-Responsive Transcription Factor SAF-1 Activity Is Linked to the Development of A

This information is current as Alpana Ray, Arvind Shakya, Deepak Kumar, Merrill D. of September 27, 2021. Benson and Bimal K. Ray J Immunol 2006; 177:2601-2609; ; doi: 10.4049/jimmunol.177.4.2601 http://www.jimmunol.org/content/177/4/2601 Downloaded from

References This article cites 49 articles, 20 of which you can access for free at: http://www.jimmunol.org/content/177/4/2601.full#ref-list-1 http://www.jimmunol.org/ Why The JI? Submit online.

• Rapid Reviews! 30 days* from submission to initial decision

• No Triage! Every submission reviewed by practicing scientists

• Fast Publication! 4 weeks from acceptance to publication

by guest on September 27, 2021 *average

Subscription Information about subscribing to The Journal of Immunology is online at: http://jimmunol.org/subscription Permissions Submit copyright permission requests at: http://www.aai.org/About/Publications/JI/copyright.html Email Alerts Receive free email-alerts when new articles cite this article. Sign up at: http://jimmunol.org/alerts

The Journal of Immunology is published twice each month by The American Association of Immunologists, Inc., 1451 Rockville Pike, Suite 650, Rockville, MD 20852 Copyright © 2006 by The American Association of Immunologists All rights reserved. Print ISSN: 0022-1767 Online ISSN: 1550-6606. The Journal of Immunology

Inflammation-Responsive Transcription Factor SAF-1 Activity Is Linked to the Development of Amyloid A Amyloidosis

Alpana Ray,1*† Arvind Shakya,* Deepak Kumar,* Merrill D. Benson,§ and Bimal K. Ray*†

Abundantly expressed serum amyloid A (SAA) under chronic inflammatory conditions gives rise to insoluble aggregates of SAA derivatives in multiple organs resulting in reactive amyloid A (AA) amyloidosis, a consequence of , Crohn’s disease, ankylosing spondylitis, familial Mediterranean fever, and Castleman’s disease. An inflammation-responsive transcription factor, SAF (for SAA activating factor), has been implicated in the sustained expression of amyloidogenic SAA under chronic inflammatory conditions. However, its role in the pathogenesis of AA amyloidosis has thus far remained obscure. In this paper we have shown that SAF-1, a major member of the SAF family, is abundantly present in human AA amyloidosis patients. To assess whether SAF-1 is directly linked to the pathogenesis of AA amyloidosis, we have developed a SAF-1 transgenic mouse

model. SAF-1-overexpressing mice spontaneously developed AA amyloidosis at the age of 14 mo or older. Immunohistochemical Downloaded from analysis confirmed the nature of the amyloid deposits as an AA type derived from amyloidogenic SAA1. Furthermore, SAF-1 transgenic mice rapidly developed severe AA amyloidosis in response to azocasein injection, indicating increased susceptibility to inflammation. Also, during inflammation SAF-1 transgenic mice exhibited a prolonged acute phase response, leading to an extended period of SAA synthesis. Together, these results provide direct evidence that SAF-1 plays a key role in the development of AA amyloidosis, a consequence of chronic inflammation. The Journal of Immunology, 2006, 177: 2601–2609. http://www.jimmunol.org/ myloidosis is a large group of incurable diseases in these different isoforms, the SAA1 of both human and mice is which deposits of fragmented or sometimes intact pro- predominant in AA amyloid deposits. A teins referred to as amyloid fibrils accumulate in the ex- A major factor responsible for the development of AA amyloid- tracellular spaces of organs, leading to dysfunction of the affected osis is the increased synthesis and subsequent degradation of the organs (1, 2). Amyloid fibrils are formed when normally soluble precursor protein SAA1 under chronic inflammatory conditions. precursor undergo conformational changes and polymer- Under normal conditions a trace amount of this protein is present, ize as insoluble fibrils, with the ␤-pleated sheet conformation vis- but in response to inflammation its level can increase up to 1000- ible under polarized light as yellowish-green birefringence (3). Al- fold. Increased synthesis of the SAA protein is primarily due to the though all organs could be affected, amyloid deposits are transcriptional induction of the SAA (reviewed in Ref. 10). by guest on September 27, 2021 predominantly seen in the spleen, , , and brain. In sec- Proinflammatory and mediators such as IL-1, IL-6, ondary or inflammation-associated amyloid A (AA)2 amyloidosis, TNF-␣, bacterial LPS, and PMA either alone or in combination inflammation plays a central role, as patients suffering from vari- induce SAA transcription (10). Multiple transcription factors, in- ous chronic inflammatory diseases including rheumatoid arthritis, cluding C/EBP (11, 12), NF-␬B (13–15), and SAF (SAA activat- Crohn’s disease, ankylosing spondylitis, familial Mediterranean ing factor) (16, 17), have been defined in the regulation of SAA fever, and Castleman’s disease, frequently develop this condition induction. Several other transcription factors, including YY1, SEF, as a secondary complication. In AA amyloidosis, the amyloid fibril AP-2, and Sp1, have also been implicated (reviewed in Ref. 10). is derived from a precursor protein identified as the acute phase Although C/EBP and NF-␬B play critical roles in tissue-specific serum AA (SAA) (2) protein (4). The SAA family is composed of expression of amyloidogenic SAA1 and SAA2 during acute inflam- a number of and proteins. The human SAA gene family matory conditions, SAF has been implicated in the expression un- contains two highly homologous SAA1 and SAA2 genes, a less der the chronic inflammatory condition that predisposes to AA related SAA4 gene, and a nonexpressible SAA3 gene (5–8). Sim- amyloidosis (18). Normally, under acute inflammatory conditions ilarly as in humans, SAA1 and SAA2 isoforms in mice are highly such as septic shock, inflammatory signals are intense but short homologous, whereas SAA3 and SAA4 are distinct (8, 9). Among lived. Thus, the effect of inflammation subsides rapidly within 24 to 48 h. The amyloidogenic SAA level, which is highly increased during this condition, also rapidly declines to a normal level. In *Department of Veterinary Pathobiology and †Department of Orthopaedic Surgery, contrast, during chronic inflammatory conditions such as rheuma- University of Missouri, Columbia, MO 65211; and ‡Department of Pathology and toid arthritis the effect of inflammation persists for an extended Laboratory Medicine, Indiana School of Medicine and Roudebush Veterans Affairs Medical Center, Indianapolis, IN 46202 period of time, causing prolonged expression of amyloidogenic SAA with the potential consequence of developing AA amyloid- Received for publication November 22, 2005. Accepted for publication May 11, 2006. osis. An earlier finding of increased SAF activity under chronic The costs of publication of this article were defrayed in part by the payment of page inflammatory conditions and its ability to promote expression of charges. This article must therefore be hereby marked advertisement in accordance amyloidogenic SAA (18) suggests that SAF could be linked to AA with 18 U.S.C. Section 1734 solely to indicate this fact. amyloidosis. To test this possibility, we have examined amyloid 1 Address correspondence and reprint requests to Dr. Alpana Ray, Department of tissues from human AA amyloidosis patients for evaluation of the Veterinary Pathobiology, University of Missouri, 126A Connaway Hall, Columbia, MO 65211. E-mail address: [email protected] level of this transcription factor. To further examine this potential 2 Abbreviations used in this paper: AA, amyloid A; SAA, serum AA; SAF, SAA causal link of SAF, we have developed a transgenic mouse model activating factor. that overexpresses SAF-1, a major isoform of this transcription

