Renal Tubulointerstitial Changes after Internal Irradiationwith �-Particle–Emitting Actinium Daughters

Jaspreet Singh Jaggi,* Surya V. Seshan,‡ Michael R. McDevitt,* Krista LaPerle,§

George Sgouros,� and David A. Scheinberg*†

*Molecular Pharmacology and Chemistry Program and †Department of Medicine, Memorial Sloan-Kettering CancerCenter, and ‡Department of Pathology and §Research Animal Resource Center, Weill Medical College of CornellUniversity, New York, New York; and �Department of Radiology, Johns Hopkins University School of Medicine,Baltimore, Maryland

The effect of external � irradiation on the kidneys is well described. However, the mechanisms of radiation nephropathy asa consequence of targeted radionuclide therapies are poorly understood. The functional and morphologic changes werestudied chronologically (from 10 to 40 wk) in mouse kidneys after injection with an actinium-225 (225Ac) nanogenerator, amolecular-sized, antibody-targeted, in vivo generator of �-particle–emitting elements. Renal irradiation from free, radioactivedaughters of 225Ac led to time-dependent reduction in renal function manifesting as increase in blood urea nitrogen. Thehistopathologic changes corresponded with the decline in renal function. Glomerular, tubular, and endothelial cell nuclearpleomorphism and focal tubular cell injury, lysis, and karyorrhexis were observed as early as 10 wk. Progressive thinning ofthe cortex as a result of widespread tubulolysis, collapsed tubules, glomerular crowding, decrease in glomerular cellularity,interstitial inflammation, and an elevated juxtaglomerular cell count were noted at 20 to 30 wk after treatment. By 35 to 40 wk,regeneration of simplified tubules with tubular atrophy and loss with focal, mild interstitial fibrosis had occurred. A lowerjuxtaglomerular cell count with focal cytoplasmic vacuolization, suggesting increased degranulation, was also observed in thisperiod. A focal increase in tubular and interstitial cell TGF-�1 expression starting at 20 wk, peaking at 25 wk, and laterdeclining in intensity with mild increase in the extracellular matrix deposition was noticed. These findings suggest thatinternally delivered �-particle irradiation–induced loss of tubular epithelial cells triggers a chain of adaptive changes thatresult in progressive renal parenchymal damage accompanied by a loss of renal function. These findings are dissimilar tothose seen after gamma or beta irradiation of kidneys.

J Am Soc Nephrol 16: 2677–2689, 2005. doi: 10.1681/ASN.2004110945

I n the past decade, there has been an increase in the use ofradioimmunotherapy (RIT) for cancer with tumor-target-ing peptides and mAb and their fragments (1). Two radio-

labeled, �-particle–emitting antibodies are Food and Drug Ad-ministration approved for the treatment of non-Hodgkin’slymphoma (2,3). �-Particle–emitting radionuclides are particu-larly advantageous for RIT because � particles have high linearenergy transfer and short path lengths (50 to 80 �m) (4,5).Therefore, a large amount of energy is deposited over a shortdistance, which renders � particles extremely cytotoxic with ahigh relative biologic effectiveness (6,7). Little collateral dam-age to surrounding normal, antigen-negative cells occurs (8).

We and others are investigating actinium-225 (225Ac), an�-emitting atomic generator, as a suitable candidate for use inRIT (9). 225Ac has a sufficiently long half-life (10 d) for feasibleclinical use, and it decays to stable bismuth-209 via six atoms

(Figure 1), yielding a net of four � particles (9). The 225Ac-antibody construct acts as a tumor-selective, molecular-sized invivo atomic generator (targetable nanogenerator) of �-particle–emitting elements (9). The nanogenerators are stable in vivo andhave been shown to be safe and potent antitumor agents inmouse models of solid prostatic carcinoma, disseminated lym-phoma, and intraperitoneal ovarian cancer and in a rat modelof meningeal neuroblastoma (9–11). The safety of 225Ac-HuM195 and 225Ac-3F8 at low doses has been demonstrated inprimates (11,12). Therefore, these agents possess the potentialto treat a variety of tumor types.

Radiation nephropathy as a result of external beam radiationis well documented (13). However, there are few reported casesof radionuclide therapy–associated nephropathy (1,14,15). Re-cent reports of radiation-induced nephropathy after the use ofradiolabeled somatostatin analogues have now raised concernsabout the safety of internal emitters as therapeutic agents(16,17). An external beam deposits a large radiation dose rela-tively uniformly to the kidney, over a short time. In contrast,radionuclide therapies result in a continuous low-dose-rateradiation, which is deposited heterogeneously in the kidneys(1,18).

There have been limited reports of radiation nephropathy

Received November 18, 2004. Accepted May 13, 2005.

Published online ahead of print. Publication date available at www.jasn.org.

Address correspondence to: Dr. David Scheinberg, Molecular Pharmacology andChemistry Program, Memorial Sloan Kettering Cancer Center, 1275 York Avenue,New York, NY 10021. Phone: 212-639-8635; Fax: 212-717-3068; E-mail:d-scheinberg@ski.mskcc.org

Copyright © 2005 by the American Society of Nephrology ISSN: 1046-6673/1609-2677

after the use of radiolabeled antibodies (15). The antibody-radionuclide construct is large and therefore is not filtered viathe glomeruli. However, the peculiarities of 225Ac nanogenera-tors that render them potent cytotoxic agents may also result inpotential nephropathy. 225Ac decays via its �-emitting daugh-ters francium-221(t1/2 � 5 min), astatine-217 (t1/2 � 32.3 ms),and bismuth-213 (t1/2 � 45.6 min) to stable, nonradioactiveBi-209 (9). The daughters that are generated and retained insidethe cancer cell after internalization of the 225Ac-labeled anti-body add to its cytotoxic effect. However, the free daughtersthat are produced in the vasculature from the circulating un-bound antibody or the antibody bound to the surface of a targetcell could be transported to distant sites via the bloodstreamand accumulate in other organs. Bismuth is known to accumu-late in the renal cortex (19–21). This has been attributed to thepresence of heavy metal binding, metallothionein-like proteinspresent in the cytoplasm of renal proximal tubular cells. Wehave also seen a rapid accumulation of francium in mice kid-neys after intravenous injection (unpublished observation).Moreover, monkeys that received an injection of escalatingdoses of the untargeted 225Ac nanogenerator developed a de-layed radiation nephropathy (12). Therefore, a possible hurdleto the development of these agents as safe and effective cancertherapeutics is likely to be their nephrotoxicity. Here, we de-scribe chronologically the pathologic and corresponding func-tional events that occur in the kidneys after 225Ac nanogenera-tor injection, in an attempt to gain insights into thepathogenesis of radiation nephropathy after radionuclide ther-apy.

