914 Volume 7 . Number 6 . 1996

Role of Tumor Necrosis Factor-a on Mesangial Cell MCP-1Expression and Monocyte Migration: MechanismsMediated by Signal Transduction1Rama Pai, Hunjoo Ha, Michael A. Kirschenbaum,2 and Vaijinath S. Kamanna

R. Pai, H. Ha, MA. Kirschenbaum, VS. Kamanna,Nephrology Section, Department of Veterans AffairsMedical Oenter, Long Beach, CA, and the Division ofNephrology, Department of Medicine, University ofCalifornia, Irvine, CA

(J, Am, Soc. Nephrol. 1996; 7:914-923)

ABSTRACTMonocyfe chemotactic protein-i (MCP-1), a specific

chemoaltractant for monocytes, has been thought toplay an important role in the recruitment and accu-mulation of monocytes within the glomerulus seen inglomerular diseases. This study examined the role oftumor necrosis factor (TNF)-cr-mediated cellular signaltransduction pathways on mesangial cell MCP-1

gene expression and monocyte migration. Incuba-tion of mesangial cells with TNF-a stimulated MOP-i

mRNA expression in a dose- and time-dependentmanner. Phorbol myristate acetate (PMA), a protein

kinase C (PKC) activator, increased MOP-i messageby mesangial cells while depleting PKC decreasedMCP-1 gene expression to control levels. Activation of

PKC-depleted mesangial cells with PMA but not withTNF-a inhibited MOP-i mRNA expression. Similarly, cal-phostin C, a PKC inhibitor, failed to inhibit TNF-a-

induced MOP-i expression. The incubation of mesan-gial cells with various protein tyrosine kinase inhibitors(P1K, e.g., herbimycin, tyrphostin, genistein) blockedTNF-a-induced MOP-i mRNA message. Additional ex-

periments examining the role of cAMP on MOP-i

expression indicated that the preincubation of mes-angial cells with various cAMP generating substances(pertussis toxin, isoproterenol, dbcAMP) did not in-duce mesangial cell MOP-i mRNA transcripts. How-ever, the coincubation of mesangial cells with TNF-aand dbcAMP completely inhibited TNF-a-inducedMCP-1 gene expression. Finally, TNF-a-activated mes-angial cell media increased monocyte transmigra-tion that could be blocked by neutralizing anti-MOP-i.These studies indicate that TNF-a facilitates monocyte

1 ReceIved August 29, 1995. Accepted February 22, 1996.

2 correspondence to Dr. MA. Klrschenbaum, Nephrology Section (111N), De-

partment of Veterans Affairs Medical center, 5901 East Seventh Street, LongBeach, CA 90822.

1046-6673/0706-0914$03.OO/OJournal of the American Society of Nephrologycopyright © 1996 by the American Society of Nephrology

transmigration into the glomerulus mediated by the

increased expression of MOP-i by mesangial cells.TNF-a-induced mesangial cell MOP-i expression isregulated by signal transduction pathways involvingPTK but not those dependent on PKC or cAMP.

Key Words: Mesangial cells, monocyte chemoatfraction,

MCP- 1, TNF-cr, signal transduction

M esangial cells, specialized vascular pericytes

within the glomemulus, perform a fundamentalrole in the homeostasis of gbomemular function. Their

unique location within the gbomemubus, their contract-

ibiity characteristics, and their ability to produce

pseudopodia that project into the glomerular capillary

lumen allow mesangial cells to interact with other

resident and bone-marrow derived cells and numer-

ous circulating chemical modulators ( 1 ). In addition,mesangial cells can express a variety of cytoregulatory

peptides, including chemokines, that attract circulat-

ing mononuclear leukocytes into the gbomerulus (2,3).

The adherence and subsequent transmigration of cir-culating monocytes into the glomemular mesangium

occurs early in the development of various forms of

renal diseases, including focal and segmental glomer-

ubosclerosis (4,5), membranoprolifemative gbomemubo-

nephmitis (6 , 7), antigbomerular basement membrane

disease (8), immune complex nephmitis (9), pumomycinaminonucleoside nephritis ( 1 0, 1 1 ), and the gbomeru-

bar injury associated with hyperlipidemia or choles-

terob feeding ( 1 2). Although the cellular and molecular

mechanisms underlying monocyte infiltration Into the

gbomemulus are poorly understood, monocyte recruit-

ment into the mesangium seems to involve a series ofcomplex receptor-mediated interactions between res-

ident glomerular cells and circulating mononuclearleukocytes. These interactions are intensified by theenhanced expression of glomerular endothelial and

mesangial cell-derived adhesion molecules and by the

elaborated expression of monocyte chemokines by

intrinsic gbomerular cells (2,3,13-15).

Monocyte chemotactic protein- 1 (MOP- 1 ) belongs to

a family ofchemokines that may be involved in inflam-matomy and tissue repair processes, including the

chemoattraction of mononuclear phagocytes into the

mesangium (3, 16). MOP- 1 is a monomeric polypeptidethat corresponds to the early response gene JE of the

mouse and is produced by a variety of cells, including

endothelial, smooth muscle, and mesangial cells and

monocyte/macrophages ( i 7-2 1 ). MOP- 1 is expressed

in response to proinflammatory factors such as inter-

leukin (IL)- 1 , tumor necrosis factor (TNF)-a, and IFN-y

( 18,22). Although original studies have indicated that

Pal et al

Journal of the American Society of Nephrology 915

MOP- i acts as a specific factor for monocyte/macro-

phage migration (23,24), recent observations suggest

that this chemokine can also act as a chemoattractant

for T-lymphocytes (25,26). Furthermore, MOP- 1 ac-

counts for nearly all of the monocyte chemotactic

activity expressed by vascular cells (27).

Although the increased expression ofmonocyte che-moattractants and the subsequent accumulation of

monocyte / macrophages in the mesangium are seen in

many experimental models of renal injury, the cellular

and molecular mechanisms associated with this in-

fiammatory process are poorly understood. In this

regard, many in vitro and in vivo studies have sug-

gested that TNF-a is a fundamental proinfiammatory

cytokine involved in the pathogenesis of gbomerular

injury (28,29). Furthermore, in response to activation,

both infiltrating macrophages and intrinsic gbomemu-

lar cells may secrete TNF-a, allowing this peptide to

serve as an important proinfiammatomy cytokine as-

sociated with the pathogenesis of gbomerular inflam-

mation and injury.Despite several studies showing that TNF-a induces

various cytomegulatomy peptides by both gbomemular

and nonglomerular cells, the cellular signaling path-

ways between TNF-a binding to its cell-surface recep-

tom and subsequent nuclear signals that regulate TNF-

a-induced gene expression in gbomerular mesangial

cells are not clearly understood. Thus, we examined

the ability of exogenous TNF-a to influence the geneexpression of MOP- 1 by murine mesangial cells in an

attempt to further elucidate the robe of TNF-a in the

pathobiobogy of gbomerular injury. Additional studies

were performed to understand the multiple signal

transduction pathways mediated through protein ki-

nase 0 (PKO), protein tyrosine kinase (PTK), and cyclicnucleotides that may be involved in TNF-a-induced

MCP- 1 gene expression. The biological significance of

increased mesangial cell synthesis and secretion of

MOP- i was assessed by its ability to modulate mono-

cyte chemotaxis.