Copyright © 2006 by The American Association of Immunologists, Inc. 0022-1767/06/$02.00 2602 SAF-1 AND AA AMYLOIDOSIS

factor. In this paper we provide evidence that increased abundance binding was blocked for 30 min at 37°C with 10% normal goat serum. of SAF-1 plays a critical role in the pathogenesis of AA Slides were incubated overnight at 4°C with primary Abs. The slides were amyloidosis. rinsed twice in the wash buffer containing 0.05% Tween 20 and then in- cubated with the secondary Ab. Bound primary Ab was detected using the HRP method with substrate-chromogen solution. Sections were counter- Materials and Methods stained with Mayer’s hematoxylin solution. In control experiments, the Animal studies primary Ab was preabsorbed using purified SAF-1 protein. SAF-1 transgenic mice were developed as described earlier (19). The CMV Analysis of amyloid fibril protein immediate-early enhancer/promoter drives transcription of the SAF-1 gene in multiple organs of the transgenic mice that exhibit elevated levels of SAF-1 Spleen and kidney tissues were homogenized in ice-cold saline using a activity as revealed by promoter-binding activity. The identities of the trans- Polytron homogenizer (Brinkmann Instruments). The homogenates were ϫ genic and nontransgenic mice were determined by PCR analysis of genomic centrifuged at 10,000 g for 30 min at 4°C. The pellets were washed in DNA isolated from tail biopsies as described before (19). Both SAF-1 trans- ice-cold saline twice and once with a buffer containing 10 mM Tris-HCl genic and nontransgenic littermates belong to a hybrid strain called B6C3F1 (pH 8.0) and 0.05 M sodium citrate. The pellets were resuspended by ϫ that is derived from C57BL/6J and C3H/HeJ inbred lines; they are heterozy- homogenization in ice-cold saline and centrifuged at 10,000 g for 30 min gous for the unique loci and homozygous for the common loci for these two at 4°C. The pellets were resuspended in ice-cold distilled water and cen- ϫ inbred lines. trifuged at 20,000 g for3hat4°C. The final pellets were resuspended in a buffer containing 250 mM Tris-HCl (pH 8.0) and 6 M guanidine HCl. Amyloid induction Aliquots were fractionated in a 16% polyacrylamide gel, and the protein bands were visualized by Coomassie brilliant blue staining. Unstained pro- Mice (8 wk old) were divided into two groups: SAF-1 transgenic and teins in the duplicate lanes were excised from the gel, subjected to tryptic nontransgenic. These two groups were subdivided into five subgroups, and

digestion, and analyzed by MALDI-TOF mass spectrometry (Perceptive Downloaded from each subgroup contained seven animals. Animals received a daily s.c. in- STR; Applied Biosystems). The peptides were identified using the MS-Fit jection of 300 ␮l of a 80 mg/ml solution of azocasein (Sigma-Aldrich) in search program. 10 mM NaHCO3 for 7, 14, 21, 35, and 42 days. Control groups of trans- genic and nontransgenic mice received 10 mM NaHCO3 only. The animals Induction of acute phase response and analysis of plasma SAA were euthanized by CO2 asphyxiation at the end of treatments. Tissues level were collected, a portion was fixed in 10% buffered formalin, and the remaining tissues were stored in liquid nitrogen until further use. All pro- Four groups of mice (n ϭ 8 per group) consisting of young (8 wk old) and

cedures used for treatment of mice in this study were approved by the aged (14 mo old) SAF-1 transgenic or nontransgenic mice were given a http://www.jimmunol.org/ University of Missouri Institutional Animal Care and Use Committee single s.c. injection of 1 ml of a 80 mg/ml solution of azocasein (Sigma- ␮ (Columbia, MO). Aldrich) in 10 mM NaHCO3. Blood samples (200 l) were collected from each mouse before azocasein injection and on days 1, 2, 3, 4, 7 and 21 after Human amyloid tissue samples injection, and plasma was isolated by centrifugation. SAA level in the plasma before and after azocasein injection was measured using an ELISA Formalin-fixed and paraffin-embedded autopsy specimens of kidney from method following the manufacturer’s protocol (BioSource International). five patients, ranging in age from 35 to 66, were used in this study. These Each sample was assayed in duplicate. The average OD reading from the patients had suffered from AA amyloidosis due to an underlying disease of duplicates sample was then plotted against a standard curve to obtain the rheumatoid arthritis, and one was paraplegic due to spinal injury. Normal concentration of SAA. tissue specimens in formalin-fixed and paraffin-embedded blocks were ob-

tained from the PathServe Autopsy and Human Tissue Bank. Northern blot analysis by guest on September 27, 2021 Analysis of amyloid Total RNA was extracted from multiple tissues by the guanidinium thio- cyanate method (23). Fifty micrograms of each RNA sample was fraction- Formalin-fixed tissues were paraffin embedded, and sections on slides were ated in a 1% agarose gel containing 2 M formaldehyde and transferred onto stained with alkaline alcoholic Congo red as described earlier (20). Briefly, a nylon membrane using capillary transfer method (24). Transferred RNA tissue sections in microscope slides were deparaffinized in xylene and hy- in the membrane was hybridized to a 32P-labeled oligonucleotide probe drated by stepwise submersion in 100, 95, 75, and 50% ethanol. The slides specific for mouse SAA1 (25). The same membrane was stripped and re- were incubated at room temperature (25°C) for 20 min in freshly prepared hybridized to 32P-labeled GAPDH cDNA probe to evaluate the relative 80% ethanol containing 3% NaCl and 0.01% NaOH. The slides were then amount of each RNA sample on the membrane. incubated at room temperature for 60 min in freshly prepared and filtered 0.2% Congo red in 80% ethanol containing 3% NaCl and 0.01% NaOH. RT-PCR analysis Excess dye was removed by washing the slides in distilled water three times. The slides were dipped in hematoxylin solution for 10 min and Total RNA from Formalin-fixed, paraffin-embedded human kidney tissue washed three times in distilled water. Tissue sections in slides were then samples was isolated using the RecoverAll total nucleic acid isolation kit dehydrated in graded (50–100%) ethanol and rinsed in xylene, and a cov- for formaldehyde- or paraformaldehyde-fixed, paraffin-embedded tissues erslip was placed with Permount. The Congo red-positive staining was obtained from Ambion. RT-PCR was performed using an RT-PCR kit fol- observed under polarized light. Each stained section was evaluated histo- lowing the manufacturer’s protocol (Invitrogen Life Technologies). One logically, and the severity scores of amyloid deposition were determined in microgram of DNase-treated RNA was used in the reverse transcription a blinded manner and rated on a scale of 0–5 following a previously de- reaction with a human SAA1-specific extension primer located in exon 4 scribed method (21). The grades used for rating are as follows: 0, no de- (5Ј-CCTGCCCCATTC ATCGGCAGCCTGATC-3Ј). PCR was performed tectable amyloid; 1, occasional amyloid specks; 2, mild amyloid deposits with two SAA1-specific primers, one located in exon 3 and the other in mainly in the perifollicular zone; 3, moderate deposits in perifollicular exon 4 (5Ј-GAGAGCCTACTCTGACATGAGAGAAG-3Ј and 5Ј-GCC zone; 4, abundant deposits but an intact follicular structure; and 5, severe AGTGAGTCCTCCGCACCATG-3Ј, respectively). DNA products were perifollicular and interfollicular deposits. resolved in a 2% agarose gel, and the identity of the amplified DNA was verified by DNA sequence analysis. Immunohistochemical analysis Western immunoblot assay Immunohistochemical staining was performed using a monoclonal anti- human AA (clone mc1; DakoCytomation) and affinity-purified anti-SAF-1 Tissues from nontransgenic and transgenic mice were homogenized in lysis rabbit IgG (22) as primary Abs. Details on the preparation of the anti- buffer (50 mM Tris-HCl (pH 7.5), 150 mM NaCl, 1 mM EDTA, 1 mM SAF-1 Ab have been described earlier (22). HRP-conjugated goat anti- PMSF, 1% Nonidet P-40, and 1% Brij96) using a Polytron homogenizer rabbit IgG and goat anti-mouse IgG were used as the secondary Abs. Tis- (Brinkmann Instruments). The cell lysates were centrifuged at 12,500 rpm sue sections were cut 5-␮m thick, deparaffinized in xylene, and rehydrated in a microcentrifuge at 4°C. Clear supernatants were used as the source of in graded ethanol solutions followed by rinsing in wash buffer (50 mM proteins whose concentrations were measured using the Bradford method Tris-HCl (pH 7.5) and 0.15 M NaCl). Endogenous peroxidase activity was (26). Fifty micrograms of protein from each sample was fractionated on

