Activation of NF-κB and expression of COX2 in association with neutrophil infiltration in systemic...

OBSTETRICS

Activation of NF-�B and expression of COX-2 in associationwith neutrophil infiltration in systemic vascular tissue ofwomen with preeclampsiaTanvi J. Shah, MS; Scott W. Walsh, PhD

OBJECTIVE: The purpose of this study was to determine whether acti-vation of NF-�B and expression of COX-2 are associated with neutro-phil infiltration in systemic vascular tissue of women withpreeclampsia.

STUDY DESIGN: Subcutaneous fat biopsies were obtained at cesareansection or abdominal surgery from preeclamptic women (n � 7), nor-mal pregnant women (n � 6), and normal nonpregnant women (n �5). Resistance-sized vessels (10 to 200 �m) in subcutaneous fat wereevaluated using immunohistochemical staining for: (1) CD66b, a neu-trophil antigen, (2) nuclear factor-�B (NF-�B), a transcription factorfor genes of inflammation, and (3) cyclooxygenase-2 (COX-2), an in-flammatory gene product.

RESULTS: The percentage of vessels which showed staining forCD66b, NF-�B, and COX-2 was significantly greater for preeclampticpatients as compared to normal nonpregnant or normal pregnant pa-

tients. In preeclamptic patients, vessel staining for NF-�B and COX-2was present in both endothelium and in vascular smooth muscle. Leu-kocytes in the lumen and adhered to endothelium also stained forNF-�B and COX-2. Activation of NF-�B and expression of COX-2 werecoincident with neutrophil flattening and adherence to endothelium andinfiltration into the intimal space.

CONCLUSION: These data demonstrate activation of NF-�B and ex-pression of COX-2 in systemic vasculature of women with preeclamp-sia, and they demonstrate that this vascular inflammation is linked withneutrophil infiltration. Neutrophil release of toxic substances, such asreactive oxygen species, TNF� and thromboxane, could be responsi-ble for vasoconstriction and vascular dysfunction. These data clearlyplace preeclampsia in the category of an inflammatory disease associ-ated with immune dysfunction.

Key words: COX-2, inflammation, NF-�B, neutrophils, preeclampsia

Cite this article as: Shah TJ, Walsh SW. Activation of NF-�B and expression of COX-2 in association with neutrophil infiltration in systemic vascular tissue ofwomen with preeclampsia. Am J Obstet Gynecol 2007;196:48.e1-48.e8.

Preeclampsia is associated with oxi-dative stress, leukocyte activation,

and endothelial cell dysfunction.1-9 Re-cently, we demonstrated flattening andadherence of neutrophils onto endothe-lium and infiltration into the intimalspace in systemic blood vessels of women

with preeclampsia.10 Neutrophil infiltra-tion was associated with markers of in-flammation in the vasculature. Intercel-lular adhesion molecule-1 (ICAM-1), anendothelial cell adhesion molecule forneutrophils, and interleukin-8 (IL-8), apotent neutrophil chemotactic agent forneutrophils, were expressed in endothe-lium, as well as in vascular smooth mus-cle in the preeclamptic women.

Oxidized lipids in the maternal circu-lation might cause inflammation of theendothelium leading to expression ofIL-8, but flattening and adherence ofneutrophils to endothelium would exac-erbate the inflammation through releaseof reactive oxygen species (ROS), tumornecrosis factor alpha (TNF�), and my-eloperoxidase. Transmission of inflam-mation to the vascular smooth musclecould be mediated by neutrophils thathave infiltrated into the intimal space.

