Neonatal dexamethasone administration causes progressive renal damage due to induction of an early...

doi: 10.1152/ajprenal.00163.2007294:F768-F776, 2008. First published 30 January 2008;Am J Physiol Renal Physiol

van der Leij, Pieter J. J. Sauer, Folkert Kuipers, Gerjan Navis and Martin H. de BorstYan Liu, Harry van Goor, Rick Havinga, Julius F. W. Baller, Vincent W. Bloks, Feike R.inflammatory responseprogressive renal damage due to induction of an early Neonatal dexamethasone administration causes

You might find this additional info useful...

39 articles, 16 of which you can access for free at: This article citeshttp://ajprenal.physiology.org/content/294/4/F768.full#ref-list-1

including high resolution figures, can be found at: Updated information and serviceshttp://ajprenal.physiology.org/content/294/4/F768.full

found at: can beAmerican Journal of Physiology - Renal Physiology about Additional material and information

http://www.the-aps.org/publications/ajprenal

This information is current as of September 12, 2012.

0363-6127, ESSN: 1522-1466. Visit our website at http://www.the-aps.org/. Rockville Pike, Bethesda MD 20814-3991. Copyright © 2008 the American Physiological Society. ISSN: volume and composition. It is published 12 times a year (monthly) by the American Physiological Society, 9650relating to the kidney, urinary tract, and their respective cells and vasculature, as well as to the control of body fluid

publishes original manuscripts on a broad range of subjectsAmerican Journal of Physiology - Renal Physiology

by g

uest o

n S

epte

mber 1

2, 2

01

2http

://ajp

renal.p

hysio

logy.o

rg/

Do

wn

loa

de

d fro

m

Neonatal dexamethasone administration causes progressive renal damage

due to induction of an early inflammatory response

Yan Liu,1,5 Harry van Goor,2 Rick Havinga,1 Julius F. W. Baller,1 Vincent W. Bloks,1

Feike R. van der Leij,1,3 Pieter J. J. Sauer,1 Folkert Kuipers,1 Gerjan Navis,4 and Martin H. de Borst2

1Center for Liver, Digestive, and Metabolic Diseases, Laboratory of Pediatrics; 2Department of Pathology and Laboratory

Medicine; University Medical Center Groningen, University of Groningen, Groningen; 3Unit Life Sciences, Van Hall

University of Applied Sciences, Leeuwarden; 4Department of Medicine, Division of Nephrology, University Medical Center

Groningen, University of Groningen, The Netherlands; and 5Department of Endocrinology, Third Hospital of Hebei Medical

University, Shijiazhuang, Hebei, China

Submitted 8 April 2007; accepted in final form 25 January 2008

Liu Y, van Goor H, Havinga R, Baller JF, Bloks VW, van der

Leij FR, Sauer PJ, Kuipers F, Navis G, de Borst MH. Neonataldexamethasone administration causes progressive renal damage dueto induction of an early inflammatory response. Am J Physiol Renal

Physiol 294: F768–F776, 2008. First published January 30, 2008;doi:10.1152/ajprenal.00163.2007.—Glucocorticoids (GCs) are widelyused to prevent chronic lung disease in immature newborns. Emergingevidence indicates that GC exposure in early life may interfere withkidney function and is associated with hypertension in later life. Inthis study, we have investigated the effect of neonatal dexamethasone(DEX) administration on renal function in rats. Male rats were treatedwith DEX in the first 3 days after birth, controls received saline(SAL). Severe renal damage associated with premature death wasfound at 50 wks upon DEX treatment, while renal function andmorphology were normal in controls. A subsequent time-course studywas performed from 2 days to 32 wks. Compared with controls,neonatal DEX administration led to significant and persistent growthretardation. Progressive proteinuria and increased systolic blood pres-sure were found from 8 wks onwards in DEX-treated animals. Renala-SMA gene expression was elevated from wk 24 onwards andmorphological fibrosis was noted at 32 wks of age following DEXtreatment. Markedly increased renal gene expression of TNF-a andMCP-1 in DEX -treated rats was observed at day 7, probably con-tributing to the permanent increase in interstitial macrophage numbersthat started at 14 days. Permanently elevated renal TGF-b geneexpression was induced by DEX administration from 4 wks onwards.Our data indicate that neonatal DEX administration in rats leads torenal failure in later life, presumably due to an early inflammatorytrigger that elicits a persistent pro-fibrotic process that eventuallyresults in progressive renal deterioration.

newborn; glucocorticoids; kidney failure; inflammation

GLUCOCORTIOIDS (GCs) are widely used in immature newbornsto prevent chronic lung disease. However, accumulating evi-dence from animal studies indicates that overexposure to GCsin early life may have long-term negative effects such ascardiac dilation (4), hyperglycemia (28), increased blood pres-sure (22), and alterations in social behavior (21).

Renal damage is one consequence of early-life GC overex-posure. Data suggest that maternal GC administration, e.g.,dexamethasone (DEX) or betamethasone, impairs nephrogen-esis and reduces glomeruli numbers, which may contribute to

hypertension in late life (17, 33). To the best of our knowledge,effects of neonatal GC administration on kidney function havenot been studied in animals. Studies in human infants are alsolimited; a high incidence of renal calcium deposition in DEX-treated newborns has been described as a clinical observation(11a, 20), but no follow-up study has been reported so far.

In the present study, we sought to establish the effect ofneonatal DEX treatment in rats on survival in relation to kidneyfunction. We found that neonatal DEX administration to ratpups causes premature death associated with the unexpectedfindings of severe progressive renal damage. Thereafter, wecarefully evaluated the sequence of events upon postnatal DEXadministration leading to fatal kidney disease. Clinical andstructural parameters were investigated, especially in the earlyphase, to elucidate the mechanism involved in the long-termdetrimental effect of GCs on the kidney. Our results indicatethat an early inflammatory response, i.e., during the first monthof life, underlies structural changes in the kidney that leads tosevere organ dysfunction in later life.

