Kidney International, Vol. 41(1992), pp. 796—804

Renin and angiotensinogen gene expression in experimentaldiabetes mellitus

RICARDO CORREA-ROTTER, THOMAS H. HOSTETTER, and MARK E. ROSENBERG

Division of Renal Diseases and Hypertension, Department of Medicine, University of Minnesota, Minneapolis, Minnesota, USA

Renin and angiotensinogen gene expression in experimental diabetesmellitus. The renin-angiotensin system may play a role in the initiationand progression of diabetic kidney disease. In this study, the localintrarenal renin-angiotensin system was examined in streptozotocin-treated rats maintained moderately hyperglycemic by daily low-doseinsulin injection. Four weeks after induction of diabetes, plasma reninactivity was significantly lower in the diabetic compared to a non-diabetic control group (diabetes: 2.30 0.30 vs. control: 6.93 1.36ngAl/ml/hr; P < 0.01). Renal tissue renin content (diabetes: 1.81 0.46vs. control: 2.05 0.27 Lg Al/mg protein/hr; P < 0.05) and renal reninmRNA (diabetes: 2.32 0.16 vs. control: 1.89 0.12 pg/Lg RNA; P =NS) were not different between diabetic and control rats. Renal andliver angiotensinogen mRNA were lower in the diabetic group. Glomer-ular renin mRNA was not different between the diabetic and shamgroup. The dissociation between systemic renin activity (a decrease),and in renal renin content or mRNA in the diabetic rats (no change),suggests a post-translational alteration in renin processing and/or reninsecretion.

Diabetic nephropathy constitutes a major cause of morbidityand mortality among diabetic patients [1, 2]. A number offactors participate in the development of microangiopathicchanges at the renal and systemic level, including alterations inglucose metabolism and changes in systemic and local levelsand responsiveness to vasoactive hormones [3—81. Renal herno-dynamic alterations play an important pathogenetic role in theinitiation and progression of diabetic nephropathy [4, 9—12].The renin-angiotensin system has an important role in thephysiologic regulation of the renal microcirculation. In conjunc-tion with other vasoactive systems, the renin-angiotensin sys-tem may contribute to the imbalance of resistances present atthe preglomerular and postglomerular sites which are responsi-ble for glomerular capillary hypertension, a major injuriousfactor in the diabetic kidney [9—14].

The relationship between experimental and clinical diabetesand the renin-angiotensin system is unclear and the availablereports vary widely [15—24]. In general, plasma renin activity(PRA) has been reported to be low in diabetic nephropathy[20—22], although some investigators have found elevations inPRA in diabetic patients with glomerular hyperflltration [23,24]. Independent from its systemic counterpart, a locally ex-pressed and regulated renin-angiotensin system in the kidneymay be involved in the control of renal function [25].

© 1992 by the International Society of Nephrology

The purpose of the present study was to examine the intra-renal renin-angiotensin system in streptozotocin-induced exper-imental diabetes. We measured PRA, tissue renin content(TRC) as well as renin and angiotensinogen mRNA levels inmoderately hyperglycemic diabetic rats. As well, we examinedthe intrarenal distribution of renin and angiotensinogen mRNAsin the diabetic rat.

Methods

Experimental designExperiment 1. Thirteen adult male Sprague-Dawley rats,

weighing 250 to 275 g, were administered either streptozotocin(STZ) in citrate buffer 65 mg/kg intravenously (i.v.) (SigmaChemical Co., St. Louis, Missouri, USA) (diabetic group, N =8) or citrate buffer i.v. (control group, N = 5). Two days laterblood glucose levels were determined and rats with bloodglucose of greater than 300 mg/dl were considered diabetic.Ultralente Iletin I U-100 insulin (Eli Lilly, Indianapolis, Indi-ana, USA) was injected daily to all diabetic rats at a dose of 1.4units subcutaneously (5 to 6 units per kg). Rats were allowedfree access to water and standard rat chow (Ralston Purina, St.Louis, Missouri, USA). Total body weight and blood glucose,measured between 9 and 10 a.m., were recorded every week.Forty-eight-hour metabolic studies were performed at two andfour weeks to assess food intake and urinary protein, ketones,sodium, and potassium excretion. At the end of the study (day30) tail vein blood was obtained for measurements of serumcreatinine and plasma renin activity (PRA). The rats were thensacrificed, kidneys were removed, snap-frozen in liquid nitro-gen, and stored at —70°C for subsequent RNA extraction.

Experiment 2. This second experiment was performed todetermine PRA, tissue renin content (TRC), and intrarenaldistribution of renin and angiotensinogen mRNA in similarlytreated diabetic and control rats. Eleven male Sprague-Dawleyrats weighing 250 to 275 g were administered either STZ incitrate buffer 65 mg/kg i.v. (Sigma) (diabetic group, N = 5) orcitrate buffer i.v. (control group, N = 6). Presence of diabetesas well as weekly measurements of body weight and bloodglucose were performed as described in the previous experi-ment. Insulin was also administered daily to the diabetic rats asdescribed. At the end of the study (day 30) tail vein blood wasobtained for PRA. The rats were then sacrificed, a portion ofthe liver was removed, snap frozen in liquid nitrogen and storedat —70°C for subsequent RNA extraction. The left kidney wasremoved, weighed, snap-frozen in liquid nitrogen, and stored at

796

Correa-R otter et a!: Renin gene expression in diabetes mellitus 797

—70°C for TRC determination. The right kidney was thenremoved, the kidney tissue was minced into small (I jm3)fragments in cold diethylpyrocarbonate-treated phosphate buff-ered saline (PBS) and passed sequentially through 250 xm, 150/zm, and then 75 jtm metal screens. The material was washedthrough the screens with PBS. Those fragments trapped on the75 im screen were employed for studies of glomeruli and thosefragments remaining on the 150 m screen were employed forstudies of tubules. Small elements of the vascular pole remainattached to approximately 7% of glomeruli. Glomeruli andtubules from rats of each group (diabetic group N = 5, controlgroup N = 6) were pooled to yield sufficient material for RNAextraction (see below).