Nuclear factor-�B (NF-�B) is a hall-mark of inflammation. It is activated by

various factors, including oxidativestress.11,12 NF-�B is a heterodimer com-posed of the p50 and p65 subunits. In itsinactive state, NF-�B is bound to inhibi-tory-�B (I-�B) in the cytosol. Upon activa-tion, the p50 and p65 subunits dissociatefrom I-�B and translocate to the nucleus,where they bind to promoter regions ofgenes involved in the inflammatory re-sponse, including IL-8, ICAM-1, andCOX-2. COX-2 is the inducible form ofcyclooxygenase and leads to increased pro-duction of prostaglandins and thrombox-ane that mediate inflammation. Throm-boxane is also a potent vasoconstrictor.COX-2 is induced during inflammatoryresponses13 and it plays an important rolein inflammatory diseases, such as rheuma-toid arthritis and osteoarthritis.14 COX-2expression in monocytes is associated withincreased circulating levels of oxidized lip-ids.15

We hypothesized that vascular infil-tration of neutrophils would be associ-

From the Departments of Obstetrics andGynecology and Physiology, VirginiaCommonwealth University Medical Center,Richmond, VA.

Received April 24, 2006; revised June 9,2006; accepted Aug. 22, 2006.

Reprints: Scott W. Walsh, PhD, VirginiaCommonwealth University Medical Center,Departments of Obstetrics and Gynecology,PO Box 980034, Richmond, VA [email protected]

Supported by grants to Dr Walsh from theNational Institutes of Health (HL069851).

0002-9378/$32.00© 2007 Mosby, Inc. All rights reserved.doi: 10.1016/j.ajog.2006.08.038

Research www.AJOG.org

48.e1 American Journal of Obstetrics & Gynecology JANUARY 2007

ated with activation of NF-�B and ex-pression of COX-2 in women withpreeclampsia. To test this, we used sub-cutaneous fat, which is highly vascular-ized, as representative of the maternalsystemic circulation. We used specificantibodies to immunostain the vesselsfor neutrophils, activated NF-�B, andCOX-2.

MATERIALS AND METHODSStudy SubjectsSubcutaneous fat biopsies were collectedfrom patients at MCV Hospitals, Vir-ginia Commonwealth University Medi-cal Center. Subcutaneous fat biopsieswere used because subcutaneous fat is ahighly vascularized tissue representativeof the systemic vasculature. Fat biopsies(approximately 1 cm � 1 cm � 1 cm)were collected at the time of cesareansection from normal pregnant patients(NP, n � 6) and preeclamptic patients(PE, n � 7) or at the time of abdominalsurgery from normal nonpregnant pa-tients (NNP, n � 5). For the preeclamp-tic patients, 4 had severe preeclampsia, 6delivered preterm, and 3 had IUGR. Fatbiopsies were placed immediately in10% neutral buffered formalin. Cesar-ean sections for normal pregnantwomen were performed because of pre-vious cesarean section or secondary tolatent herpes simplex virus or fetal mal-position. Surgeries for normal nonpreg-nant women were performed for re-moval of fibroids or for tissue biopsies.Exclusion criteria were chorioamnioni-tis, maternal infection, active STDs, dia-betes, and smoking. All patients werematched for body mass index (BMI) andwere not in labor. Informed consent wasobtained before surgery. The Office ofResearch Subjects Protection of VirginiaCommonwealth University approvedthis study.

ImmunohistochemistryTissue was paraffin embedded and cut in10 �m sections. Vessels were stained forCD66b, NF-�B, and COX-2. CD66b is agranulocyte-specific membrane antigenthat is upregulated and secreted upongranulocyte activation. CD66b stainingprimarily represents neutrophils be-

cause they comprise 96% of the granulo-cyte population. Neutrophils were alsoidentified by staining for naphthol AS-Dchloroacetate esterase activity. This ac-tivity was determined using a commer-cial kit according to manufacturer’s in-structions (Sigma Diagnostics, St Louis,MO). The NF-�B antibody was directedagainst epitopes on the p65 subunit,which are masked when bound to I-�B.When activated the epitopes are un-masked, so our staining reflects activatedNF-�B.