MATERIALS AND METHODS

Animal model. Experiments were in accord with institutional andlegislative regulations and approved by The Local Committee forAnimal Experiments. Pregnant Wistar rats (270–300 g) were housedindividually and kept under conventional housing conditions with freeaccess to food and water. Pups were born on day 21 or 22 of gestation.On the day of birth (day 0), male pups were selected and randomlydivided into treatment and control groups. Room temperature andhumidity were kept constant, and the rats had free access to food andwater. An artificial 12:12-h light-dark cycle was employed. Rat pupsin the treatment group were injected (ip) with dexamethasone 21-phosphate (DEX) on day 1 (0.5 mg/g body wt), day 2 (0.3 mg/g bodywt), and day 3 (0.1 mg/g body wt) after birth in the morning between0900 and 1200, as described previously (4, 22, 28). The dosages ofDEX used in the present study were similar to those used in immaturehuman babies with chronic lung disease. The Committee on Fetus andNewborn of the American Academy of Pediatrics reported that, inmost cases, DEX is given in a dosage of 0.5 mg zkg body wt21

zday21

for 3 days, followed by a tapering course of 0.3, 0.1, and 0.05 mg zkgbody wt21

zday21 each for 3 days (1). Controls received equalvolumes (10 ml/g body wt) of sterile pyrogen-free saline (SAL).

Experiment 1. This first experiment was designed to establisheffects of neonatal DEX administration on survival. On the day of

Address for reprint requests and other correspondence: H. van Goor, Dept.of Pathology and Laboratory Medicine, Univ. Medical Center Groningen,Univ. of Groningen, Hanzeplein 1, 9713 GZ Groningen, The Netherlands(e-mail: [email protected]).

The costs of publication of this article were defrayed in part by the paymentof page charges. The article must therefore be hereby marked “advertisement”in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

Am J Physiol Renal Physiol 294: F768–F776, 2008.First published January 30, 2008; doi:10.1152/ajprenal.00163.2007.

0363-6127/08 $8.00 Copyright © 2008 the American Physiological Society http://www.ajprenal.orgF768

by g

uest o

n S

epte

mber 1

2, 2

01

2http

://ajp

renal.p

hysio

logy.o

rg/

Do

wn

loa

de

d fro

m

birth (day 0) male pups were selected and randomly divided intotreatment and control groups (SAL, n 5 22; DEX, n 5 30). Bodyweight and survival were measured weekly. At 50 wk of age, 24-hurine was collected in a metabolic cage, and urine volume as well asconcentrations of creatinine (Cr) and urinary protein were measured.Serum Cr and blood glucose were determined in blood collected at theend of the experiment prior to euthanasia. After the rats were killed,the kidneys were collected without perfusion for periodic acid-Schiff(PAS) staining. Procedure details are described below.

Experiment 2. This second experiment was designed to explore themechanism underlying renal damage induced by DEX. We performeda time course study in rats aged from 2 days to 32 wk. On day 0, malepups were selected and randomly divided into treatment and controlgroups (n 5 6 per group). Rats pups received the same treatmentdescribed in experiment 1. Rats were killed at indicated ages from 2days to 32 wk in the morning between 0900 and 1200.

Systolic blood pressure measurement. It has been suggested thatearly-life GC overexposure may cause increased blood pressure inadult life (15, 22, 28). Therefore, systolic blood pressure (SBP) wasmeasured in these animals. SBP was measured from age 4 wk to 32wk after 2 wk of daily training before the first measurement. Anautomated multichannel system was used with tail cuffs and photo-electric sensors to detect the tail pulse (Apollo 179; IITC Life Science,Woodland Hills, CA) as described previously (37). The rats wereplaced in a test chamber in restrainers while the temperature wasmaintained at 27–29°C. For each rat, the value was calculated fromthe mean of three or four consecutive measurements.

Urinary protein, blood glucose, and Cr measurements. Twenty-four-hour urine was collected in a metabolic cage from 4- to 50-wk-old rats. Urinary Pr and Cr concentrations were measured by use of anautoanalyzer (Merck Mega, Darmstadt, Germany). Blood sampleswere collected from abdominal artery under anesthesia. Blood glucosewas measured with a Medisense Precision glucose meter (Abbott,Amersfoort, The Netherlands). Serum Cr was determined by autoana-lyzer (Merck Mega).

Tissue processing and renal morphology. Kidneys were perfusedwith saline under anesthesia after blood samples were taken. Animalswere then killed, and kidneys were harvested for histological exami-nation. Tissue was fixed in 4% paraformaldehyde and processed forparaffin embedding. Sections (3 mm) were stained with PAS andexamined by light microscopy.

Estimation of nephron numbers. It has been reported that maternalGC administration may lead to reduced nephron numbers in offspring(17, 29). In rats, nephrogenesis is not complete until at least 1 wk afterbirth (25). An estimation of the nephron number in adult animals (8wk and 24 wk of age) was performed by counting representative

glomerular numbers in microscopic sections after PAS staining under3100 magnification (n 5 5–6 rats per group at every age, 20 fieldsper rat). Only glomeruli with intact capsule in the cortical area werecounted.

Immunohistochemistry. An antibody against a-smooth muscle cellactin (a-SMA), a marker of myofibroblast transformation, was used todetect prefibrotic changes in the kidney, and the monocyte/macro-phage marker ED1 was used as marker for macrophage recruitment,an early marker of renal damage (27). Furthermore, various studiessuggest involvement of the c-Jun NH2-terminal kinase (JNK) signal-ing pathway in inflammatory response by modulating monocytechemotactic protein-1 (MCP-1) (13, 38). Therefore, phosphorylatedJNK (pJNK) protein was studied in newborn rats to reveal whetherthis pathway might be involved in our model. Paraffin sections (3 mm)were dewaxed and subjected to heat-induced antigen retrieval byovernight incubation in 0.1 M Tris zHCl buffer at 80°C. Endogenousperoxidase was blocked with 0.03% H2O2 in PBS for 30 min. Thensections were stained with a mouse monoclonal antibody againsta-SMA (clone 1A4; Sigma Chemical, St. Louis, MO), a polyclonalrabbit antibody against pJNK (Cell Signaling Technology, Denver,CO), or a mouse monoclonal antibody against ED1 (Serotec, Kidling-ton, UK). Antibodies were incubated for 60 min at room temperatureexcept for the pJNK antibody, which was incubated for 120 min.Binding of the antibody was detected using sequential incubationswith peroxidase-labeled (PO-) secondary and tertiary antibodies inaccordance with the species in which the primary antibody was raised(Dako Cytomation, Glostrup, Denmark) for 30 min. Antibody dilu-tions were made in PBS supplemented with 1% BSA and 1% normalrat serum. Peroxidase activity was developed by using 3,39-diamino-benzidine tetrachloride for 10 min. Sections were counterstained withhematoxylin. For all immunohistochemical staining, controls in whichthe primary antibody was replaced by PBS and appropriate isotypecontrols were included; these were consistently negative.