RNA extraction and Northern blot hybridizationTotal RNA from the whole kidney or sieved glomeruli and

tubules was isolated using a modification of the guanidinium-isothiocyanate/cesium chloride method 1261. The RNA wasdissolved in sterile water and RNA concentrations determinedby absorbance readings at 260 nm. Aliquots (20 gig) of totalRNA were electrophoresed in a 1% agarose gel containing 20mM MOPS, 1 m'vi EDTA, 5 m'vi sodium acetate pH 7.0, and 2.2M formaldehyde, and transferred to nylon membranes (DuralonUVTM Stratagene, La Jolla, California, USA). In each gel,equivalent loading of RNA, absence of degradation, and theposition of the 28 S and 18 S ribosomal RNA was determined byethidium bromide staining. RNA was fixed to the nylon mem-brane by ultraviolet crosslinking (StratalinkerTM, Stratagene).The membranes were prehybridized at 60°C for four hours in abuffer containing 5x SSC, Sx Denhardt's reagent, 50 mMTris-hydrochioride, pH 7.5, 0.1% sodium pyrophospate, 0,2%SDS, 200 tg/ml sonicated, denatured salmon testes DNA, and100 ig/ml yeast tRNA. The membranes were then hybridized at42°C with oligomer-primer labelled cDNA probes (see below)for 16 to 18 hours in a buffer containing 50% deionized form-amide, 5x SSC, lx Denhardt's reagent, 50mM Tris-hydrochlo-ride, pH 7.5, 0.1% sodium pyrophosphate, 1% SDS, 100 Lg/mlsalmon testes DNA, and 100 ig/ml yeast tRNA. The mem-branes were washed for 45 minutes in 2x SSC and 0.1% SDStwice at room temperature and once at 60°C, and were thenwashed in 0.2x SSC and 0.1% SDS at 60°C for 45 minutes.Autoradiographs (Kodak XAR-5 film, Eastman Kodak Co.,Rochester, New York, USA) were obtained and quantified bycomputer assisted videodensitometry [271.

Solution hybridizationThe presence of a single band for renin mRNA (1600 bp) on

hybridization of the nylon filters, permitted quantitation ofrenin mRNA by solution hybridization based on a modificationof the technique of Durnam and Palmiter [281. This procedurewas performed to confirm results obtained by Northern analysisand to more quantitatively assess total kidney renin mRNA indiabetic and control rats. Total kidney RNA (25 to 60 g) washybridized for approximately 18 hours with 20,000 cpm of a35S-labelled renin cRNA at 80°C in 20 d of 0.6 M NaCl, 10mMTris-HC1, pH 6.5, 5 mrvi EDTA, 2.5% ethanol, and 0.1% SDS.The samples were digested with 1 ml RNase solution [RNase A(25 tg/ml), RNase TI (250 U/mI), 0.3 M NaCI, 10mM Tris-HC1,pH 7.5, 2 mrs'i EDTA, and 75 tg salmon testes DNAI to digestunhybridized probe, precipitated with trichloroacetic-acid, and

quantified by liquid scintillation counting. A standard curve wasconstructed and run with each assay using mRNA transcribedby SP6 RNA polymerase after linearization of the renin plasmidby Hind III, (2 to 10,000 pg).

cDNAs and preparation of the cDNA and cRNA probes

The following cDNA probes were used: rat renin (pRen44.ceb) [291 rat angiotensinogen (pRANG 6) [301, (cDNAs werea gift of K. R. Lynch), and mouse a-actin cDNA (pAM9l) [31].For preparation of the cDNA probes, we utilized the method ofrandom oligomer-primer labelling (Promega, Madison, Wiscon-sin, USA) with 32P-dCTP (6000 Ci/mmol; New England Nu-clear, Dupont, Wilmington, Delaware, USA). Only gel isolatedinsert eDNA was used in the labelling reactions. The specificactivity of the probes was 1 to 2 x io cpmIg DNA. Forpreparation of the cRNA probe, the renin plasmid was firstlinearized with the restriction enzyme BamilI. 250 Ci of 35S-UTP (1320 Ci/mmol; New England Nuclear) was evaporated todryness in a microfuge tube and the following added: 1 tg of thelinearized cDNA, 2.5 d water, 2.0 .d 5x buffer (200 mMTris-HCI, pH 7.5, 30 mi'i MgCl2, 10mM spermadine), 1.0 d T7RNA polymerase. The mixture was incubated for 60 minutes at37°C. The reaction mixture was DNase-treated and unincorpo-rated nucleotides removed by chromatography on a SephadexG50 column.

Biochemical methods

Serum creatinine was measured with a creatinine autoana-lyzer (Beckman Instruments, Brea, California, USA). Urinaryprotein was measured from aliquots of two consecutive 24-hoururine collections using the Coomassie blue dye method (Bio-Rad Laboratories, Richmond, Virginia, USA). Sodium andpotassium concentrations were measured also from aliquots oftwo consecutive 24-hour urine collections, using a flame pho-tometer.