Materials and MethodsAnimals

Female BALB/c mice, 6 wk old and weighing 18 to 20 g, wereobtained from Taconic (Germantown, NY). All animal studies wereconducted according to the National Institutes of Health’s Guide for theCare and Use of Laboratory Animals and were approved by the Institu-tional Animal Care and Use Committee at Memorial Sloan KetteringCancer Center.

Preparation and Quality Control of 225Ac-LabeledAntibodies

HuM195 (humanized IgG1 that recognizes the extracellular domainon CD33) was a gift from Protein Design Labs (Fremont, CA). 225Ac wasconjugated to HuM195 antibody and quality controlled, as describedpreviously (22).

Radioimmunoconjugate Administration to MiceThe mice were anesthetized by intraperitoneal injection of ketamine

(100 mg/kg) and xylazine (10 mg/kg) and received an injection of 0.35

�Ci (720 kBq/kg) of [225Ac]HuM195 via the retro-orbital venous plexus.The specific activity of the radioimmunoconjugate was 0.08 Ci/g.

Clinical and Anatomic PathologyUrine protein was estimated weekly using Uristix reagent strips

(Uristix; Bayer, Elkhart, IN). Mice were killed by carbon dioxide as-phyxiation at 10, 20, 25, 30, 35, and 40 wk after injection with 225Ac-HuM195 (three to five per time point), and their blood urea nitrogen(BUN) and serum creatinine were estimated (IDEXX Laboratories, To-towa, NJ). Tissues were fixed in 10% neutral-buffered formalin, pro-cessed by routine methods, and embedded in paraffin. Sections (2 to 3�m) were stained with hematoxylin and eosin, periodic acid-Schiff, andMasson’s trichrome and evaluated with an Olympus BX45 light micro-scope. Age-matched normal BALB/c mice served as controls for thecomparison of gross and microscopic features at each time point.

ImmunohistochemistryParaffin-embedded kidney sections (2 to 3 �m) were immunostained

with polyclonal rabbit anti-human TGF �1 antibody (Promega, Madi-son, WI; 1:200 dilution), polyclonal rabbit anti-plasminogen activatorinhibitor-1 (PAI-1) antibody (Santa Cruz Biotechnology, Santa Cruz,CA; 1:200 dilution), rabbit anticleaved caspase-3 antibody (Cell Signal-ing Technology, Beverly, MA; 1:50 dilution), and rabbit anti–Wilms’tumor 1 (WT1) antibody (Santa Cruz; 1:50 dilution), using methodsdescribed previously (23–26).

Electron MicroscopyPieces of renal cortical tissue were fixed in 4% paraformaldehyde,

postfixed in 1% osmium tetroxide, and later embedded in Epon. Ultra-thin sections (200 to 400 Å) were cut on nickel grids, stained with uranylacetate and lead citrate, and examined using an electron microscope(Hitachi H-7500, Pleasanton, CA).

Evaluation of Renal Pathologic DamageThe glomeruli were assessed for decrease in size and cellularity. The

percentage of glomeruli that showed shrinkage (50% of control size)was evaluated by counting the total number of glomeruli in midcoronalsections of the kidneys (three to four per mouse). The number ofjuxtaglomerular (JG) cells per glomerular hilus and the extent of glo-merular involvement (percentage of glomeruli involved) were calcu-lated. The arterial and arteriolar medial (smooth muscle) cell vacuol-ization was estimated semiquantitatively on the basis of the percentageof cells involved per vessel cross-section (�25% � �/�; 25 to 50% �

1�; 50 to 75% � 2�; �75% � 3�). The extent of tubular cell vacuol-ization, tubulolysis, loss of brush border, and tubular atrophy wasexpressed as the mean percentage of total number of tubules in thekidney sections. The extent of basement membrane thickening in thelysed tubules was examined in trichrome-stained sections. The inten-sity (1� or 2� or 3�) and the extent (percentage of preserved andtubulolytic tubules) of immunoperoxidase staining for TGF-�1 andPAI-1 were evaluated. Fifty glomeruli were counted for epithelial cellsthat showed positive nuclear staining for WT1, and the average numberof WT1-positive cells per glomerulus was calculated. The average num-ber of cells that showed positive staining for cleaved caspase-3 perkidney cross-section was also noted.

Estimation of Renal Absorbed DoseThe mean absorbed dose to the kidneys was estimated using a

method described previously (12,27). The adult renal cortex-to-totalkidney mass ratio (0.69) was used to obtain the renal cortical mass (28).The absorbed dose to the renal cortex was calculated by assuming that

Figure 1. Simplified actinium-225 (Ac-225) generator to Bi-213decay scheme, yielding four net �. The half-lives are shown initalics.

2678 Journal of the American Society of Nephrology J Am Soc Nephrol 16: 2677–2689, 2005

the entire measured radioactivity in the kidneys was confined to therenal cortex.

ResultsGeneral Examination, Serum Chemistry, and GrossMorphology of the Kidneys

Examination of 225Ac-treated mice disclosed loss of bodyweight, starting 20 wk after 225Ac injection and progressivelyworsening with time. At 40 wk after injection, the weight of225Ac-treated mice was approximately 50 to 60% lower thanthat of age-matched mice of the same strain. Increased urinaryprotein was seen in some mice as early as 13 wk, and by 20 wk,all mice showed significant proteinuria. The BUN increasedprogressively, starting at 20 wk after injection (Figure 2). Serumcreatinine, electrolytes, and hepatic function tests were withinnormal limits for mice. The kidneys became progressivelysmaller from 10 wk onward, and by 40 wk after injection, theywere reduced to approximately 50% of the normal size withmarked pallor and focal, fine surface granularity. No significantchanges were observed in other organs at any time point.

Light Microscopic FindingsAt 10 wk after treatment with 225Ac nanogenerator, the renal

parenchymal architecture (cortex and medulla) was generallypreserved, having normocellular glomeruli, thin capillarywalls, and patent lumina. The nuclei of all types of glomerularcells (epithelial, mesangial, and endothelial) and cortical andmedullary tubules appeared prominent with marked hyper-chromasia (Figure 3, A and B). On close examination, scatteredproximal tubules disclosed isolated cell swelling and lysis ac-companied by karyorrhexis (Figure 3A). Focal, fine cytoplasmicvacuolization and loss of luminal brush border were observedin individual cells in �10% of the tubules.

