Kallikrein-binding protein levels are reduced in the ... fileReduced Retinal Kallikrein-Binding Protein in Diabetic Rats 659 with diabetes show potential involvement of the tissue

Kallikrein-Binding Protein Levels Are Reduced in theRetinas of Streptozotocin-Induced Diabetic Rats

Heather C. Hatcher * Jian-Xing Ma* Julie Chao,~\ Lee Chao,~\ and Anna Ottlecz%

Purpose. To determine the involvement of rat kallikrein-binding protein (RKBP) in the devel-opment of diabetic retinopathy.

Methods. Diabetes was induced by streptozotocin (STZ) (55 mg/kg body weight in 0.05 Mcitrate buffer, pH 4.5) in male Sprague-Dawley rats (150 to 175 g, 6 weeks old) as confirmedby hyperglycemia and reduced body weight. Retinas were dissected from animals at 1, 2, and4 months of diabetes. The functional activity of RKBP in retinal homogenates was determinedby its complex formation with tissue kallikrein. Immunoreactive RKBP levels were measuredby enzyme-linked immunosorbent assay. The RKBP messenger RNA (mRNA) levels in theretina were measured by Northern blot analysis using the RKBP complementary DNA probe.The activity of total Na+,K+-ATPase was determined by a radioassay. Total protein concentra-tion was determined by a protein assay.

Results. The kallikrein-binding activity was reduced in the retinas of STZ-diabetic rats at 1(59%), 2 (50%), and 4 (38%) months of diabetes compared to those of age-matched controlsubjects. Levels of immunoreactive RKBP were significandy lower in the diabetic animals ateach time point examined compared to those of control subjects. At 1 and 2 months ofdiabetes, RKBP levels (nanogram/milligram protein) were decreased significandy to 6.9 ±0.7 (n = 8) and 10.6 ± 1.0 (n = 8), respectively, compared to those of age-matched controlsubjects (14.1 ± 0.7, n = 8, P < 0.001, and 14.1 ± 1.2, n = 8, P < 0.01). At 4 months ofdiabetes, retinal RKBP levels were lower in both control and diabetic groups, but RKBP levelsin diabetic groups were significandy lower (5.8 ± 0.6, n = 8) than those of die age-matchedcontrol subjects (8.4 ± 0.9, n = 8, P < 0.01). Similarly, Northern blot analysis showed thatRKBP mRNA levels were reduced in the retina of each group of STZ-diabetic rats, suggestingthat the decrease in RKBP occurred at die level of transcription.

Conclusions. The results show that STZ-induced diabetic rats have decreased retinal RKBP;moreover, this suggests diat RKBP may contribute to diabetic retinopathy. Invest OphthalmolVisSci. 1997; 38:658-664.

J. he tissue kallikrein-kinin system consists of tissuekallikrein, kallikrein-binding protein (KBP), kinins, ki-nin-precursor kininogens, kinin receptors, and ki-ninases.1 Tissue or glandular kallikrein is a serine pro-teinase that releases kinins from kininogen substrates.Kinins are potent vasoactive peptides that interact withtheir receptors on the cell surface and mediate vasodi-lation and blood pressure reduction, and they regu-

From the Departments of* Ophthalmology and f Biochemistry and Molecular Biology,Medical University of South Carolina, Charleston, South Carolina, and the%Collegeof Optometry, University of Houston, Houston, Texas.Supported by Research to Prevent Blindness (JM), the National Institutes of Healthgrants HL29397 and HL44083 (JC), and the University of Houston, College ofOptometry Vision Research Support Grant (AO).Submitted for publication August 19, 1996; revised October 14, 1996; acceptedOctober 15, 1996.Proprietary interest category: N.Reprint requests: Jian-xing Ma, Department of Ophthalmology, Medical Universityof South Carolina, 171 Ashley Avenue, Charleston, SC 29425.

late local blood flow and rate of tissue metabolism.Kinins also induce smooth muscle contraction, cellproliferation, and inflammatory responses.1"4 Tissuekallikrein is regulated at the transcriptional level bymultiple hormones5 and at the posttranslational levelby its modulator, KBP.6"8 KBP binds covalently withtissue kallikrein and inhibits its activities in vitro.7'9 Inaddition, KBP has a profound effect on the clearancerate of tissue kallikrein in vivo because it prolongs thehalf-life of kallikrein.10 These results suggest that KBPis a modulator of tissue kallikrein's activity and bio-availability.11 The physiological function of KBP is cur-rently unknown; however, several lines of evidenceindicate that it may participate in blood pressure regu-lation, inflammatory response, growth, and develop-

ment.11-14

Studies in diabetic animal models and patients

658Investigative Ophthalmology & Visual Science, March 1997, Vol. 38, No. 3Copyright © Association for Research in Vision and Ophthalmology

