tis, IgA nephropathy, as well as membranoprolifera- tive glomerulonephritis, variants of focal sclerosis, lupus nephritis, and possibly diabetic nephropa- thy.1`3 In ...
American Journal of Pathology, Vol. 14 7, No. 5, November 1995 Copyright © American Society for Investigative Pathology
Autocrine Growth Regulation of Human Glomerular Mesangial Cells Is Primarily Mediated by Basic Fibroblast Growth Factor
Aleksandar Francki,* Peter Uciechowski,* Juergen Floege,t Juliane von der Ohe,* Klaus Resch,* and Heinfried H. Radeke* From the Institute of Clinical Molecular Pharmacology,* and Institute of Nephrology,t Medical School Hannover, Hannover, Germany
For various forms of human glomerulonephritis a close relationship between inflammatory injury and a local mesangialproliferative response has been described. Herein, we used primary cultures of human glomerular mesangial ceUs (HMCs) from five different donors to determine the autocrine growth-inducing capacity of their supernatants after stimulation with different cytokines and lipopolysaccharide (LPS) to determine whether this effect is due to basicflbroblast growth factor (bFGF). The basal growth-inducing capacity of supernatants coUected from serum-free cultured HMC and concentrated 100fold above a cut-off size of 10 kd was significantly increased by interleukin (IL)-1$8, platelet-derived growth factor (PDGF), and LPS up to 15-fold, but not by IL-1a , IL-6, or bFGF. An anti-human bFGF antibody blocked the majority of IL-1 or LPS-induced proliferative effect of supernatants; complete inhibition was achieved by a combination of anti-human bFGF- and anti-human platelet-derived growth factor antibodies. HMCs express different isoforms of bFGF (18, 21.5, and 24 kd) in membrane, cytosolic, and nuclear fractions. AU isoforms of bFGF were found in the nuclear fraction of HMC, whether stimulated or not. Immunoblots for bFGF protein of HMC demonstrated that only a -16 kd bFGF protein was released into HMC supernatants after stimulation with IL-1 (3, platelet-derived growth factor-BB, and LPS. The 18 kd isoform of bFGF accumulated in the membranes but was not released after stimulation with IL- ae, IL-6, and bFGF, suggesting that its release was a prerequisite for autocrine growth stimulation. By
means of reverse transcription polymerase chain reaction controUled by Southern blots, bFGF-mRNA expression of HMC was enhanced by IL-1 a, IL-1 (3, and LPS. Finaly, we were able to show that HMCs are expressing bFGF receptors. In summary, our data demonstrate for the first time that the autocrine proliferative response of HMC to major infZammatoryfactors may primarily be mediated by bFGF. (Am J Pathol 1995, 147:1372-1382)
Proliferation of intrinsic glomerular mesangial cells is a key feature of various human renal diseases, including the world's most common glomerulonephritis, IgA nephropathy, as well as membranoproliferative glomerulonephritis, variants of focal sclerosis, lupus nephritis, and possibly diabetic nephropathy.1`3 In vitro studies have identified a large number of mediators with a mitogenic potential on cultured human glomerular mesangial cells (HMCs) such as basic fibroblast growth factor (bFGF); platelet-derived growth factor (PDGF); insulin-like growth factor (IGF-1); mediators with a concentration-dependent mitogenic or anti-proliferative effect, such as transforming growth factor (TGF); as well as mediators with an unknown effect on proliferation, such as interleukin (IL)-1 and IL-6.1 4-7 First in vitro studies have confirmed some of these observations and, more importantly, have shown that therapeutic interference, eg, with PDGF or TGF, can ameliorate the course of an experimental mesangioproliferative glo-
merulonephritis.a Given the large number of mediators affecting HMC growth, it appears crucial from a potential therSupported by a grant of the Deutsche Forschungsgemeinschaft SFB 244/B1 and the Graduiertenforderung of the Medical School Hannover (AF). Accepted for publication August 2, 1995. Address reprint requests to Dr. Aleksandar Francki, Institute of Clinical Molecular Pharmacology, Medical School Hannover, Konstanty-Gutschow-Strasse 8, D-30623 Hannover, Germany.