At 20 wk, the renal cortical tissue was mildly shrunken withearly glomerular crowding. Whereas the glomerular size re-mained within normal limits, a mild decrease in cellularity,

particularly as a result of loss of glomerular epithelial cells, wasevident. The cortical tubules (50 to 60%) showed considerableloss of lining cells, leaving bare, intact tubular basement mem-branes and leading to collapse (Figure 3D). In addition, areas oftubular cell flattening, extensive loss of brush border, cytoplas-mic swelling, and vacuolization were noted. The remainingcortical and medullary tubules displayed varying degrees ofshrinkage and simplification. There was no significant intersti-tial inflammation or fibrosis, except for focal, small aggregatesof plasma cells and lymphocytes in the perivascular areas.Scattered tubules contained proteinaceous casts in the lumina.The nuclei of tubular cells were prominent with coarse chro-matin pattern, focal distortion, pyknosis, and karyorrhexis.

After 25 wk after injection, the cortical thickness was reducedfurther by one third compared with that at 20 wk. Increasedglomerular crowding and occasional subcapsular glomerulo-sclerosis (segmental or global) with only a mild compromise incapillary lumina were observed. No glomerular endothelialswelling, mesangiolysis, or inflammatory cell infiltration wasseen. As compared with 20 wk (Figure 3D), there was a morewidespread cortical tubulolysis (�75%), leaving empty tubularprofiles and thickening and wrinkling of basement membranes(Figure 3G). Increased cellular protein resorption droplets wereobserved in the remaining islands of mildly affected tubules,with focal microcalcification in some simplified or atrophictubules. There was focal perivascular interstitial inflammationcomposed primarily of lymphocytes and smaller proportion ofplasma cells. Simplification and mild atrophy were seen in�50% of the medullary collecting ducts with diminished peri-tubular vascular spaces. Other blood vessels showed increasedvascular smooth muscle vacuolization and dilation of corticalperitubular capillaries with margination of inflammatory cells.

The renal parenchymal findings at 30 wk were similar tothose observed at 25 wk, except for progression in the extent ofthe tubular cell damage, lysis, and atrophy, with focal evidenceof regenerative change in the lining epithelium. The interstitialinflammatory reaction was conspicuously increased with focaltubulitis composed of lymphocytes and plasma cells (Figure4A). In addition, mild to moderate dilation of the Bowmanspace was seen in 10 to 30% of the glomeruli.

At 35 wk, further cortical (subcapsular) atrophy, dilation ofthe glomerular Bowman space with mild collapse of the capil-lary tufts, and reduction in the number of glomeruli wereevident. Twenty-five to 30% of tubules showed reactive nuclearmorphology and cellular regenerative change with early recov-ery of low brush border (Figure 4D). The rest of the tubulesshowed widespread simplification and focal atrophy. Completeloss of tubular lining secondary to cell lysis and residual thick-ened duplicated basement membranes persisted, involving50% of the tubules. The interstitial inflammation was markedlyreduced, affecting just 5% of the cortex. The proteinaceous castswere noted in 20 to 25% of both the cortical and the medullarytubules, with more frequent tubular microcalcification re-stricted to the medullary tissue. Whereas focal mild medialhypertrophy and narrowing of the vascular lumina were noted,smooth muscle cell vacuolization had mostly disappeared.

The renal pathologic findings at 40 wk demonstrated com-

Figure 2. Blood urea nitrogen (BUN) levels in mice at differenttime points after injection with 0.35 �Ci of 225Ac-HuM195. BUNlevels increased progressively, starting 20 wk after injection.

J Am Soc Nephrol 16: 2677–2689, 2005 �-Particle–Mediated Renal Toxicity 2679

plete subcapsular cortical atrophy. Occasional narrow or ste-notic glomerulotubular neck was recognized. Regenerativechanges with attenuated tubules and restoration of the brushborder with minimal interstitial fibrosis were noted in approx-imately 30% of cortical tubules (Figure 4G). Moderate atrophyand tubular loss were also seen in the medulla.

Age-matched control BALB/c mice showed no significantchanges at 10, 20, 25, 30, and 35 wk time points (Figure 4, J andK). Mild to moderate glomerular mesangial matrix thickeningand mild arteriosclerosis were seen at 40 wk, which is consis-tent with normal aging process (Figure 4J).

Trichrome FindingsTrichrome-stained sections disclosed progressive basement

membrane thickening, wrinkling, and collapse with aggrega-

tion of denuded tubular profiles with only minimal or noincrease in interstitial fibrosis from 20 to 40 wk after treatment(Figures 3, E and H, and 4, B, E, and H). Progressive increase inJG cell hyperplasia and granularity involving 10 to 20% ofglomerular hilar areas was seen. Arteriolar medial cell meta-plasia with increased granularity was noted at 20, 25, and 30wk, which declined to fewer cells and considerable degranula-tion at 35 to 40 wk.

Immunohistochemical FindingsThe immunohistochemical staining for TGF-�1 displayed an

exclusively tubulointerstitial pattern. Whereas no significantstaining was observed in control (untreated) mice (Figure 4L)and those at 10 wk after treatment (Figure 3C), scattered clus-ters of preserved tubular cells displayed positive (1 to 2�)

Figure 3. Representative photomicrographs of morphologic changes in mouse kidneys after injection with 225Ac-labeled antibody.(A through C) At 10 wk, isolated tubular cell swelling, lysis, and karyorrhexis (arrows); normal juxtaglomerular (JG) cells(arrowheads); and negative TGF-�1 staining. (D through F) At 20 wk, varying tubulolysis (arrows), increased JG cells (arrow-heads), and focal TGF-�1–positive tubular and interstitial cells (*). (G through I) At 25 wk, widespread tubulolysis with emptytubular basement membrane (TBM) profiles (arrows), increased JG cells (arrowheads), and increased TGF-�1—positive cells (*).Magnification, �400 in A and B; �160 in C through F, H, and I.

2680 Journal of the American Society of Nephrology J Am Soc Nephrol 16: 2677–2689, 2005

staining, affecting approximately 25 and 50% of cells at 20 and25 wk, respectively (Figure 3, F and I). A lower intensity (trace-1�) was seen in �5% of tubular cells at 30, 35, and 40 wk after

treatment (Figure 4, C, F, and I). Similarly, there was intense,focal staining for TGF-�1 in the interstitial cells at 20 and 25 wkthat declined progressively in extent and intensity after 30 wk

Figure 4. Representative photomicrographs of morphologic changes in mouse kidneys after injection with 225Ac-labeled antibody.(A through C) At 30 wk, tubulolysis, interstitial inflammation (arrow), tubular casts, decreased JG cells (arrowheads), anddecreased TGF-�1–positive cells. (D through F) At 35 wk, tubular regeneration and simplification (arrows), thickened TBM,degranulated JG cells (arrowheads), and focal TGF-�1–stained cells (*). (G through I) At 40 wk, simplified tubular regeneration,focal brush border (arrows), thickened and collapsed TBM (arrowheads), and scattered TGF-�1–positive cells (*). (J through L) Acontrol mouse kidney with preserved renal parenchyma, normal JG cells (arrowheads), and negative TGF-�1 staining. Magnifi-cation, �160 in A through F and H through L; �80 in G.