Downloaded From: http://iovs.arvojournals.org/pdfaccess.ashx?url=/data/journals/iovs/933197/ on 04/28/2017

Reduced Retinal Kallikrein-Binding Protein in Diabetic Rats 659

with diabetes show potential involvement of the tissuekallikrein-kinin system in diabetes mellitus.15'16 Instreptozotocin (STZ)-diabetic rats, tissue kallikrein lev-els in the submandibular gland, kidney, and plasmaare reduced.17"19 Urinary active kallikrein was re-duced, which resulted in the decreased ratio of active-to-total kallikrein.20 Conversely, mice with diabetesand patients with diabetes were found to have in-creased renal kallikrein levels.21"24 In addition, therenal kallikrein activity was increased significantly ingenetic C57B\/Ks]-mdb mice with diabetes. Therefore,the alterations of kallikrein levels may be dependenton the mechanism or type of diabetes.22

The mechanism linking the tissue kallikrein-kininsystem to diabetic complications remains unclear. Tis-sue kallikrein is prevalent in vascular tissues.25 Thehigh efficiency with which kinins regulate vasomotionimplies that they are regulated tightly.26 Abnormalitiesin the tissue kallikrein-kinin system, which involveoverproduction of kinin, have been implicated in vaso-dilation, vascular impairment, and hyperfiltration indiabetes.20 Moreover, kinin has been shown to havea proliferation-stimulating effect.1 Our recent studieshave identified high levels of kallistatin, a humancounterpart of rat kallikrein-binding protein (RKBP),in various ocular tissues, including retinal neuronalcells and retinal capillary endothelial cells.27 More-Over, vitreous kallistatin concentrations in patientswith diabetes were significantly lower than those innondiabetic vitreous fluids.27 To elucidate the role ofRKBP in the pathogenesis of diabetic retinopathy, wehave identified the expression and quantified RKBPlevels in the retina of a diabetic rat model.

METHODS

Materials

Adenosine triphosphate (ATP; Tris-salt, vanadiumfree), ethylenediaminetetraacetic acid (EDTA),EGTA, Hanks' balanced salt solution, Hepes, andcharcoal-activated and glucose oxidase kits were fromSigma Chemical (St. Louis, MO). [y-32P]ATP and [a-32P] dCTP were purchased from New England Nuclear(Boston, MA). Protein assay kits were purchased fromBio-Rad Laboratories (Richmond, CA). All otherchemical agents were of analytical grade and were pur-chased from Sigma.

Animals

Sprague-Dawley rats (male albino, 150 to 175 g, 6weeks old) were purchased from Harlan Sprague-Dawley (Houston, TX), housed in stainless steel cages,and fed standard rat chow and tap water ad libitum.The animals were in a room on a 12-hour light-darkcycle with an ambient temperature of 22° ± 1°C. Allexperimental and animal care procedures were per-

formed in compliance with the Guide for the Careand Use of Laboratory Animals (DHHS PublicationNo. NIH 85-23, revised 1985) and adhered to theARVO Statement for the Use of Animals in Ophthal-mic and Vision Research. Animals were injected intra-peritoneally with STZ (55 mg/kg body weight in 0.05M citrate buffer, pH 4.5) or citrate buffer alone aftera 20-hour fast. Diabetes was verified using the Eliteglucometer from Bayer (Elkhart, IN) 3 days after theinjection of STZ. Animals with a blood glucose level>300 mg/dl and a body weight reduction >70g com-pared to those of control subjects were considereddiabetic. The diabetic animals did not receive any ex-ogenous insulin. Body weight was measured and ani-mals were killed by decapitation at 1, 2, and 4 monthsafter injection. Blood samples were collected to mea-sure serum glucose levels using a glucose oxidase kit(Sigma), and the animals' retinas were dissectedwithin 30 seconds of death.

Preparation of Retinal Homogenates

Retinas were rinsed in Hanks' balanced salt solution(pH 7.4). One retina from each animal was placedinto 800 fj\ of ice cold 0.05 M Hepes buffer (pH 7.4)and homogenized in a glass homogenizer (Teflon;Thomas Scientific, Swedesboro, NJ) using 20 strokesfor enzyme-linked immunosorbent assay (ELISA) andadenosinetriphosphatase (ATPase) assay. The otherretina from each rat was placed into GIT buffer (4.0M guanidine isothiocyanate, 0.1 M Tris-HCl, 1% fi-mercaptoethanol, pH 7.5) for RKBP messenger RNA(mRNA) studies. Samples were kept frozen at — 70°Cuntil use.

Protein Assay

Protein concentration was measured using the Bio-Rad protein assay kit (Bio-Rad Laboratories, Rich-mond, CA) and bovine serum albumin as a standardaccording to the protocol provided by the manufac-turer.