quenched by immersion in 3% H2O2 in methanol for 20 min, followed by 12% SDS-polyacrylamide gels and transferred onto a nitrocellulose rinses in the wash buffer. The slides were then incubated in 0.1% trypsin membrane using an electroblotter. After transfer, relative amounts of

solution with 0.1% CaCl2 for 60 min at 37°C to unmask Ags. Nonspecific proteins in each lane were verified by staining with Ponceau S solution The Journal of Immunology 2603

(Sigma-Aldrich). Immunoblotting was performed with an anti-SAF-1 Ab. SAF-1 Ab with purified SAF-1 protein completely abrogated the Bands were detected using a chemiluminiscence detection system (Amer- staining of the tissue sections (data not shown), indicating that the sham Biosciences). staining in Fig. 1A, e and f, was SAF-1-specific. To assess whether EMSA the kidney tissue expressed SAA in the pathogenic condition of amyloidosis, RNA was extracted from the paraffin-embedded tis- Nuclear extracts were prepared from various tissues of mice following the method described earlier (17). Protein content was measured by the Brad- sues and analyzed for the presence of SAA transcripts by RT-PCR. ford method (26). EMSA was performed with equal amounts of proteins A DNA product of predicted 180 bp, based upon the two PCR using a method described previously (17). A radiolabeled probe containing primers chosen, was detected (Fig. 1B). DNA sequence analysis of the SAF-binding element of the SAA promoter was prepared by using the PCR product further verified the identity of the DNA as SAA1. ␣ 32 [ - P]dCTP as the substrate to label the double-stranded oligonucleotide As shown in Fig. 1B, compared with a normal kidney the amyloid- probe (17). Two complementary oligonucleotides, 5Ј-CCCTTCCTCTC CACCCACAGCCCC-3Ј and 3Ј-GGGAAGGAGAGGTGGGTGTCGGG laden kidney contained higher levels of SAA1, suggesting that GG-5Ј, representing sequences from Ϫ254 to Ϫ230 of the rabbit SAA2 SAA1 is overexpressed in the amyloidotic kidney tissue. promoter, were annealed to prepare the double-stranded SAF-binding el- ement. Oligonucleotides containing nucleotide sequences from Ϫ190 to Ϫ150 and Ϫ98 to Ϫ78 of the mouse SAA1 promoter (9) were used as Development and characterization of SAF-1 transgenic mice C/EBP- and NF-␬B-binding elements, respectively, in the EMSA. These DNA molecules were radiolabeled using [␣-32P]dCTP as the substrate. A high level of SAF-1 in the affected tissues of AA amyloidosis DNA-protein complexes were separated in a nondenaturing 6% polyacryl- patients raises the possibility that SAF-1 might play a critical role amide gel. For competition analysis, 25- and 50-fold molar excesses of the in the pathogenesis of the disease. To further investigate this pos- Ϫ unlabeled double-stranded SAA SAF-1 oligonucleotide ( 254/-230) were Downloaded from used. As a nonspecific oligonucleotide, an Epstein-Barr nuclear Ag-1 sibility, SAF-1 transgenic mice were constructed in which a full- DNA-binding element containing the sequence 5Ј-TATCTGGGTAG length SAF-1 cDNA was placed under the control of CMV im- CATATGCTATCCTAAT-3Ј was used. For an Ab interaction study, anti- mediate-early enhancer/promoter (Fig. 2A). The presence of the SAF1 and anti-SAF2 Abs were added to the reaction mixture during a SAF-1 gene in the founder mice was verified by PCR amplification preincubation period of 30 min on ice. of DNA obtained from tail clippings using primers specific for the Statistical analysis SAF-1 cDNA and the CMV enhancer/promoter DNA. A band 574

bp in size was seen in transgenic mice (Fig. 2B). As the CMV http://www.jimmunol.org/ Incidence of in different groups of mice and the histological amyloid grades described in Table I were assessed by using the Mann- enhancer/promoter is capable of directing expression in multiple Whitney U test. A two-way ANOVA was performed to determine the tissues, the SAF-1 transgene was found to be expressed in various significant differences between the transgenic and nontransgenic mice in organs, albeit at different levels (19). the azocasein-treated and untreated groups.

Results SAF-1 transgenic mice spontaneously develop AA amyloid in Presence of SAF-1 protein in human AA amyloidotic kidney multiple organs tissue To examine whether SAF-1 transgenic mice contain amyloid de- Serial sections of amyloid deposit-containing kidney tissue from posits in different organs, both SAF-1 transgenic and nontrans- by guest on September 27, 2021 human AA amyloidosis patients were examined using Congo red genic littermates were sacrificed, and liver, kidney, and spleen tis- dye stain followed by visualization under polarized light. A rep- sues were tested. Tissue sections of mice of different ages, from 4 resentative sample revealed the presence of amyloid deposits in the wk to 18 mo, were stained with Congo red dye and visualized kidney of an AA amyloidosis patient (Fig. 1Aa). Immunohisto- under cross-polarized light. Up to the age of 6 mo there was no chemical analysis using an anti-AA Ab characterized the amyloid noticeable amyloid deposit in either the nontransgenic or the deposits as derivatives of amyloidogenic SAA (Fig. 1Ac). A high SAF-1 transgenic mice (data not shown). However, 12- to 14-mo- level of SAF-1 was also detected in the AA amyloid-laden tissues old SAF-1 transgenic mice were found to contain amyloid deposits (Fig. 1Ae). However, both AA and SAF-1 levels were found to be in spleen, kidney, and liver (Fig. 3, b, d, and f), whereas no such much lower in the normal tissues (Fig. 1A, d and f). Similar high deposits were seen in the age-matched nontransgenic animals (Fig. levels of AA protein and SAF-1 were detected in all other speci- 3, a, c, and e). The spleens of SAF-1 transgenic mice appeared to mens of human AA amyloidosis patients. Preabsorption of anti contain the most abundant amyloid deposits.