MATERIALS AND METHODS

Materials

Dulbecco’s minimal essential media (DMEM), RPM! 1640,dibutyryl (db)cAMP, phorbol myristate acetate (PMA), andother tissue culture reagents were obtained from SigmaChemical Company, St. Louis, MO. Fetal bovine serum (FBS)was obtained from Hyclone Laboratories Inc. , Logan. UT.Human recombinant TNF-a was purchased from R & DSystems. Minneapolis, MN. cDNA probes for murine MOP-iand human glyceraldehyde-3-phosphate dehydrogenase(GAPDH) were procured from American Type Culture Collec-tion (ATCO, Rockville, MD). MSITM nylon transfer membraneswere obtained from Fisher Scientific, Tustin, CA. Oalphostmn0, isoproterenol, pertussis toxin, genistein, tyrphostin B46,herbimycin A, and isobutyl- 1 -methylxanthine (IBMX) were

purchased from Calbiochem, La Jolla, CA. Murine anti-MOP- 1 was obtained from Pharmingen, San Diego. CA. All

other chemicals used were of analytical grade.

Mesangial Cell Culture

A murine mesangial cell line (MES- 13, cloned from micetransgenic for the early region of SV-40 virus, Passage 25)

was obtained from ATCC, Rockville, MD. The mesangial cellswere grown in DMEM containing 5% FBS, 1% glutamine-

streptomycmn-penidilhin mixture. 44 mM NaHCO3. and 14mM N-hydroxyethylpiperazine-N’-2-ethanesufforic acid

(HEPES) in an atmosphere of 5% 002 and 95% air at 37#{176}Cina humidified incubator. Subcultures were made from conflu-

ent stock cultures by trypsinization in phosphate-bufferedsaline (PBS) containing 0.5 mM EDTA and 0.025% trypsin. A

murine monocyte/macrophage cell line (J774.A1 ) was ob-tamed from ATCC and cultured in RPMI 1 640 media supple-mented with L-glutamine, penicillin/streptomycin, and 10%FBS.

Monocyfe Chemotaxis Assay

The ability ofmesangial cell conditioned media in response

to TNF-a activation to modulate monocyte migration was

examined by in vitro monocyte chemotaxis assay as de-scribed earlier by Falk et al. (30). TNF-cr activated mesangial

cell conditioned media was obtained from mesangial cellsincubated with varying concentrations of TNF-a (0 to 200ng/mL) for 3 h at 37#{176}C.An aliquot of this medium (five-foldconcentrailon) was placed in the lower compartment of a

microchemotaxis assembly (Neuro Probe, Inc. , Cabin John,MD) with two compartment chambers separated by a 5-smpolycarbonate membrane filter. Murine monocyte I macro-phages (J774.Ai, 6 x i04 cells) were added to the upper

chamber and incubated at 37#{176}Cfor 3 h in a humidified 5%

002 incubator. After incubation, the membrane filter wasremoved, stained with Duff-quick solution (Fisher Scientific,

Tustin, CA) and mounted on a glass slide. The migratedmonocytes were counted on an oil immersion microscope

with a 20X magnification containing 10-mm2 counting grid.Four grid areas were counted per sample and averaged.

Similar chemotaxis assay was performed with control mes-

angial cell conditioned media and compared with the results

obtained by utilizing TNF-a activated mesangial cell condi-

tioned media. Appropriate exogenous TNF-a controls were

used to normalize the data obtained from cytokmne-activatedcell supernates. Neutralizing anti-MOP- 1 was used in thechemotaxis assay medium to determine the biological spec-ificity of MOP- 1 secreted by mesangial cells to stimulatemonocyte migration.

Mesangial Cell MCP-i Gene Expression:Response of TNF-a Dose and Incubation Time

Mesangial cells (5 x 106) were grown in DMEM media

containing 5% FBS at 37#{176}Cfor 24 h to attain 75 to 80%

confluence. The media was replaced with DMEM + 5% FBS

containingvarious concentrations ofTNF-a (0 to 200 ng/mL)

and Incubated for 3 h at 37#{176}C.After incubation, the cellmonolayer was washed and used for RNA isolation. The

incubation time kinetics studies were performed by growingmesangial cells (5 x 106) in DMEM media containing 5% FBS

for 24 h. The media was replaced with fresh DMEM + 5%FBS containing 25 ng/mL of TNF-a and incubated at 37#{176}O

TNF-a Activation of Mesangial Cell MOP-i

916 Volume 7 . Number 6 - 1996

for varying periods (0.5 to 24 h). After incubation, cellmonolayers were washed and used for RNA isolation.

Mesangial Cell MCPI Gene Expression inResponse to TNF-a Activation: RegulationThrough Protein Kinases

These experiments were performed to determine the regu-latory role of various protein kinases and cyclic-AMP (cAMP)

on TNF-a-induced gene expression by mesangial cells. Theeffect of PKC on mesangial cell gene expression was evalu-ated by incubating cells with either inducers or inhibitors ofPKC. Mesangial cells (5 X 106) were grown in DMEM contain-ing 5% FBS for 24 h to attain 75 to 80% confluence. Themedia was replaced with fresh DMEM + 5% FBS containingPMA (50 nM), an activator of PKC, and incubated for 3 h at37#{176}C.In other experiments, intracellular PKC was depletedby incubating mesangial cells with 50 nM PMA for 24 h at37#{176}C.After predepletion ofintracellular PKC, mesangial cellswere stimulated by incubating them with either PMA (50 nM)or TNF-cx (10 ng/mL) for 3 h at 37#{176}C.After the incubation,mesangial cell monolayers were washed and used for RNA

Isolation. Further experiments were performed to assess theeffect of PKC on cellular gene expression by preincubatingmesangial cells (5 x 106) at 37#{176}Cfor 1 h with a PKC inhibitor,caiphostin C I 100 nM). After preincubation with calphostinC, mesangial cells were stimulated for 3 h at 37#{176}Cwith 10ng/mL TNF-a. Mter the incubation, cell monolayers werewashed and used for RNA isolation.

Additional studies were designed to examine the role ofcAMP-mediated cellular signalling pathways in mesangial

cell MOP- 1 expression. Mesangial cells (5 x 106) were grown

in DMEM + 5% FBS for 24 h at 37#{176}Cto attain 75 to 80%confluence. The media was replaced with fresh DMEM + 5%FBS containing various cAMP generating agents. includingpertussis toxin (50 ng/mL), isoproterenol (5 j.�m) or dbcAMP( 1 mM), and incubated for 3 h at 37#{176}C.Experiments were alsoperformed to determine the effect ofcAMP on TNF-a-inducedmesangial cell MCP- 1 gene expression. Mesangial cells werecoincubated with TNF-a and dbcAMP for 3 h. Additionalstudies were done to examine the role of Intracellular cAMPdegradation in the observed changes in MCP- 1 gene expres-sion. Mesangial cells were incubated for 3 h with 0.5 mMIBMX alone or in the presence of cAMP-generating sub-stances. After the incubation, mesangial cell monolayer waswashed, RNA isolated and used for Northern blot analysis.

The role of PTK was examined by preincubating cells withthree different PTK inhibitors. Because these structurallydifferent PTK inhibitors used in the study may have different

modes of action on tyrosine phosphorylation (3 1-33), wehave used three structurally distinct PTK inhibitors to assessthe role ofTNF-a in MOP- 1 expression. Additionally, becauseoftheir structural dissimilarity. previous studies have shownthat each inhibitor requires different incubation times topenetrate cells for their optimal activity. Therefore, we haveused these different inhibitors at various incubation times asreported earlier by others (34). Mesangial cells (5 x 106) were

grown in DMEM containing 5% FBS for 24 h at 37#{176}Cto attain75 to 80% confluence. The media was replaced with freshDMEM + 5% FBS and preincubated with genistein (25�g/mL) for 15 mm, herbimycin A I 10 �M) for 1 h, or tyrphos-tin B46 ( 100 �M) for 2 h at 37#{176}C.After preincubation withPTK inhibitors, mesangial cells were stimulated with TNF-a(10 ng/mL) for 3 h at 37#{176}C,cells were washed, and RNA

isolated for Northern hybridization.