Na+,K+-ATPase Assay

After thawing, cellular debris was removed by centrifu-gation (1000g, 5 minutes), plasma membranes werepermeabilized by freezing and thawing (three cycles),and aliquots of supernatants were used for ATPasedeterminations. Total ATPase activity was determinedin retinal homogenates by measuring inorganic phos-phate (Pi) liberated from [y-32P]ATP according toKowluru et al28 with slight modifications. Briefly, theassay mixture contained 50 mM Hepes, 100 mM NaCl,10 mM potassium chloride, 4 mM magnesium chlo-ride, 1 mM EDTA, 0.25 mM EGTA (pH 7.4), and 10fi\ retinal homogenate (5 to 7 /zg of protein) in eitherthe absence or presence of 10"3 M ouabain in a finalvolume of 50 fil. The reaction was initiated by theaddition of ATP [y-32P]ATP: -30,000 cpm/nmol, 2


660 Investigative Ophthalmology & Visual Science, March 1997, Vol. 38, No. 3

mM final concentration of unlabeled ATP and pro-ceeded for 20 minutes at 37°C.

Tissue Kallikrein-Rat Kallikrein-BindingProtein Complex Formation

Retina tissue homogenates were pooled in eachgroup, and protein concentrations were measured.Tissue homogenates (160 fig of protein) were incu-bated with [l25I]-labeled rat tissue kallikrein (10,000cpm/sample) in 10 mM sodium phosphate buffer, pH7.0, at 37°C for 1 hour as described previously.12 Thebinding was stopped by adding one-third volume of 3X sodium dodecyl sulfate (SDS) sample buffer (0.125M Tris-Cl, pH 6.8, 30% glycerol and 5% SDS) andboiling for 5 minutes. The mixture then was resolvedby sodium dodecyl sulfate-polyacrylamide gel electro-phoresis under nonreducing conditions and stainedwith Coomassie blue. The gel was dried and exposedto Kodak X-Omat film (Eastman Kodak, Rochester,NY).

Enzyme-Linked Immunosorbent Assay

ELISA was modified from a method described pre-viously.29 A 96-well microtiter plate was coated withnonlabeled anti-RKBP immunoglobulin G (1 /Ltg/mlin phosphate-buffered saline (PBS), 100/il/well) over-night at 4°C. The plate then was blocked with 200 \A/well of PBS (10 mM sodium phosphate, pH 7.4, 150mM NaCl) containing 1% bovine serum albumin at37°C for 1 hour and washed three times with PBScontaining 0.1% Tween-20 (washing solution). Dilu-tions of purified RKBP (0.4 to 25 ng/ml) and retinahomogenates, which were diluted in 100 \A of freshlymade dilution buffer (PBS containing 0.05% Tween-20 and 0.5% gelatin), were added to individual wells.The plate was incubated at 37°C for 90 minutes andwashed three times with the washing solution. Biotin-labeled anti-RKBP antibody30 was added into each wellat a concentration of 1 //g/ml in a total volume of 100fj.\. The plate was incubated at 37°C for 1 hour andwashed three times with the washing solution andonce with PBS. Freshly prepared substrate solution(10 ml of 0.1 M citrate buffer, pH 4.3, 3 mg of 2,2'-azino-bis[3-ethylbenzthiazoline-sulfonic acid], and 10fj\ of 3% hydrogen peroxide) was added into eachwell (100 //I/well). After 30 minutes, the plate wasmeasured with an ELISA reader at 414 nm for ab-sorbance.

Extraction of RNA From Rat Retina

The dissected retinas were homogenized in the GITbuffer with an ultrasonicator, and the resulting ho-mogenate was layered on a 750 fA cushion of 5.7 Mcesium chloride in a clear ultracentrifuge tube. Thesample was centiifuged at 135,000gin a TLA-100 rotorat 20°C for 4 hours. After centrifugation, the superna-tant was removed carefully, and the RNA pellet was

washed three times with cold 70% ethanol. The pelletwas dried in a Speed-Vac (Savant, Farmingdale, NY)and resuspended in diethyl pyrocarbonate-treated wa-ter (30 (A) and transferred to a microcentrifuge tube.