For glomerular measurement of ED1, the number of positive cellsper glomerulus was counted (average of 50 glomeruli). For interstitialmeasurement of ED1, the number of positive cells per square milli-meter was counted (average of 30 fields per kidney): vessels andglomeruli were excluded from measurements. For measurement ofpJNK, the number of positive cells per per square millimeter wascounted by a blinded observer (average of 30 fields per kidney).

RNA isolation and real-time quantitative PCR. A number of proin-flammatory factors that may trigger inflammatory response and fibro-sis have been measured by real-time quantitative PCR. For RNAisolation, kidneys tissues from 2-day- to 32-wk-old animals wererapidly excised and immediately frozen in liquid nitrogen followingthe procedure mentioned above. RNA was extracted from kidneysamples (including cortex and medulla) by using the Tri Reagentmethod (Sigma). The integrity of total RNA was assessed usingLab-on-a-Chip 2100 Bioanalyzer (Agilent). Total RNA was reversetranscribed as described previously (30). Real-time quantitative PCRwas performed using an ABI PRISM 7700 sequence detector (Applied

Fig. 1. Neonatal dexamethasone (DEX) administration affects survival atolder age. Kaplan-Meier survival curves demonstrate that neonatal DEXadministration results in significantly increased death rate. Saline (SAL) group,n 5 22; DEX group, n 5 30. *P , 0.05 by log rank test.

Table 1. Characteristics of 50-wk-old rats upon neonataltreatment with SAL or DEX

SAL DEX

Body weight (BW), g 625639 559661*Kidney weight (KW), g 1.6960.13 1.6060.26KW/BW ratio,mg/g 2.7260.22 2.7660.79Blood glucose, mmol/l 6.7161.32 5.3560.31*Serum Cr, mmol/l 67.065.8 102.0630.0*Volume of 24-h urine, ml 15.063.5 39.6620.0*Urinary Pr/Cr, g/mmol 0.2260.06 0.5560.23*

Data are expressed as means 6 SD; n 5 8 per treatment group. SAL, salinetreated; DEX, dexamethasone treated; Cr, creatine; Pr, protein. *P , 0.05DEX vs. SAL by Mann-Whitney U-test.

F769NEONATAL DEXAMETHASONE TREATMENT CAUSES RENAL DAMAGE

AJP-Renal Physiol • VOL 294 • APRIL 2008 • www.ajprenal.org

by g

uest o

n S

epte

mber 1

2, 2

01

2http

://ajp

renal.p

hysio

logy.o

rg/

Do

wn

loa

de

d fro

m

Biosystems, Foster City, CA) according to the manufacturer’s instruc-tions. Primers were obtained from Invitrogen (Carlsbad, CA). Fluoro-genic probes, labeled with 6-carboxyfluorescein and 6-carboxytetra-methyl-rhodamine, were made by Eurogentec (Seraing, Belgium).The following primers and probes were used, in addition to 36B4,which has been published before (36): TNF-a (NM_012675, forwardGTA GCC CAC GTC GTA GCA AAC, reverse AGT TGG TTG TCTTTG AGA TCC ATG, probe CGC TGG CTC AGC CAC TCC AGC);TGF-b (X52498, forward GGG CTA CCA TGC CAA CTT CTG,reverse GAG GGC AAG GAC CTT GCT GTA, probe CCT GCCCCT ACA TTT GGA GCC TGG A); MCP-1( CCL2, NM_031530forward TGT CTC AGC CAG ATG CAG TTA AT, reverse CCGACT CAT TGG GAT CAT CTT, probe CCC CAC TCA CCT GCTGCT ACT CAT TCA). Gene expression data were subsequentlystandardized for 36B4 mRNA, which was quantified in separateruns.

Statistics. Data are expressed as means 6 SD. Differences betweentwo age-matched groups were assessed by Mann-Whitney U-test.Survival analysis was investigated by the Kaplan-Meier method, and

the difference between two groups was examined by the log-rank test.The level of significance was set at P , 0.05. Analyses wereperformed using SPSS for Windows software (SPSS, Chicago, IL).

RESULTS

Neonatal DEX administration leads to reduced life span andsevere kidney disease in surviving rats at 50 wk of age. Ratsneonatally exposed to DEX showed a significantly increaseddeath rate, with a survival percentage at 50 wk of only 83%,and all deceased rats had severe kidney disease. No singledeath was observed in the SAL-treated group at 50 wk(Fig. 1).

Compared with the SAL group, DEX-treated rats that sur-vived up to 50 wk of age showed a significantly reduced bloodglucose concentration. Serum Cr concentration, volume of24-hr urine, and urinary protein/Cr ratio were all significantlyelevated in 50-wk-old rats treated with DEX (Table 1).

Fig. 3. Effects of neonatal DEX administration on bodyweight and relative kidney weight. A: neonatal DEX admin-istration led to significant growth retardation during theexperiment. Open symbols, SAL groups; closed symbols,DEX groups (n 5 6–35 per group every age). B: neonatalDEX treatment led to increased ratio of kidney weight tobody weight in rats of 2 and 7 days old; thereafter, this ratiowas reduced from 14 days to 24 wk of age compared withage-matched controls. Open bars, SAL groups; filled bars,DEX groups. Data are expressed as means 6 SD; n 5 6 pergroup at every age. *P , 0.05 DEX vs. age-matched SALby Mann-Whitney U-test.