Mesangial Cell MCP-1 Gene Expression inResponse to TNF-a Activation: Effect ofCycloheximide

Additional studies were undertaken to determine the reg-

ulatory role of protein synthesis on TNF-a-induced mesan-gial cell MCP- 1 gene expression. Mesangial cells (5 X 106)

grown in DMEM containing 5% FBS for 24 h to attain 75 to80% confluence. The media was replaced with DMEM + 5%FBS, preincubated with cycloheximide ( 10 j.�g/mL) for 2 hand stimulated with TNF-a for 3 h. After incubation, cellmonolayers were washed, and RNA was isolated and used for

Northern blot analysis.

Northern Blot Analysis

After the media was removed, mesangial cells were col-lected and washed three times with PBS. Total RNA wasisolated from the cells using the protocol described by Chom-czynski and Sacchi (35). In brief, cells were homogenizedwith 4 M guanidium thiocyanate total protein and DNAextracted with acid phenol, and RNA precipitated with iso-propanol. After they were washed with ethanol, samples weredried under vacuum centrifugation and the amount of RNA

was quantitated by measuring the absorbance at 260 am

using a spectrophotometer. Thirty �g of total RNA was loadedinto individual wells of a 1 .2% agarose gel containing form-aldehyde and electrophoresis performed (36). The RNA fromthe gel was transferred onto MSI nylon membranes using acapillary transblotting technique. The nylon membrane was

ultraviolet (UV)-linked using a LW Crosslinker (Fisher Scien-

tific, Pittsburgh, PA). The cDNA probes for MOP- 1 andGAPDH were used for hybridization after [32PIdOTP-labelingby random oligonucleotide priming. The membrane was hy-

bridized using a 32P-labeled murine MOP- 1 cDNA probe. Themembranes were washed three times for 30 min: first in 2X

standard saline citrate (SSO) with 0. 1% SDS at 25#{176}C;second

in 0.2X SSO with 0. 1% SDS at 25#{176}C:and third in 0.2X 550with 0. 1% SDS at 55#{176}C.Autoradiography was performed byexposing the blots to Kodak X-ray film (Eastman Kodak,

Rochester, NY) with intensifying screens at - 70#{176}C.Blotswere then rehybridized with a [32Pl-labeled human GAPDHcDNA probe as an internal control to assess RNA quantityand integrity. Quantitation of mRNA signals was performedby densitometric scanning of autoradiographs and normal-

ized with the GAPDH mRNA signal.

Statistical Analysis

Results are presented as mean values ± standard error(SE) for 3 to 4 separate experiments. The t test was used tocompare the means and a P value of less than 0.05 wasconsidered significant.

RESULTS

In this study, we have used SV-40 transformedmumine mesangial cells as an in vitro model to examinethe role of TNF-a on gbomemular mesangial cell MCP- 1

gene expression. These transformed cells were devel-oped by cloning mesangial cells from tmansgenic mice

for the early region of SV-40 virus and have beenshown to exhibit similar characteristics to those of

primary cultures of murine mesangial cells (37). Inaddition to maintaining normal differentiation char-acteristics, transformed mesangial cells exhibit con-

tractile properties in response to vasoactive sub-

-4.

0.5

- +

3

MCP-1 (0.9 kb) MCP-1 (0.9 kb)

Pal et al

Journal of the American Society of Nephrology 917

stances and produce various cytokines analogous to

primary cultures of mesangial cells (38,39), including

MOP-i.Utilizing these transformed murine mesangial cells,

studies were performed to examine the effect of exog-

enous TNF-a on MOP- 1 steady-state mRNA expres-

sion. The incubation of mesangial cells with TNF-a for3 h stimulated MOP- 1 mRNA expression in a dose-

dependent fashion (Figure 1 ). TNF-a, as low as 0.5ng/mL, induced mesangial cell MOP- 1 mRNA tran-

scripts and reached a maximal level at 25 ng/mL of

TNF-a in the incubation media. Quantitative analysisof MCP- 1 gene expression was performed by densito-

metric scanning of autoradiographs and was normal-

ized with a GAPDH mRNA signal. This showed a

dose-dependent increase in MCP- 1 mRNA gene ex-

pression by TNF-a-treated mesangial cells: concentra-tion of TNF-a, 0, 0.5, 1 , 5, 25 ng/mL; respective

arbitrary densitometric values were 0. 1 1 ± 0.001,0.73 ± 0.005 (P < 0.0001), 0.85 ± 0.006 (P = 0.003),

1.06 ± 0.006 (P < 0.0001), 1.09 ± 0.16 (P < 0.003),1 .45 ± 0.004 (P < 0.000 1); all P values were calculated

by comparing the experimental values with control. At

higher TNF-a concentrations (50 to 200 ng/mL), nofurther induction of mesangial cell MOP- 1 mRNA ex-pression was seen when compared with 25 ng/mL ofTNF-a (data not shown). Although a significant induc-

tion of MOP- 1 was noted at 30 mm (arbitrary densito-

metric values for 30 mm control, 0.07 ± 0.005; TNF-a,

0.22 ± 0.00 1 ; P = 0.004), the maximal effect of TNF-a

on mesangial cell MCP- 1 mRNA gene expression was

observed at 3 h (arbitrary densitometric values for 3 h

control, 0.13 ± 0.004; TNF-a, 1.24 ± 0.03, P

0.000 1) of incubation and the message remained sta-ble at least up to 24 h of incubation (Figure 2, arbi-

trary densitometric values for 24 h control, 0.04 ±

0.0001; TNF-a, 1.0 ± 0.07; P 0.0001). On the basis

of these observations, mesangial cell-activation with

MCP-1 (0.9 kb)

GAPDH (1 .3 kb)

- + - + TNF-cc

6 24 incubation time (h)

Figure 2. Effect of TNF-a on the murine mesangial cell timekinetics of steady-state MCP-i mRNA expression by Northernblot analysis. Mesangial cells were incubated in the pres-ence or absence of TNF-a for varying time periods (0.5 to24 h) and cells were analyzed for MCP-i mRNA expression asdescribed In the Methods section. Northern blot Is represen-tative of three separate experiments.

TNF-a was performed for 3 h in all subsequent stud-

ies.

Additional studies were performed to identify the

cellular signal transduction pathways associated with

TNF-a-induced mesangial cell MCP- 1 gene expression.

Incubation of mesangial cells with PMA, an activator

ofPKO, for 3 h induced the steady-state mRNA expres-

sion of MOP- 1 (Figure 3, Lane 3 as compared withLane 1 : arbitrary densitometmic values for control,0.17 ± 0.02; PMA, 1.17 ± 0.15; P 0.003). The

depletion of mesangial cell PKO by prolonged incuba-

tion with PMA (24 h) completely blocked the MOP-i

mRNA message and the signal was comparable to

control cells without any activation (Figure 3, Lane 4

compared with Lane 3: arbitrary densitometric values

for PKO-depletion, 0.30 ± 0.002; PMA, 1 . 1 7 ± 0. 15;

P = 0.005). Additional studies indicated that the

activation of PKO-depleted mesangial cells with PMA

did not induce mesangial cell MOP- 1 expression, sug-gesting the regulatory role of PKO in mesangial cell

C 0.5 1 5 25

TNF-ot (nglml)

GAPDH (1 .3 kb)

Figure 1 . Effect of TNF-a on murine mesangial cell steady-state mRNA expression of MCP-i by Northern blot analysis.After incubation of mesangial cells with TNF-a (0 to 25 ng/mL)for 3 h, cell monolayers were washed with PBS and RNAisolated, electrophoresed, and transferred onto nylon mem-branes. Northern blots were hybridized with 32P-labeled mu-rine cDNA probe for MCP-i and then rehybridized with cDNAprobe for GAPDH. Autoradiographs were developed byexposing blots to X-ray films. Northern blot is representativeof three separate experiments.