Table I. Histologic evaluation of amyloid deposits in the spleens of 8-wk-old mice

Amyloid Positive Severity Scoreb Days of Treatmenta Injection Total No. 012345

SAF-1-Tg 7 7 610000 SAF-1-Tg 14 7 133000 SAF-1-Tg 21 7 013300 SAF-1-Tg 35 7 003310 SAF-1-Tg 42 7 000322 Non-Tg 7 7 700000 Non-Tg 14 7 700000 Non-Tg 21 7 241000 Non-Tg 35 7 043000 Non-Tg 42 7 023200

a Tg, transgenic. b Values are the number of mice with the indicated severity score (0, no detectable amyloid; 1, occasional amyloid specks; 2, mild amyloid deposits in the perifollicular zone; 3, moderate perifollicular deposits; 4, abundant deposits but an intact follicular structure; 5, severe perifollicular and interfollicular deposits). 2604 SAF-1 AND AA AMYLOIDOSIS Downloaded from

FIGURE 2. Configuration of SAF-1 transgene and identification of SAF-1 transgenic mice. A, Schematic showing CMV-SAF-1 transgene. A full-length SAF-1 cDNA was placed under the control of a CMV promoter

using the pCI-Neo mammalian expression vector (Promega). The arrows http://www.jimmunol.org/ indicate the positions of transgene specific primers used for PCR to identify the SAF-1 transgene. B, PCR assay to identify the transgenic animals. A 574-nucleotide product is amplified only when the animals contain the SAF-1 transgene. Lanes 1 and 3 show PCR-amplified products from SAF-1 transgenic animals, while lanes 2 and 4 show the absence of such DNA in nontransgenic animals. Lane M shows molecular size markers.

tissues of SAF-1 transgenic mice (Fig. 4, d and h). Preabsorption of anti-AA and anti-SAF1 Abs abrogated staining, indicating AA by guest on September 27, 2021 and SAF1-specific staining in Fig. 4, b, d, f, and h (data not shown). Together, the data correlated well with the results ob- tained from human AA amyloidosis patients shown in Fig. 1. The AA amyloid protein in the amyloid fibril was further ana- lyzed by fractionation of the protein and subsequent mass spec- trometry analysis using MALDI-TOF. A peptide corresponding to FIGURE 1. Presence of SAF-1 protein in the kidney tissue of a human the amino-terminal region of mouse SAA1 was identified, indicat- AA amyloidosis patient. A, Kidney tissue sections from an AA amyloidosis ing that the AA amyloid is derived from the SAA1 isoform (data patient (AA) and normal cadaver (N) were stained with Congo red and not shown). No SAA2-specific peptides were detected in the mass viewed under polarized light (a and b, respectively). The bright stains in a spectrometry analysis. represent highly abundant amyloid deposits. In c and d immunohistochem- ical staining with anti-AA Abs for amyloidosis and normal kidney tissue SAF-1 and SAA1 levels are high in the affected organs of SAF-1 sections, respectively, is shown. In e and f is shown IHC staining with transgenic mice anti-SAF-1 Abs for amyloidosis and normal kidney tissue sections, respec- tively. Magnification, ϫ200. B, RT-PCR analysis of kidney for SAA ex- To test whether the increased level of SAF-1 detected in the AA pression. RNA was isolated from the paraffin-embedded kidney tissues as amyloid tissue might play a role in the increased SAA1 expression described in Materials and Methods. Following RT-PCR, the products in the amyloidotic organs of aged SAF-1 transgenic mice, we ex- were fractionated in a 2% agarose gel. Lanes 1 and 2 contain samples amined the activity level of SAF-1 and the expression of SAA1, derived from two amyloidotic kidney tissues, and lanes 3 and 4 contain which is the predominant SAA isoform in the AA amyloid. The ϳ those derived from normal kidney tissues. A major DNA product of 180 functional activity of SAF-1 in different tissues was determined by bp in size is identified. Lane M contains DNA size markers. using DNA-binding assay. Spleens, kidneys, and liver tissues from transgenic mice showed high levels of DNA-binding activity (Fig. 5A), which was characterized as SAF-1 by competition analysis To assess whether the amyloid deposits are composed of AA, and Ab ablation assay (Fig. 5B). Among the three tissues, spleens serial sections of spleen and kidneys from 14-mo-old SAF-1 trans- showed the highest level of SAF-1 activity. SAF-1 protein level genic and age-matched nontransgenic mice were further examined was also the highest in the spleen tissue, whereas its level in non- using a monoclonal anti-AA Ab. Results shown in Fig. 4, a, b, e, transgenic mice was much lower (Fig. 5C). It is worthy of mention and f, revealed that the amyloid deposits in the spleen and kidneys in this context that the spleen tissue of SAF-1 transgenic mice of SAF-1 transgenic mice are indeed composed of AA amyloid displayed the most amyloid deposition. Northern blot analysis re- protein. Immunohistochemistry with anti-SAF-1 Ab showed vealed high levels of SAA1 mRNA presence in the spleen, kidney, higher levels of SAF-1 protein in the amyloid deposit-containing and liver tissues of SAF-1 transgenic mice (Fig. 5D). These data The Journal of Immunology 2605

major inducer of SAA1 expression in the amyloidotic tissues of the transgenic mice. SAF-1 transgenic mice are highly susceptible to develop AA amyloidosis in response to inflammatory stimulus Our finding that only aged SAF-1 transgenic mice spontaneously develop AA amyloidosis (Figs. 3 and 4) raises the question of whether younger SAF-1 transgenic mice are resistant or whether they might develop AA amyloidosis if exposed to additional in- flammatory stimulus. To test these possibilities, we injected azo- casein, a known inflammatory stimulus (28, 29), into the SAF-1 transgenic mice and compared their response with the nontrans- genic mice. Amyloid-enhancing factor, in conjunction with an in- flammatory stimulus, is traditionally used for the rapid induction of experimental amyloidosis in mice. We chose not to use amyloid- enhancing factor, because it markedly enhances the process and therefore holds the possibility of blurring the rate of amyloid de- velopment. Eight-week-old SAF-1 transgenic and nontransgenic

mice received daily azocasein injections for different lengths of Downloaded from time, namely 7, 14, 21, and 42 days. At the end of each treatment period mice were sacrificed, and serum, spleen, kidney, liver, and other tissues were collected and stored at Ϫ70°C. Paraffin-embed- ded sections of kidney and spleen tissues were stained with Congo red dye and examined for amyloid deposition (Fig. 6). As shown

