1992, Vol. 140, No. 3

Thus, it is conceivable that, due to the inherent suscep-tibility of the CL to proteolysis under physiologic condi-tions,38 the fragmented light-chain constituents recov-ered in extracts of light-chain-associated amyloid are theresult of tissue proteolysis that occurred after the amyloidwas formed.

The ability to reproduce in an animal model the bio-logical and chemical features of human light-chain-associated amyloidosis provides a novel means to deter-mine the amyloidogenic potential of human monoclonallight chains. Furthermore, the availability of an in vivomodel to investigate the pathogenesis of AL amyloidosisand factors that accelerate (e.g., amyloid enhancing fac-tor:394), prevent, or reverse its formation is of obviousclinical importance.

Acknowledgments

The authors thank Ms. Teresa Williams, Ms. Bobbie Rush, Ms.Mildred Conley, Dr. David Gerard, Ms. Glenys Tennent, and Ms.Melanie Simpson for their assistance with the experimental as-pects of this study, and Ms. Julie Ottinger for manuscript prep-aration.

References

1. Osserman EF, Takatsuki K, Talal N: Multiple myeloma. Thepathogenesis of "amyloidosis." Studies on the role of abnor-mal gamma globulins and gamma globulin fragments of theBence Jones (L-polypeptide) type in the pathogenesis of"sprimary" and "secondary amyloidosis," and the "amyloid-osis" associated with plasma cell myeloma. Semin Hematol1964,1:3-86

2. Isobe T, Osserman EF: Pathologic conditions associatedwith plasma cell dyscrasias: A study of 806 cases. Ann NYAcad Sci 1971, 190:507-518

3. Isobe T, Osserman EF: Patterns of amyloidosis and theirassociation with plasma-cell dyscrasia, monoclonal immu-noglobulins and Bence-Jones proteins. N EngI J Med 1974,290:473-477

4. Glenner GG, Terry W, Harada M, lsersky C, Page D: Amy-loid fibril proteins: Proof of homology with immunoglobulinlight chains by sequence analyses. Science 1971,172:1150-1151

5. Stone MJ: Amyloidosis: A final common pathway for proteindeposition in tissues. Blood 1990, 75:531-545

6. Glenner GG: Amyloid deposits and amyloidosis: The 1-fi-brilloses. N Engl J Med 1980, 302:1283-1292

7. Pepys MB: Amyloidosis. Immunological Diseases, 4th Ed.,Vol. 1. Edited by M Samter, DW Talmage, MM Frank, KFAusten, HN Claman. Boston, Little Brown and Co., 1988, pp631-674

8. Baltz ML, Caspi D, Evans DJ, Rowe IF, Hind CRK, PepysMB: Circulating serum amyloid P component is the precur-sor of amyloid P component in tissue amyloid deposits. ClinExp Immunol 1986, 66:691-700

9. Hawkins PN, Lavender PJ, Pepys MB: Evaluation of sys-temic amyloidosis by scintigraphy with 1231-labeled serumamyloid P component. N Engl J Med 1990, 323:508-513

10. Zschiesche W, Jakob W: Pathology of animal amyloidoses.Pharmacol Ther 1989, 41:49-83

11. Koss MN, Pirani CL, Osserman EF: Experimental BenceJones cast nephropathy. Lab Invest 1976, 4:579-591

12. Solomon A, Weiss DT, Kdttine AA: Nephrotoxic potential ofBence Jones proteins. N Engl J Med 1991, 324:1845-1851

13. Solomon A, Weiss DT: Experimental production of humanamyloidosis AL. Amyloid and Amyloidosis. Edited by NatvigJB, F0rre 0, Husby G, Husebekk A, Skogen B, Sletten K,Westermark P. Dordrecht, The Netherlands, Kluwer Aca-demic Publishers, 1991, pp 193-196

14. Solomon A: Light chains of human immunoglobulins. MethEnzymol 1985,116:101-121

15. Solomon A, Weiss DT, Macy SD, Antonucci RA: Immunocy-tochemical detection of kappa and lambda light chain Vregion subgroups in human B-cell malignancies. Am JPathol 1990, 137:855-862

16. Pepys MB: Isolation of serum amyloid P component (proteinSAP) in the mouse. Immunology 1979, 37:637-641

17. Pras M, Shubert M, Zucker-Franklin D, Rimon A, FranklinEC: The characterization of soluble amyloid prepared in wa-ter. J Clin Invest 1968, 47:924-933

18. Solomon A, Frangione B, Franklin EC: Bence Jones proteinsand light chains of immunoglobuins. Preferential associationof the V A, v subgroup of human light chains with amyloidosisAL(A). J Clin Invest 1982, 70:453-460

19. Solomon A, Kyle RA, Frangione B: Light chain variable re-gion subgroups of monoclonal immunoglobulins in amyloid-osis AL. Amyloidosis. Edited by Glenner GG, Osserman EF,Benditt, EP, Calkins E, Cohen AS, Zucker-Franklin D. NewYork, Plenum Press, 1986, 449-462