12345678

GAPDH (1 .3 kb)

Figure 3. Role of PKC in basal and TNF-a-induced mesangialcell MCP-1 gene expression. Mesangial cells were incu-bated with the following substances: control (Lane 1), TNF-afor 3 h (Lane 2), PMA for 3 h (Lane 3), PMA for 24 h (Lane 4),PMA for 24 h and activation with TNF-a for 3 h (Lane 5), PMAfor 24 h and activation with PMA for 3 h (Lane 6), TNF-a for 3 h(Lane 7), and TNF-a plus calphostin C for 3 h (Lane 8). Alllanes in this figure were derived from a single gel andgraphically rearranged for clarity. For more details, seeMethods section. After incubation, mesanglal cell monolay-ers were washed, RNA isolated, and used for Northern blotanalysis. The Northern blot is representative of four separateexperiments.

TNF-a Activation of Mesangial Cell MOP-i

918 Volume 7 - Number 6 . 1996

MCP- 1 gene expression (Figure 3, Lane 6 as compared

with Lane 4: arbitrary densitometric values for PKO

depletion, 0.30 ± 0.002; PKO-depletion + PMA, 0.50

± 0.08). Howevem, the incubation of PKO-depletedmesangial cells with TNF-a had no effect on the inhi-

bition of mesangial cell MOP- 1 mRNA signal and the

message was similar to TNF-a activation of non-PKO-

depleted cells (Figure 3, Lanes 5 and 2: arbitrary

densitometric values for PKO-depletion + TNF-a, 0.75

± 0.33; TNF-a, 1 . 14 ± 0. 13). Further confirmatory

experiments regarding the role of PKO on TNF-a-

induced mesangial cell MOP- 1 expression were per-

formed by utilizing calphostin 0, a specific PKO inhib-itom. Data obtained from these studies indicated that

the preincubation ofmesangial cells with calphostin 0

did not significantly inhibit TNF-a-induced mRNA ex-pression of MOP- 1 when compared with the stimula-

tion with TNF-a (Figure 3, Lanes 8 and 7, respectively:

arbitrary densitometric values for TNF-a, 1 . 14 ± 0.13;

TNF-a + calphostin 0, 1 .083 ± 0.03).

The mole of PTK in mesangial cell mRNA expressionof MOP- 1 was examined by utilizing various PTK

inhibitors (e.g. , genistein, hembimycin, and tymphos-

tin). As shown in Figure 4, preincubation of mesangial

cells with genistein, herbimycin, or tyrphostin caused

a marked inhibition of TNF-a-induced MOP- 1 expres-sion (Figure 4, Lanes 3, 4, and 5 as compared with

Lane 2: arbitrary densitometmic values for TNF-a, 2.17± 0.12; genistein + TNF-a, 0.14 ± 0.01; herbimycin +

TNF-a, 0.15 ± 0.02; tyrphostin + TNF-a, 0.13 ±

0.006; P < 0.0001 for all experimental values corn-pared with TNF-a control).

We further investigated the role of cAMP as a second

messenger in mesangial cell MOP- 1 mRNA message by

utilizing cAMP generating substances (i.e. , pertussis

toxin or isoproterenol) and the cell permeable analogof cAMP (dbcAMP). Preincubation of mesangial cells

1 2 3 4 5

MCP-1 (0.9 kb)

GAPDH (1.3 kb)

Figure 4. Effect of PTK inhibitors on TNF-a--induced MCP-imRNA expression by mesangial cells. Lanes represent theincubation of mesangial cells for 3 h (Lane 1) and TNF-a(Lane 2). Mesangial cells were preincubated with genisteinfor 15 mm (Lane 3), herbimycin for 1 h (Lane 4), andtyrphostin for 2 h (Lane 5) before TNF-a activation for 3 h. Afterincubation, mesangial cell monolayers were washed, RNAisolated, and used for Northern blot analysis as outlined inthe Methods section. All lanes in this figure were derived froma single gel and graphically rearranged for clarity. TheNorthern blot is representative of three separate experi-ments.

with pertussis toxin, isoprotemenol, or dbcAMP for 3 h

did not cause any induction in mesangial cell MOP-i

message when compared with controls (Figure 5,

Lanes 3, 4, 5 as compared with Lane 1 , arbitrary

densitometmic values for control, 0.70 ± 0.005; per-

tussis toxin, 0.55 ± 0.03; isoprotemenol, 0.30 ± 0.002;

dbcAMP, 0.30 ± 0.03). Furthermore, studies were

designed to examine whether an increase in intracel-lular cAMP would have any effect on TNF-a-induced

mesangial cell MOP- 1 mRNA expression. Ooincuba-

tion of mesangial cells with dbcAMP and TNF-a for 3 h

resulted in an inhibition of TNF-a-induced mesangial

cell MOP- 1 message (Figure 5, Lane 6 as comparedwith Lane 2: arbitrary densitometmic values for TNF-a

+ dbcAMP, 0.13 ± 0.01; TNF-a, 1.22 ± 0.16; P

0.002). Additional experiments were performed to ex-

amine the effects of cAMP-generating substances in

the presence of IBMX, a phosphodiestemase inhibitor,

to determine whether cAMP degradation may have anyinfluence on mesangial cell MOP- 1 expression. The

data from these experiments showed that the incuba-

tion of mesangial cells with cAMP generating sub-stances also did not stimulate mesangial cell MOP-i

mRNA transcripts in the presence of IBMX (data not

shown).

To examine whether de novo protein synthesis is

necessary for TNF-a-induced MOP- 1 gene expression,mesangial cells were preincubated with cycloheximide

( 10 �g/mL) for 2 h before the activation of these cellswith TNF-a for 3 h. The inhibition of cellular protein

synthesis by cycboheximide inhibited TNF-a-induced

mesanglal cell MCP- 1 gene expression without alter-

ing GAPDH message (Figure 6: arbitrary densitometmic

values for TNF-a, 1 .05 ± 0. 1 1 ; TNF-a + cycbohexi-

mide, 0.070 ± 0.002; P 0.0009).

Because the results of the Northern hybridization

indicated that the activation of mesangial cells with

TNF-a stimulate the steady-state mRNA expression of

MOP- 1 , further studies were performed to examine the

ability of TNF-a-mediated secretion of MOP- 1 to stim-

123456

MCP-1 (0.9 kb)

GAPDH (1 .3 kb)

Figure 5. Effect of cAMP-generating substances on basal andTNF-a-stimulated mesangial cell MCP-i gene expression.Mesangial cells were incubated for 3 h with the followingsubstances: control (Lane 1), TNF-a (Lane 2), pertussis toxin(Lane 3), isoproterenol (Lane 4), dbcAMP (Lane 5), anddbcAMP plus TNF-a (Lane 6). After incubation, mesangial cellmonolayers were washed, RNA isolated, and used for North-em blot analysis as outlined in the Methods section. All lanesin this figure were derived from a single gel and graphicallyrearranged for clarity. The Northern blot is representative ofthree separate experiments.

ated with biological activity to induce monocyte migra-

80

60

40

*

Cl)w

C.)0C0Ea)C

a)E

I,-0

a).0

EC

*

20 * **

*

0

*

Pal et al

Journal of the American Society of Nephrology 919

., MCP-1 (0.9 kb)

..#{149} GAPDH(1.3kb)

- #{247} Cycloheximide (10 pg/mI)

+ + TNF-a (25 ng/mI)