Fig. 2. Renal PAS staining of 50-wk-old rats. At 50 wk of age, normal kidney structure was found in SAL-treated animals (A and C), whereas severe renaldamage was present in DEX-treated animals (B and D). Arrow, glomerular sclerosis; n 5 6 per group. Magnification: A and B, 3400; C and D, 3100.

F770 NEONATAL DEXAMETHASONE TREATMENT CAUSES RENAL DAMAGE


by g

uest o

n S

epte

mber 1

2, 2

01

2http

://ajp

renal.p

hysio

logy.o

rg/

Do

wn

loa

de

d fro

m

All surviving 50-wk-old rats that had received DEX treat-ment presented with renal disease, and PAS staining indicatedsevere nephropathy with extensive scaring, glomerulosclerosis,and dilation of the tubular system. Accumulation of lymphocytesand macrophages indicated inflammation (Fig, 2, B and D). Noovert renal pathological alterations were found in the SALcontrol group.

Neonatal DEX administration leads to body and kidneygrowth retardation. Subsequently, a time course study wasinitiated to allow identification of primary events in the cascadeleading to kidney failure. Growth retardation was persistentduring the entire experiment in DEX-treated animals comparedwith age-matched controls. No catch-up growth was observedin this study up to 32 wk of age (Fig. 3A).

Compared with SAL-treated rats, a significantly higher ratioof average kidney weight vs. body weight was found in 2-day-and 7-day-old rats upon DEX treatment (P , 0.05). Thereafter,this ratio was either similar or significantly lower in DEX-treated rats (Fig. 3B).

Blood pressure increases after neonatal DEX administra-tion. SBP of SAL- and DEX-treated rats slowly increased from8 wk to 32 wk of age. Compared with controls, a significantlyincreased SBP was observed in DEX-treated rats from 8 wkonwards (Fig. 4).

Serum Cr and proteinuria. No differences in serum Crconcentrations between SAL- and DEX-treated rats werenoted up to 32 wk (Table 2). Compared with the SAL group,a slight increase in 24-h urine volume was noted from 12 wkonward upon DEX treatment, which was statistically signif-icant at 20 wk (data not shown). Urinary protein/Cr ratio isa sensitive marker for protein loss in urine (10). Comparedwith SAL treatment, neonatal DEX treatment led to a highlysignificant increase in urine protein/Cr ratio from 8 wk ofage onward, reaching a fourfold difference between thegroups at week 32 (Fig. 5).

Pathological changes in kidneys induced by neonatal DEXexposure. PAS staining indicated no pathological alterations inkidneys of 2 day-, 14 day-, and 8-wk-old animals of SAL- andDEX-treated groups. Glomerular sclerosis was found in 32-wk-old rats upon DEX treatment (Fig. 6F). In contrast to SAL-treated rats, increased tubulointerstitial staining of a-SMA wasobserved in rats treated with DEX at 32 wk, indicating myo-

fibroblast transformation, an early indication of kidney fibrosis(Fig. 7, A and B).

Nephron numbers were found slightly but significantly re-duced in rats both at 8 and 24 wk with DEX treatmentcompared with age-matched controls (Fig. 8, A and B).

Numbers of macrophages in glomeruli were increased at 8wks and 32 wk of age in rats treated with DEX (see Fig. 10A),which matched the time course of increased urinary proteinloss. Compared with SAL-treated rats, the total number ofmacrophages in the tubulointerstitium was significantly re-duced in DEX-treated 2-day-old rats (Fig. 9, A and B andFig. 10B). Significant and persistent increases in the numbersof tubulointerstitial ED1-positive macrophages were observedfrom day onward in DEX-treated rats compared with age-matched controls (Fig. 9, C–F and Fig. 10B).

Time-dependent changes in renal gene expression upon neo-natal DEX treatment. Compared with SAL-treated animals,renal gene expression of TNF-a was significantly suppressedin DEX-treated 2-day-old rats, with a subsequent increase at 7days and 4 wks of age (Fig. 11A). Significantly increased renalgene expression levels of MCP-1 were noted in DEX-treatedanimals at 7 days, 14 days, and 32 wk of age compared with theage-matched controls (Fig. 11B).

Compared with SAL-treated rats, TGF-b gene expressionwas significantly suppressed in the DEX group at 2 and 7 daysof age. Thereafter, a slight but significantly increased TGF-bgene expression were noted from 4 wk to 32 wk of age in ratsupon DEX treatment, which may promote progressive kidneyfibrosis (Fig. 11C).

Fig. 4. Neonatal DEX administration results in increased systolic bloodpressure (SBP). DEX treatment led to significantly elevated SBP from 8 to 32wk of age vs. age-matched controls. Open symbols, SAL groups; closedsymbols, DEX groups. Data are expressed as means 6 SD; n 5 6–29 pergroup every age. *P , 0.05 DEX vs. age-matched SAL by Mann-WhitneyU-test.

Fig. 5. Neonatal DEX administration increases urinary protein-to-creatine(Pr/Cr) ratio from 8 wk of age. Compared with SAL-treated rats, DEXadministration in newborn rats led to significantly increased urinary Pr/Cr ratiofrom 8 to 32 wk of age, indicating impaired kidney function. Open symbols,SAL groups; closed symbols, DEX groups. Data are expressed as means 6SD; n 5 6 per group every age. *P , 0.05 DEX vs. SAL by Mann-WhitneyU-test.