in Fig. 6A,amyloid deposits were not detected in nontransgenic http://www.jimmunol.org/ mice until day 21 of azocasein treatment, and the deposits were FIGURE 3. Histopathological analysis of the tissues from aged (14 mo much milder compared with that of SAF-1 transgenic mice. The old) SAF-1 transgenic (SAF-1 Tg) mice for amyloid deposits. Tissue sec- tions were stained with Congo red dye, and the amyloid deposits were rate of amyloid deposition was highly accelerated in SAF-1 trans- viewed under polarized light as bright white stains in the picture. a, c, and genic mice and, although both groups of mice developed amyloids e, typical sections from spleen, kidney, and liver tissues of nontransgenic in multiple organs (Fig. 6, A and B), at 42 days the severity score (Non-Tg) mice, and b, d, and f represent spleen, kidney, and liver tissues in the SAF-1 transgenic group was much higher (Table I). Immu- of SAF-1 transgenic mice, respectively. Eight mice in each group were nohistochemical analysis indicated that amyloid-laden tissues con- examined. The aged SAF-1 transgenic mice contained amyloid deposits in tained AA amyloids that were highly prevalent in SAF-1 trans- the spleen, kidney, and liver tissues, whereas no such deposition was seen genic mice (Fig. 6C). by guest on September 27, 2021 in the age-matched nontransgenic mice. Magnification, ϫ200 Acute phase response is prolonged in SAF-1 transgenic mice To examine whether increased and/or prolonged acute phase re- correlated well with the mass spectrometry analysis of purified sponse is accountable for the increased propensity of AA amyloid amyloid fibril protein, which identified it as SAA1 (described deposition in SAF-1 transgenic mice, we measured the plasma above). It was interesting to note that although the nuclear extracts level of total SAA after a single injection of azocasein (n ϭ 8in from nontransgenic mice contained noticeable DNA-binding ac- each group). Results of this experiment indicated that SAF-1 trans- tivity of SAF (Fig. 5A), it did not contribute toward SAA1 expres- genic mice express higher levels of SAA than the nontransgenic sion because there was no detectable SAA1 mRNA in nontrans- mice (Fig. 7). Moreover, whereas the SAA level in the nontrans- genic mice tissues (Fig. 5D). This finding could be due to the genic mice almost returned to control values on day 21, SAF-1 binding of SAF-2, a splice variant that inhibits the transactivation transgenic mice continued to have higher levels. When compared potential of SAF-1 (27), to the SAF-binding element of the SAA between old and young SAF-1 transgenic and nontransgenic ani- promoter. This possibility was evaluated in a DNA-binding assay. mals, 14-mo-old animals exhibited more robust SAA expression Ablation of the DNA-protein complex by SAF-2 Ab (Fig. 5B, lane than young 8-wk-old mice. This result suggested that aged SAF-1 18) indicated the presence of SAF-2 in the spleen nuclear extract transgenic mice experience a much more increased and prolonged of nontransgenic mice. Similar levels of SAF-2 were present in acute phase response, which accounts for the increased SAA levels other tissues (data not shown). Overexpression of SAF-1 in the in the circulations of these mice. SAF-1 transgenic mice significantly increased the nuclear pool of SAF-1 (Fig. 5C), thus overcoming the inhibitory effect of SAF-2 Discussion and activating SAA1 expression. Secondary or reactive AA amyloidosis is a fatal consequence of Because SAA1 expression is known to be induced by C/EBP many chronic inflammatory diseases. Increased synthesis due to and NF-␬B, EMSA was performed to measure the level these tran- transcriptional induction of the amyloidogenic SAA1 gene during scription factors in the liver, kidney, and spleen tissues of SAF-1 chronic inflammatory condition plays a central role in the devel- transgenic mice. Neither C/EBP nor NF-␬B activities were in- opment of AA amyloidosis. The transcriptional induction mecha- duced in any of these tissues of the SAF-1 transgenic mice (data nism of SAA has therefore been intensely investigated. These stud- not shown). This result is consistent with an earlier finding, which ies led to the identification of the SAF-1 transcription factor, which revealed that chronic inflammatory condition is not associated with is one of the primary regulators of SAA expression. To investigate NF-␬B or inducible C/EBP activities (18). To the contrary, SAF whether SAF-1 plays any role in the pathogenesis of AA amyloid- activity was found to remain high and contributed to persistent osis, we have developed a SAF-1 transgenic mouse model. Here, SAA expression (18). Together, the data indicate that SAF-1 is a we show that Ͼ75% of aged SAF-1 transgenic mice spontaneously 2606 SAF-1 AND AA AMYLOIDOSIS

FIGURE 4. Immunohistochemical analysis of the amyloid-laden spleen and kidney tissues of aged (14 mo old) SAF-1 transgenic mice (SAF-1 Tg) in comparison with age-matched nontransgenic (Non-Tg) mice. Thin sections from spleen (a–d) and kidney (e–h) of nontransgenic (a, c,e, and g) and SAF-1 transgenic (b, d, f, and h) mice were immunostained using anti-AA (a, b, e, and f) and anti-SAF-1 (c, d, g, and h) Ab, respectively. Magnification is ϫ200. Intense dark stains (indicated by arrows) depict positive staining for amyloid A and SAF-1. Downloaded from develop AA amyloidosis. Furthermore, compared with their non- nificant effect on the severity of amyloidosis developed in the af- transgenic counterparts, SAF-1 transgenic mice more rapidly and fected organ. This latter aspect is further discussed below. with increased severity develop experimentally induced AA amy- Numerous studies on SAA induction indicate that the regulatory loidosis in response to amyloid-inducing agents. We also observed process is highly complex as evidenced by variable inducibility that amyloid A deposits are highly prevalent in organs where ranging from 10- to 1000-fold, depending upon the signaling http://www.jimmunol.org/ SAA1 expression is very high. This latter finding correlated well agent. Variations in the induction level of SAA are due to the with that for human AA amyloidosis patients, where amyloid- activation of distinct pathways regulated by the involved transcrip- laden kidney tissue was found to overexpress amyloidogenic tion factors, including SAF-1 (18, 30). Multiple in vitro studies SAA1. Together, these findings indicate that: 1) SAF-1 is a po- indicate that the activation of SAF-1 in response to cytokines in the tential risk factor whose overexpression predisposes host animals cells results in increased SAA transcription in both hepatic and to develop AA amyloidosis; 2) persistent expression of SAF-1, nonhepatic cells (16, 17, 31). Constitutive synthesis of amyloido- endured by aged SAF-1 transgenic mice, also leads to spontaneous genic SAA1 at detectable levels in multiple organs of SAF-1 trans- development of AA amyloidosis; and 3) increased overexpression genic mice, but not in the nontransgenic littermates, correlated of amyloidogenic SAA1 in different organs has a discrete and sig- with the higher levels of the SAF-1 protein and the increased by guest on September 27, 2021

FIGURE 5. DNA-binding activity of SAF-1 and SAA levels in SAF-1 transgenic (SAF-1 Tg) and nontransgenic (nonTg) mice. A, Measurement of SAF-1 DNA-binding activity. Nuclear extracts (10 ␮g of protein) prepared from kidney (lanes 1 and 2), liver (lanes 3 and 4), and spleen (lanes 5 and 6) from 14-mo-old SAF-1 transgenic mice and age-matched nontransgenic mice were incubated with a 32P-labeled SAF-binding element of rabbit SAA2 promoter. The resultant DNA-protein complexes were resolved in a 6% native polyacrylamide gel. Asterisk denotes a nonspecific band that was only seen to be formed with spleen nuclear extract. B, Characterization of the DNA-protein complex. Nuclear extracts (10 ␮g of protein) prepared from kidney (lanes 7–11) and spleen (lanes 12–16) tissues of SAF-1 transgenic mice were incubated with the 32P-labeled SAF-binding probe. In addition, lanes 8, 9, 13, and 15 contained 25- and 50-fold molar excess of unlabeled SAF-binding oligonucleotide (SAF-1 oligo), lanes 10 and 15 contained 100-fold molar excess of nonspecific oligonucleotide (nonsp. oligo), and lanes 11 and 16 contained anti-SAF-1 Ab. In lanes 17 and 18 nuclear extracts of nontransgenic mice spleen tissues were used. Lane 18 also contained anti-SAF-2 Ab. C, Western blot analysis. Fifty micrograms of protein prepared from the indicated tissues of nontransgenic and SAF-1 transgenic animals as shown in A were fractionated by SDS-PAGE, transferred to nitrocellulose membrane, and probed with anti-SAF-1 Ab. D, Northern blot analysis. Total RNA (50 ␮g) from spleen, kidney, and liver tissues of nontransgenic and SAF-1 transgenic mice, as indicated, was separated in a 1.0% agarose gel containing 2 M formaldehyde, transferred to a nylon membrane, and probed with a 32P-labeled, SAA1- specific oligonucleotide. The same blot was stripped and rehybridized using 32P-labeled GAPDH cDNA. The Journal of Immunology 2607