20. Solomon A, Weiss DT: Serologically defined V region sub-groups of human X light chains. J Immunol 1987,139:824-830

21. Solomon A, Weiss DT: A perspective of plasma cell dyscra-sias: clinical implications of monoclonal light chains in renaldisease. The Kidney in Plasma Cell Dyscrasias. Edited byMinetti L, D'Amico G, Ponticelli C. Dordrecht, The Nether-lands, Kluwer Academic Publishers, 1988, pp 3-18

22. Hawkins PN, Myers MJ, Epenetos M, Caspi D, Pepys MB:Specific localization and imaging of amyloid deposits in vivousing 1231-labeled serum amyloid P component. J Exp Med1988,167:903-913

23. Dische FE, Wernstedt C, Westermark GT, Westermark P,Pepys MB, Rennie JA, Gilbey SG, Watkins PJ: Insulin as anamyloid-fibril protein at sites of repeated insulin injections ina diabetic patient. Diabetologia 1988, 31:158-161

24. Baltz ML, Gomer K, Davies AJS, Evans DJ, Klaus GGB,Pepys MB: Differences in the acute phase responses ofserum amyloid P component (SAP) and C3 to injections ofcasein or bovine serum albumin in amyloid-susceptible andresistant mouse stains. Clin Exp Immunol 1980, 39:355-360

25. Snel FWJJ, Niewold ThA, Baltz ML, Hol PR, van Ederen AM,Pepys MB, Gruys E: Experimental amyloidosis in the ham-

Experimental Production of Human AL Amyloidosis 637AJP March 1992, Vol. 140, No. 3

ster: correlation between hamster female protein levels andamyloid deposition. Clin Exp Immunol 1989, 76:296-300

26. Hawkins PN, Wootton R, Pepys MB: Metabolic studies ofradioiodinated serum amyloid P component in normal sub-jects and patients with systemic amyloidosis. J Clin Invest1990, 86:1862-1869

27. Bellotti V, Merlini G, Bucciarelli E, Perfetti V, Quaglini S, As-cari E: Relevance of class, molecular weight and isoelectricpoint in predicting human light chain amyloidogenicity. Br JHaematol 1990, 74:65-69

28. Benson MD, Dwulet FE, Madura D, Wheeler G: Amyloidosisrelated to a XIV immunoglobulin light chain protein. Scand JImmunol 1989, 29:175-179

29. Solomon A: Light chains of immunoglobulins: Structural-genetic correlates. Blood 1986, 68:603-610

30. Glenner GG, Eanes ED, Bladen HA, Terry W, Page DL: Cre-ation of "amyloid" fibrils from Bence Jones proteins in vitro.Science 1971, 174:712-714

31. Shirahama T, Benson MD, Cohen AS, Tanaka A: Fibrillarassemblage of variable segments of immunoglobulin lightchains: An electron microscopic study. J Immunol 1973,110:21-30

32. Epstein WV, Tan M, Wood IS: Formation of "amyloid" fibrilsin vitro by action of human kidney lysosomal enzymes onBence Jones proteins. J Lab Clin Med 1974, 84:107-1 10

33. Linke RP, Zucker-Franklin D, Franklin EC: Morphologic,chemical, and immunologic studies of amyloid-like fibrilsformed from Bence proteins by proteolysis. J Immunol 1973,111:10-23

34. Eulitz M, Linke R: Amyloid fibrils derived from V-region to-gether with C-region fragments from a All-immunoglobulin

light chain (HAR). Biol Chem Hoppe Seyler 1985,366:907-915

35. Eulitz M, Breuer M, Linke RP: Is the formation of AL-typeamyloid promoted by structural peculiarities of immunoglob-ulin L-chains? Primary structure of an amyloidogenic X-L-chain (BJP-ZIM). Biol Chem Hoppe Seyler 1987, 368:863-870

36. Eulitz M, Breuer M, Eblen A, Weiss DT, Solomon A: Produc-tion of amyloidogenic peptides from human immunoglobulinlight (L)-chains. Amyloid and Amyloidosis 1990. Edited byNatvig JB, F0rre 0, Husby G, Husebekk A, Skogen B, Slet-ten K, Westermark P. Dordrecht, The Netherlands, KluwerAcademic Publishers, 1991, 505-510

37. Buxbaum J: Aberrant immunoglobulin synthesis in lightchain amyloidosis. Free light chain and light chain fragmentproduction by human bone marrow cells in short-term tissueculture. J Clin Invest 1986, 78:798-806

38. Solomon A, McLaughlin CL: Bence-Jones proteins and lightchains of immunoglobulins. I. Formation and characteriza-tion of amino-terminal (variant) and carboxyl-terminal (con-stant) halves. J Biol Chem 1969, 244:3393-3404

39. Axelrad MA, Kisilevsky R, Willmer J, Chen SJ, Skinner M:Further characterization of amyloid-enhancing factor. LabInvest 1982, 47:139-146

40. Kisilevsky R, Boudreau BA. Kinetics of amyloid deposition. I.The effects of amyloid-enhancing factor and splenectomy.Lab Invest 1982, 48:53-59

41. Baltz M, Caspi D, Hind CRK, Feinstein A, Pepys MB: Isola-tion and characterization of amyloid enhancing factor (AES).Amyloidosis. Edited by Glenner GG, Osserman EF, BendittEP, Calkins E, Cohen AS, Zucker-Franklin D. New York, Ple-num Press, 1986, pp 123-128