Northern Blot Analysis

The RNA extracted from rat retina was electropho-resed through a 1.2% agarose gel containing formal-dehyde. The samples were prepared by mixing thefollowing in a sterile microfuge tube: 4.5 fi\ (20 //g)RNA, 2.0 fi\ 5 X MOPS running buffer (1 X MOPSrunning buffer = 20 mM 3-[N-morpholino] pro-panesulfonic acid, pH 7.0, 8 mM sodium citrate, 1 mMEDTA), 3.5 fj\ formaldehyde, and 10.0 fj\ formamide.The samples were heated for 15 minutes at 65°C andthen chilled on ice. After a brief centrifugation, 2 fj,\of formaldehyde gel-loading buffer (50% glycerol, 1mM EDTA, 0.25% bromophenol blue, 0.25% xylenecyanole FF) was added to each tube. The gel was runsubmerged in 1 X MOPS buffer at 3 to 4 V/cm forapproximately 3 hours. At the end of the run, themolecular weight marker was cut from the gel, stainedwith ethidium bromide for 30 minutes, and photo-graphed under ultraviolet light. The RNA from thegel was transferred to a nylon membrane by capillarytransfer using 20 X SSC (1 X SSC = 0.15 M sodiumchloride, 15 mM sodium citrate, pH 7.0). The RNAwas ultraviolet-crosslinked to the nylon membrane.Full-length RKBP complementary DNA was nick-trans-lated with [aT32P]dCTP and hybridized with the mem-brane in a hybridization solution containing 5 X SSPE(1 X SSPE = 0.18 M NaCl, 10 mM monosodium acidphosphate, 1 mM EDTA, pH 7.4), 5 X Denhardt'ssolution (0.1% Ficoll 400, 0.1% polyvinylpyrrolidone,0.1% bovine serum albumin), 0.5% SDS, and 100 //g/ml denatured herring sperm DNA at 60°C overnight.31

The membrane was washed in a final solution of 6 XSSPE, 0.1% SDS twice at 60°C, and exposed to KodakX-Omat film (Eastman Kodak, Rochester, NY) at-70°C.

Data Analysis

Retinal RKBP levels were averaged within a group andpresented as mean ± standard error of the mean. Thesignificance of the difference between two groups wasdetermined by Student's Hest.

RESULTS

Streptozotocin-Induced Diabetes and Alterationof the Total Na+,K+-ATPase Activity in theRetina

Body weight and blood glucose levels (Table 1) wereused to assess the general condition of STZ-induceddiabetic rats. Diabetic rats displayed significant growthretardation that was reflected in their lower body


Reduced Retinal Kallikrein-Binding Protein in Diabetic Rats m.TABLE i. Body Weight, Blood Glucose Level, and Retinal Na , K -ATPase

Total Na+, K* -ATPase-fGroup Month Injection n Body Weight* (g) Blood Glucose* (mg/dl) (nmol PJminute • mg protein)

ControlDiabeticControlDiabeticControlDiabetic

112244

878888

325 ±218 ±411 ±314 ±531 ±388 ±

2.810.24.86.57.69.2

104 ± 2.8506 ±21.1106 ± 5.1527 ± 25.3

92 ± 1.0486 ± 16.6

216 ± 7.69211 ± 8.20220 ± 3.86188 ± 2.62J222 ± 3.97190 ± 7.41$

n = number of rats per group.* Values represent mean ± SEM at the time of death.•f The total Na+, K+-ATPase activity was the difference between the activities recorded in the absence and presence of IO~;l M ouubain.X P < 0.05 versus age-matched controls.

weight compared to that of age-matched control sub-jects. Blood glucose levels were fourfold to fivefoldhigher in diabetic animals than in age-matched con-trol subjects. The total Na+,K+-ATPase activity was sig-nificantly lower in the retinas of diabetic rats after 2(16%) and 4 months (15%) of the disease comparedto that of age-matched control subjects (Table 1).

Analysis of Kallikrein-Binding Activity in theRetina

The RKBP is present in retina tissue homogenatesfrom all groups as it formed an SDS-stable complexwith rat tissue kallikrein. The molecular weight of thecomplex was 92 kDa, which is identical to that of thecomplex formed from purified RKBP and tissue kalli-krein,7 indicating the existence of a functional RKBPin rat retina. As measured by densitometry, the quanti-ties of the high molecular weight complex were de-creased in the retinas of diabetic rats at 1 (59%), 2(50%), and 4 (38%) months of diabetes compared totheir age-matched control subjects, when the sameamount of protein was used for the complex forma-tion assay (Fig. 1). These results indicate that the func-tional RKBP is reduced in retinas of diabetic rats.

Measurements of the Immunoreactive RatKallikrein-Binding Protein in the Retina

Immunoreactive RKBP levels in the retina were mea-sured by ELISA specific to RKBP. Retina tissue homog-enates were diluted serially and absorbance values at414 nm (A,|M) were measured and plotted as the func-tion of dilutions. As shown in Figure 2, serial dilutioncurves of the retinal homogenates of both control anddiabetic rats were parallel to that of die RKBP stan-dard, suggesting their immunologic identity.