FIGURE 6. Histopathological analysis of amyloid deposits in azocasein-injected SAF-1 transgenic (SAF-1 Tg) mice. Eight-week-old nontransgenic (Non-Tg) and SAF-1 transgenic mice were given daily s.c. injection of azocasein. Mice were sacrificed at different days as indicated, and the tissues were processed for detection of amyloid deposition. A, Paraffin-embedded sections of kidney tissues from SAF-1 transgenic and nontransgenic mice were collected at specified time points, stained with Congo red dye, and visualized under cross-polarized and UV light. Magnification, ϫ200. The presence of Downloaded from amyloid deposits is revealed by the appearance of bright white color streaks. B, Paraffin-embedded sections of spleen tissue from SAF-1 transgenic and nontransgenic mice collected at day 42 of azocasein treatment were stained with Congo red dye and visualized under cross-polarized and UV light as indicated. Magnification, ϫ200. Presence of amyloid deposits is revealed by the bright white streaks. C, Typical sections of spleen and kidney tissues of azocasein-injected nontransgenic and SAF-1 transgenic mice at day 42 were immunostained using anti-AA Ab. Intense dark stains indicate positive staining for amyloid A. http://www.jimmunol.org/

DNA-binding activity to the SAA promoter (Fig. 5, A and C), con- than that of age-matched nontransgenic mice (see Fig. 7, bars firming the role of SAF-1 as a regulator of SAA1 expression in marked 0d for day 0). It is possible that such an increase in the vivo (Fig. 5D). Consistent with this finding, the amyloid-laden circulating level of SAA is at least partly responsible for the amy- kidney tissue of human amyloid patients was also found to over- loidosis detected in the transgenic mice (Fig. 3). Another attractive express amyloidogenic SAA (Fig. 1B). Together, these data speculation is that overexpression of SAA in the affected tissue showed that, during the pathogenic condition of AA amyloidosis, may have contributed to the severity of amyloidosis. Our finding of SAA is expressed in the affected organs such as the kidney. It is the SAA overexpression in spleen and kidney tissues in addition to not known yet whether such local synthesis of SAA plays any the liver (Fig. 5) raises such a possibility. It is noteworthy to men- by guest on September 27, 2021 significant role in amyloid deposition. Contrarily, earlier studies tion in this regard that previous reports have also shown overex- have shown that amyloid deposits are derived from circulating pression of SAA in extrahepatic tissues in pathogenic conditions SAA in complex with high-density and that these SAA (34, 35). Compelling evidence of local synthesis of SAA in amy- molecules are synthesized in the liver (32, 33). In the present loid deposition is provided by the finding of brain-specific expres- study, the circulating level of SAA in the amyloidosis-affected, sion of mouse SAA1 (36, 37). Our data, presented here, do not rule aged SAF-1 transgenic mice in the absence of any inflammatory out the contribution of circulating SAA in amyloid formation. The stimulus was found to be ϳ100 ␮g/ml, which is only 2-fold higher answer to the question of whether both local production and in- creased circulation of SAA are necessary for amyloid deposition is yet to be established. The present finding, however, suggests that local production most likely increases the propensity, because SAF-1 transgenic mice spontaneously developed amyloidosis in organs overexpressing SAA. Together, our findings of high levels of SAF-1 along with AA amyloid protein deposits in amyloid- laden human kidney tissue (Fig. 1) are highly consistent with the proposed role of SAF-1 in the pathogenesis of AA amyloidosis. An important result of the current study is the surprising ap- pearance of significant AA deposits in Ͼ75% of the aged SAF-1 transgenic mice (Figs. 3 and 4). Amyloidosis in aged SAF-1 trans- genic mice was not due to normal age-related changes in the cy- tokine profile or metabolic processes, because age-matched non- transgenic littermates showed very few amyloid deposits. An explanation for this apparent age-related phenomenon in sponta- FIGURE 7. Plasma SAA concentration of young and aged nontrans- neous AA amyloid development can be obtained from the results genic and SAF-1 transgenic mice after acute phase induction. Four groups presented in Fig. 7. Acute phase response, as assessed by the per- ϭ of mice (n 8) consisting of young (8 wk old) and aged (14 mo old) sistent higher plasma SAA level following an inflammatory stim- SAF-1 transgenic or nontransgenic mice were given a single s.c. injection ulus, was prolonged in the SAF-1 transgenic mice compared with of 1.0 ml of 80 mg/ml azocasein solution. Blood samples were collected before azocasein injection and subsequently 1, 2, 4, 7, and 21 days after the nontransgenic mice. Based on this finding, we hypothesize that injection. Concentration of SAA was measured by ELISA in duplicate for periodic and extended acute phase responses throughout the life- each sample, and the average was calculated. time of the mice lead to an increased and extended synthesis of 2608 SAF-1 AND AA AMYLOIDOSIS amyloidogenic SAA that crosses the threshold level sanction dep- genic Core Facility, Cincinnati, OH) for help in the development of SAF-1 osition of amyloid fibrils. Interestingly, old nontransgenic mice transgenic founder mice. contained the same plasma level of SAA as young transgenic mice at the early stages of acute phase response up to day 4, but the level Disclosures rapidly declined thereafter in the old nontransgenic mice as com- The authors have no financial conflict of interest. pared with young transgenic mice, where the higher level persisted for extended period (Fig. 7). A persistent higher level of SAA References appears to be a key factor in the development of AA amyloidosis, 1. Sipe, J. D. 1994. Amyloidosis. Crit. Rev. Clin. Lab. Sci. 31: 325–324. and SAF-1 plays a role in this process (18). The rapid decline of 2. Kisilevsky, R. 2000. Amyloidogenesis — unquestioned answers and unanswered questions. J. Struct. Biol. 130: 99–108. SAA levels in aged nontransgenic mice may explain why these 3. Cooper, J. H. 1974. Selective amyloid staining as a function of amyloid compo- mice did not develop AA amyloidosis. sition and structure. Histochemical analysis of the alkaline Congo red, standard- In reactive AA amyloidosis, although the contribution of a pre- ized toluidine blue, and iodine methods. Lab Invest. 31: 232–238. 4. Rosenthal, C. J., E. C. Franklin, B. Frangione, and J. Greenspan. 1976. Isolation cursor SAA1 protein is clear, it remains unknown why some pa- and partial characterization of SAA — an amyloid-related protein from human tients develop amyloidosis while others with similar high levels of serum. J. Immunol. 116: 1415–1418. SAA1 in the system remain free of this pathology. It has been 5. Betts, J. C., M. R. Edbrooke, R. V. Thakker, and P. Woo. 1991. The human acute-phase serum amyloid A gene family: structure, evolution, and expression in speculated that additional inflammation-responsive molecular hepatoma cells. Scand. J. Immunol. 