On the day of staining, tissue sectionswere incubated in 3% H2O2 in methanolfor 30 minutes to quench endogenoustissue peroxidase. Antigen retrieval wasperformed using a beaker filled with 600mL of 10 mmol/L citrate buffer heatedon a hot plate to a gentle boil. Slides wereplaced in the solution for 15 minutes.Tissue was stained with: (1) a mouse IgMantihuman monoclonal antibody spe-cific for CD66b (1:50, BD BioSciences,San Diego, CA); (2) a rabbit antihumanpolyclonal antibody specific for the p65

subunit of NF-�B (1:200, Zymed Labo-ratories, San Francisco, CA); and (3) amouse IgG antihuman monoclonal anti-body specific for COX-2 (1:400, ZymedLaboratories). Negative controls forCD66b were stained with a mouse IgMmonoclonal isotype control (1:50, BDBioSciences) and negative controls forNF-�B and COX-2 were stained with amouse IgG monoclonal isotype control(prediluted, Zymed Laboratories). Neu-trophils isolated from whole bloodby Histopaque density centrifugationserved as a positive control for CD66b.Lymphoma tissue was used as the NF-�Bpositive control and colon carcinomatissue was used as the COX-2 positivecontrol. Immunohistochemical stainingwas performed at room temperature us-ing a Zymed SuperPicture Kit (ZymedLaboratories), according to the manu-facturer’s instructions. Slides were pro-cessed by hand and tissue was counter-stained with alcian blue and methylgreen. DAB reagent resulted in brownstaining of antigens.

TABLE 1Clinical data for patient groups

Normalnonpregnant(n � 5)

Normalpregnant(n � 6)

Preeclamptic(n � 7)

Maternal age 42.4 � 6.0* 22.3 � 2.7 25.4 � 4.7..............................................................................................................................................................................................................................................

Race.....................................................................................................................................................................................................................................

Hispanic 1 2.....................................................................................................................................................................................................................................

White 1 3.....................................................................................................................................................................................................................................

Black 4 3 4..............................................................................................................................................................................................................................................

Prepregnancy BMI 28.9 � 4.0 30.8 � 5.5 31.2 � 9.0..............................................................................................................................................................................................................................................

Pregnancy BMI NA 36.6 � 3.0 38.0 � 2.8..............................................................................................................................................................................................................................................

Systolic bloodpressure (mmHg)

126.8 � 10.4 126.3 � 19.9 176.9 � 11.0†

..............................................................................................................................................................................................................................................

Diastolic bloodpressure (mmHg)

73.6 � 11.2 75.2 � 14.1 111.0 � 9.6†

..............................................................................................................................................................................................................................................

Proteinuria (mg/24 h) ND ND 670 � 352.8 (n � 3)..............................................................................................................................................................................................................................................

Dipstick ND ND 3.0 � 1.0 (n � 4)..............................................................................................................................................................................................................................................

Parity NA 1.0 � 0.6 1.0 � 1.3..............................................................................................................................................................................................................................................

Gestational age (wk) NA 38.8 � 1.5 33.3 � 3.8‡

..............................................................................................................................................................................................................................................

Infant birthweight (g) NA 3187 � 631.2 1864 � 901.5‡

..............................................................................................................................................................................................................................................

Values are mean � SD. NA, not applicable; ND, not determined.

* P � .001 compared to normal pregnant or preeclamptic.† P � .001 compared to normal pregnant or normal nonpregnant.‡ P � .01 compared to normal pregnant.

www.AJOG.org Obstetrics Research

JANUARY 2007 American Journal of Obstetrics & Gynecology 48.e2

Data AnalysisWe analyzed vessels between 10 �m and200 �m. Lumen width was measured foreach vessel. Vessel staining for CD66b,NF-�B, and COX-2 was graded using avisual score ranging from 0 to 3 (absentto intense staining) by an investigatornot blinded to the patient group. ForCD66b, the score included neutrophilinvolvement in the vessel. Objectivity ofthe visual score was verified by a secondinvestigator and by optical density mea-surements of staining using image anal-ysis software (IP Lab, Scanalytics, Inc,Fairfax, VA). Staining was also evaluatedfor percent of vessels with staining forneutrophils; percent of vessels with neu-trophils flattened and adhered to endo-thelial cells; and percent of vessels withneutrophils infiltrated into the intimalspace. NF-�B and COX-2 staining wereevaluated for percent of total vessels withstaining, percent of vessels with stainingin vascular smooth muscle, and percentof vessels that also had leukocytes stainedfor NF-�B or COX-2.