The RKBP concentration in the retinal homoge-nate from each rat was determined by ELISA and nor-malized by the total protein concentration as mea-sured by the Bio-Rad protein assay. The RKBP levelswere averaged in each group, and a comparison wasmade between diabetics and their correspondent con-trol subjects (Table 2). As listed in Table 2, retinal

RKBP levels in STZ-diabetic rats were reduced by 51%(P< 0.001) and by 25% (P< 0.01) at 1 and 2 monthsof diabetes, respectively, compared to those of the age-matched control subjects. After 4 months of diabetes,RKBP levels were decreased in both control (8.4 ±0.9 ng/mg protein, n = 8) and diabetic rats (5.8 ±0.6, n = 8) compared to those of the control animalsat 1 and 2 months. Interestingly, the RKBP levels inthe diabetic retinas still were significantly lower thanthose in the control subjects (31%, P < 0.01) at thesame time point.

Rat KaUikrein-Binding Protein mRNA Levels inthe RetinaTotal RNA (20 p,g) from each group was subjected toNorthern blot analysis using a full-length RKBP com-

I Month 2 Months 4 Monthspnst-KT/ post - S I / pnst-STZ( D ( II ( I)

— Cnriipli'v

— free kiillikrein

211-

FIGURE 1. Rat kallikrein-binding protein-kallikrein complexformation in the rat retina. Tissue homogenates of retinasfrom control and STZ-diabetic rats were incubated with l2ril-labeled rat tissue kallikrein and resolved by sodium dodecylsulfate-polyacrylamide gel electrophoresis. The free kalli-krein and the high molecular weight complex formed be-tween kallikrein and rat kallikrein-binding protein, indi-cated to the right, were visualized by autoradiography. Lane1 = control, 1 month after injection; lane 2 = diabetic, 1month after injection; lane 3 = control, 2 months afterinjection; lane 4 = diabetic, 2 months after injection; lane5 = control, 4 months after injection; lane 6 — diabetic, 4months after injection; lane 7 = iasI-labeled tissue kallikreinalone.



3.2

2.4

0.4 -

0.2

1:64! _

Dilution of Tissue Homogenate

1:161:32I

1:8

• Purified RKBP

o Control Rat Retina

D Diabetic Rat Retina

0.4I

0.8 1.6I

3.2

RKBP (ng/ml)

I6.4

I12.5

I25.0

FIGURE 2. Enzyme-linked immunosorbent assay of rat retinahomogenates and purified rat kallikrein-binding protein.The standard curve of rat kallikrein-binding protein rangesfrom 0.39 to 25.0 ng/ml. Retina homogenates from controland diabetic rats were diluted serially. A polyclonal antibodyto rat kallikrein-binding protein was used in the assay.

1 Monthpost-STZC D

2 Monthspost-STZC D

4 .Monthspost-STZC D

1.8 kb

1 2 3 4 5 6

FIGURE 3. Northern blot analysis of rat kallikrein-bindingprotein messenger RNA in the rat retina. The same amountsof total RNA (20 //g) from retinas of control and diabeticrats were hybridized with a 32P-labeled rat kallikrein-bindingprotein complementary DNA probe. Lane 1 = control, 1month after injection; lane 2 = diabetic, 1 month after injec-tion; lane 3 = control, 2 months after injection; lane 4 =diabetic, 2 mondis after injection; lane 5 = control, 4months after injection; lane 6 = diabetic, 4 months afterinjection. The rat kallikrein-binding protein messenger RNA(1.8 kb) is indicated at left.

piementary DNA probe. The result showed that RKBPmRNA levels in the retinas of the STZ-diabetic ratswere decreased after 1, 2, and 4 months of diabetescompared to those of control subjects (Fig. 3). Densi-tometry of the specific RKBP mRNA bands showed a56%, 30%, and 41% reduction in the STZ-diabetic ratsafter 1, 2, and 4 months of diabetes, respectively (Fig.3). This result showed that RKBP is synthesized in theretina, and the decrease of RKBP in the retina oc-curred at the transcription level.

DISCUSSION

In the current study, we have identified high levels offunctional RKBP in the retina. Retinal RKBP levelswere reduced significantly in STZ-diabetic rats after 1,2, and 4 months of diabetes as analyzed by kallikrein-binding assay, ELISA, and Northern blot analysis. Si-multaneously, decreased Na, K-ATPase activity was

TABLE 2. RKBP Levels in Rat Retina

Month After RKBP*Group Injection n (ng/mg protein) Significance

ControlDiabeticControlDiabeticControlDiabetic

112244

878888

14.1 ± 0.76.9 ± 0.7

14.1 ± 1.210.6 ± 1.08.4 ± 0.95.8 ± 0.6

P < 0.001

P< 0.01

P < 0.01

n = number of rats per group.* Values represent the mean ± SEM.RKBP = rat kallikrein-binding protein.

measured in the retinas of diabetic rats at each timepoint.