PRA was determined at pH 6.0 by radioimmunoassay using acommercially available kit (New England Nuclear). For deter-mination of tissue renin content (TRC), the kidney aliquotswere kept at —70°C until their renin content was measured. Thekidney was homogenized in a buffer containing 2.6 mi EDTA,1.6 miv dimercaprol, 3.4 mtvi 8-OH quinoline sulfate, 0.2 mMPMSF, and 5 mM ammonium acetate, spun at 5000 rpm, and thesupernate was removed and frozen and thawed four times, aprocedure which does not activate prorenin [32]. TRC wasdetermined on an aliquot of supernatant after dilution to 1:1000by the quantitation of generated angiotensin I (Rianen AssaySystem, Du Pont Co., Billerica, Massachusetts, USA). Twohundred d of the sample was incubated for one hour with 100d of plasma obtained from nephrectomized male rats, 50 d 4%EDTA, 10 d dimercaprol, 10 d 8-HO quinoline sulfate, and 630d maleate buffer (pH 6.0). The protein concentration of thesupernatant was measured by the Coomassie blue dye method.TRC per mg of protein was determined by dividing tissue renincontent by the protein concentration of the aliquot of kidneyassayed.

Statistical analysis

Statistical significance is defined as P < 0.05 and the resultsare presented as mean standard error of the mean (±sEM).

798 Correa-R otter et al: Renin gene expression in diabetes mellitus

Day 0 1 2 3 4—2

Time, weeks

Fig. 1. A. Total body weight of diabetic (E) and control (•) rats at 0,1, 2, 3, and 4 weeks after injection of STZ or citrate buffer. *p < 0.01.B. Blood glucose of diabetic and control animals at 0, 1, 2, 3, and 4weeks. P < 0.01.

The significance of the differences was analyzed by the Stu-dent's t-test of unpaired data.

Results

Experiment IGeneral parameters and metabolic studies. As expected,

body weight in the diabetic rats decreased during the first twoweeks, then stabilized during the third week, and increasedthereafter. At all time points, control rats showed a significantlyhigher body weight than diabetic rats. (Fig. 1A). Diabetic ratswere moderately hyperglycemic throughout the study (Fig. 1B).Since blood glucose was only measured between 9 and 10 a.m.,it is possible that higher glucose levels occurred at other timesof the day. No urinary ketones were present in the 24-hoururine samples obtained at week 2 and week 4. Table I shows theresults of the 48-hour metabolic studies performed at two andfour weeks of the study. Food intake, urinary volume, urinaryNa and K excretion and urinary protein excretion were allsignificantly increased in the diabetic rats as compared tocontrol animals, at both time points (Table 1). The left kidneyweight was also significantly increased in the diabetic comparedto the control group when expressed as both absolute weight(diabetic: 1.35 0.03 vs. control: 1.17 0.01 g; P <0.001) andas kidney weight/body weight ratio (diabetic: 5.06 0.08 vs.control: 3.51 0.07 x i03; P < 0.001). Serum creatinine was

Table 1. Metabolic studies

Diabetes Control(N=8) (N=5)

Food intake g 2 weeks4 weeks

36.5 l.6a39.8 1.9'

20.5 0.818.4 1.1

Urinary protein 2 weeks 44.6 2.7a 18.3 1.8mg124 hr 4 weeks 41.8 46' 20.0 2.4

Urinary Na 2 weeks 3.1 0.3' 1.5 0.1mEq/24 hr 4 weeks 2.5 0.2' 1.6 0.1

Urinary K 2 weeks 5.7 0.3' 2.9 0.3,nEq/24 hr 4 weeks 5.7 0.4' 3.5 0.2

Values are mean SEM. Results of 48-hour metabolic studiesperformed at 2 and 4 weeks. Food intake, urinary protein, Na and Kexcretion were all significantly increased in the diabetic rats at bothtime points.

'P < 0.001

not different at the end of the study between the two groups(diabetic: 0.5 0.01 vs. control: 0.5 0.04 mg/dl; P NS).

Northern hybridization. Analysis of Northern blots con-firmed a single band for renal renin mRNA (1600 bp), and renaland liver angiotensinogen mRNA (1800 bp), and a predominantband for ct-actin mRNA (2200 bp). Figure 2 shows an autorad-iograph of a northern blot of total kidney RNA extracted fromdiabetic and control rats, which has been hybridized with arenin eDNA probe. Each lane corresponds to one rat. Quanti-tation by computer assisted videodensitometry demonstratedthat renin mRNA was not different between the diabetic and thecontrol group (diabetes: 435 45 vs. control: 432 36 O.D.units; P = NS; Fig. 2). The same Northern blot was hybridizedwith an angiotensinogen eDNA probe, after allowing sufficienttime for the initial radioactivity to decay. AngiotensinogenmRNA was significantly lower in the diabetic group (diabetes:294 22 vs. control: 508 53 O.D. units; P < 0.01; Fig. 3). Anidentical Northern blot was hybridized with an cs-actin eDNAprobe to serve as a control for the integrity of the RNA. a-ActinmRNA was not different between the two groups (diabetes:1286 96 vs. control: 1250 135 O.D. units; P = NS).