Table 2. Concentration of serum Cr in rats neonatallytreated with SAL or DEX

Age, wk SAL DEX

48

16 56.8610.4 57.8610.424 61.869.5 57.265.332 62.767.2 62.767.2

Data are expressed as means 6 SD; n 5 6 per treatment group every age.*P , 0.05 DEX vs. SAL by Mann-Whitney-U test.



by g

uest o

n S

epte

mber 1

2, 2

01

2http

://ajp

renal.p

hysio

logy.o

rg/

Do

wn

loa

de

d fro

m

Transient changes in renal pJNK immunostaining uponneonatal DEX treatment. Activated JNK (pJNK) was found ina number of tubular epithelial cells of both SAL- and DEX-treated rats. In 2-day-old rats, neonatal DEX administration

resulted in significantly decreased pJNK-positive nuclei num-bers compared with SAL-treated animals (Fig. 12, A–C). At 7days of age, on the other hand, significantly increased numbersof pJNK-positive nuclei were noted in rats treated with DEXcompared with same-age controls. At 14 days, there was nolonger any difference between the two groups (Fig. 12C).

Fig. 8. Nephron numbers. Estimation of nephron numbers were performed inPAS-stained slides from rats of 8 wk (A) and 24 wk old (B). Total numbers ofglomeruli were counted under 3100 magnification, and 20 fields per animalwere scored. Open bars, SAL groups; filled bars, DEX groups. Data areexpressed as means 6 SD; n 5 5–6 rats per group every age. *P , 0.05 DEXvs. SAL by Mann-Whitney U-test.

Fig. 6. Renal histology at 2 days, 8 wk, and 32 wk. Representative images of PAS staining show normal kidney structure of SAL-treated rats at age of 2 days,8 wk, and 32 wk (A, C, and E). DEX treatment does not result in apparent structural renal damage in rats aged 2 days or 8 wk (B and D). However, glomerularsclerosis was observed at 32 wk in DEX-treated rats (F); n 5 6 per group every age. Magnification: A–F, 3400. Arrow, glomerular sclerosis.

Fig. 7. Renal a-SMA staining of 32 wk of age. Increased interstitial a-smoothmuscle cell actin (a-SMA) staining (brown precipitate) was observed at 32 wkof age with DEX treatment (B) in contrast to age-matched controls (A, SALgroup); n 5 6 per group. Magnification, 3400.



by g

uest o

n S

epte

mber 1

2, 2

01

2http

://ajp

renal.p

hysio

logy.o

rg/

Do

wn

loa

de

d fro

m

DISCUSSION

In the present study, neonatal DEX administration wasshown to result in severe and progressive kidney disease inrats, likely contributing to the reduced life span observed underthese conditions. A time course study demonstrated that neo-natal DEX treatment led to persistent growth retardation andreduced kidney weight. Increased blood pressure and protein-uria were observed from 8 wk of age onward, leading toextensive renal fibrosis and glomerular sclerosis at 32 wk ofage. An early renal inflammatory response was observed in ratpups, after withdrawal of DEX, that may have triggered thesubsequent persistent and progressive renal fibrotic processthat ultimately led to progressive renal impairment.

While this paper was being prepared, a study in a similaranimal model of neonatal DEX exposure was published show-ing that neonatal DEX administration is associated with pre-mature death and kidney failure (22), which is consistent withour findings described in the present paper. Furthermore, Ortizet al. (29) show that there are specific times of DEX adminis-tration that seem to have an adverse effect. We found that DEXadministration shortly after birth might lead to reduced glo-merular numbers later in life. The study by Ortiz et al. was onthe maternal situation, the dosage of DEX was relatively lower,especially the real dose of DEX that could reach the fetus. Yetour data could be in line with their finding that apparently thereare time windows critical for nephrogenesis, which makes thefetus more sensitive to stimuli or inhibitors in that particularperiod. The first week after birth might be a crucial time periodfor renal nephrogenesis.

As described previously (26), postnatal DEX administra-tion in rats results in persistent growth retardation, whichmight be due to the suppression of growth hormones and ofthe insulin-like growth factor axis (18, 34). A significantlyincreased ratio of kidney vs. body weight was observed with

DEX at 2 and 7 days, which might be largely due to waterretention since this ratio was increased acutely and nomorphological hypertrophy was observed at those ages. Thesubsequent persistently lower relative kidney weight mayindicate that neonatal DEX administration permanently in-hibited kidney growth or may reflect the process of fibroticscarring.

DEX administration led to a significantly increased systolicblood pressure from 8 wk of age onward. This is in accord witha recent study using the same model showing that increasedblood pressure was present at 3 and 11 mo of age (22). Themolecular mechanism underlying the “programmed” hyperten-sion induced by early GC exposure is still obscure. It has beenspeculated that altered renal angiotensin system activity, re-duced nephron numbers, and an overactive brain angiotensionsystem might all participate in the pathogenesis of hyperten-sion induced by prenatal DEX exposure (16). In rats, thekidney is still immature at birth, and nephrogenesis is notcomplete until at least 1 wk after birth (25). Although themethod of counting nephron numbers in the present study waslimited compared with previous methods described (5, 12), it islikely that neonatal DEX administration in the first 3 days afterbirth leads to reduced nephron numbers in adult life by inhi-bition of nephronegenesis, which has also been found inprenatal studies (14, 17). In addition, in our study, DEX-treatedrats had increased urinary protein loss from 8 wk onward. Wespeculate that reduced nephron number and /or renal damagemay both be involved in increased blood pressure in adulthood.It should be noted that SAL-treated rats in the current study hadhigher blood pressure levels compared with data from theliterature (11), which might be due to increased endogenousGC levels at the neonatal stage induced by stress of handling.