We recendy have identified kallistatin, the coun-terpart of RKBP in humans,9 by in situ hybridizationin multiple ocular tissues, including the retina neu-ronal and vascular tissues.27 The RKBP, which is amember of the serine proteinase inhibitor superfam-ily,713 forms a heat- and SDS-resistant complex withtissue kallikrein.6 In this study, we showed that theretina contains kallikrein-binding activity. In the bind-ing assay, the 68-kDa albumin was visualized with la-beled tissue kallikrein alone (Fig. 1, lane 7). Albumindisappears after incubation with the retinal homoge-nates, which suggests that it was degraded by protein-ases present in the retinal homogenates. It is possiblethat the retina contains other serine proteinase inhibi-tors that may form high molecular weight complexeswith tissue kallikrein32; however, retinal RKBP was con-firmed to be immunologically identical to purifiedRKBP from serum by ELISA (Fig. 2). Consistent withthe results from human retina, the current study con-firms that RKBP mRNA is synthesized endogenouslyin the rat retina.

Although a typical proliferative diabetic retinopa-thy does not develop in STZ-diabetic rats, similar tothat which develops in humans, these rats do exhibitsome retinal vascular abnormalities.33"35 The tissuekallikrein-kinin system participates in the regulationof vascular function and has been implicated in thediabetic complications of animal models and pa-tients.1719'21 Our results show that both functional andimmunoreactive RKBP levels are reduced in STZ-dia-betic rats as early as 1 month of diabetes and contin-


Reduced Retinal Kallikrein-Binding Protein in Diabetic Rats 663

lied to decrease after 2 and 4 months of diabetes.Thus, the decrease in retinal RKBP levels at an earlystage of diabetes suggests that RKBP may play a rolein the pathogenesis of diabetic retinopathy and maynot be a secondary consequence of other diabetic ab-normalities. These findings are consistent with the re-sults in humans, which showed that vitreous kallistatinis decreased significantly in patients with diabetic reti-nopathy.27 Interestingly, reduced RKBP levels were ob-served in both the control and diabetic animals at 4months after injection (Table 2), whereas reductionwas seen only in the diabetic animals at 1 and 2 monthsafter injection. Because RKBP expression is depen-dent on growth hormone,13'36 age-related changes ingrowth hormone levels may explain the decrease ofRKBP in the older animals.

The role of RKBP in the development of diabeticretinopathy currently is unclear. Previously, we haveidentified and localized components of the tissue kalli-krein-kinin system in the eye, including tissue kalli-krein, kininogen, kallistatin, and kinin receptors inphotoreceptor cells, bipolar cells and retinal ganglioncells of the human retina, and endothelial cells ofblood vessels.27 Furthermore, components of the re-nin-angiotensin system appear to be synthesized lo-cally in the eye.37 The tissue kallikrein-kinin system,in counterbalance with the renin-angiotensin system,regulates vascular function and affects blood pressureand local blood flow.115 A disturbance in the balanceof these two systems has been implicated in hyperten-sion and microvascular complications of diabetes mel-litus.1'15'26'37'38 Intramuscular delivery of the RKBPgene increases systemic blood pressure significantly inhypotensive transgenic mice that overexpress kalli-krein, suggesting that RKBP inhibits tissue kallikrein'sactivity in vivo.14 Thus, the decreased level of retinalRKBP in STZ-diabetic rats may allow accumulation ofkinin due to increased kininogenase activity of tissuekallikrein.

Kinin possesses a potent vasorelaxant effect thathas been linked to the release of nitric oxide andprostacyclin from endothelial cells.1'39 Excess kininmay contribute to vasodilation and increased retinalblood flow, which are early events in diabetic retinopa-thy40'41; furthermore, it may stimulate proliferation ofendothelial cells through its proliferation-stimulatingactivities or through the induction of other growthfactors2'42'43 and also may inhibit Na, K-ATPase activitythrough the stimulation of phosphoinositide turn-over.44 Future studies may elucidate whether de-creased retinal RKBP facilitates vasodilation of retinalcapillaries associated with diabetic retinopathy.

Key Words

diabetic retinopathy, enzyme-linked immunosorbent assay,kallikrein, kallikrein-binding protein, Northern blot, strep-tozotocin

Acknowledgments

The authors thank Dr. Rosalie K. Crouch and her laboratorystaff in the Department of Ophthalmology at the MedicalUniversity of South Carolina for technical support, Dr. JimNicholson in the Department of Pathology for performingdensitometry, Zhirong Yang in the Department of Biochem-istry at the Medical University of South Carolina for per-forming ELISA, and Thouria Bensaoula at the University ofHouston College of Optometry for blood glucose measure-ments.