Statistical AnalysisVisual score data were analyzed byKruskal-Wallis test with Dunn’s post-hoc test to determine differences be-tween patient groups. For optical densitymeasurements, one-way analysis of vari-ance (ANOVA) was used with Newman-Keuls post-hoc test. Regression analysiswas performed to correlate visual scoreswith optical density measurements. Per-cent of vessels with staining was analyzedby one-way ANOVA with Newman-Keuls post-hoc test. For CD66b staining,analysis was also performed for the per-cent of vessels with staining in the fol-lowing vessel locations: (1) adhered tothe endothelium and (2) within the in-tima. Bar graph data are reported asmean � SE. A statistical software appli-cation was used (Prism, GraphPad Soft-ware, Inc, San Diego, CA).

RESULTSClinical data for normal nonpregnant,normal pregnant, and preeclamptic pa-tients are summarized in Table 1.

Staining scores for CD66b, NF-�B,and COX-2 are shown in Figure 1. The

staining score for CD66b for preeclamp-tic women was significantly higher thanfor normal pregnant or normal non-pregnant patients (1.5 � 0.3 vs 0.3 � 0.1vs 0.1 � 0.05, respectively, P � .01). Op-tical density of staining for CD66b wasalso significantly greater for preeclamp-tic women (141.4 � 18.0 vs 98.9 � 13.7vs 84.4 � 7.4 OD, respectively, P � .05,with background optical density approx-imately 60 OD). Staining scores forNF-�B and COX-2 were significantlygreater for preeclamptic than normalpregnant or normal nonpregnant pa-

tients (NF-�B, 1.9 � 0.2 vs 0.6 � 0.1 vs0.5 � 0.3, respectively, P � .001 andCOX-2, 1.6 � 0.2 vs 0.4 � 0.1 vs 0.3 �0.2, respectively, P � .001). OD mea-surements for NF-�B and COX-2 werealso significantly greater for preeclamp-tic than for normal pregnant or normalnonpregnant patients (NF-�B, 151.0 �5.9 vs 107.5 � 5.4 vs 104.7 � 13.2 OD,respectively, P � .01 and COX-2, 151.0� 6.1 vs 90.2 � 7.7 vs 88.5 � 6.3 OD,respectively, P � .001).

The percent of vessels that stained forCD66b, NF-�B, and COX-2 is shown inFigure 2. The percentage of vesselsstained for CD66b was significantlygreater for preeclamptic patients than

FIGURE 1Staining intensity of resistance-sized vessels in subcutaneousfat for CD66b, NF-�B, and COX-2

Staining intensity for all 3 antigens was mark-edly greater for patients with preeclampsia (PE)than for normal nonpregnant (NNP) or normalpregnant (NP) patients. Optical density (ODs)measurements were highly correlated with stain-ing scores (CD66b, r � 0.85; NF-�B, r � 0.89;COX-2, r � 0.92). Data are expressed as mean� SE. **P � .01, ***P � .001.

FIGURE 2Percent of vessels stained forCD66b, NF-�B, and COX-2

Preeclamptic patients had significantly morevessels stained for all 3 antigens compared withnormal nonpregnant and normal pregnant pa-tients. **P � .01, ***P � .001.

Research Obstetrics www.AJOG.org


normal pregnant patients or normalnonpregnant patients (72.5 � 10.8 vs33.8 � 10.3 vs 17.1 � 8.8%, respectively,P � .01). In preeclamptic patients, ascompared with normal pregnant pa-tients and normal nonpregnant patientsthere was significantly greater adherenceand flattening of neutrophils along theendothelium (51 � 9 vs 25 � 7 vs 16 �8%, respectively, P � .05) and there wassignificantly more infiltration into theintima (42 � 12 vs 11 � 4 vs 4 � 3%,respectively, P � .05).