J Am Soc Nephrol 16: 2677–2689, 2005 �-Particle–Mediated Renal Toxicity 2681

after injection. However, no significant staining for PAI-1was observed in the 225Ac nanogenerator–treated or controlmice at any time point (data not shown). No significantdifference in the number of WT1-positive cells per glomer-ulus was seen (Figures 5, A B, and C; and 6, A through D)between the untreated control and the radioimmunoconju-gate-treated mice at any time point (approximately nine to 12cells per glomerulus at each time point). The number of cellsper kidney cross-section that showed positive perinuclear orcytoplasmic staining for cleaved caspase-3 was mildly in-creased at 10 wk (approximately three to four cells per

kidney cross-section); peaked at 20, 25, and 30 wk (approx-imately 18 to 20 per-cross section); and was moderatelylower at 35 and 40 wk (approximately 10 per cross-section)after 225Ac injection (Figures 5, D, E, and F; and 6, E, F, andG). No staining for cleaved caspase-3 was seen in untreated(negative control) mice (Figure 6H).

Electron Microscopic FindingsElectron microscopic study of kidneys from control mice

showed normal glomerular and tubular architecture (Figure7A). Ultrastructural findings at 10 wk was limited to focal,

Figure 5. Representative photomicrographs of immunohistochemical staining for Wilms’ tumor 1 (WT1) and cleaved caspase-3 inmouse kidney sections at 10, 20, and 25 wk after injection with 225Ac-labeled antibody. (A, B, and C) Ten, 20, and 25 wk,respectively: No significant decrease in the number of WT1-positive cells per glomerulus from 10 to 25 wk. Increase in the numberof caspase-3–positive cells (arrows) at 20 and 25 wk (E and F, respectively) as compared with 10 wk (D) after injection.

2682 Journal of the American Society of Nephrology J Am Soc Nephrol 16: 2677–2689, 2005

isolated tubular cell injury characterized by swelling or vacu-olization of the cytoplasm with focal disruption of the cellmembranes, karyorrhexis, loss or attenuation of luminal brushborder, and degeneration of cytoplasmic organelles (Figure 7B).The nuclei were mildly enlarged in most tubular cells, contain-ing irregular heterochromatin and occasional nucleoli. The glo-merular capillary architecture was preserved.

The electron microscopic findings of affected tubules at 20,25, and 30 wk after injection disclosed widespread, moderate tosevere tubular cell injury ranging from cellular flattening toextensive loss of cell cytoplasm and accompanied by markedly

thickened, multilayered, wrinkled, and denuded basementmembranes (Figure 7, C and D). However, interstitial collagendeposition was minimal.

At 35 and 40 wk, there were varying degrees of regenerationof simplified tubular cell lining on preexisting thickened andreduplicated basement membranes (Figure 7, E and F). Thesecells lacked basolateral membrane infoldings, contained dis-torted mitochondria, and displayed segmental to circumferen-tial appearance of luminal brush border. Tubules with severetubulolysis demonstrated no significant restoration of the lin-ing cells.

Figure 6. Representative photomicro-graphs of immunohistochemical stainingfor WT1 and cleaved caspase-3 in mousekidney sections at 30, 35, and 40 wk afterinjection with 225Ac-labeled antibody. Thenumber of WT1-positive cells per glomer-ulus is unchanged between 30, 35, and 40wk after injection (A, B, and C, respec-tively) and compared with untreated con-trols (D). The number of caspase-3–posi-tive cells (arrows) decreased moderatelyfrom 30 wk (E) to 35 and 40 wk (F and G,respectively) after injection. No stainingfor cleaved caspase-3 was seen in un-treated control mice (H).

J Am Soc Nephrol 16: 2677–2689, 2005 �-Particle–Mediated Renal Toxicity 2683

Summary of Pathologic FindingsIn summary, glomerular and tubular cell nuclear hyperchro-

masia and isolated tubular cell lysis with karyorrhexis wereobserved as early as 10 wk after injection (Tables 1 and 2).However, the general renal parenchymal architecture was pre-served. From 20 to 30 wk after injection, widespread tubuloly-sis and tubular collapse contributed to progressive thinning ofthe renal cortex, resulting in glomerular crowding. A relativedecrease in glomerular cellularity, variable interstitial inflam-mation and hyperplasia, and increased granularity of JG cells

was noted during this period. By 35 wk, focal regenerativetubular changes began to emerge with epithelial simplification,reactive nuclei, and early recovery of brush border but withonly focal minimal increase in interstitial collagen. Thickenedand reduplicated tubular basement membranes leaving emptyprofiles (as a result of tubulolysis) as well as some regenerated,simplified tubules were seen at 40 wk after injection. Areas oftubular regeneration alternated with segmental cortical scar-ring and moderate medullary atrophy at 40 wk. Microcysticdilation of the glomerular Bowman capsules was seen in the

Figure 7. Representative electron micrographs of ultrastructural changes in mouse kidneys after injection with 225Ac-labeledantibody. (A) A control mouse kidney showing proximal tubules with normal basement membranes, preserved luminalbrush borders, basolateral infoldings (arrows), and cytoplasmic organelles. (B) At 10 wk, focal tubular cell swelling, reactivenuclei, and loss of surface brush border (arrows). (C) At 20 wk, proximal tubules in various stages of tubulolysis, with twoshowing complete cell lysis and bare basement membranes (arrows). (D) At 25 wk, complete tubulolysis, simplified tubule,and interstitial lymphocytes (arrows). (E) At 35 wk, regenerating tubular cells with partial brush border (arrows) and absenceof basolateral infoldings on thickened and duplicated basement membranes. (F) At 40 wk, degenerative tubular changes withthickened and wrinkled basement membranes (arrows) and small, scattered bundles of interstitial collagen. Magnification,�1900.

2684 Journal of the American Society of Nephrology J Am Soc Nephrol 16: 2677–2689, 2005

outer cortex. A lower JG cell count with focal cytoplasmicdegranulation was also noted at this time point.

Renal DosimetryThe estimated absorbed dose to mouse kidneys is 27.6 Gy,

with absorbed dose contributions of 0.6, 7.1, 7.9, and 12 Gy,from 225Ac, 221Fr, 217At, and 213Bi, respectively (Table 3). Theestimated dose to the renal cortex is 39 Gy.