Figure 4 shows an autoradiograph of a Northern blot of totalliver RNA extracted from diabetic (N = 6) and control (N = 5)rats, hybridized with an angiotensinogen eDNA probe. Quan-titation by computer assisted videodensitometry demonstratedthat angiotensinogen mRNA was lower in the diabetic group(diabetes: 1199 60 vs. control: 2462 119 O.D. units; P <0.01). No difference in liver a-actin was detected between thediabetic and control group (57.9 7 vs. 54.9 6.4 O.D. units;P = NS).

Solution hybridization. Renin mRNA was quantitated bysolution hybridization in the diabetic and control rats. ReninmRNA was not different between the two groups (diabetes: 2.32

0.16 vs. control: 1,89 0.12 pg/pg RNA; P NS),confirming the results of northern analysis.

PRA measurement. PRA was significantly lower in the dia-betic rats (diabetes: 4.29 0.63 vs. control: 7.00 0.76 ngAl/ml/hr; P < 0.02).

Experiment 2Northern hybridization of glomeru/ar and tubular RNA.

Renin mRNA was detected in glomerular and tubular fractionsof both diabetic and control animals. Although renin-like activ-ity has been detected in proximal tubular cells [331, the renin

A

340

320

300-cC)

260

240

220

B

400

E 300200

0C.)

0

0 1 2 3 4

Time, weeks

ThOObp'

Th00bp

600

500

400

300

200

100

0

Diabetes Control

t0

C

da

600

400

200

0Diabetes ControlN=8 N=5

Diabetes Control

0tW

1/Diabetes

-Control

N=B N5

Correa-Rotter et al: Renin gene expression in diabetes mellitus 799

Fig. 2. Autoradiograph of Northern blot oftotal kidney RNA (20 p.g) extracted from eightdiabetic and five control rats. A 32P randomoligomer-primer labelled renin cDNA probewas used for hybridization. Quantitation byvideodensitometry showed that the relativerenin mRNA level was not different betweenthe two groups (P = NS).

Fig. 3. Autoradiograph of Northern blot oftotal kidney RNA (20 pg) extracted from eightdiabetic and five control rats and hybridizedwith an angiotensinogen cDNA probe.Angiotensinogen mRNA level wassignificantly lower in the diabetic rats (P <0.01).

mRNA present in the tubular fraction may represent arteriolarcontamination. Since the samples were pooled, no statisticalanalysis was possible but the respective values as determinedby videodensitometry were similar between the diabetic and thecontrol group (diabetic glomerular renin mRNA, 1042 vs.control glomerular renin mRNA, 974; diabetic tubular reninmRNA, 607 vs. control tubular renin mRNA, 904 O.D. units;

Fig. 5A). Angiotensinogen mRNA was detected in tubular RNAof both diabetic and control animals and was higher in thecontrol group (diabetic tubular angiotensinogen mRNA, 71 vs.control tubular angiotensinogen mRNA, 182 O.D. units; (Fig.5B). a-Actin mRNA was higher in glomerular than in thetubular samples in both diabetic and control rats. The levels ofcs-actin were similar for both glomerular and tubular fractions

Diabetes ControlI I

Anglo1800 bp

0)4-.C

0

2000

1000

0

•n'.

Diabetes ControlN=6 N=5

800 Correa-Rotter et a!: Renin gene expression in diabetes mellitus

Fig. 4. Autoradiograph of Northern blot oftotal liver RNA (20 aug) extracted from eightdiabetic and five control rats and hybridizedwith an angiotensinogen cDNA probe.Angiotensinogen mRNA level wassignificantly lower in the diabetic rats (P <0.01).

between the diabetic and the control group (diabetic glomerulara-actin mRNA, 1898 vs. control glomerular a-actin mRNA,1996; diabetic tubular a-actin mRNA, 737 vs. control tubulara-actin mRNA 926, O.D. units; Fig. 5C). Equivalent loadingand integrity of the RNA was determined by ethidium bromidestaining (Fig. 5D).

PRA and TRC measurement. Figure 6 shows the results ofPRA and TRC obtained in this experiment. PRA, as seen in theprevious experiment, was significantly lower in the diabetic rats(diabetes: 2.30 0.30 vs. control: 6.93 1.36 ng Al/ml/hr; P <0.01; Fig. 6A). In contrast, TRC was not different between thetwo groups (diabetes, 1.81 0.46 vs. control, 2.05 0.27 gAl/mg protein/hr; P < 0.05; Fig. 6B).

Discussion

The present study examined components of the systemic andlocal renin-angiotensin system four weeks after induction ofdiabetes by STZ in rats maintained moderately hyperglycemicwith exogenous insulin. In comparison to a non-diabetic controlgroup, PRA was significantly lower in the diabetic animals. Inthe kidney, TRC and renin mRNA (a marker of local reninsynthesis) were not different in the diabetic compared to thenon-diabetic animals. No increase in glomerular renin mRNAwas observed between the diabetic and control rats. Angioten-sinogen mRNA was lower in whole kidney, the renal tubularRNA fraction, and livers of the diabetic rats. Thus, no evidencewas found for activation of the local renal renin-angiotensin

system in diabetes, although some discrepancy between thesystemic and intrarenal systems was observed for renin, withdecreased PRA yet no fall in TRC or renal renin mRNA in thediabetic compared to the non-diabetic controls.