References

1. Bhoola KD, Figueroa CD, Worthy K. Bioregulation ofkinins: Kallikreins, kininogens, and kininases. Pharma-col Rev. 1992; 44:1-80.

2. Schachter M. Kallikreins (kininogenases)—a groupof serine proteases with bioregulatory actions. Pharma-col Rev. 1980;31:l-17.

3. Madeddu P, Chao J, Chao L, et al. Urinary levels ofkallikrein and kallistatin in pregnancy-induced hyper-tension. Hypertension in Pregnancy. 1995; 14:1-12.

4. Linz W, Wiemer G, Gohlke P, Unger T, Scholkens BA.Contribution of kinins to the cardiovascular actions ofangiotensin-converting enzyme inhibitors. PharmacolRev. 1995; 47:25-49.

5. Clements JA. The glandular kallikrein family of en-zymes: Tissue specific expression and hormonal regu-lation. EndocrRev. 1989; 10:393-419.

6. Chao J, Tillman DJ, Wang M, Margolius HS, Chao L.Identification of a new tissue-kallikrein-binding pro-tein. BiochemJ. 1986;239:325-331.

7. Chao J, Chai K-X, Chen L-M, et al. Tissue kallikrein-binding protein is a serpin. / Biol Chem. 1989;265:16394-16401.

8. Murray SR, Chao J, Lin F-K, Chao L. Kallikreinmultigene families and the regulation of their expres-sion. J Cardiovasc Pharmacol. 1990; 15:S7-S16.

9. Zhou G, Chao L, Chao J. Kallistatin: A novel humantissue kallikrein inhibitor. J Biol Chem. 1992;267:25873-25880.

10. Xiong W, Tang CQ, Zhou GX, Chao L, Chao J. Invivo catabolism of human kallikrein-binding proteinand its complex with tissue kallikrein. Tohoku J ExpMed. 1992; 171:135-144.

11. Chao J, Chao L. Biochemistry, regulation and poten-tial function of kallistatin. Biol Chem Hoppe Seyler.1995;376:7O5-713.

12. Chao J, Chao L. A major difference of kallikrein-bind-ing protein in spontaneously hypertensive versus nor-motensive rats. /Hypertens. 1988;6:551-557.

13. Yoon J-B, Towle HC, Seelig S. Growth hormone in-duces two mRNA species of the serine proteinase in-hibitor gene family in rat liver. / Biol Chem. 1987;262:4284-4289.

14. Ma J-X, Yang Z, Chao J, Chao L. Intramuscular deliv-ery of rat kallikrein-binding protein gene reverses hy-potension in transgenic mice expressing human tissuekallikrein. J Biol Chem. 1995; 270:451-455.

15. Margolius HS. Tissue kallikreins and kinins: Regula-tion and roles in hypertensive and diabetic diseases.Annu Rev Pharmacol Toxicol. 1989;29:343-364.



16. Mayfield RK, Margoliuos HS, Bailey GS, et al. Urinaryand renal tissue kallikrein in the streptozotocin-dia-betic rat. Diabetes. 1985; 34:22-28.

17. Jaffa AA, Miller DH, Bailey GS, et al. Abnormal regula-tion of renal kallikrein in experimental diabetes./C/inInvest. 1987; 80:1651-1659.

18. Jaffa AA, Chai KX, Chao J, Chao L, Mayfield RK. Ef-fects of diabetes and insulin on expression of kalli-krein and renin genes in the kidney. Kidney Int.1992;41:789-795.

19. Handa H, Skurama S, Nakagawa S, Yasukouchi T, Ska-moto W, Iz'umi H. Glandular kallikrein, renin, andangiotensin converting enzyme of diabetic and hyper-tensive rats. Adv Exp Med Biol. 1989;247:443-448.

20. Yasujima M, Yoshida K, Kanazawa M, et al. Urinaryactive kallikrein excretion and diabetic renal impair-ments in streptozotocin-treated rats. TohokuJExp Med.1993;171:135-144.

21. Harvey JN, Edmundson AW, Jaffa AA, Martin LL, May-field RK. Renal excretion of kallikrein and eicosanoidsin patients with Type I (insulin-dependent) diabetesmellitus. Relationship to glomerular and tubular func-tion. Diabetologia. 1992; 35:857-862.

22. Catanzaro OL, Pivetta OH, Zuccollo A, et al. Studieson kallikrein-kinin system and enzyme activity in or-gan and urine from diabetic mice models. Adv ExpMed Biol. 1989; 247:593-597.

23. Manto A, Cotroneo P, Porcelli G, et al. Urinary kalli-krein excretion in Type I (insulin-dependent) dia-betic mellitus. Diabetologia. 1993; 36:423-427.