The percentage of vessels stained forNF-�B (Figure 2) was significantlygreater for preeclamptic patients thannormal pregnant patients or normalnonpregnant patients (87.1 � 4.9 vs 39.2� 10.7 vs 26.7 � 12.7%, respectively, P� .01). Neutrophils, as well as other leu-kocytes present in the lumen and ad-

hered to endothelium stained forNF-�B. When there was vessel stainingfor NF-�B, leukocytes that were stainedfor NF-�B were present 76-78% of thetime. Preeclamptic patients had signifi-cantly more vessels with leukocytesstained for NF-�B than normal preg-nant or normal nonpregnant patients(66.4 � 4.6 vs 30.5 � 4.0 vs 20.1 �13.1%, P � .01, Figure 3). Vesselsstained for NF-�B had staining in vas-cular smooth muscle cells, as well as inendothelial cells. Preeclamptic patientshad significantly more vessels withNF-�B staining in vascular smoothmuscle than normal pregnant or nor-mal nonpregnant patients (72.2 � 4.5vs 33.3 � 5.3 vs 22.6 � 4.8%, P � .001,Figure 3).

The percentage of vessels stained forCOX-2 was significantly higher for pre-

eclamptic patients than normal pregnantor normal nonpregnant patients (92.9 �2.9 vs 21.1 � 8.2 vs 11.2 � 9.1%, respec-tively, P � .001, Figure 2). Leukocytesalso stained for COX-2. When there wasvessel staining for COX-2, leukocytesthat were stained for COX-2 werepresent 75-81% of the time. Preeclamp-tic patients had significantly more vesselswith leukocytes stained for COX-2 thannormal pregnant or normal nonpreg-nant patients (69.4 � 5.4 vs 16.6 � 3.2 vs9.1 � 1.0%, P � .001, Figure 3). Vesselsstained for COX-2 had staining in vascu-lar smooth muscle cells, as well as in en-dothelial cells. Preeclamptic patients hadsignificantly more staining for COX-2 invascular smooth muscle than normalpregnant or normal nonpregnant pa-tients (77.7 � 4.3 vs 18.1 � 10.3 vs 11.2� 1.0%, P � .001, Figure 3).

FIGURE 3Percent of vessels with neutrophils and vascular smooth muscle stained for NF-�B and COX-2

Preeclamptic patients had significantly more vessels with neutrophils stained for NF-�B and COX-2 as compared to normal nonpregnant or normalpregnant patients. Preeclamptic patients also had significantly more vessels with vascular smooth muscle stained for NF-�B and COX-2 as comparedto normal nonpregnant or normal pregnant patients. **P � .01, ***P � .001.



Figure 4 shows representative stainingresults for CD66b. There was no stainingfor CD66b in the IgM negative control(A). There was little or no staining in ves-sels of normal nonpregnant (B) or nor-mal pregnant (C) women. In contrast,there was extensive brown staining forCD66b in vessels of preeclamptic women(D-G). Staining for naphthol AS-D chlo-roacetate esterase activity shown in panelH further confirmed extensive involve-ment of neutrophils in vessels of pre-eclamptic women.

Figure 5 shows representative stain-ing results for NF-�B. There was nostaining for NF-�B in the IgG negativecontrol (A). There was little or nostaining in vessels of normal nonpreg-nant (B) or normal pregnant (C)women. In contrast, there was intensebrown staining for NF-�B in pre-eclamptic patients (D-F) in endothelialcells and in vascular smooth musclecells. Leukocytes stained for NF-�Bwere also apparent in the lumen andflattened and adhered to endothelium.Under high power (1000�), NF-�Bstaining could be discerned in nuclei ofvascular smooth muscle cells (F).

Figure 6 shows representative stainingresults for COX-2. There was no stainingfor COX-2 in the IgG negative control(A) and little to no staining in vessels ofnormal nonpregnant (B) or normalpregnant (C) women. In contrast, therewas intense brown staining of COX-2along the endothelium, in the vascularsmooth muscle and in leukocytes of pre-eclamptic women (D).