The renin-angiotensin system has an important role in theregulation of the renal microcirculation. Angiotensin II (AngII), a potent renal vasoconstrictor, has preferential effects onthe efferent arteriole [13, 14]. Therefore, increased action ofAng II would be manifested by increased efferent arteriolarconstriction with consequent elevation in glomerular capillarypressure. Glomerular hypertension has been implicated in theprogression of several glomerular diseases including diabeticnephropathy [4, 9—121. Marked renal hemodynamic changesaccompany experimental diabetic renal disease. In the moder-ately hyperglycemic rat, elevations in single nephron plasmaflow and glomerular pressure lead to elevations in glomerularfiltration rate [9]. These hemodynamic changes are secondaryto the imbalance of resistances at the pre- and postglomerularlevels with a greater dilatation occurring in the afferent com-pared to the efferent arteriole. Treatment of diabetic rats withangiotensin converting enzyme inhibitors (CEI) or Ang IIblockers, but not other antihypertensives, decreases glomerularpressure and is associated with amelioration of renal injury[10—12, 15]. The greater decrease in afferent compared toefferent resistance and the beneficial response to maneuverswhich decrease Ang II, suggest a role for the renin-angiotensin

A Renin mRNA B Angiotensinogen mRNA

Sham Diabetes Sham Diabetes

C a-Actin mRNA

Sham Diabetes

g t g

t g t

D EthBrstained gel

t g t g

t

Correa-R otter et al: Renin gene expression in diabetes mellitus 801

Fig. 5. A. Autoradiograph of Northern blot ofpooled samples of glomerular and tubularRNA (20 sag) extracted from diabetic (N = 5)and control rats (N = 6) and hybridized witha renin cDNA probe. B. Autoradiograph ofsame Northern blot, hybridized with anangiotensinogen cDNA probe. C.Autoradiograph of same Northern blot,hybridized with an s-actin cDNA probe. D.Photograph of ethidium bromide stainedNorthern gel of glomerular and tubular RNA.

system in the development and progression of diabetic renaldisease.

The local intrarenal renin-angiotensin system may participatein the regulation of renal function independent of its systemiccounterpart [25]. Preliminary studies examining the renin-angi-otensin system in the diabetic kidney have yielded conflictingresults. Anderson and coworkers reported increased renal renincontent and angiotensinogen mRNA in insulin-treated diabeticrats six to eight weeks following STZ-induced diabetes. In thatstudy, chronic Ang II blockade decreased glomerular pressureindependent of changes in systemic pressure [15]. Sajid andcollaborators examined the time course of renin and angioten-sinogen expression following induction of diabetes by STZ.Renin mRNA and renal and liver angiotensinogen mRNA were

suppressed from days 1 to 42 and then normalized [16]. Sechiand coworkers reported no change in angiotensinogen and reninmRNA in untreated diabetic rats compared to insulin-treateddiabetic or control rats 12 days after STZ. The untreateddiabetic group did have a decrease in All receptors and anincrease in prorenin compared to the other two groups [17]. Inspontaneously diabetic BB rats, Everett and coworkers dem-onstrated an initial increase in renal renin and angiotensinogenmRNA, and renin stained cells per juxtaglomerular apparatus,which later decreased during the more chronic phase of diabe-tes [181. Finally, Jaffa and coworkers demonstrated that threeweeks after the induction of diabetes with STZ, severelyhyperglycemic rats had low renin mRNA while moderatelyhyperglycemic (insulin treated) rats had a renin mRNA level not

802 Correa-Rotter et a!: Renin gene expression in diabetes mellitus

Diabetes ControlN=5 N=6

Fig. 6. A. PRA in diabetic and control rats. PRA was significantlylower in rats with diabetes as compared to control rats (P < 0.02). B.TRC in diabetic and control rats. TRC was not different between thetwo groups (P = NS).

different from control. This same group reported that theinsulin-treated diabetic rats had a plasma renin content notdifferent from that of the control group, however, renal renincontent was significantly increased [191. All of these discrepan-cies are not readily explained given the preliminary nature of allof these reports. However, differences may be related to thevarying time periods at which the rats were studied, the degreeof glycemic control, dietary intake, and the diabetic modelstudied.

Based on measurements of components of the systemic andintrarenal renin-angiotensin system, we could find no evidencefor activation of this system in our insulin-treated diabetic ratscompared to control rats. However, the dissociation betweenthe fall in plasma renin activity with no decrease in renal renincontent or renin mRNA suggests a lack of suppression of theintrarenal system compared to its systemic counterpart. Renalrenin mRNA was expressed in the conventional fashion as aproportion of total RNA. Since hypertrophy occurs in thediabetic kidney, the total amount of renal RNA, and thereforerenin mRNA, would be greater in the diabetic compared to the

control rat, further emphasizing the discrepancy between sys-temic and renal renin. These findings are compatible with adecrease in prorenin conversion to active renin that couldcontribute to the increased plasma levels of prorenin demon-strated in diabetic rats and patients [17, 34—37]. A prorenin torenin conversion defect could explain the lower active reninlevel in the face of normal to elevated levels of renal reninmRNA. An abnormal prorenin response to diuretics has beendemonstrated in some diabetic patients, and this phenomenon,as well as hyporeninemic hypoaldosteronism often found indiabetic patients, may both derive from some defect in activa-tion of renin [361. An alternative explanation for this dissocia-tion between a low PRA and no change in TRC or renin mRNAcould be a reduced amount of substrate (angiotensinogen)available for renin action in the assay. Although renin substrateavailability is not thought to be rate limiting in the generation ofangiotensin I, the lower liver angiotensinogen mRNA present inthe diabetic rats raises this possibility. Finally, a more distaldefect in renin secretion and/or increased peripheral reninclearance might also account for the lower plasma compared torenal renin, however, there is no clear reason to suspect thelatter possibility [381. The examination of glomerular versustubular expression of renin mRNA was performed in thepresent study to determine if renin synthesis shifted intodiabetic glomeruli. We felt it was important to test for thispossibility as we have previously documented a shift of reninsynthesis into the glomeruli of rats following subtotal nephrec-tomy [39], another model characterized by low plasma reninactivity yet exhibiting a beneficial renal response to angiotensinconverting enzyme inhibition. No difference in glomerular reninmRNA was found in the diabetic compared to the non-diabeticrats speaking against any diabetes-related intraglomerular shiftof renin synthesis.