Autocrine Growth Regulation of Human Mesangial Cells
AJP November 1995, Vol. 14 7, No. 5
apeutic point of view to identify those with a central and/or initial role in the induction of cell proliferation. To this respect bFGF merits consideration. In rat mesangial cells it has recently been shown that these cells, like many other cell types of mesenchymal origin, constitutively express bFGF in their cytoplasm both in vitro and in vivo.9 bFGF is a particular cytokine, in that it lacks a signal peptide and is therefore not secreted but in most instances is released upon sublethal or lethal cell injury.10 Indeed, in rat mesangial cells in vivo and in vitro, bFGF release after antibody- and complement-mediated injury has been documented.9 Evidence has also been provided to suggest that in rats the released bFGF may be bioactive and may participate in the induction of regenerative mesangial cell proliferation after cell injury.9 These features, namely the constitutive expression and immediate release after severe cell injury render bFGF an attractive candidate as one of the earliest inducers of mesangial cell proliferation. Based on the above observations, we have undertaken the following study in cultured HMC to address three questions: 1) Do HMC similar to rat mesangial cells produce and/or constitutively express bFGF? 2) Can bFGF release from HMC be induced by other than lethal stimuli, in particular by stimulation with inflammatory mediators such as IL-1, IL-6, lipopolysaccharide (LPS), PDGF, or bFGF itself? 3) Do any of these latter mediatorinduce alterations of the subcellular compartmentalization of bFGF protein expression, eg, a translocation or increased expression of bFGF in HMC membranes? This third question appears important, because it has been shown that bFGF is expressed in different isoforms and localized in different cellular compartments such as nucleus, cytoplasm, and membrane.1 1-4 Furthermore it has been described that those different isoforms have different functions and bioactivities 1,12; ie, the low molecular 18 kd isoform, lacking a signaling sequence, is stored in the membrane and secreted after lethal or sublethal injury, whereas the high molecular isoforms (21.5, 24 kd) are found in the nucleus region and seem to have transcriptioncontrolling functions.15 In this study we show that bFGF is an autocrine factor for HMC and is released after stimulation with IL-1, PDGF, and LPS.
Materials and Methods Materials Human recombinant IL-ia (IL-ia; 2 x 107 U/mg) kindly provided by Dainippon Co. (Osaka,
Japan), recombinant human (rh) IL-1 (IL-1; 2 x 107 U/mg) was a gift from Dr. H. Gallati (Hoffmann LaRoche, Basel, Switzerland), and FPLC-purified rh IL-6 (IL-6) was donated by Dr. W. Fiers (Gent, Belgium). rh PDGF-AA, -AB, and -BB were kindly donated by Dr. J. Hoppe (WOrzburg, Germany). rh bFGF and polyclonal, neutralizing rabbit anti-rh bFGF antibodies were a gift from the Pharma Biotechnologie Hannover (PBH, Hannover, Germany). rh epidermal growth factor (EGF), LPS from Escherichia coli, phosphate-buffered saline, (PBS), TrisEDTA/Tris-acetate EDTA (TE/TAE) buffer, MCDB302 medium, insulin (bovine), and transferrin (human) were purchased from Sigma Chemical Co. (Deisenhofen, Germany). RPMI 1640, non-essential amino acids, L-glutamine, sodium pyruvate, mycoplasma-free fetal calf serum (FCS, 20Q19), trypsin/EDTA (T/E) all were from GIBCO BRL (Eggenstein, Germany). Cell culture plastic material was from Nunc (Wiesbaden, Germany). All other materials were obtained commercially at the highest quality available.
HMC Preparation and Characterization For the present experimental series selected fresh, tumor-free, healthy tissue of kidneys from five different donors undergoing tumor nephrectomy (who had given their informed consent) was obtained with the help of the Department of Urology, Medical School Hannover. HMCs were prepared as described.16'17 Cells obtained from the different specimens showed no significant differences in growth, proliferation, and morphology. Characterization by immunofluorescence staining showed a positive reaction with smooth muscle cell myosin and actin, vimentin, fibronectin, desmin, and collagen type IV; and a negative reaction with antihuman keratin, factor VIII, and major histocompatibility complex class 11 antigen antisera. There was no morphological evidence of the presence of macrophages (pseudopodia), endothelial- or epithelial-like cells (cobblestone) or fibroblasts. Bone marrow-derived resident macrophages and endothelial cells could be excluded by the following criteria: HMCs are major histocompatibility complex class 11-negative2; showed a typical profile of prostanoids6 18; and in contrast to macrophages, did not express 5-lipoxygenase-mRNA or produce leukotrienes6; and, under non-stimulated conditions, released no reactive oxygen species. 16,19 Furthermore, the use of HMC after passage 3 excludes macrophage as well as endothelial cell contaminations, as these cell types do not survive
Francki et al
AJP November 1995, Vol. 147, No. 5
multiple passaging under the described medium conditions.