COMMENTIn this study we report significant activa-tion of NF-�B and expression of COX-2in endothelium and vascular smoothmuscle of systemic blood vessels inwomen with preeclampsia. Vascular ac-tivation of NF-�B and expression ofCOX-2 were associated with significantinfiltration of neutrophils as identifiedby the neutrophil marker, CD66b. Leu-kocytes in blood vessels were also appar-ent by their activation of NF-�B and ex-pression of COX-2 in women withpreeclampsia. NF-�B and COX-2 stain-ing of leukocytes primarily represented

FIGURE 4Representative sections of vessels in subcutaneous fatimmunostained for CD66b, a neutrophil marker

A, IgM negative control. B, Normal nonpregnant. C, Normal pregnant. D-H, Preeclamptic. Vesselsof normal nonpregnant and normal pregnant patients had little or no staining for CD66b. Vesselsof preeclamptic patients showed extensive brown staining of neutrophils in the lumen, adhered andflattened along the endothelium, infiltrated into the intimal space and present within the vascularsmooth muscle (VSM). Diffuse staining was also evident, which probably represented CD66bsecreted by neutrophils or neutrophils undergoing apoptosis. Staining for naphthol AS-D chloro-acetate esterase activity in neutrophils further confirmed neutrophil activity in preeclamptic patients(H). A, adipocyte; VL, vessel lumen; VSM, vascular smooth muscle. All images are 400�, exceptH, which is 1000� magnification. Scale bars indicate 50 �m



neutrophils because they are the majorleukocyte class that infiltrates the vascu-lature in preeclampsia.16 Neutrophilswere observed to be flattened and ad-hered to endothelium and present in theintimal space.

Other workers have demonstratedactivation of leukocytes in the circula-tion of women with preeclampsia bymeasuring surface markers or produc-

tion of ROS,4-8 but this is the first re-port that NF-�B is activated and thatleukocytes express COX-2. Leukocyteactivation likely occurs as they circu-late through the intervillous space andare exposed to oxidized lipids secretedby the placenta.2,17 Circulating oxi-dized lipids2,17 or other inflammatorymediators, such as TNF�,18,19 mightinitially activate endothelial produc-

tion of IL-8 to attract neutrophils, butonce adhered, neutrophil productionof TNF�, myeloperoxidase, and ROSwould exacerbate inflammation anddysfunction of the endothelium.

Infiltration of neutrophils to the inti-mal space may be responsible for inflam-mation and dysfunction of the vascularsmooth muscle. We found activation ofNF-�B and expression of COX-2 and in aprevious study expression of IL-8 andICAM-110 by vascular smooth muscle inpreeclamptic women. These findingsdemonstrate inflammation of systemicvascular smooth muscle in women withpreeclampsia. Neutrophils may be themediators of this inflammation throughtheir production of TNF�, myeloperox-idase, and ROS. Neutrophil release ofROS may also contribute to hyperten-sion because ROS in combination withlinoleic acid, which is elevated in womenwith preeclampsia,20 stimulate vascularsmooth muscle production of throm-boxane, a potent vasoconstrictor.21 Acti-vated neutrophils produce thrombox-ane, so infiltrated neutrophils might alsocontribute to vasoconstriction and hy-pertension of preeclampsia. Neutrophilproduction of thromboxane is inducedby oxidative stress and blocked by spe-cific inhibition of COX-2.22 Anothersource of thromboxane in preeclampsiais the placenta. Placental production ofthromboxane is significantly increasedin preeclampsia,23 and it is mainly de-rived from the trophoblast cells.24 Tro-phoblast cells express COX-2 in pre-eclampsia and specific COX-2 inhibitionsignificantly reduces trophoblast pro-duction of thromboxane.25,26

These new findings present novel ap-proaches for the treatment of pre-eclampsia, such as using neutralizing an-tibodies against cell adhesion moleculesto prevent neutrophil infiltration, usinginhibitors of NF-�B or using inhibitorsof COX-2. Potential efficacy of using aCOX-2 inhibitor to treat women withpreeclampsia is suggested by the abilityof low-dose aspirin, which inhibits bothCOX-2 and COX-1, to significantly re-duce the incidence of preeclampsia forwomen who are compliant in takingtheir low-dose aspirin.27