DiscussionRadiation nephropathy is a major adverse effect of total body

irradiation used in the preparation for bone marrow trans-plants (13). The increase in the use of radionuclide therapies forcancer is also leading to an increased incidence of radiationnephropathy. In contrast to external beam irradiation, a radio-nuclide-carrier conjugate delivers a radiation dose to specificsegments of the nephron, depending on the pharmacokineticsof the conjugate and the emission characteristics of the radio-nuclide (1). Radiolabeled peptides, low molecular weight pro-teins, and antibody fragments are filtered through the glomer-ulus and reabsorbed in the tubules and subsequently undergolysosomal degradation in the proximal tubular cells (29). Ra-diolabeled full-length antibodies, such as an 225Ac-labeled an-tibody (225Ac nanogenerator), are large enough to escape glo-merular filtration. However, the �-particle–emitting daughtersthat are generated in the blood stream from circulating 225Acnanogenerators can accumulate in and irradiate the kidneys.Although the effective half-lives of many antibodies in humansare usually measured in several days, the 10-d half-life of 225Acand constant generation of daughters may lead to a moreprolonged irradiation of the renal tubules. Therefore, an under-standing of the pathogenetic mechanisms of radiation nephrop-athy, based on systematic renal pathologic examination, is cen-tral to the design of interventions to abrogate the potential renaltoxicity of 225Ac nanogenerators.

Injection of mice with 225Ac nanogenerators led to a progres-

sive decline in renal function, measured as an increase in theBUN. However, the serum creatinine levels remained withinnormal limits, possibly as a result of loss of muscle mass in225Ac-treated mice. Pallor of abdominal organs was first seen at20 wk after injection and increased progressively over time.Renal injury has been shown to reduce the potential for eryth-ropoietin gene expression in interstitial fibroblast-like cells (30).Therefore, the pallor may be due to a reduction in erythropoi-etin production by the interstitial fibroblasts as a consequenceof radiation damage, resulting in a decrease in the hematocrit.However, hematocrit or total red cell count could not be mea-sured in these mice.

Glomerular capillary endothelial and mesangial cells havebeen shown to play an important role in the pathogenesis ofradiation nephropathy, and glomerular injury is considered tobe central to the development of radiation nephropathy afterexternal beam irradiation (31,32). Glomerular thrombotic mi-croangiopathy has been reported in patients who were treatedwith yttrium-90–labeled (a �-particle emitter) somatostatin an-alogue (17). In our mouse model of radiation nephropathy,progressive, severe, and widespread tubular injury (lysis, sim-plification, and atrophy) is seen in the absence of any specificglomerular pathology or major interstitial fibrosis. This is ac-companied by a significant reduction in the renal function.Therefore, it seems that the mechanism of development ofradiation nephropathy in this system is different from the pro-cesses previously described with external beam radiation (13).The dilation of the Bowman spaces may be secondary to thetubular obstruction by casts or loss of tubules (atubular glo-meruli). Because � particles have short path lengths (5), theradiation dose is deposited in close vicinity of the sites ofdaughter accumulation in the nephron (tubules and peritubularendothelial and interstitial cells), and the glomeruli are rela-tively spared of the radiation dose. Our findings further sup-port the evidence on tubular epithelial cells and interstitial

Table 1. Salient renal pathologic findings after injection with 225Ac nanogenerators: Glomerular and vascularfeaturesa

C 10 Wk 20 Wk 25 Wk 30 Wk 35 Wk 40 Wk

Glomerular featuresreduction in size and cellularityb � � 10 20 20 30 30dilated Bowman spaceb � � 10 10 to 20 10 to 30 �50 50 to 75abnormal nuclei � � � � � � �glomerulosclerosis � � Rare,

subcapsularRare,

subcapsularRare,

subcapsularFocal

corticalSubcapsular,focal cortical

Vascular featuresprominent endothelial cells � � � � � � �medial cell vacuolizationc � � 1� 2� 1� 1� �peritubular capillaries Normal Normal Normal Inflammatory

cellsInflammatory

cells10–20% loss 20–30% loss

Juxtaglomerular cellsc

no. of cells per glomerulus 1 to 2 1 to 2 1 to 4 1 to 3 1 to 2 1 to 2 0 to 1granularity Normal Normal Increased Increased Decreased Decreased Decreased% of glomeruli 10 10 10 to 20 10 to 20 10 5 �5arteriolar metaplasia � � � � Rare � � �

a225Ac, Actinium-225; C, untreated control.bPercentage of glomeruli.cTrichrome stain.

J Am Soc Nephrol 16: 2677–2689, 2005 �-Particle–Mediated Renal Toxicity 2685

Tab

le2.

Salie

ntre

nal

path

olog

icfi

ndin

gsaf

ter

inje

ctio

nw

ith

225 A

cna

noge

nera

tors

:Tub

uloi

nter

stit

ial

feat

ures

and

TG

F-�

1ex

pres

sion

C10

Wk

20W

k25

Wk

30W

k35

Wk

40W

k

Tub

uloi

nter

stit

ial

feat

ures

abno

rmal

/re

acti

venu

clea

rch

ange

�M

ildM

oder

ate,

foca

lka

ryor

rhex

is

Foca

lka

ryom

egal

y,Se

vere

kary

orrh

exis

Foca

lka

ryom

egal

y,Se

vere

kary

orrh

exis

Rea

ctiv

ech

ange

,M

ildka

ryor

rhex

iscy

topl

asm

icva

cuol

izat

ion

(%of

cells

)�

25�

50�

10�

10�

—tu

bulo

lysi

sw

ith

colla

pse

(%of

tubu

les)

�O

ccas

iona

lce

lls�

50�

75�

7550

30–5

0lo

ssof

brus

hbo

rder

(%of

tubu

les)

�O

ccas

iona

lce

lls70

75�

7560

60ba

sem

ent

mem

bran

eth

icke

ning

a�

��

�2�

wit

hla

yers

2�w

ith

laye

rs2

to3

�w

ith

laye

rsat

roph

y(%

oftu

bule

s)�

��

1010

10—

shri

nkag

e/si

mpl

ific

atio

n(%

ofre

sid

ual

tubu

les)

��

5060

7050

30

inte

rsti

tial

infl

amm

atio

n�

��

Foca

l,m

ildpe

riva

scul

ar

20to

25%

Foca

l,m

ild5%

Foca

l,m

ild—

inte

rsti

tial

fibr

osis

��

��

Foca

l,pe

riva

scul

ar�

—m

edul

lary

tubu

les

Nor

mal

Nor

mal

Nor

mal

Mild

atro

phy,

5%ca

sts

Mild

atro

phy,

foca

lm

icro

calc

ific

atio

n

Mod

erat

eat

roph

y,fo

cal

mic

roca

lcif

icat

ion

Mod

erat

eat

roph

y,fo

cal

mic

roca

lcif

icat

ion

tubu

lar

cast

s(%

ofco

rtic

altu

bule

s)�

�R

are

2020

20to

25T

GF-

�1

imm

unos

tain

ing

pres

erve

dtu

bule

s(i

nten

sity

/%

ofce

lls)

��

Foca

l1–

2�/

25Fo

cal

1to

2�/

50Fo

cal,

trac

e–1�

/10

Foca

l,tr

ace/

5Fo

cal,

trac

e/

5tu

bulo

lyti

clin

ing

(int

ensi

ty/

%of

cells

)�

�1

to3�

/50

1to

3�/

50T

race

–1�

/25

Tra

ce–1

�/

20Fo

cal,

trac

e–1�

/20

foca

lin

ters

titi

alce

lls�

�1

to2�

2�T

race

–1�

Tra

ceT

race

aT

rich

rom

est

ain.