34: 471–482. components are necessary for the amyloid deposition. In correla- 6. Whitehead, A. S., M. C. deBeer, D. M. Steel, M. Rits, J. M. Lelias, W. S. Lane, tion, SAA transgenic mice overexpressing high levels of SAA1 did and F. C. deBeer. 1992. Identification of novel members of the serum amyloid A protein super family as constitutive of high density lipoprotein. not develop AA amyloidosis without an additional inflammatory J. Biol. Chem. 267: 3862–3867. Downloaded from stimulus (36, 37). Conversely, no such stimulus was needed for 7. Kluve-Beckerman, B., M. L. Drumm, and M. D. Benson. 1991. Non-expression amyloid development in the aged SAF-1 transgenic mice. We pos- of the human serum amyloid A three (SAA3) gene. DNA Cell Biol. 10: 651–661. 8. Sipe, J. D. 1999. Revised nomenclature for serum amyloid A (SAA). Amyloid 6: tulate that in addition to the increased level of amyloidogenic SAA 67–70. via transcriptional induction of this gene, inflammation-responsive 9. Lowell, C. A., D. A. Potter, R. S. Stearman, and J. F. Morrow. 1986. Structure of the murine serum amyloid A gene family. Gene conversion. J. Biol. Chem. stimulus provides an additional component(s) for the onset of amy- 261: 8442–8452. loid deposits. The increased propensity for amyloid development 10. Uhlar, C. M., and A. S. Whitehead. 1999. Serum amyloid A, the major vertebrate http://www.jimmunol.org/ in the young SAF-1 transgenic mice in response to the amyloid- acute-phase reactant. Eur. J. Biochem. 265: 501–523. 11. Ray, A., and B. K. Ray. 1994. Serum amyloid A gene expression under acute- inducing stimulus provided by azocasein injection supports this phase conditions involves participation of inducible C/EBP-␤ and C/EBP-␦ and notion. Consistent with our finding, AA amyloidosis development their activation by phosphorylation. Mol. Cell. Biol. 14: 4324–4332. in IL-6 transgenic mice was found to be age-dependent, with AA 12. Huang, J. H., and W. S. Liao. 1994. Induction of the mouse serum amyloid A3 gene by cytokines requires both C/EBP family proteins and a novel constitutive deposits being first evident at 3 mo of age and increasing over the nuclear factor. Mol. Cell. Biol. 14: 4475–4484. next 6 mo (38). Interestingly, IL-6 has been shown to activate 13. Edbrooke, M. R., D. W. Burt, J. K. Cheshire, and P. Woo. 1989. Identification of SAF-1 (16–18), which raises the possibility that one of the effects cis-acting sequences responsible for phorbol ester induction of human serum amyloid A gene expression via a nuclear factor ␬B-like transcription factor. Mol. of this in the IL-6 transgenic mice could be the activation Cell. Biol. 9: 1908–1916. of endogenous SAF-1. Thus, SAF-1 appears to be a crucial mol- 14. Li, X., and W. S. Liao. 1992. Cooperative effects of C/EBP-like and NF ␬ B-like by guest on September 27, 2021 binding sites on rat gene expression in liver cells. Nucleic ecule responsible for the development of AA amyloidosis. Acids Res. 20: 4765–4772. Tissue-specific expression of SAA has been implicated as play- 15. Ray, A., M. Hannink, and B. K. Ray. 1995. Concerted participation of NF-␬ B ing a determining role in the pathogenesis of a number diseases, and C/EBP heteromer in lipopolysaccharide induction of serum amyloid A gene expression in liver. J. Biol. Chem. 270: 7365–7374. including Alzheimer’s disease, arthritis, and , where 16. Ray, A., and B. K. Ray. 1996. A novel cis-acting element is essential for cyto- SAA is overexpressed in the specific cell types of the brain (39), kine-mediated transcriptional induction of the serum amyloid A gene in nonhe- synovium (40–42), and artery (43, 44). The regulation of SAA patic cells. Mol. Cell. Biol. 16: 1584–1594. 17. Ray, A., and B. K. Ray. 1998. Isolation and functional characterization of cDNA genes in these diverse cell types most likely involves different of serum amyloid A-activating factor that binds to the serum amyloid A pro- combinations of transcription factors that are active in the cells of moter. Mol. Cell. Biol. 18: 7327–7335. the affected tissues. A paradigm is the elucidation of the tissue- 18. Ray, A., and B. K. Ray. 1999. Persistent expression of serum amyloid A during experimentally induced chronic inflammatory condition in rabbit involves differ- specific expression of SAA1 and SAA2 in hepatic and epithelial ential activation of SAF, NF-␬ B, and C/EBP transcription factors. J. Immunol. cells (45) and SAA1 expression in aortic smooth muscle and he- 163: 2143–2150. 19. Ray, A., Kumar, D., Shakya, A., Brown, C. R., Cook, J. L., and Ray, B. K. 2004. patic cells (46). High levels of C/EBP activity in the liver have Serum amyloid activating factor-1 transgenic mice are prone to develop a severe been found to be beneficial for SAA expression in the hepatic cells form of inflammation-induced arthritis. J. Immunol. 173: 4684–4691. (11, 12, 14, 46), and such action of C/EBP is further potentiated by 20. Puchtler, H., F. Sweat, and M. Levine. 1962. On the binding of Congo red by ␬ amyloid. J. Histochem. Cytochem. 10: 355–354. NF- B activity (13–15, 47). Interestingly, SAF-1 is found to be 21. Botto, M., P. N. Hawkins, C. M. Bickerstaff, J. Herbert, A. E. Bygrave, very active in many tissue-specific cells, including synoviocytes A. McBride, W. L. Hutchinson, G. A. Tennent, M. J. Walport, and M. B. Pepys. and chondrocytes of the arthritic joint (22, 48) and macrophage 1997. Amyloid deposition is delayed in mice with targeted deletion of the serum amyloid P component gene. Nat. Med. 3: 855–859. cells of the atherosclerotic plaque (49). Recent evidence also im- 22. Ray, A., K. Kuroki, J. L. Cook, B. S. Bal, K. Kenter, G. Aust, and B. K. Ray. plicates SAF-1 in the human SAA1 and SAA2 expression in the 2003. Induction of matrix metalloproteinase 1 gene expression is regulated by trophoblast cells during early fetal development (50). inflammation-responsive transcription factor SAF-1 in osteoarthritis. Arthritis Rheum. 48: 134–145. The data presented in this paper establish that SAF-1 is involved 23. Chomczynski, P., and N. Sacchi. 1987. Single-step method of RNA isolation by in the development of AA amyloidosis by inducing amyloidogenic acid guanidinium thiocyanate-phenol-chloroform extraction. Anal. Biochem. 162: SAA1 expression in multiple tissues. Further studies will deter- 156–159. 24. Sambrook, J., and D. W. Russell. 2001. Molecular Cloning: A Laboratory Man- mine the potential impact of tissue-specific expression mecha- ual, 3rd Ed. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY. nisms in controlling the severity of the pathogenesis. 25. Meek, R. L., and E. P. Benditt. 1986. Amyloid A gene family expression in different mouse tissues. J. Exp. Med. 164: 2006–2017. 26. Bradford, M. M. 1976. A rapid and sensitive method for the quantitation of microgram quantities of protein utilizing the principle of protein-dye binding. Acknowledgments Anal. Biochem. 72: 248–254. 27. Ray, B. K., R. Murphy, P. Ray, and A. Ray. 2002. SAF-2, a splice variant of We thank Howard Wilson for assistance in preparing photomicrographs SAF-1 acts as a negative regulator of transcription. J. Biol. Chem. 277: and figures. We also thank Brad Wagner (University of Cincinnati, Trans- 46822–46830. The Journal of Immunology 2609