In the early phase of renal disease, in general there is aninflux of inflammatory cells, i.e., macrophages, into the kidney

Fig. 9. Renal macrophage accumulation. Representative images of renal immunohistochemistry for macrophage marker ED1. A and B show less interstitialED11 cells in 2-day-old rats with DEX treatment vs. age-matched SAL group; A, SAL group 2 days; B, DEX group 2 days. Increased numbers of interstitialED11 cells were observed in DEX group at 14 days in contrast to control; C, SAL group, 14 days; D, DEX group, 14 days. ED11 cells were scarce in 32-wk-oldSAL-treated rats (E). In contrast, large interstitial amounts of ED11 macrophages were found in rats with DEX treatment at 32 wk (F). Magnification A–F, 3400.Arrows, ED11 cells (group size n 5 6 at every age).



by g

uest o

n S

epte

mber 1

2, 2

01

2http

://ajp

renal.p

hysio

logy.o

rg/

Do

wn

loa

de

d fro

m

irrespective the initial insult (27). This process can be stimu-lated by proinflammatory factors such as TNF-a and chemo-kines like MCP-1. Recent data suggest that TNF-a plays acrucial role in development of kidney disease (23) by inducingmacrophage accumulation and fibrogenesis. In our study, weobserved suppression of TNF-a gene expression at 2 days ofage, i.e., during the DEX administration, which is most likelydue to the anti-inflammatory effects of GCs. Remarkably, afterwithdrawal of GCs, we observed markedly increased renalTNF-a gene expression at 7 days and 4 wks of age. This isconsistent with an inflammatory response that could potentiallybe involved as a trigger for the onset of renal damage that weobserved subsequently.

MCP-1 is expressed at sites of injury and inflammation todirect macrophage recruitment. It has been suggested thatMCP-1 is predominantly expressed by tubular epithelial cellsand not by glomeruli and promotes tubular epithelial cells andnot glomerular damage (31, 35). We found increased kidneygene expression of MCP-1 in 7- and 14-day-old rats after DEXtreatment. In the present study, stress MCP-1 gene expressionmight be a crucial factor in the development of progressive

kidney damage in later life by its stimulation of inflammatorypathways during a pivotal stage of kidney development.

To further address the cellular mechanisms involved in thisinflammatory response, we evaluated the activation state of theJNK pathway, which has been suggested to modulate MCP-1expression (2, 13, 38, 39). In the present experiment, we foundthat the extent of renal JNK activation (as determined byexpression of pJNK) was significantly reduced at 2 days of ageupon DEX treatment. This seems also reasonable since previ-ous studies have indicated that GCs may inhibit activation ofthe JNK pathway (8, 9). Remarkably, we found significantlyincreased pJNK in 7-day-old rats after DEX withdrawal, con-sistent with rebound inflammatory effects. Our studies thussuggest involvement of the JNK pathway in DEX-inducedrenal injury, although further studies including interventions

Fig. 10. ED11 cell numbers in tubulointerstitium and glomeruli. A: comparedwith SAL-treated rats, markedly increased glomerular ED11 cells wereobserved from 8 to 32 wk of age in DEX-treated animals; n 5 3–6 per groupevery age. Open bars, SAL groups; filled bars, DEX groups. Data areexpressed as means 6 SD. *P , 0.05 DEX vs. SAL by Mann-Whitney U-test.B: tubulointerstitial ED11 cell numbers were reduced in 2-day-old DEX-treated rats vs. age-matched controls, whereas markedly increased tubuloin-terstitial ED11 cells were observed from 14 wk of age and persisted thereafterduring the experiment; n 5 3 per group every age. Open bars, SAL groups;filled bars, DEX groups.

Fig. 11. Effects of neonatal DEX administration on renal gene expression ofTNF-a, monocyte chemotactic protein-1 (MCP-1), and TGF-b. A: Comparedwith SAL-treated control, neonatal DEX administration led to significantlydecreased TNF-a gene expression at 2 days of age and significantly increasedgene expression at 7 days and at 4 wk of age separately. B: compared withage-matched SAL-treated rats, neonatal DEX treatment resulted in significantincrease in MCP-1 gene expression at 7 days, 14 days, and 32 wk of age.C: compared with age-matched controls, neonatal DEX administration resultedin significantly reduced TGF-b gene expression at 2 and 7 days; thereafter, thisgene is significantly and persistently expressed in DEX-treated rats from 4 to32 wk of age. Open bars, SAL groups; filled bars, DEX groups. Geneexpression data were normalized by 36B4 mRNA analyzed in separate runs.Data are expressed as means 6 SD; n 5 5–6 per group every age. *P , 0.05DEX vs. age-matched SAL control by Mann-Whitney U-test.



by g

uest o

n S

epte

mber 1

2, 2

01

2http

://ajp

renal.p

hysio

logy.o

rg/

Do

wn

loa

de

d fro

m

with specific JNK inhibitors are required to further investigateits role.

GCs are widely used in immunosuppressive therapy ininflammatory kidney diseases with well-established inhibitory

effects on macrophages recruitment (19). This is consistentwith our finding of reduced numbers of macrophages in thetubulointersitium at 2 days of age during DEX treatment.Surprisingly, we found persistently increased macrophagenumbers in tubulointerstitium from 14 days of age onward,which strongly suggests the existence of life-long renal inflam-matory reactions after postnatal DEX treatment. Our data alsosuggest that macrophage accumulation in tubulointerstitiumand not in glomeruli participates predominantly in develop-ment of early kidney damage. In recent years, it has becomewidely accepted that interstitial inflammation plays a centralrole in progression to end-stage renal failure. In this process,interstitial macrophages are involved in both the initiation andcontinuation of the inflammatory response (32).

Fibrosis is characteristic of progressive renal damage. Ele-vated a-SMA expression has been regarded as an importantearly marker for renal fibrosis (3). Neonatal DEX treatmentresulted in increased a-SMA protein levels at 32 wk of age,i.e., when renal damage is extensive. TGF-b is a key regulatorymolecule in the control of the activity of fibroblasts and hasbeen implicated in several disease states characterized byexcessive fibrosis (24). In the present study DEX adminis-tration suppressed TGF-b gene expression in 2-day-old rats,which is consistent with previous findings in rat hepaticstellate cells in vitro (7). Although reduced TGF-b expres-sion slowly restored during the first 2 wk of life in DEX-treated animals, it is highly expressed in the rest of life afterweaning, contributing to a persistent profibrotic environ-ment in the kidney.

GCs are widely used for their anti-inflammatory effectsfor many different disease conditions. However, our dataindicate that the transient inhibition of renal inflammatoryparameters during DEX in neonatal rats is followed by arebound of proinflammatory factors, leading to a permanentincrease in interstitial macrophage accumulation after with-drawal, followed by progressive renal damage in later life.Therefore, early lifetime GC administration, especially dur-ing the crucial period of nephrogenesis, should not be takenlightly.