2686 Journal of the American Society of Nephrology J Am Soc Nephrol 16: 2677–2689, 2005

fibroblasts as the targets in the development of radiation ne-phropathy (33). The more pronounced involvement of corticaltubules as compared with medullary tubules was an expectedfinding, considering the known affinity of free bismuth forproximal tubular cells.

Radiation-induced renal fibrosis involves complex and dy-namic interactions among glomerular, tubular, and interstitialcells (13). Various biologic mediators such as TGF-�1, PAI-1,and others have been thought to play a role in this process (13).In vitro irradiation of renal tubular cells resulted in a significantenhancement of TGF-�1, PAI-1, and collagen-1 gene expression(34). TGF-�1, a profibrogenic cytokine, is considered to be oneof the most important mediators in renal fibrosis (35). TGF-�1

regulates the transdifferentiation of tubular epithelial cells to�-smooth muscle actin–expressing myofibroblasts, a cell typethat is believed to be the main source of increased extracellularmatrix deposition in renal fibrosis (36). We did not observe adetectable increase in TGF-�1 expression until 20 to 25 wk afterinjection. An increase in TGF-�1 expression generally occurs inlate phases of radiation nephropathy (24). Moreover, radiation-induced renal pathologic changes appear much later in micethan in other species (13). PAI-1 is the major inhibitor of tissueplasminogen activator in vivo. PAI-1 inhibits matrix degenera-tion and promotes extracellular matrix deposition and fibrosisvia inhibition of plasmin production from plasminogen (37).The source of PAI-1 is glomerular capillary endothelial cells(24), which are relatively spared of the radiation dose in ourmodel of radiation nephropathy. Therefore, in contrast to ex-ternal beam radiation, PAI-1 does not seem to play a major rolein this model of radiation nephropathy.

WT1 is a zinc finger protein that plays a role in transcriptionas well as posttranscriptional processing of RNA (26). In adultkidneys, WT1 is expressed by the visceral epithelial cells of theglomeruli. The absence of a change in the number of WT1-positive cells between different time points after injection andcompared with untreated controls suggests that the observedmild decrease in glomerular cellularity was not a result ofvisceral epithelial cell loss. It may be due to the loss of parietalepithelial cells, which are more likely to fall in the range of�-particle radiation as a result of their anatomic location. Thedetection of cleaved caspase-3 immunostaining alludes to therole of apoptosis as a mechanism of the observed renal tubularcell death. Caspase-3 is a key effector caspase in the process ofapoptosis and is activated via proteolytic cleavage (25). Cleavedcaspase-3 immunostaining is detected only transiently in thecells that undergo apoptosis and is lost once the cells haveapoptosed.

The renin-angiotensin system has been implicated in thepathogenesis and progression of radiation nephropathy(13,24,38). Angiotensin II (Ang II) is a growth promoter and isthought to cause mitotic cell death of renal epithelial cells byinducing the lethally irradiated cells to divide (39). Ang II hasalso been shown to trigger the production of profibrogenicmediators such as TGF-�1 (40) and PAI-1 (41) in the kidneys.However, Cohen et al. (38) did not observe any increase inplasma renin activity or blood Ang II levels in the first 10 wkafter irradiation and suggested that normal activity of the re-nin-angiotensin system could be detrimental to the irradiatedkidney. Capillary endothelial damage from external beam ra-diation results in a decreased nitric oxide production (42).Therefore, a decreased modulation of the effects of Ang II bynitric oxide may be the reason for enhanced sensitivity of thekidney to the effects of Ang II (43). We have observed hyper-trophy and hypergranularity of the JG apparatus cells at 20 and25 wk after 225Ac injection, which diminished by 30 wk andreturned to normal by 35 wk as the JG cells get degranulated,leaving behind vacuolated cells. Because of experimental sam-ple limitations, the plasma renin activity could not be measuredin these mice. An increase in the JG cell index at 15 wk afterlocal kidney irradiation in rats was reported previously (44).

On the basis of a mechanism of progressive nephropathysuggested by Remuzzi et al. (45), we propose that radiation-induced nephron loss produces adaptive responses involvingincreased Ang II production and glomerular-capillary hyper-tension, which result in enhanced filtration of plasma proteinsinto the tubular fluid. The reabsorption of filtered proteins bythe remaining tubular cells activates intracellular pathways forupregulation of inflammatory and vasoactive genes, the prod-ucts of which cause varying degrees of interstitial inflammationand fibroblast activation. Because tubular cells are the sites forradioactive daughter accumulation, the radiation-induced lossof TGF-�1–expressing tubular and adjoining interstitial cellsmay account for the reduced extent and intensity of TGF-�1

staining and minimal fibrosis than what has been reportedpreviously in studies with external beam irradiation.

In the case of radiation nephropathy after bone marrowtransplantation, 10 Gy of total body irradiation with x-rays as asingle dose or 14 Gy fractionated over 3 d results in radiationnephropathy (13). In our model, the absorbed dose estimatesfor the whole kidney and renal cortex are 27.6 and 39 Gy,respectively. The dose estimates are expressed in units of Gyrather than Sv because the exact relative biologic effectivenessvalue of � particles for kidney cells in vivo is not known. Thedose of [225Ac]HuM195 used in the experiments (0.35 �Ci), ona weight basis, is higher than the planned human dose.