As expected, the diabetic rats had higher food intakes thanthe non-diabetic controls. Since both dietary sodium and pro-tein modulate renin synthesis, the changes in renin may beinfluenced by these dietary components [40, 41]. For example,isolated increases in sodium intake would be expected todecrease renal renin and angiotensinogen, while a higher pro-tein intake is associated with increased renal renin mRNA butno change in angiotensinogen mRNA [40, 41]. The interactionsbetween these dietary components and the diabetic state werenot specifically dissected in this current study, but the patternof results does not suggest that either increased protein intakeor salt imbalance are solely responsible for the changes in thecomponents of the renin-angiotensin system measured.

Although quantitation of renin protein or mRNA does notprovide definitive evidence for increased Ang II production,renin is the rate-limiting enzyme in the formation of Ang II.Furthermore, measurement of renal renin provides informationabout local renin synthesis in the kidney [25]. Definitive evi-dence for a pathogenetic role of the renin-angiotensin systemrests on the results of studies examining specific blockade ofthis system. Studies of this nature suggest that the renin-angiotensin system is activated in diabetes [11, 12, 15]. How-ever, most of the components of this system including systemicand renal renin and renin mRNA, renal angiotensinogenmRNA, liver angiotensinogen mRNA, and Ang II receptornumber and affinity are either unchanged or decreased indiabetic compared to non-diabetic animals [16—22, 42]. This

A

8.0

6.0

4.0

2.0

0.0Diabetes ControlN=5 N=6

.

I0crI—

B54.

3,

2'

1•

0'ni

Correa-R otter et a!: Renin gene expression in diabetes mellitus 803

apparent discrepancy may be explained by post-receptorevents. Alternatively, a number of other diverse vasoactivesystems may be playing a role in the modulation of intrarenalhemodynamics in diabetes including increased glomerular pros-taglandin production [5, 6], activation of the kallikrein system[8], increased atrial natriuretic peptide [7], and decreasedsensitivity of tubuloglomerular feedback conditioned by glu-cose delivery to the macula densa [43]. Therefore, even withoutmarked elevations in the components of the renin-angiotensinsystem, removing the effects of this system may alter thecritical balance between vasoconstrictor and vasodilator sys-tems.

Evidence implicating the renin-angiotensin system in thepathogenesis and progression of diabetic nephropathy in hu-mans derives primarily from studies of CEI therapy and frommeasurements of renin and prorenin. Treatment with CEIreduces proteinuria in both normotensive and hypertensivediabetics and is associated with a decrease in the rate of declinein renal function [43—48]. In diabetic patients, PRA has beenreported as being low, normal and high [21—24]. The highervalues have been noted in diabetics with glomerular hyperfil-tration [231. Since early increases in GFR are associated withlater development of nephropathy [49], the increase in PRAmay be playing a pathogenetic role in the development ofdiabetic nephropathy. Elevations in circulating prorenin havebeen demonstrated in diabetics [34—37]. These increased prore-nm values, independent of changes in active renin, have beenregularly associated with the presence and development ofdiabetic microvascular complications [34, 35, 37]. As noted, thehigh prorenin and normal to low active renin found in somehuman diabetics, taken together with our findings of a decreasein PRA but no difference in renal renin content or renin mRNA,raises the possibility of a defect in conversion of prorenin intoactive renin. Furthermore, one might speculate that such adefect could contribute to the heightened prorenin levels notedin complicated human diabetes.

In conclusion, insulin-treated diabetic rats had lower PRAbut no change in TRC or renal renin mRNA compared tonon-diabetic controls. Both renal and liver angiotensinogenmRNA were lower in the diabetic group. The dissociationbetween a decrease in systemic but no change in renal renin inthe diabetic rats is consistent with alterations in intrarenal reninprocessing and/or secretion by the diabetic kidney.

Acknow'edgments

We thank David Chmielewski and Stefan Kren for technical assis-tance, and K.R. Lynch for the renin and angiotensinogen cDNAs. Thiswork was supported by a US Public Health Service Grant (AM-31437)(T.H.H.) and by a Young Investigator Grant from the National KidneyFoundation (M.E.R). Dr. Correa-Rotter is a recipient of JuvenileDiabetes Foundation International Postdoctoral Research Fellowship,and is also partially supported by the 'Instituto Nacional de IaNutriciOn Salvador Zubirhn", Mexico City.

Reprint requests to Ricardo Correa-Rotter, M.D., University ofMinnesota, Department of Medicine, Box 736 UMHC, 516 DelawareSt. SE., Minneapolis, Minnesota 55455, USA.