So far, unfortunately, human data are lacking. Since a hugenumber of newborns have been treated with DEX and otherGCs from the 1980s on, however, one might consider assessingblood pressure, proteinuria, and kidney function in this popu-lation to establish whether the GCs might have elicited dele-terious long-term effects.

In conclusion, data provided in this study suggest thatneonatal DEX administration leads to end-stage renal diseasein rats in later life. The accumulation of inflammatory factorsin the tubulointerstitium induced by DEX at early-life agemight participate in kidney function impairment, fibrosis, andkidney failure. Furthermore, although it is dangerous to extrap-olate from animal experiment to human situations, we proposethat follow-up studies on kidney functions in humans thatreceived neonatal DEX are necessary.

ACKNOWLEDGMENTS

We thank Fjodor van der Sluijs for primer design and Marian Bulthuis for

technical assistance.

Fig. 12. Renal JNK activation following DEX administration. Neonatal DEXadministration results in reduced numbers of renal phosphorylated (p)JNK-positive nuclei in 2-day-old rats vs. SAL-treated control. A and B: represen-tative images of pJNK immunohistochemistry in 2-day-old SAL (A) and DEX(B)-treated rats; n 5 6 per treatment group. Magnification, 3400. Arrows,pJNK-positive nuclei. C: neonatal DEX treatment significantly reduced num-bers of pJNK-positive nuclei in 2-day-old rats and significantly increasedpJNK-positive nuclei in 7-day-old rats vs. age-matched controls. No differencein pJNK-positive nuclei numbers was found at 14 days of age. Open bars, SALgroups; filled bars, DEX groups. Data are expressed as means 6 SD; n 5 6 pergroup every age. *P , 0.05 DEX vs. age-matched SAL control by Mann-Whitney U-test.



by g

uest o

n S

epte

mber 1

2, 2

01

2http

://ajp

renal.p

hysio

logy.o

rg/

Do

wn

loa

de

d fro

m

REFERENCES

1. Committee on Fetus and Newborn Infants. Postnatal corticosteroids totreat or prevent chronic lung disease in preterm infants. Pediatrics 109:330–338, 2002.

2. Arndt PG, Suzuki N, Avdi NJ, Malcolm KC, Worthen GS. Lipopo-lysaccharide-induced c-Jun NH2-terminal kinase activation in humanneutrophils: role of phosphatidylinositol 3-kinase and Syk-mediated path-ways. J Biol Chem 279: 10883–10891, 2004.

3. Badid C, Vincent M, Fouque D, Laville M, Desmouliere A. Myofibro-blast: a prognostic marker and target cell in progressive renal disease. Ren

Fail 23: 543–549, 2001.4. Bal MP, de Vries WB, van der Leij FR, van Oosterhout MF, Berger

RM, Baan J, van der Wall EE, van Bel F, Steendijk P. Neonatalglucocorticosteroid treatment causes systolic dysfunction and compensa-tory dilation in early life: studies in 4-week-old prepubertal rats. Pediatr

Res 58: 46–52, 2005.5. Bertram JF. Analyzing renal glomeruli with the new stereology. Int Rev

Cytol 161: 111–172, 1995.7. Bolkenius U, Hahn D, Gressner AM, Breitkopf K, Dooley S, Wickert

L. Glucocorticoids decrease the bioavailability of TGF-beta which leads toa reduced TGF-beta signaling in hepatic stellate cells. Biochem Biophys

Res Commun 325: 1264–1270, 2004.8. Bruna A, Nicolas M, Munoz A, Kyriakis JM, Caelles C. Glucocorticoid

receptor-JNK interaction mediates inhibition of the JNK pathway byglucocorticoids. EMBO J 22: 6035–6044, 2003.

9. Caelles C, Gonzalez-Sancho JM, Munoz A. Nuclear hormone receptorantagonism with AP-1 by inhibition of the JNK pathway. Genes Dev 11:3351–3364, 1997.

10. Carroll MF, Temte JL. Proteinuria in adults: a diagnostic approach. Am

Fam Physician 62: 1333–1340, 2000.11. Coelho MS, Passadore MD, Gasparetti AL, Bibancos T, Prada PO,

Furukawa LL, Furukawa LN, Fukui RT, Casarini DE, Saad MJ, Luz

J, Chiavegatto S, Dolnikoff MS, Heimann JC. High- or low-salt dietfrom weaning to adulthood: effect on body weight, food intake and energybalance in rats. Nutr Metab Cardiovasc Dis 16: 148–155, 2006.

11a.Cranefield DJ, Odd DE, Harding JE, Teele RL. High incidence ofnephrocalcinosis in extremely preterm infants treated with dexamethasone.Pediatr Radiol 34: 138–142, 2004.

12. Cullen-McEwen LA, Kett MM, Dowling J, Anderson WP, Bertram

JF. Nephron number, renal function, and arterial pressure in aged GDNFheterozygous mice. Hypertension 41: 335–340, 2003.

13. de Borst MH, Prakash J, Melenhorst WB, van den Heuvel MC, Kok

RJ, Navis G, van Goor H. Glomerular and tubular induction of thetranscription factor c-Jun in human renal disease. J Pathol 213: 219–228,2007.

14. Dickinson H, Walker DW, Wintour EM, Moritz K. Maternal dexa-methasone treatment at midgestation reduces nephron number and altersrenal gene expression in the fetal spiny mouse. Am J Physiol Regul Integr

Comp Physiol 292: R453–R461, 2007.15. Dodic M, May CN, Wintour EM, Coghlan JP. An early prenatal

exposure to excess glucocorticoid leads to hypertensive offspring in sheep.Clin Sci (Lond) 94: 149–155, 1998.

16. Dodic M, Moritz K, Koukoulas I, Wintour EM. Programmed hyper-tension: kidney, brain or both? Trends Endocrinol Metab 13: 403–408,2002.