A wide range of preventive pharmacologic strategies is beinginvestigated with an aim of attenuating the radiation-inducedrenal damage. Because the development of radiation nephrop-athy is dose dependant (33), a reduction in the radiation dose tothe kidneys should be protective. We have shown that chelationof free metal, diuresis, and competitive metal blockade canreduce the accumulation of 225Ac daughter elements and, there-fore, the renal radiation dose (27). The use of an enzyme-cleavable linker to link chelated radiometal to the antibody is

Table 3. Kidney and renal cortex absorbed doses

Radionuclide Whole Kidney (Gy) Renal Cortex (Gy)

225Ac 0.6 0.8221Fr 7.1 10217At 7.9 11213Bi 12 17Total 27.6 38.8

J Am Soc Nephrol 16: 2677–2689, 2005 �-Particle–Mediated Renal Toxicity 2687

another potentially useful approach. After the enzymatic cleav-age of the linker, the 225Ac-chelate should clear rapidly; there-fore, the radiation dose to the kidneys from 225Ac daughteraccumulation could be diminished. Pharmacologic modifica-tion of the renin-angiotensin-aldosterone system in the inter-vening period between irradiation and renal failure has beenshown to protect against radiation-induced renal damage in anumber of animal studies (13). Angiotensin-converting enzymeinhibitors, angiotensin receptor antagonists, and spironolactone(aldosterone receptor antagonist) ameliorated the functionaland morphologic changes in the kidneys after external beamradiation (23,46). Therefore, similar interventions may be usefulin reducing the toxicity of internal radio-emitters.

In summary, we show the chronological sequence of mor-phologic and functional changes in the renal tissue after injec-tion of mice with 225Ac nanogenerators. The observed renalhistopathologic changes are different from those reported withexternal beam irradiation, which reflects the differences in thetwo models with regard to the pattern and rate of renal radia-tion dose deposition. An understanding of the pathogeneticmechanisms of �-particle–mediated renal toxicity should beuseful in designing of interventions for its amelioration.

AcknowledgmentsThis research was supported by grants R01-CA 55349 and P01-33049

from the National Institutes of Health, the Joseph LeRoy and Ann C.Warner Fund, and the William and Alice Goodwin CommonwealthFoundation for Cancer Research. D.S. is a Doris Duke DistinguishedClinical Scientist.

We thank Drs. Eric Cohen and Carlos Flombaum for careful reviewand valuable comments on the manuscript. We thank Michael Gangerand Steven Bowe for valuable help with electron microscopy andpreparation of illustrations, respectively, and Yi-Fang Liu for assistancein immunohistochemistry.

References1. Lambert B, Cybulla M, Weiner SM, Van De Wiele C, Ham

H, Dierckx RA, Otte A: Renal toxicity after radionuclidetherapy. Radiat Res 161: 607–611, 2004

2. Ghobrial I, Witzig T: Radioimmunotherapy: A new treat-ment modality for B-cell non-Hodgkin’s lymphoma. On-cology (Huntingt) 18: 623–630; discussion 633–634, 637–638,640, 2004

3. Hernandez MC, Knox SJ: Radiobiology of radioimmuno-therapy: Targeting CD20 B-cell antigen in non-Hodgkin’slymphoma. Int J Radiat Oncol Biol Phys 59: 1274–1287, 2004

4. Chang CH, Sharkey RM, Rossi EA, Karacay H, McBride W,Hansen HJ, Chatal JF, Barbet, J, Goldenberg DM: Molecu-lar advances in pretargeting radioimmunotherapy withbispecific antibodies. Mol Cancer Ther 1: 553–563, 2002

5. Kozak RW, Atcher RW, Gansow OA, Friedman AM, HinesJJ, Waldmann TA: Bismuth-212-labeled anti-Tac monoclo-nal antibody: Alpha-particle-emitting radionuclides as mo-dalities for radioimmunotherapy. Proc Natl Acad Sci U S A83: 474–478, 1986

6. Bethge WA, Wilbur DS, Storb R, Hamlin DK, Santos EB,Brechbiel MW, Fisher DR, Sandmaier BM: Selective T-cellablation with bismuth-213-labeled anti-TCRalphabeta as

nonmyeloablative conditioning for allogeneic canine mar-row transplantation. Blood 101: 5068–5075, 2003

7. Yao Z, Garmestani K, Wong KJ, Park LS, Dadachova E,Yordanov A, Waldmann TA, Eckelman WC, Paik CH,Carrasquillo JA: Comparative cellular catabolism and re-tention of astatine-, bismuth-, and lead-radiolabeled inter-nalizing monoclonal antibody. J Nucl Med 42: 1538–1544,2001

8. Waldmann TA: Immunotherapy: Past, present, and future.Nat Med 9: 269–277, 2003

9. McDevitt MR, Ma D, Lai LT, Simon J, Borchardt P, FrankRK, Wu K, Pellegrini V, Curcio MJ, Miederer M, BanderNH, Scheinberg DA: Tumor therapy with targeted atomicnanogenerators. Science 294: 1537–1540, 2001

10. Borchardt PE, Yuan RR, Miederer M, McDevitt MR, Schein-berg DA: Targeted actinium-225 in vivo generators fortherapy of ovarian cancer. Cancer Res 63: 5084–5090, 2003

11. Miederer M, McDevitt MR, Borchardt P, Bergman I,Kramer K, Cheung NK, Scheinberg DA: Treatment of neu-roblastoma meningeal carcinomatosis with intrathecal ap-plication of alpha-emitting atomic nanogenerators target-ing disialo-ganglioside GD2. Clin Cancer Res 10: 6985–6992,2004

12. Miederer M, McDevitt MR, Sgouros G, Kramer K, CheungNK, Scheinberg DA: Pharmacokinetics, dosimetry, andtoxicity of the targetable atomic generator, 225Ac-HuM195,in nonhuman primates. J Nucl Med 45: 129–137, 2004

13. Cohen EP, Robbins ME: Radiation nephropathy. SeminNephrol 23: 486–499, 2003

14. Behr TM, Sharkey RM, Sgouros G, Blumenthal RD, DunnRM, Kolbert K, Griffiths GL, Siegel JA, Becker WS, Gold-enberg DM: Overcoming the nephrotoxicity of radiometal-labeled immunoconjugates: Improved cancer therapy ad-ministered to a nude mouse model in relation to theinternal radiation dosimetry. Cancer 80: 2591–2610, 1997

15. Bunjes D: 188Re-labeled anti-CD66 monoclonal antibodyin stem cell transplantation for patients with high-riskacute myeloid leukemia. Leuk Lymphoma 43: 2125–2131,2002

16. Cybulla M, Weiner SM, Otte A: End-stage renal diseaseafter treatment with 90Y-DOTATOC. Eur J Nucl Med 28:1552–1554, 2001

17. Moll S, Nickeleit V, Mueller-Brand J, Brunner FP, MaeckeHR, Mihatsch MJ: A new cause of renal thrombotic mi-croangiopathy: Yttrium 90-DOTATOC internal radiother-apy. Am J Kidney Dis 37: 847–851, 2001

18. Flynn AA, Pedley RB, Green AJ, Dearling JL, El-Emir E,Boxer GM, Boden R, Begent RH: The nonuniformity ofantibody distribution in the kidney and its influence ondosimetry. Radiat Res 159: 182–189, 2003