FIGURE 5Representative sections of vessels in subcutaneous fatimmunostained for the p65 subunit of NF-�B

The antigenic sites of p65 are exposed upon activation of NF-�B, so our results indicateactivated NF-�B. A, IgG negative control. B, Normal nonpregnant. C, Normal pregnant. D-F,Preeclamptic. Vessels of normal nonpregnant and normal pregnant patients had little or nostaining for NF-�B. Vessels of preeclamptic patients showed intense brown staining for NF-�B.NF-�B staining was evident in the endothelium and vascular smooth muscle. Leukocytes alsostained for NF-�B (bold white arrows in E). Staining for NF-�B was also present within thenuclei of vascular smooth muscle (bold white arrows, F). A, adipocyte; EC, endothelial cells;VL, vessel lumen; VSM, vascular smooth muscle. All images are 400�, except F, which is1000� magnification. Scale bars indicate 50 �m.



In summary, we found activation ofNF-�B and expression of COX-2 associ-ated with neutrophil infiltration in sys-temic blood vessels of women with pre-eclampsia. Neutrophils also demon-strated activation of NF-�B and expres-sion of COX-2. These data show that vas-cular inflammation is associated withneutrophil infiltration in women withpreeclampsia. f

REFERENCES1. Walsh SW, Vaughan JE, Wang Y, RobertsLJ, 2nd.Placental isoprostane is significantly in-creased in preeclampsia. FASEB J 2000;14:1289-96.2. Walsh SW. Maternal-placental interactions ofoxidative stress and antioxidants in preeclamp-sia. Semin Reprod Endocrinol 1998;16:93-104.3. Hubel CA. Oxidative stress in the pathogen-esis of preeclampsia. Proc Soc Exp Biol Med1999;222:222-35.

4. Clark P, Boswell F, Greer IA. The neutrophiland preeclampsia. Semin Reprod Endocrinol1998;16:57-64.5. Greer IA, Haddad NG, Dawes J, JohnstoneFD, Calder AA. Neutrophil activation in preg-nancy-induced hypertension. BJOG 1989;96:978-82.6. Tsukimori K, Maeda H, Ishida K, Nagata H,Koyanagi T, Nakano H. The superoxide gener-ation of neutrophils in normal and preeclampticpregnancies. Obstet Gynecol 1993;81:536-40.7. Gervasi MT, Chaiworapongsa T, Pacora P,Naccasha N, Yoon BH, Maymon E, et al. Phe-notypic and metabolic characteristics of mono-cytes and granulocytes in preeclampsia. Am JObstet Gynecol 2001;185:792-7.8. Sacks GP, Studena K, Sargent K, RedmanCW. Normal pregnancy and preeclampsia bothproduce inflammatory changes in peripheralblood leukocytes akin to those of sepsis. Am JObstet Gynecol 1998;179:80-6.9. Roberts JM, Taylor RN, Musci TJ, RodgersGM, Hubel CA, McLaughlin MK. Preeclampsia:an endothelial cell disorder. Am J Obstet Gy-necol 1989;161:1200-4.