Figure 6. Northern blot analysis of the effect of cyclohexi-mide on TNF-a-induced mesangial cell MCP-i mRNA expres-sion. Mesangial cells were preincubated with cyclohexi-mide (10 �g/mL) for 2 h before TNF-a-activation for 3 h.Additional descriptions are given in the text. Northern blot Isrepresentative of three separate experiments.

ulate the migration of monocytes by using a chemo-taxis assay. Oonditioned media obtained by incubat-ing mesangial cells with TNF-a increased monocyte

migration in a dose-dependent manner (Figure 7). Theaddition of neutralizing anti-MOP- 1 to the chemotaxischamber attenuated the migration of monocytes dur-ing basal as well as TNF-a-activated conditions (Fig-

ure 7). Additional experiments were performed to ex-amine the effect of exogenous TNF-a on monocytechemotaxis. Results from these experiments indicated

that the addition of TNF-a (25 ng/mL) to the chemo-

taxis incubation media did not alter monocyte migra-

tion when compared with the appropriate controls

(data not shown). These data indicated that any exog-

enous TNF-a present in the conditioned media used

for determining monocyte chemotaxis was not associ-

*

I TNF, 0 ng/mlI TNF, 0.5 nglml� TNF, 2 nglmlE1 TNF, 5 nglml� TNF, 25 nglml

TNF-a TNF-ct + anti-MCP-1Figure 7. Effect of conditioned media obtained by incubat-ing mesangial cells with TNF-a on monocyte chemotaxis.Mesangial cells were incubated with TNF-a (0 to 25 ng/mL)for 3 h, the media was removed and used for the monocytechemotaxis assay as described in the Methods section.Similar studies were performed in the presence of saturatingconcentrations of neutralizing anti-MCP-i . Data represent theresults of three separate experiments. � � < 0.01.

tion.

DISCUSSION

In this study, we have delineated mesangial cell

signal transduction pathways involved in TNF-a-me-diated MOP- 1 expression and monocyte migration.Increased gbomerular localization of MOP- 1 mRNA

message and protein was shown to be associated withmonocyte/macrophage accumulation within the mes-

angium of human and experimental glomerular dis-

eases (40,4 1 ). The chemoattraction of circulatingmononuclear phagocytes within the mesangium, an

early pathobiological cellular event seen in gbomerular

injury, often precedes the development of functional

and structural changes associated with both immune

and nonimmune-mediated renal injury (4-12). Fur-thermore, monocyte infiltration into the gbomerulushas been shown to be associated with proteinumia and

declining renal function in human subjects and ex-perimental animals (6,42,43). In addition to the asso-

ciation between the increased accumulation of mono-

cytes and gbomerular injury, the depletion of renalcortical macrophages in experimental animals by es-

sential fatty acid diet or by carefully-timed systemic

X-irradiation decreased the progression of gbomemulardisease suggesting a critical regulatory role for mono-

cytes in the pathogenesis of renal disease (8,44). Onthe basis of both in vivo and in vitro studies, it was

suggested that monocyte-mediated renal injury oc-

curs, at least in part, through the enhanced synthesis

and secretion of various cytoregulatory factors (e.g.,cytokines, prostanoids, reactive oxygen species, pro-teolytic enzymes, etc.) that regulate glomerular cell

proliferation, extracellular matrix production, and the

functional and structural integrity of the glomerulus(4,45).

Although various endogenous mediators can initi-ate cellular activation within the glomerulus, TNF-ahas been considered a primary proinflammatory cyto-kine involved in the pathogenesis of glomemular injury(46,47). Supporting evidence for the involvement ofTNF-a in the pathogenesis of glomemubonephritis have

been derived primarily from in vivo studies indicatingthat the administration of exogenous TNF-a or agentsthat induce the release of endogenous TNF-a (e.g.,endotoxin) increases the severity of experimental gb-merular injury (28). Furthermore, the use of specific

neutralizing antibodies or receptor-antagonists to in-

hibit the action of TNF-a has been shown to signifi-

cantly attenuate the progression of glomerular disease(28,48,49). Although these studies suggest a patho-

biological role for TNF-a in the genesis of glomerubo-

sclerosis, the role of this cytokine in the cellular

responses within the glomerulus that activate signal

transduction pathways and stimulate various cellular

genes associated with monocyte infiltration are notclearly defined.

Because MOP- 1 acts as a specific chemoattractant

TNF-a Activation of Mesanglal Cell MOP-i

920 Volume 7 . Number 6 ‘ i996

for the migration ofmonocytes into the gbomerubus, weexamined the cellular signal transduction pathways

associated with TNF-a-mediated mesangial cellMOP- 1 gene expression and monocyte migration. Theactivation of mesangial cells with TNF-a induced

MOP- 1 mRNA transcripts in a dose- and time-depen-

dent manner. The results obtained from these studies

supported earlier observations made in this and other

laboratories that the activation of a variety of differentcell types (including mesangial, smooth muscle, endo-thelial, tubular cells, etc.) in the presence ofTNF-a caninduce prototypic cytomegulatomy peptides (e.g. , adhe-sion molecules, colony-stimulating factors, andmonocyte chemoattractants) associated with mono-cyte adhesion and subsequent transmigration into theglomemulus (34,50-53). Additionally, the conditionedmedia obtained from TNF-cr-activated mesangial cellsincreased monocyte migration as determined by achemotaxis assay. Because the monocyte migrationcould be attenuated by neutralizing anti-MOP- 1 , thesedata suggested that MOP- 1 was a primary factorresponsible for the observed monocyte chemotacticactivity induced by TNF-a-activation of mesangialcells. The pathobiological effects of the increased ex-pression of MOP- 1 associated with TNF-a-mediatedmesangial cell activation may partially explain the

increased glomerular adhesion and accumulation ofmonocytes seen in rabbits that received intravenousinfusions of TNF-a or in nephrotoxic nephmitis ratstreated with exogenous TNF-a (49,55).

Although the cellular mechanisms that regulateglomerular cytokine synthesis and secretion are not

clearly understood, it appears that TNF-a can besecreted by a variety of cells in response to eitherimmune- or nonimmune-mediated stimuli, such asbacterial lipopolysaccharide, interferon-y, and im-mune complexes (28,46). We have recently observedthat atherogenic lipoproteins may also induce mesan-gial cell TNF-cr expression (H. Ha, R. Pal, V.5. Ka-manna, M.A. Kirschenbaum, unpublished observa-tions). Thus, the pleiotropic nature of TNF-a derivedfrom activated resident gbomerular cells and from

infiltrating mononuclear cells may synergistically ini-tiate a cascade of cellular events culminating in thedevelopment of glomerubar injury.