19. Russ GA, Bigler RE, Tilbury RS, Woodard HQ, Laughlin JS:Metabolic studies with radiobismuth. I. Retention and dis-tribution of 206Bi in the normal rat. Radiat Res 63: 443–454,1975

20. Slikkerveer A, de Wolff FA: Pharmacokinetics and toxicityof bismuth compounds. Med Toxicol Adverse Drug Exp 4:303–323, 1989

21. Szymanska JA, Mogilnicka EM, Kaszper BW: Binding ofbismuth in the kidneys of the rat: The role of metallothio-nein-like proteins. Biochem Pharmacol 26: 257–258, 1977

22. McDevitt MR, Ma D, Simon J, Frank RK, Scheinberg DA:

2688 Journal of the American Society of Nephrology J Am Soc Nephrol 16: 2677–2689, 2005

Design and synthesis of 225Ac radioimmunopharmaceuti-cals. Appl Radiat Isot 57: 841–847, 2002

23. Brown NJ, Nakamura S, Ma L, Nakamura I, Donnert E,Freeman M, Vaughan DE, Fogo AB: Aldosterone modu-lates plasminogen activator inhibitor-1 and glomeruloscle-rosis in vivo. Kidney Int 58: 1219–1227, 2000

24. Cohen EP, Bonsib SA, Whitehouse E, Hopewell JW, Rob-bins ME: Mediators and mechanisms of radiation nephrop-athy. Proc Soc Exp Biol Med 223: 218–225, 2000

25. Olney JW, Tenkova T, Dikranian K, Muglia LJ, Jermakow-icz WJ, D’Sa C, Roth KA, Yang Y, Jeanpierre C, DresslerGR, Lacoste M, Niaudet P, Gubler MC: Ethanol-inducedcaspase-3 activation in the in vivo developing mouse brain.Neurobiol Dis 9: 205–219, 2002

26. Yang Y, Jeanpierre C, Dressler GR, Lacoste M, Niaudet P,Gubler MC: WT1 and PAX-2 podocyte expression inDenys-Drash syndrome and isolated diffuse mesangialsclerosis. Am J Pathol 154: 181–192, 1999

27. Jaggi JS, Kappel BJ, McDevitt MR, Sgouros G, FlombaumCD, Cabassa C, Scheinberg D: Efforts to control the errantproducts of a targeted in vivo generator. Cancer Res 65:4888–4895, 2005

28. Bouchet LG, Bolch WE, Blanco HP, Wessels BW, Siegel JA,Rajon DA, Clairand I, Sgouros G: MIRD Pamphlet No 19:Absorbed fractions and radionuclide S values for six age-dependent multiregion models of the kidney. J Nucl Med44: 1113–1147, 2003

29. Behr TM, Goldenberg DM, Becker W: Reducing the renaluptake of radiolabeled antibody fragments and peptidesfor diagnosis and therapy: Present status, future prospectsand limitations. Eur J Nucl Med 25: 201–212, 1998

30. Maxwell PH, Ferguson DJ, Nicholls LG, Johnson MH, Rat-cliffe PJ: The interstitial response to renal injury: Fibro-blast-like cells show phenotypic changes and have reducedpotential for erythropoietin gene expression. Kidney Int 52:715–724, 1997

31. Robbins ME, Jaenke RS, Bywaters T, Golding SJ, RezvaniM, Whitehouse E, Hopewell JW: Sequential evaluation ofradiation-induced glomerular ultrastructural changes inthe pig kidney. Radiat Res 135: 351–364, 1993

32. Stephens LC, Robbins ME, Johnston DA, Thames HD,Price RE, Peters LJ, Ang KK: Radiation nephropathy in therhesus monkey: Morphometric analysis of glomerular andtubular alterations. Int J Radiat Oncol Biol Phys 31: 865–873,1995

33. Robbins ME, O’Malley Y, Zhao W, Davis CS, Bonsib SM:The role of the tubulointerstitium in radiation-inducedrenal fibrosis. Radiat Res 155: 481–489, 2001

34. Zhao W, O’Malley Y, Wei S, Robbins ME: Irradiation of rattubule epithelial cells alters the expression of gene prod-ucts associated with the synthesis and degradation of ex-tracellular matrix. Int J Radiat Biol 76: 391–402, 2000

35. Border WA, Noble NA: TGF-beta in kidney fibrosis: Atarget for gene therapy. Kidney Int 51: 1388–1396, 1997

36. Fan JM, Ng YY, Hill PA, Nikolic-Paterson DJ, Mu W,Atkins RC, Lan HY: Transforming growth factor-beta reg-ulates tubular epithelial-myofibroblast transdifferentiationin vitro. Kidney Int 56: 1455–1467, 1999

37. Brown NJ, Vaughan DE, Fogo AB: Aldosterone and PAI-1:Implications for renal injury. J Nephrol 15: 230–235, 2002

38. Cohen EP, Fish BL, Moulder JE: The renin-angiotensinsystem in experimental radiation nephropathy. J Lab ClinMed 139: 251–257, 2002

39. Moulder JE, Fish BL, Regner KR, Cohen EP: Angiotensin IIblockade reduces radiation-induced proliferation in exper-imental radiation nephropathy. Radiat Res 157: 393–401,2002

40. Datta PK, Moulder JE, Fish BL, Cohen EP, Lianos EA:TGF-beta 1 production in radiation nephropathy: Role ofangiotensin II. Int J Radiat Biol 75: 473–479, 1999

41. Nakamura S, Nakamura I, Ma L, Vaughan DE, Fogo AB:Plasminogen activator inhibitor-1 expression is regulatedby the angiotensin type 1 receptor in vivo. Kidney Int 58:251–259, 2000

42. Cohen EP, Fish BL, Moulder JE: The role of nitric oxide inradiation nephropathy. Arch Physiol Biochem 104: 200–206,1996

43. Millatt LJ, Abdel-Rahman EM, Siragy HM: Angiotensin IIand nitric oxide: A question of balance. Regul Pept 81: 1–10,1999

44. Fisher ER, Hellstrom HR: Pathogenesis of hypertensionand pathologic changes in experimental renal irradiation.Lab Invest 19: 530–538, 1968

45. Remuzzi G, Bertani T: Pathophysiology of progressive ne-phropathies. N Engl J Med 339: 1448–1456, 1998

46. Molteni A, Moulder JE, Cohen EP, Fish BL, Taylor JM,Veno PA, Wolfe LF, Ward WF: Prevention of radiation-induced nephropathy and fibrosis in a model of bonemarrow transplant by an angiotensin II receptor blocker.Exp Biol Med (Maywood) 226: 1016–1023, 2001

J Am Soc Nephrol 16: 2677–2689, 2005 �-Particle–Mediated Renal Toxicity 2689