10. Leik CE, Walsh SW. Neutrophils infiltrate re-sistance-sized vessels of subcutaneous fat inwomen with preeclampsia. Hypertension2004;44:72-7.11. Baldwin AS, Jr.The NF-kappa B and Ikappa B proteins: new discoveries and insights.Annu Rev Immunol 1996;14:649-83.12. Barnes PJ, Karin M. Nuclear factor-kappaB: a pivotal transcription factor in chronic in-flammatory diseases. N Engl J Med 1997;336:1066-71.13. Willoughby DA, Moore AR, Colville-NashPR. COX-1, COX-2, and COX-3 and the futuretreatment of chronic inflammatory disease.Lancet 2000;355:646-8.14. Crofford LJ, Lipsky PE, Brooks P, Abram-son SB, Simon LS, van de Putte LB. Basic bi-ology and clinical application of specific cyclo-oxygenase-2 inhibitors. Arthritis Rheum 2000;43:4-13.15. Bonaterra GA, Hildebrandt W, Bodens A,Sauer R, Dugi KA, Deigner HP, et al. Increasedcyclooxygenase-2 expression in peripheralblood mononuclear cells of smokers and hyper-lipidemic subjects. Free Radic Biol Med2005;38:235-42.16. Cadden KA, Walsh SW. Comparison of leu-kocyte classes most likely to cause vasculardysfunction in preeclampsia. J Soc Gynecol In-vest 2006;13(Suppl):228A.17. Walsh SW. Lipid peroxidation in pregnancy.Hypertens Pregnancy 1994;13:1-32.18. Kupferminc MJ, Peaceman AM, Wigton TR,Rehnberg KA, Socol ML. Tumor necrosis fac-tor-alpha is elevated in plasma and amnioticfluid of patients with severe preeclampsia. Am JObstet Gynecol 1994;170:1752-7.19. Williams MA, Farrand A, Mittendorf R, So-rensen TK, Zingheim RW, O’Reilly GC, et al.Maternal second trimester serum tumor necro-sis factor-alpha-soluble receptor p55 (sT-NFp55) and subsequent risk of preeclampsia.Am J Epidemiol 1999;149:323-9.20. Lorentzen B, Drevon CA, Endresen MJ,Henriksen T. Fatty acid pattern of esterified andfree fatty acids in sera of women with normaland pre-eclamptic pregnancy. BJOG 1995;102:530-7.21. Leik CE, Walsh SW. Linoleic acid, but notoleic acid, upregulates production of interleu-kin-8 by human vascular smooth muscle cellsvia arachidonic acid metabolites under condi-tions of oxidative stress. J Soc Gynecol Investig2005;12:593-8.22. Vaughan JE, Walsh SW. Neutrophils frompregnant women produce thromboxane andtumor necrosis factor-alpha in response to lino-leic acid and oxidative stress. Am J Obstet Gy-necol 2005;193:830-5.23. Walsh SW. Preeclampsia: an imbalance inplacental prostacyclin and thromboxane pro-duction. Am J Obstet Gynecol 1985;152:335-40.24. Walsh SW, Wang Y. Trophoblast and pla-cental villous core production of lipid perox-

FIGURE 6Representative sections of vessels in subcutaneous fatimmunostained for COX-2

A, IgG negative control. B, Normal nonpregnant. C, Normal pregnant. D, Preeclamptic. Vessels ofnormal nonpregnant patients showed little or no staining of COX-2. C shows some light brownstaining in a normal pregnant patient. Vessels of preeclamptic patients showed intense brownstaining along the endothelium and in the vascular smooth muscle (D). Leukocytes also stained forCOX-2 (bold red arrow in D). A, adipocyte; EC, endothelial cells; VL, vessel lumen; VSM, vascularsmooth muscle. All images are 400�. Scale bars indicate 50 �m.



ides, thromboxane, and prostacyclin in pre-eclampsia. J Clin Endocrinol Metab 1995;80:1888-93.25. Johnson RD, Sadovsky Y, Graham C, An-teby EY, Polakoski KL, Huang X, et al. The ex-pression and activity of prostaglandin H syn-

thase-2 is enhanced in trophoblast fromwomen with preeclampsia. J Clin EndocrinolMetab 1997;82:3059-62.26. Woodworth SH, Li X, Lei ZM, Rao CV, Yuss-man MA, Spinnato JA, 2nd, et al. Eicosanoid bio-synthetic enzymes in placental and decidual tis-

sues from preeclamptic pregnancies: increasedexpression of thromboxane-A2 synthase gene.J Clin Endocrinol Metab 1994;78:1225-31.27. Walsh SW. Eicosanoids in preeclampsia.Prostaglandins Leukot Essent Fatty Acids2004;70:223-32.



Activation of NF-κB and expression of COX2 in association with neutrophil infiltration in systemic...

Documents

Transcript of Activation of NF-κB and expression of COX2 in association with neutrophil infiltration in systemic...