References

1. ANDERSEN AR, SANDAHL-CHRISTIANSEN J, ANDERSEN JK,KREINER 5, DECKERT T: Diabetic nephropathy in type I (insulin

dependent) diabetes: An epidemiological study. Diabetologia 25:496—501, 1983

2. KROLEWSKI AS, WARRAM JH, RAND LI, KAHN CR: Epidemiologicapproach to the etiology of type I diabetes mellitus and its compli-cations. NEnglJMed3l7:1390—1398, 1987

3. Ros.isi'ocs J, RASKIN P: Relationship of diabetic nephropathy toglycemic control, in The Kidney in Diabetes Mellitus, edited byBRENNER BM, STEIN JH, New York, Churchill-Livingstone, 1989,pp. 1—18

4. HOSTETTER TH, RENNKE HG, BRENNER BM: The case for intra-renal hypertension in the initiation and progression of diabeticnephropathy and other glomerulopathies. Am J Med 72:375—380,1982

5. SHAMBELAN M, BLAKE S: Increased prostaglandin production byglomeruli isolated from rats with streptozotocin-induced diabetesmellitus. J Cli,, Invest 75:404—412, 1985

6. KASISKE BL, O'DONNELL MP, KEANE WF: Glucose-induced in-creases in renal hemodynamic function. Possible modulation byrenal prostaglandins. Diabetes 34:360—364, 1985

7. ORTOLA FV, BALLERMAN BJ, ANDERSON S, MENDEZ RE, BREN-NER BM: Elevated plasma atrial natriuretic peptide levels in dia-betic rats. Potential mediator of hyperfiltration. J Clin Invest80:670—674, 1987

8. JAFFA AA, MILLER DH, BAILEY GS, CHAO J, MARG0LIUS HS,MAYFIELD RK: Abnormal regulation of renal kallikrein in experi-mental diabetes. Effects of insulin on prokallikrein synthesis andactivation. J Clin Invest 80:1651—1659, 1987

9. HOSTETTER TH, TROY JL, BRENNER BM: Glomerular hemodynam-ics in experimental diabetes mellitus. Kidney mt 19:410—415, 1981

10. ZATZ R, MEYER TW, ANDERSON S, RENNKE HG, BRENNER BM:Predominance of hemodynamic rather than metabolic factors in thepathogenesis of diabetic glomerulopathy. Proc Nat Acad Sci (USA)82:5963—5967, 1985

11. ZATZ R, DUNN BR, MEYER TW, ANDERSON B, RENNKE HG,BRENNER BM: Prevention of diabetic giomerulopathy by pharma-cologic amelioration of glomerular capillary hypertension. J GunInvest 77:1925—1930, 1986

12. ANDERSON 5, RENNKE HG, GARCIA DL, BRENNER BM: Short andlong term effects of antihypertensive therapy in the diabetic rat.Kidney mt 36:526—536, 1989

13. EDWARDS RM: Segmental effects of norepinephrine and angioten-sin II on isolated renal microvessels. Am J Physiol 244:F526—F534,1983

14. YUAN BH, ROBINETTE JB, CONOER JD: Effect of angiotensin 11and norepinephrine on isolated rat afferent arterioles. Am J Physiol258:F741—F750, 1990

15. ANDERSON 5, BOUYOUNES B, CLAREY LE, INGELFINGER JRIntrarenal renin-angiotensin system (RAS) in experimental diabe-tes. (abstract) JAm Soc Nephrol 1:621, 1990

16. SAJID M, GRAVES K, HOLLAND B, RAJARAMAN 5: The tissuerenin-angiotensin system (RAS) in experimental diabetes mellitus(DM). (abstract) J Am Soc Nephrol 1:640, 1990

17. SECHI LA, GRIFFIN C, KALINYAK JE, SHAMBELAN M: Character-ization of the intrarenal renin-angiotensin system (RAS) in strepto-zotocin-induced diabetes mellitus. (abstract) J Am Soc Nephrol1:426, 1990

18. EVERETT AD, SCOTT J, MARINO B, ROSENKRANZ R, GOMEZ AR:Renal renin and angiotensinogen expression during the evolution ofdiabetes. (abstract) J Am Soc Nephrol 1:414, 1990

19. JAFFA AA, CHAI KX, CI-IAO J, CHAO L, MAYFIELD RK: Reningene expression is altered by diabetes and insulin treatment.(abstract) Clin Res 39:307A, 1991

20. CHRISTLIEB AR: Renin, angiotensin and norepinephrine in alloxandiabetes. Diabetes 23:962—970, 1974

21. BURDEN AC, THURSTON H: Plasma renin activity in diabetesmellitus. Clin Sd 56:255—259, 1979

22. BELL GM, BERNSTEIN RK, LARAGH JH, ALLAS SA, JAMES GD,PECHER MS, SEALY SE: Increased plasma atrial natriuretic factorand reduced plasma renin in patients with poorly controlled diabe-tes mellitus. Gun Sci 77:177—182, 1989

23. WI5EMAN MJ, DRURY PL, KEEN H, VLBERTI GC: Plasma reninactivity in insulin-dependent diabetes with raised glomerular filtra-tion rate. Cli,, Endocrinol 21:409—414, 1984

804 Correa-Rotter et at: Renin gene expression in diabetes mellitus

24. SOLERTE SB, FIORAVARTI M, PETRAGLIA F, FACHINETTI F,APRILE C, GENAzzI AR, FERRARI E: Circulating opiods and plasmarenin activity in insulin-dependent diabetics with renal hemody-namic alterations. Nephron 46:194—198, 1987