Although TNF-a receptors have not been clearlyshown to possess intrinsic protein kinase activity, thephosphorylation of several distinct proteins in various

cell types was shown to occur within minutes after theexposure of cells to TNF-a (47,56). These cellularsignals in response to TNF-a activation are likely theresult of the activation of several major cellular ki-nases including PKO, cAMP-dependent protein kinaseA, PTK, and mitogen-activated protein kinases. De-

spite these studies, relatively little is known of theTNF-a-mediated signal transduction pathways in-

volved in mesangial cell MOP- 1 expression.Utilizing gene analysis techniques, endothelial cells

were shown to exhibit a phorbol ester-responsive ele-ment at the 5’ -flanking region of the MCP- 1 gene,

suggesting an Involvement of PKO in the regulation of

MCP- 1 gene expression (57). In agreement with these

studies, we have also shown in the study presented

here that the activation of mesangial cells with PMA,

an activator of PKO, stimulated mRNA transcripts for

MOP- 1 indicating the regulatory role of PKO in MCP- 1

gene expression. Additionally, the depletion of mesan-

gial cell PKO by prolonged incubation with PMA nearlycompletely inhibited MOP- 1 message further confirm-ing the mole of PKO in mesangial cell MOP- 1 expres-sion. Prolonged treatment with PMA in endothelial

and other cell types resulted in the down regulation ofPKO and a concomitant loss of phombol ester respon-siveness that rendered MOP- 1 gene expression insen-

sitive to PMA stimulation (58). Similarly, we have

noted that mesangial cells treated with PMA for a

prolonged period of time were insensitive to PMA

activation, suggesting that the phorbol ester-relatedmechanisms leading to MCP- 1 gene expression werealso operating in mesangial cells. These PKO-desensi-

tized cells, unresponsive to PMA stimulation, were

used to distInguish between PKO-dependent or PKO-independent effects of TNF-ct-mediated mesangial cell

MCP- 1 gene expression. Utilizing this approach, weexamined whether TNF-a-induced mesangial cellMOP- 1 expression was mediated through PKO or

through other PKO-independent mechanisms. The in-

cubation of PKO-desensitized mesangial cells with

TNF-a for 3 h did not inhibit MOP- 1 mRNA transcriptswhen compared with TNF-a-activation ofcontrob mes-angial cells. These data suggested that although PKO

activation stimulated mesangial cell MOP- 1 mRNAmessage, the TNF-a-induced MOP- 1 mRNA tran-

scripts were independent of PKO-mediated pathways.

Similarly, recent studies in mesangial cells alsoshowed PKO-independent mechanisms for interleu-kin- 1 (IL- 1)-induced MOP- 1 expression (34). In con-

trast, TNF-a-mediated MOP- 1 message appeared to be

PKO-dependent in fibroblasts, liver fat-storing cells,

and vascular endothelial cells (59,60), suggesting thatmesangial cells may utilize PKO-independent path-

ways to regulate cytokine gene expression.Because the activation of cells with inflammatory

cytokines (including TNF-a and IL- 1 ) stimulates pro-

tein tyrosine phosphorylation (6 1 ), we proposed thatcellular mechanisms dependent on PTK may serve as

second messenger pathways for TNF-a-induced mes-

angial cell MOP- 1 expression. In the study presented

here, the preincubation of mesangial cells with van-ous PTK inhibitors markedly inhibited TNF-a-stimu-

bated mesangial cell MCP- 1 gene expression. These

results indicated that mesangial cell MOP- 1 expres-

sion in response to TNF-a activation was regulatedthrough PTK-mediated pathways. In agreement with

these observations, it has been shown that IL- 1-

induced mesangial cell MOP- 1 expression was medi-

ated through PTK pathways (34). Apart from the role of

PTK in the ability of mesangial cells to produce mono-

cyte chemoattractants, the activation of monocytes by

various biologically active molecules (e.g. , LTB4, GM-

Pal et al

Journal of the American Society of Nephrology 921

OSF, TNF-a, etc.) also appears to be mediated through

PTK-dependent cellular pathways (62-64). Thus, cel-

lular signal transduction pathways involving PTK may

play a critical role in monocyte/macrophage pathobi-obogy and activation processes including oxidative

burst, adherence and chemotaxis and the upregula-

tion of various cellular gene products associated with

early inflammatory reactions (65-67).

We further examined the mole of cAMP as a secondmessenger to stimulate mesangial cell MOP- 1 expres-

sion. The preincubation of mesangial cells with per-

tussis toxin (a cAMP generating substance that cata-

byze ADP-ribosylation ofa subunit ofthe G protein and

activates adenylate cyclase leading to elevated intra-

cellular cAMP) did not stimulate mesangial cell MOP-imRNA expression. Additionally, isoproterenob, an ac-tivator of receptors coupled to G� protein and adenyl-

ate cyclase, and dbcAMP, a cell permeable analog of

cAMP, also had no effect on mesangial cell MOP-i

message. As intracellular cAMP may undergo degra-dation, further experiments were performed to exam-me the effect of cAMP-generating substances on mes-angial cell MOP- 1 expression in the presence of IBMX,a potent inhibitor of the phosphodiesterase that de-

grades cAMP. The data from these studies indicatedthat the presence of IBMX alone or IBMX plus cAMP

generating substances did not influence mesangialcell MOP- 1 signal when compared with their respec-

tive controls. Thus, the results from these studies

indicated that increased mesangial cell cAMP had no

stimulatory effect on MOP- 1 steady-state mRNA ex-

pression. Because the basal mesangial cell MOP-isteady-state mRNA message was minimally ex-

pressed, we examined whether cAMP would InfluenceTNF-a-induced MCP- 1 gene expression. The coincu-

bation of dbcAMP and TNF-a for 3 h with mesangial

cells completely abolished the TNF-a-induced mesan-gial cell MOP- 1 mRNA signal. Similarly, earlier studies

also showed inhibitory effects of forskolin, a cAMP-

generating substance , on TNF- a-induced mesangial

cell MOP- 1 mRNA expression (68). Taken together,

these data indicate that cAMP has an Inhibitory role in

TNF-a-induced MOP- 1 expression.Apart from these signal transduction mechanisms,

recent studies have indicated that reactive oxygenspecies can act as second messengers In TNF-a-in-

duced MOP- 1 and M-OSF expression (38). In this

regard, Satriano et al. (38) showed that increasedmesangial cell MOP- 1 expression in response to TNF-aactivation can be attenuated by free-radical scaven-gers. These investigators also indicated that TNF-a

induced MOP- 1 expression was inhibited by a NADPH-

dependent oxidase inhibitor, suggesting an involve-ment of reactive oxygen species in TNF-a stimulated

mesangial cell MCP- 1 gene expression. Furthermore,

Satriano and Schlondorff (69) reported that the in-creased reactive oxygen species in TNF-a-stimulatedmesangial induced MCP- 1 gene expression through

activation of NF-KB, a transcriptional factor involved

in gene transcription.

Additional studies examining the role of protein

synthesis on TNF-a-induced mesangial cell MOP-i

expression indicated that the preincubation of mes-

angial cells with cycboheximide, an inhibitor of protein

synthesis, completely blocked TNF-a-mediated in-

crease in mesangial cell MOP- 1 mRNA signal. suggest-ing that intact protein synthetic machinery was re-quired for TNF-a-lnduced MOP- 1 expression.

In summary, we have shown that the activation of

mesangial cells with TNF-a stimulated MOP- 1 mRNA

transcripts and that the MOP- 1 secreted into the

culture media enhanced monocyte migration as as-

sessed by the use of specific neutralizing antibodies.

The TNF-a-induced mesangial cell MCP- 1 gene ex-

pression was regulated by signal transduction path-

ways mediated by PTK but not PKO. Increasing intra-cellular cAMP by the use of activators of adenylate

cyclase and a cell-permeant analog of cAMP did not

induce basal mesangial cell MOP- 1 expression. How-ever, elevated Intracellular cAMP induced by the incu-bation with dbcAMP completely inhibited TNF-a-in-duced mesangial cell MOP- 1 mRNA message. On the

basis of these and previous studies, we suggest thatagents that augment intracellular cAMP may provide

potential tools to downregulate the sustained expres-

sion of adhesion molecules and monocyte chemoat-

tractants and limit monocyte recruitment and accu-

mulation within the gbomerulus in inflammatorycytokine-mediated glomerular injury.

ACKNOWLEDGMENTSThis investigation was supported by a Merit Review from the Depart-

ment ofVeterans Affairs and an Institutional Fellowship Award fromthe National Kidney Foundation of Southern California. The authors

thank Ms. Susan Hama and Dr. Mahamad Navab (Department of

Medicine. University ofCalifornia, Los Angeles) for their assistance inperforming the chemotaxis assay. and Ms. Kathy Arndt for hersecretarial assistance in the preparation of the manuscript.