Culture Conditions and Stimulation of HMC As described,16'19 HMCs after the third passage were grown in culture medium for proliferating cells consisting of RPMI-1640, nonessential amino acids, (1 ml/di), L-glutamine (2 mmol/L), sodium pyruvate (2 mmol/L), transferrin (5 ,ug/ml), insulin (125 U/ml), and FCS (10%). For passaging, HMC were detached by T/E (0.125%/0.01%, w/v) and split 1:2. Serum-free culture of HMC was performed using MCDB-302 medium, which was supplemented as the culture medium but without any FCS, leading to a growth arrest of HMC after 48 hours.18'20 After 48 hours the resting medium was removed, and cells were washed twice with PBS and cultivated for various stimulation periods in resting medium alone (control) or with the addition of optimized concentrations of IL-1,B (10 ng/ml), IL-la (10 ng/ml), IL-6 (10 ng/ml), bFGF (10 ng/ml), PDGF (10 ng/ml), TGF-3 or LPSE. Coli (10 ,ug/ml) for 2, 6, or 16 hours before mRNA or for 24 hours before protein preparation of HMC. Concentration and time-dependent effects of these stimuli on [3H]thymidine incorporation or on the induction of autocrine proliferative activity have been investigated previously'1618'20 or are described below.
Concentration of HMC Supernatants HMC supernatants (HMC-SNs) were collected from serum-free cultured HMC, which were seeded in an equal cell number in all tests. HMCSNs were pre-cleared from cellular debris by a centrifugation at 150 x g and were then concentrated at 40C 100-fold by centrifugation with Centriprep concentration filter units (Amicon, Beverly, MA) with a cut-off size of 10 kd and subsequently stored frozen at -80°C. The autocrine activity was determined in concentrations ranging from 0.1 to 10% of the 100-fold concentrated HMC-SNs under serum-free conditions as described below.
Proliferation Assays with HMC Cell proliferation was assessed by measuring the DNA incorporation of [3H]-labeled thymidine essentially as described.18'20 Cells were plated in 96-well plates at a density of 10,000 cells/well and cultured for 24 hours in culture medium with 10% FCS. After reaching subconfluence the cells were washed with PBS three times and switched to se-
rum-free MCDB-302 medium for 48 hours. According to our previous studies HMCs have been growth-arrested and showed no significant thymidine incorporation after 48 hours.18'20 Subsequently, the cells were incubated for 24 hours with fresh MCDB-302 medium and the respective stimuli, followed by pulsing and harvesting as described. 18'20 Incorporation of [3H]thymidine into DNA was determined by liquid scintillation-counting. Comparisons between the different agonists were made in the same and across all five mesangial cell preparations used in the experiments. No significant variations of the stated effects induced by all agonists were observed among the five different HMC specimens.
Preparation of HMC Compartments Nuclear, cytosolic, and membrane fractions of HMC were prepared according to Radeke et al.19 Briefly, after collection of the SNs the cells were washed twice with ice-cold PBS, carefully scraped and resuspended in PBS, and centrifuged 10 minutes, 40C, at 100 x g. The pelleted intact cells were resuspended in PBS containing 1 mmol/L phenyl methyl sulfonyl fluoride, iodoacetamide, benzamidine, and leupeptin and sonicated for 30 seconds, 50 W. After centrifugation for 10 minutes, 40C, at 400 x g, the pelleted nuclei were resuspended in 10 mmol/L HEPES buffer, pH 7.4, with proteinase inhibitors. Membrane and cytosolic fraction were separated at 100,000 x g for 30 minutes, 40C. The cytosolic fraction was collected and the membranes resuspended in HEPES buffer. The protein concentration of all cell fractions and HMC-SN were determined using the Bradford microassay, and the fractions were kept frozen (-80°C) until usage.