17. Figueroa JP, Rose JC, Massmann GA, Zhang J, Acuna G. Alterationsin fetal kidney development and elevations in arterial blood pressure inyoung adult sheep after clinical doses of antenatal glucocorticoids. Pediatr

Res 58: 510–515, 2005.18. Huysman MW, Hokken-Koelega AC, Hop WC, Sauer PJ. Effect of

dexamethasone treatment on serum GH, IGF-I, and the binding proteinsIGFBP-1 and -3 in ventilated very preterm infants. Pediatr Res 54: 37–43,2003.

19. Ikezumi Y, Atkins RC, Nikolic-Paterson DJ. Interferon-gamma aug-ments acute macrophage-mediated renal injury via a glucocorticoid-sen-sitive mechanism. J Am Soc Nephrol 14: 888–898, 2003.

20. Kamitsuka MD, Williams MA, Nyberg DA, Fox KA, Lee DL, Hickok

D. Renal calcification: a complication of dexamethasone therapy in pre-

term infants with bronchopulmonary dysplasia. J Perinatol 15: 359–363,1995.

21. Kamphuis PJ, Croiset G, Bakker JM, Van Bel F, Van Ree JM,

Wiegant VM. Neonatal dexamethasone treatment affects social behaviourof rats in later life. Neuropharmacology 47: 461–474, 2004.

22. Kamphuis PJ, de Vries WB, Bakker JM, Kavelaars A, van Dijk JE,

Schipper ME, van Oosterhout MF, Croiset G, Heijnen CJ, van Bel F,

Wiegant VM. Reduced life expectancy in rats after neonatal dexametha-sone treatment. Pediatr Res 61: 72–76, 2007.

23. Khan SB, Cook HT, Bhangal G, Smith J, Tam FW, Pusey CD.

Antibody blockade of TNF-alpha reduces inflammation and scarring inexperimental crescentic glomerulonephritis. Kidney Int 67: 1812–1820,2005.

24. Liu Y. Renal fibrosis: new insights into the pathogenesis and therapeutics.Kidney Int 69: 213–217, 2006.

25. Moritz KM, Wintour EM. Functional development of the meso- andmetanephros. Pediatr Nephrol 13: 171–178, 1999.

26. Neal CR Jr, Weidemann G, Kabbaj M, Vazquez DM. Effect ofneonatal dexamethasone exposure on growth and neurological develop-ment. Am J Physiol Regul Integr Comp Physiol 287: R375–R385, 2004.

27. Noronha IL, Fujihara CK, Zatz R. The inflammatory component inprogressive renal disease–are interventions possible? Nephrol Dial Trans-

plant 17: 363–368, 2002.28. Nyirenda MJ, Welberg LA, Seckl JR. Programming hyperglycaemia in

the rat through prenatal exposure to glucocorticoids-fetal effect or mater-nal influence? J Endocrinol 170: 653–660, 2001.

29. Ortiz LA, Quan A, Zarzar F, Weinberg A, Baum M. Prenatal dexa-methasone programs hypertension and renal injury in the rat. Hypertension

41: 328–334, 2003.30. Plosch T, van der Veen JN, Havinga R, Huijkman NC, Bloks VW,

Kuipers F. Abcg5/Abcg8-independent pathways contribute to hepatobili-ary cholesterol secretion in mice. Am J Physiol Gastrointest Liver Physiol

291: G414–G423, 2006.31. Prodjosudjadi W, Gerritsma JS, Klar-Mohamad N, Gerritsen AF,

Bruijn JA, Daha MR, van Es LA. Production and cytokine-mediatedregulation of monocyte chemoattractant protein-1 by human proximaltubular epithelial cells. Kidney Int 48: 1477–1486, 1995.

32. Sean Eardley K, Cockwell P. Macrophages and progressive tubulointer-stitial disease. Kidney Int 68: 437–455, 2005.

33. Slotkin TA, Seidler FJ, Kavlock RJ, Bartolome JV. Fetal dexametha-sone exposure impairs cellular development in neonatal rat heart andkidney: effects on DNA and protein in whole tissues. Teratology 43:301–306, 1991.

34. Smink JJ, Koedam JA, Koster JG, van Buul-Offers SC. Dexametha-sone-induced growth inhibition of porcine growth plate chondrocytes isaccompanied by changes in levels of IGF axis components. J Endocrinol

174: 343–352, 2002.35. Tesch GH, Schwarting A, Kinoshita K, Lan HY, Rollins BJ, Kelley

VR. Monocyte chemoattractant protein-1 promotes macrophage-mediatedtubular injury, but not glomerular injury, in nephrotoxic serum nephritis.J Clin Invest 103: 73–80, 1999.

36. van der Veen JN, Kruit JK, Havinga R, Baller JF, Chimini G,

Lestavel S, Staels B, Groot PH, Groen AK, Kuipers F. Reducedcholesterol absorption upon PPARdelta activation coincides with de-creased intestinal expression of NPC1L1. J Lipid Res 46: 526–534, 2005.

37. Waanders F, Greven WL, Baynes JW, Thorpe SR, Kramer AB, Nagai

R, Sakata N, van Goor H, Navis G. Renal accumulation of pentosidinein non-diabetic proteinuria-induced renal damage in rats. Nephrol Dial

Transplant 20: 2060–2070, 2005.38. Wang N, Verna L, Hardy S, Forsayeth J, Zhu Y, Stemerman MB.

Adenovirus-mediated overexpression of c-Jun and c-Fos induces intercel-lular adhesion molecule-1 and monocyte chemoattractant protein-1 inhuman endothelial cells. Arterioscler Thromb Vasc Biol 19: 2078–2084,1999.

39. Xiao J, Chodosh J. JNK regulates MCP-1 expression in adenovirus type19-infected human corneal fibroblasts. Invest Ophthalmol Vis Sci 46:3777–3782, 2005.



by g

uest o

n S

epte

mber 1

2, 2

01

2http

://ajp

renal.p

hysio

logy.o

rg/

Do

wn

loa

de

d fro

m

Neonatal dexamethasone administration causes progressive renal damage due to induction of an early...

Documents

Transcript of Neonatal dexamethasone administration causes progressive renal damage due to induction of an early...