REFERENCES1 . Brenner BM, Zatz R, Ichikawa I: The renal circulation.

In: Brenner BM, Rector FO, Eds. The Kidney. 3rd Ed.Philadelphia: WB Saunders; 1986:107.

2. Sterzel RB, Schulze-Lohoff, Marx M: Oytokines andmesangial cells. Kidney Int 1993;43ESuppl 391:�26-�31.

3. Schlondorff D: The role of chemokines in the initiationand progression of renal disease. Kidney mt 1995;47[Suppl 491:544-547.

4. Schreiner GF: The role of the macrophage in glomerularinjury. Semin Nephrol 1 99 1 ; 11:268-275.

5. Magil AB, Cohen AR: Monocytes and focal glomerulo-sclerosis. Lab Invest 1989;6l :404-409.

6. Hooke DH, Hancock WW, Gee DC, Kraft N, Atkins RC:Monoclonal antibody analysis of glomerular hypercellu-laity in human glomerulonephrltis. Olin Nephrol 1984;22:163-168.

7. Nolasco FEB, Cameron JS, Hartley B, Coelho A, Hil-dreth G, Reuben R: Intraglomerular T cells and mono-cytes in nephritis: Study with monoclonal antibodies.Kidney Int l987;3l:l 160-1166.

8. Schreiner GF, Rovin B, Lefkowith JB: The anti-inflammatory effects of essential fatty acid deficiency inexperimental glomerulonephritis. J Immunol 1989; 143:3 192-3 199.

9. Schreiner GF, Cotran RS, Pardo V. Wnanue ER: A mono-nuclear cell component in experimental immunologicalglomerulonephritis. J Exp Med 1978; 147:369-384.

TNF-a Activation of Mesangial Cell MOP-i

922 Volume 7 . Number 6 . 1996

10. Diamond JR, Pesek I, Ruggieri S. Karnovsky MJ: Es-sential fatty acid deficiency during acute puromycinnephrosis ameliorates late renal injury. Am J Physiol1989;257:F789-F807.

1 1 . Saito T, Atkins RC: Contribution of mononuclear leuko-cytes to the progression of experimental focal glomeru-losclerosis. Kidney Int 1990;37: 1076-1083.

12. Diamond JR, Pesek I, McCarter MD, Karnovsky MJ:Alterations in rat macrophage function during nephro-515: Synergistic effects of hypercholesterolemia. AmJ Pathol 1989; 135:71 1-7 18.

13. Brady HR: Leukocyte adhesion molecules and kidneydiseases. Kidney Int 1994;45: 1285-1300.

1 4. Briscoe DM, Cotran RS: Role of leukocyte endothelialcell adhesion molecules in renal inflammation. In vitroand in vivo studies. Kidney Int 1993;44lSuppl 421:527-534.

15. Brady HR, Denton MD, Jimenez W, Takata S. Palliser D,Brenner BM: Chemoattractant provoke monocyte adhe-sion to human mesangial cells and mesangial cell injury.Kidney Int l992;42:480-487.

1 6. Leonard EJ, Yoshimura T: Human monocyte chemoat-tractant protein- 1 (MOP- 1). Immunol Today 1990; 12:97-101.

17. Rollins BJ, Stier P, Ernst T, Wong GG: The humanhomolog of the JE gene encodes a monocyte secretoryprotein. Mol Oell Biol 1989;9:4687-4695.

18. Stricter RM, Wiggins R, Phan SH, et aL: Monocytechemotactic protein gene expression by cytokine-treatedhuman fibroblasts and endothellal cells. Biochem Bio-phys Res Oommun 1989; 162:694-700.

19. Cushing SD, Fogelman AM: Monocytes may amplifytheir recruitment into inflammatory lesions by inducingmonocyte chemotactic protein. Arterioscler Thromb1992; 12:78-82.

20. Mazzone T, Jensen M, ChaitA: Human arterial wail cellssecrete factors that are chemotactic for monocytes. ProcNatl Acad Sci USA 1983;80:5094-5097.

2 1 . Rovin BH, Yoshiumura T, Tan L: Oytokine-induced pro-duction of monocyte chemoattractant protein- 1 by cub-tured human mesangial cells. J Immunol 1992; 148:2148-2153.

22. Rollin BJ, Yoshimura T, Leonard EJ, Pober JS: Cyto-kine-activated human endothelial cells synthesize andsecrete a monocyte chemoattractant, MOP- 1 /JE. AmJ Pathol l990;l36:1229-l233.

23. Yoshimura T, Leonard EJ: Identification of high affinityreceptors for human monocyte chemoattractant pro-tein- 1 on human monocytes. J Immunol 1990; 145:292-297.

24. Gerrity RG, Naito IlK, Richardson M, Schwartz CJ:Dietary-induced atherogenesis in swine: Morphology ofthe intima in prelesion stages. Am J Pathol l979;95:775-786.

25. Carr MW, Roth SJ, Luther E, Rose SS, Springer TA:Monocyte chemoattractant protein 1 acts as a T-lymphocyte chemoattractant. Proc Natl Acad Sci USA1 994;9 1:3652-3656.

26. Taub DD, Proost P, Murphy WJ, Anver M, Longo DL,Damme JV. Oppenheim JJ: Monocyte chemotactic pro-tein- 1 (MOP- 1), -2, and -3 are chemotactic for humanT-lymphocytes. J Olin Invest 1995;95: 1370-1376.

27. Cushing SD, Berliner JA, Valente A.J, et at: Minimally-modified low density lipoprotein induces monocyte che-motactic protein- 1 in human endothelial cells andsmooth muscle cells. Proc Natl Acad Sci USA l990;87:5134-5138.

28. Egido J, Gomez-Chiarri M, Ortiz A, et at.: Role of tumornecrosis factor-a in the pathogenesis of gbomerular dis-eases. Kidney Int l993;43[Suppl 391:559-564.

29. Remick DG: Significance of in vivo detection of tumornecrosis factor. Lab Invest 199 1 ;65:259-26 1.

30. Faik W, Goodwin RHJ, Leonard EJ: A 48-well microchemotaxis assembly for rapid and accurate measure-ment of leukocyte migration. J Immunol Methods 1980;33:239-247.

3 1 . Levltzki A, Gazit A, Osherov N, Posner I, Gilon C:Inhibition of protein-tyrosine kinases by tyrphostins.

Meth Enzymol 199l;201:347-36l.32. Akiyama T, Ogawara H: Use and specificity of genistein,

and inhibitor of protein-tyrosine kinase. Meth Enzymol1991;201:362-370.

33. Uehara Y, Fukazawa H: Use and selectivity of herbimycinA, an inhibitor of protein-tyrosine kinases. Meth Enzy-mol l991;201:370-379.

34. Rovin BH, Tan LC: Role of protein kinase pathways inIL- 1 -induced chemoattractant expression by humanmesangial cells. Kidney Int l994;46: 1059-1068.

35. Chomczynski P, Sacchi N: Single-step method of RNAisolation by acid guanidinium thiocyanate-phenol-chloroform extraction. Anal Biochem 1989; 162:156-160.

36. Sambrook J, Fritsch EF, Maniatis T: Extraction, purifi-cation and analysis of mRNA. In: Sambrook J, FritschEF, Maniatis T, Eds. Molecular Cloning. New York: ColdSpring Harbor Laboratory; 1989:7.3-7.84.

37. MacKay K, Striker LI, Elliot S, Pinkert CA, Brinster RL,Striker GE: Gbomerular epithelial. mesangial and endo-thelial cell lines from transgenic mice. Kidney Int 1988;33:677-684.