24. Jaffa AA, Rust PF, Mayfield RK. Kinin, a mediator ofdiabetic-induced glomerular hyperfiltration. Diabetes.1995;44:156-160.

25. Nolly HL, Lama MC, Carretero OA, Scicli AG. Thekallikrein-kinin system in blood vessels. Agents ActionsSuppl. 1992;38(Suppl. III):l-9.

26. Mombouli J-V, Vanhoutte PM. Kinins and endothelialcontrol of vascular smooth muscle. Annu Rev Pharma-col Toxicol. 1995; 35:679-705.

27. Ma JX, King LP, Yang Z, Crouch RK, Chao L, ChaoJ. Kallistatin in human ocular tissues: Reduced levelsin vitreous fluids from patients with diabetic retinopa-thy. CurrEyeRes. 1996; 15:1117-1123.

28. Kowluru R, Bitensky MW, Kowluru A, Dembo M, Kea-ton PA, Buican T. Reversible sodium pump defect andswelling in the diabetic rat erythrocytes: Effects onfilterability and implications for microangiopathy. ProcNatlAcadSci USA. 1989;86:3327-3331.

29. Chen L-M, Song Q, Chao L, Chao J. Cellular localiza-tion of tissue kallikrein and kallistatin mRNAs in hu-man kidney. Kidney Int. 1995;48:690-697.

30. Guesdon JL, Ternyck T, Avrameas S. The use of avidin-biotin interaction in immunoenzymatic techniques. /Histochem Cytochem. 1979; 27:1131-1139.

31. Chai KX, MaJ-X, Murray SR, Chao J, Chao L. Molecu-lar cloning and analysis of the rat kallikrein-bindingprotein gene. /Biol Chem. 1991; 266:16029-16036.

32. Chen L-M, Chao L, Mayfield RK, Chao J. Differentialinteractions of human kallikrein-binding protein and«i antitrypsin with human tissue kallikrein. BiochemJ.1990;267:79-84.

33. Leuenberger P, Cameron D, Stauffacher W, et al. Ocu-lar lesions in rats rendered chronically diabetic withstreptozotocin. Ophthalmic Res. 197l;2:189-204.

34. Leuenberger P, Beauchemin ML, Babel J. Experimen-tal diabetic retinopathy. Arch Ophthalmol. 1974; 34:289-302.

35. Sosula L, Beaumont P, Hollows F, et al. Dilatatin andendothelial proliferation of retinal capillaries in strep-tozotocin-diabetic rats. Invest Ophthalmol Vis Sci.1972; 11:926-935.

36. Le Cam A, Pages G, Auberger P, et al. Study of agrowth hormone-regulated protein secreted by rat he-patocytes: cDNA cloning, anti-protease activity andregulation of its synthesis by various hormones. EMBOJ. 1987; 6:1225-1232.

37. Danser AHJ, Van Den Dorpel MA, Deinum J, et al.Renin, prorennin, and immunoreactive retinin in vit-reous fluid from eyes with and without diabetic reti-nopathy. / Clin Endocrinol Metab. 1989; 68:160-167.

38. Ottlecz A, Bensaoula T, Peterson RG, Eichberg J. An-giotensin converting enzyme activity in retinas ofstreptozotocin-induced and Zucker diabetic rats: Theeffect of angiotensin II on Na+,K+-ATPase activity. In-vest Ophthalmol Vis Sci. 1996; 37:1633-1641.

39. Gryglewski RJ, Moncada S, Palmer RM. Bioassay ofprostacyclin and endothelium-relaxing factor (EDRF)from porcine aortic endothelial cells. Br J Pharmacol.1986;87:685-694.

40. Garner A. Histopathology of diabetic retinopathy inman. Eye. 1993; 7:250-253.

41. Kohner EM, Patel V, Rassam SMB. Role of blood flowand impaired autoregulation in the pathogenesis ofdiabetic retinopathy. Diabetes. 1995;44:603-607.

42. EmeisJJ, Tranquille N. On the role of bradykinin insecretion from vascular endothelial cells. Agents Ac-tions Suppl. 1992;38(Suppl. II):285-291.

43. Sharma JN, Buchanan WW. Pathogenic responses ofbradykinin system in chronic inflammatory rheuma-toid disease. Exp Toxicol Pathol. 1994; 46:421-433.

44. Derian CK, Moskowitz MA. Polyphosphoinositide hy-drolysis in endothelial cells and carotid artery seg-ments. JBiol Chem. 1986;261:3831-3837.


Documents

Kallikrein-binding protein levels are reduced in the ... fileReduced Retinal Kallikrein-Binding Protein in Diabetic Rats 659 with diabetes show potential involvement of the tissue