28. Wohlgethan, J. R., and E. S. Cathcart. 1981. Amyloidosis in (CBA/J ϫ A/J)F2 41. Strissel, K. J., M. T. Girard, J. A. West-Mays, W. B. Rinehart, J. R. Cook, mice: correlation of amyloid resistance and low mitogenic response to con- C. E. Brinckerhoff, and M. E. Fini. 1997. Role of serum amyloid A as an inter- canavalin A. J. Immunol. 127: 1003–1007. mediate in the IL-1 and PMA-stimulated signaling pathways regulating expres- 29. Janigan, D. T., and R. L. Druet. 1966. Experimental amyloidosis. Role of anti- sion of rabbit fibroblast collagenase. Exp. Cell Res. 237: 275–287. genicity and rapid induction. Am. J. Pathol. 48: 1013–1025. 42. Ray, A., H. Schatten, and B. K. Ray. 1999. Activation of Sp1 and its functional 30. Ray, A., and B. K. Ray. 1997. Serum amyloid A gene expression level in liver in co-operation with SAF in synoviocyte cells trigger synergistic action of IL-1 and response to different inflammatory agents is dependent upon the nature of acti- IL-6 in serum amyloid A gene expression. J. Biol. Chem. 274: 4300–4308. vated transcription factors. DNA Cell Biol. 16: 1–7. 43. Kumon, Y., J. D. Sipe, C. E. Brinckerhoff, and B. M. Schreiber. 1997. Regulation 31. Ray, B. K., and A. Ray. 1997. Involvement of an SAF-like transcription factor in of extrahepatic serum amyloid A (ApoSAA) gene expression by the activation of serum amyloid A gene in monocyte/macrophage cells by lipo- -1 ␣ alone: synthesis and secretion of ApoSAA by cultured aortic polysaccharide. Biochemistry 36: 4662–4668. smooth muscle cells. Scand. J. Immunol. 46: 284–291. 32. Husebekk, A., B. Skogen, G. Husby, and G. Marhaug. 1985. Transformation of 44. Meek, R. L., S. Urieli-Shoval, and E. P. Benditt. 1994. Expression of apolipopro- amyloid precursor SAA to protein-AA and incorporation in amyloid fibrils in tein serum amyloid A mRNA in human atherosclerotic lesions and cultured vas- vivo. Scand. J. Immunol. 21: 283–287. cular cells: implications for serum amyloid A function. Proc. Natl. Acad. Sci. 33. Tape, C., R. Tan, M. Nesheim, and R. Kisilevsky. 1988. Direct evidence for USA 91: 3186–3190. circulating apoSAA as the precursor of tissue AA amyloid deposits. Scand. J. Im- 45. Thorn, C. F., Z.-Y. Lu, and A. S. Whitehead. 2003. Tissue-specific regulation of munol. 28: 317–324. the human acute-phase serum amyloid A genes. SAA1 and SAA2, by glucocor- ticoids in hepatic and epithelial cells. Eur. J. Immunol. 33: 2630–2639. 34. Urieli-Shoval, S., L. Reinhold, and M. Yaacov. 2000. Expression and function of 46. Kumon, Y., T. Suehiro, D. J. Faulkes, T. Hosakawa, Y. Ikeda, P. Woo, J. D. Sipe, serum amyloid A, a major acute-phase protein, in normal and disease states. Curr. and K. Hashimoto. 2002. Transcriptional regulation of serum amyloid A1 gene Opin. Hematol. 7: 64–69. expression in human aortic smooth muscle cells involves CCAAT/enhancer bind- 35. O’Hara, R., E. P. Murphy, A. S. Whitehead, O. FitzGerald, and B. Bresnihan. ing proteins (C/EBP) and is distinct from HepG2 cells. Scand. J. Immunol. 56: 2000. Acute-phase serum amyloid A production by rheumatoid arthritis synovial 504–511. tissue. Arthritis Res. 2: 142–144. 47. Betts, J. C., J. K. Cheshire, S. Akira, T. Kishimoto, and P. Woo. 1993. The role 36. Guo, J. T., J. Yu, D. Grass, F. C. de Beer, and M. S. Kindy. 2002. Inflammation- of NF-␬B and NF-IL6 transactivating factors in the synergistic activation of dependent cerebral deposition of serum amyloid A protein in a mouse model of human serum amyloid A gene expression by interleukin-1 and interleukin-6. Downloaded from amyloidosis. J. Neurosci. 22: 5900–5909. J. Biol. Chem. 268: 25624–25631. 37. Kindy, M. S., and F. C. de Beer. 1999. A mouse model for serum amyloid A 48. Ray, A., B. S. Bal, and B. K. Ray. 2005. Transcriptional induction of MMP-9 in amyloidosis. Methods Enzymol. 309: 701–704. the chondrocyte and synoviocyte cells is regulated via a novel mechanism: evi- 38. Solomon, A., D. T. Weiss, M. Schell, R. Hrncic, C. L. Murphy, J. Wall, dence for functional cooperation between SAF-1 and AP-1. J. Immunol. 175: M. D. McGavin, H. J. Pan, G. W. Kabalka, and M. J. Paulus. 1999. Transgenic 4039–4048. mouse model of AA amyloidosis. Am. J. Pathol. 154: 1267–124. 49. Ray, B. K., A. Shakya, J. R. Turk, A. S. Apte, and A. Ray. 2004. Induction of the 39. Kindy, M. S., J. Yu, J. T. Guo, and H. Zhu. 1999. Apolipoprotein serum amyloid MMP-14 gene in macrophages of the atherosclerotic plaque: role of SAF-1 in the

A in Alzheimer’s disease. J. Alzheimers Dis. 1: 155–167. induction process. Circ. Res. 95: 1082–1090. http://www.jimmunol.org/ 40. Miyamae, T., D. E. Malehorn, B. Lemster, M. Mori, T. Imagawa, S. Yokota, 50. Kovacevic, A., A. Hammer, M. Sundl, B. Pfister, A. Hrzenjak, A. Ray, B. K. Ray, W. L. Bigbee, M. Welsh, K. Klarskov, N. Nishomoto, et al. 2005. Serum protein W. Sattler, and E. Malle. 2006. Expression of serum amyloid A transcripts in profile in systemic-onset juvenile idiopathic arthritis differentiates response ver- human trophoblast and fetal-derived trophoblast-like choriocarcinoma cells. sus nonresponse to therapy. Arthritis Res. Ther. 7: R746–R755. FEBS Lett. 580: 161–167. by guest on September 27, 2021