38. Satriano JA, Shuldiner M, Hora K, Xing Y, Shan Z,Schiondorif D: Oxygen radicals as second messengersfor expression of the monocyte chemoattractant protein,JE/MOP- 1 , and the monocyte colony-stimulating factor,OSF- 1 , in response to tumor necrosis factor-a and im-munogbobulin G. J Olin Invest 1993;92: 1564-1571.

39. Pal R, Kirschenbaum MA, Kamanna VS: Low densitylipoprotein stimulates the expression of macrophagecolony-stimulating factor in glomerular mesangial cells.Kidney Int 1995;48:1254-1262.

40. Rovin BH, Rumancik M, Tan L, Dickerson J: Glomeru-lar expression of monocyte chemoattractant protein- 1 inexperimental and human glomerulonephritis. Lab Invest1994;7 1:536-542.

4 1 . Stahl RAK, Thaiss F, Disser M, Helnichen U, Hora K,Schlondorff D: Increased expression of monocyte che-moattractant protein- 1 in anti-thymocyte antibody in-duced gbomerulonephritis. Kidney Int 1993;44: 1036-1047.

42. Bolton WK, hines DJ Jr. Sturgill BC, Kaiser DL: T-cellsand macrophages in rapidly progressive glomerulone-phritis. Clinicopathologic correlations. Kidney Int 1987;32:869-876.

43. Schreiner GF, Cotran RS, Unanue ER: Macrophages andcellular immunity in experimental glomerubonephritis.Springer Semin Immunopathol l982;5:25 1-267.

44. Pesek-Diamond I, Ding G, Frye J, Diamond JR: Macro-phages mediate adverse effects of cholesterol feeding inexperimental nephrosis. Am J Physiol i992;263:F776-F783.

45. El Nahas AM: Growth factors and glomerular sclerosis.Kidney Int 1992;4ilSuppl 361:515-520.

46. Sherry B, Cerami A: Cachectin/tumor necrosis factorexerts endocrine, paracrine, and autocrine control ofinflammatory responses. J Cell Biol 1988; 107:1269-1277.

47. Vilcek J, Lee TI!: Tumor necrosis factor. New insightsinto the molecular mechanisms of its multiple actions.J Biol Ohem l991;266:73l3-73l6.

48. Hruby ZW, Shirota K, Jothy 5, Lowry RP: Antiserumagainst tumor necrosis factor alpha and protease inhib-itor reduce immune gbomerular injury. Kidney Int 1991;40:43-51.

49. Tomosugi NI, Cashman SJ, Hay H, Pusey CD, Evans DJ,Shaw A, Rees AJ: Modulation of antibody-mediatedglomerular injury in vivo by bacterial lipopolysacchamide,tumor necrosis factor and IL- 1 . J Immunol 1989; 142:3083-3090.

50. Pal V. Pal R, Kamanna VS, Kirschenbaum MA: Tumornecrosis factor (TNF)-a stimulates monocyte (MO) adhe-sion to mesangial cells (MO) mediated by lOAM- 1 expres-sion [Abstracti. J Am Soc Nephrol 1994;5:727A.

5 1 . Zoja C, Wang JM, Bettoni 5, et at.: Interleukin- 1 /3 andtumor necrosis factor induce gene expression and pro-duction of leukocyte chemotactic factors, colony-stimulating factors, and interleukin-6 in human mesan-

Pal et al

Journal of the American Society of Nephrology 923

gial cells. Am J Pathol 199 l;138:991-l003.52. Couffinhal T, Duplaa C, Labat L, et at.: Tumor necrosis

factor-a stimulates lOAM- 1 expression in human vascu-lar smooth muscle cells. Arterioscler Thromb 1993; 13:407-4 14.

53. Wertheimer SJ, Myers CL, Wallace RW, Parks TP: Inter-cellular adhesion molecule- 1 gene expression in humanendothelial cells. Differential regulation by tumor necro-sis factor-a and phorbol myristate acetate. J Biol Ohem1992;267: 12030-12035.

54. Rovin BH, Yoshimura T, Tan L: Oytokine-induced pro-duction of monocyte chemoattractant protein- 1 by cul-tured human mesangial cells. J Immunol 1992; 148:2148-2 153.

55. Bertani T, Abbate M, Zoja C, et at: Tumor necrosisfactor induces glomerular damage in the rabbit. AmJ Pathol l989;l34:4l9-430.

56. Heller PA, Kronke M: Tumor necrosis factor receptormediated si�nalling pathways. J Oell Biol 1994; 126:5-9.

57. Shyy Y-Z, Li Y-S, Kolattukudy PE: Structure of humanmonocyte chemotactic protein gene and its regulation byTPA. Biochem Biophys Res Oommun 1990; 169:346-351.

58. Hepler JR. Earp HS, Harden TK: Long-term phorbolester treatment down-regulates protein kinase 0 andsensitizes the phosphoinositide signaling pathway tohormone and growth factor stimulation. J Biol Ohemi988;263:76 10-7619.

59. Shyy Y-Z, Li Y-S, Kolattukudy PE: Activation of MOP-igene expression is mediated through multiple signallingpathways. Biochem Biophys Res Oommun 1993; 192:693-699.

60. Maim F, Valente AJ, Pinzani M, Abboud HE: Oulturedhuman liver fat-storing cells produce monocyte chemo-tactic protein- 1 . Regulation by proinflammatory cyto-kines. J Olin Invest l993;92: 1674-1680.

61. Guy GR, Chua SP, Wong NS, Ng SB, Tan YH: Interleu-kin- 1 and tumor necrosis factor activate common mul-tiple protein kinases in human fibroblasts. J Biol Ohem

1991 ;266: 14343-14352.

62. Berkow RI, Dodson RW: Tyrosine-specific protein phos-phorylation during activation of human neutrophils.Blood i990;75:2445-2452.

63. Yuo A, Kitagawa 5, Azuma E, et at.: Tyrosine phosphor-ylation and intracellular alkalinization are early eventsin human neutrophils stimulated by tumor necrosisfactor. granulocyte-macrophage colony-stimulating fac-tor and granulocyte colony- stimulating factor. BiochemBiophys Acta 1993; 1156:197-203.

64. Fuortes M, Jin WW, Nathan C: Adhesion-dependentprotein tyrosine phosphorylation in neutrophils treatedwith tumor necrosis factor. J Oell Biol 1993; 120:777-784.

65. Kusunokl T, Higashi H, Hosoi S, et at.: Tyrosine phos-phorylation and its possible role in superoxide produc-tion by human neutrophils stimulated with FMLP andIgG. Biochem Biophys Res Oommun 1992; 183:789-796.

66. Gaudiy M, Caon AC, Gilbert C, Lille S, Naccache PH:Evidence for the involvement of tyrosine kinases in thelocomotory responses of human neutrophils. J Leuko-cyte Biol 1992;51:103-108.

67. McCoil SR, Diperslo JF, Caon AC, Ho P, Naccache PH:Involvement of tyrosine kinases in the activation ofhuman peripheral blood neutrophils by granulocyte-macrophage colony-stimulation factor. Blood 1 99 1 ;78:1842-1852.

68. Satriano JA, Hora K, Shan Z, Stanley ER, Mon T,Schlondorff D: Regulation of monocyte chemoattractantprotein- 1 and macrophage colony-stimulating factor-iby IFN-y, tumor necrosis factor-a, IgG aggregates. andcAMP in mouse mesangial cells. J Immunol 1993; 150:197 1-1978.

69. Satriano J, Schiondorif D: Activation and attenuation oftranscription factor NF-KB in mouse glomerular mesan-gial cells in response to tumor necrosis factor-a, immu-nogbobulin G, and adenosine 3:5-cyclic monophos-phate. J Olin Invest l994;94:1629-l636.