25. DZAU VJ, BURT DW, PRATT RE: Molecular biology of the renin-angiotensin system. Am J Physiol 255:F563—F573, 1988

26. CHIRGWIN JJ, PRZBYLA AE, MACDONALD RJ, RUTTER Wi: Isola-tion of biologically active ribonucleic acid from sources enriched inribonucleases. Biochemistry 18:5294—5299, 1979

27. MARIASH CN, SEELIG S, OPPENHEIMER JH: A rapid, inexpensive,quantitative technique for the analysis of two-dimensional electro-phoretograms. Anal Biochem 121:388—394, 1982

28. DURNAM DM, PALMITER RD: A practical approach for quantitatingspecific mRNAs by solution hybridization. Anal Biochem 131:385—393, 1983

29. BURNHAM CE, HAWELU-JOHNSON CL, FRANK BM, LYNCH KR:Molecular cloning of rat renin cDNA and its gene. Proc Nat! AcadSci USA 84:5605—5609, 1987

30. LYNCH KR, SUMNAD VI, BEN-AR! ET, GARRISON JC: Localizationof pre-angiotensinogen messenger RNA sequences in the rat brain.Hypertension 8:540—543, 1986

31. MINTY AJ, CARAVATTI M, ROBERT B, COHEN A, DAUBAS P,WEYDERT A, GROS F, BUCKINGHAM ME: Mouse actin messengerRNAs. Construction and characterization of recombinant plasmidmolecule containing a complementary DNA transcript of mouseu-actin mRNA. JBiol Chem 256:1008—1014, 1981

32. SEALEY JE, MOON C, LARAGH JH, ALDERMAN M: Plasma prore-nm: Cryoactivation and relationship to renin substrate in normalsubjects. Am J Med 61:731—738, 1976

33, YANGAWA N, CAPPARELLI AW, Jo OD, FRIEDAL A, BARRETT JD,EGGENA P: Production of angiotensinogen and renin-like activityby rabbit proximal tubular cells in culture. Kidney tnt 39:938—941,1991

34. LUETSCHER JA, KRAEMER FB, WILSON DM, SCHWARZ HC,BRYER-ASH M: Increased plasma inactive renin in diabetes mclii-tus. NEnglJMed3l2:14l2—1417, 1985

35. LUETSCHERJA, KRAEMER FB, WILSON DM: Prorenin and vascularcomplications of diabetes. Am J Hypertens 2:382—386, 1989

36. BRYER-ASII M, FRAZE EB, LUETSCHER JA: Plasma renin andprorenin (inactive renin) in diabetes mellitus: Effects of intravenousfurosemide. J Cliii Endocrinol Metab 66:454—458, 1988

37. WILSON DM, LUETSCHER JA: Plasma prorenin activity and com-plications in children with insulin dependent diabetes mellitus. NEnglJ Med 323:1101—1 106, 1990

38. KIM S, INAO H, NAKAMURA N, IKEMOTO F, YAMAMOTO K: Fateof circulating renin in rats. Am J Physiot 252:E136—El46, 1987

39. ROSENBERG ME, CORREA-ROTTER R, INAGAMI T, KREN SM,HOSTETTER TH: Glomerular renin in the remnant kidney. Kidneyliii 40:677—683, 1991

40. INGELFINGER JR. PRATT RE, ELLISON K, DZAU VJ: Sodiumregulation of angiotensinogen mRNA expression in rat kidneycortex and medulla. J Clin Invest 78:1311—1315, 1986

41. ROSENBERG ME, CHMIELEWSKI D, HOSTETTER TH: Effect ofdietary protein on rat renin and angiotensinogen gene expression. JClin Invest 85:1144-1149, 1990

42. BALLERMAN BJ, SKORECKI KL, BRENNER BM: Reduced glomeru-lar angiotensin II receptor density in early untreated diabetesmellitus in the rat. Am J Physiol 247:Fl 10-Fl 16, 1984

43. Woos LL, MIGELLE HL, HALL SE: Control of renal hemodynam-ics in hyperglycemia: Possible role of tubuloglomerular feedback.Am J Physiol 252:F65—F73, 1987

44. TAGUMA Y, KITAMOTO Y, FUTAKI G, UEDA H, MONMA H,ISHAZAKI M, TAKAHASHI H, SEKINO H, SASAKI Y: Effect ofcaptopril on heavy proteinuria in azotemic diabetics. N Engi JMed313:1617—1620, 1985

45. BJORCK 5, NYBERO G, MULEC H, GRANERUS G, HERLITZ H,AURELL M: Beneficial effects of angiotensin converting enzymeinhibition on renal function in patients with diabetic nephropathy.Br Med J 293:471—474, 1986

46. MARRE M, LEBLANC H, SUAREZ L, GUYENNE T-T, MENARD J,PASSA P: Converting enzyme inhibition and kidney function innormotensive diabetic patients with persistent microalbuminuria.Br MedJ 294:1448—1452, 1987

47. INSUA A, RIBSTEIN J, MIMRAN A: Comparative effect of captopriland nifedipine in normotensive patients with incipient diabeticnephropathy. Postgrad Med J 64(S3):59—62, 1988

48. BJORCK S, MULEC H, JOHNSEN SA, NYBERG G, AURELL M:Contrasting effects of enalapril and metoprolol on proteinuria indiabetic nephropathy. Br Med J 300:904—907, 1990

49. MOGENSEN CE, CHRISTENSEN CK: Predicting diabetic nephrop-athy in insulin-dependent patients. N Eng! J Med 3 11:89—93, 1984