SDS-PAGE/Western Blotting Cultured HMC were lysed and fractionated as described above. The samples were boiled in Laemmli buffer under reducing conditions with 1 mmol/L dithiothreitol and electrophoresed at 0.1% sodium dodecyl sulfate (SDS) by 12.5% SDS-polyacrylamide gel electrophoresis (SDS-PAGE). The lanes were loaded with 10 jig protein from nuclei, cytosols, membranes, and SNs. For the analysis of concentrated HMC-SNs 100 ,tg protein were loaded per lane. The electrophoresed proteins were transferred to nitrocellulose filters by electroblotting, incubated with a specific primary polyclonal rabbit anti-human bFGF antibody, and
Autocrine Growth Regulation of Human Mesangial Cells 1375 AJP November 1995, Vol. 14 7, No. 5
immunostained with a biotinylated secondary antibody and the streptavidin-alkaline phosphatase system as described.1"
RNA Preparation Total cellular RNA for reverse transcription polymerase chain reaction (RT-PCR) was prepared from HMC cultured under resting (serum-free, MCDB-302 medium, supplements) conditions after stimulation by a modification of Chomczynski's method using RNAzolTIB (WAK-Chemie, Bad Homburg, Germany).17 To increase the purity of the RNA samples we extended the basic method described by an additional LiCI2 (4 mmol/L) precipitation step.17
Analytical RT-PCR As described,17 we used a "hot start" RT-PCR protocol modified from the method described for the GeneAmpR RNA PCR Kit (Perkin Elmer Cetus, Norwalk, CT). Primer pairs specific for human bFGF cDNA (sense: 5'-GGCTTCTTCCTGCGCATCCA, antisense: 5'-GCTCTTAGCAGACATTGGAAG; product size: 354 bp),21 for the membranespanning domain common for human bFGF receptor cDNAs (sense: 5'-GACAAAGAGATGGAGGTGCT, antisense: 5'-GTTGTAGCAGTATTCCAGCC; product size: 802 bp),22 and for a universal 3-tubulin cDNA7 were selected and purchased from MWG-BIOTECH (Ebersberg, Germany). To establish conditions that allow the comparison of the amount of cDNA produced by RT-PCR (as a semiquantitative measure of the initial mRNA level), the cycle number was varied from 24 to 40 cycles. As an internal standard we amplified ,B-tubulin-mRNA, a ubiquitously expressed gene. The amounts of the expressed f3-tubulin mRNA were unchanged in unstimulated versus stimulated cells. After electrophoresis of 1/10 of the PCR reaction products the ethidium bromide-stained ,B-tubulin bands became visible after 24 PCR cycles, and those of bFGF and bFGF receptor cDNA after 28 cycles. Both PCR products reached a plateau past 32 cycles. To reach submaximal amplification levels for the control and both bFGF and bFGF receptor-mRNA we used 28 PCR cycles throughout our experimental series. Subsequently, the analysis of PCR products was completed by Southern blot hybridization with the respective oligonucleotides as described.17
Results Growth Regulators of HMC and bFGFDependent Autocrine Effects of SNs from Stimulated HMC With primary HMC cultures prepared from five different donors our first experimental series confirmed earlier results of our group5'20 and others23'24 that the human recombinant cytokines IL-la, IL-1,B, IL-6, and LPS alone had no significant proliferative effect on growth-arrested, synchronized HMC (Figure 1A). rh PDGF-AA, -AB, and -BB with increasing potency enhanced DNA synthesis of HMC time dependently with a maximum after 24 to 36 hours. rh EGF even at 50 ng/ml had only minor proliferative potency, whereas TGF-f at low concentrations (1 ng/ml) inhibited HMC growth.20 rh bFGF induced a time- and dose-dependent increase of [3H]thymidine incorporation up to 208 ± 7.5% compared with unstimulated control cells after 36 hours (Figure 1A). Dose-response experiments (0.1 pg/ml to 100 ng/ml) revealed that the rh bFGF effect reached a plateau at 25 ng/ml (data not shown). Effects of cytokines on growth behavior have been described in the literature. To analyze whether this is due to the induction of low amounts of autocrine growth factors the following experiments have been designed to evaluate which autocrine mechanisms exist for HMC to regulate growth. After 24-hour periods of stimulation with the respective cytokines cells were washed and the SNs collected for the subsequent experiments under serum-free conditions. To compensate for the high volume of fluid compared with cell numbers in culture flasks we then concentrated the SNs at average 100-fold (see Materials and Methods) and used it with the same HMC preparation at a later passage (or with one from another donor) to determine whether "autocrine" growth regulators had been induced by the respective stimuli. The results of these experiments in Figure 1 B showed that adding a 10% preparation of the 100fold concentrated SNs of unstimulated HMCs already had a significant proliferative effect of 200 ± 8%, which was not significantly altered when HMC had been stimulated by rh IL-la, IL-6, or bFGF (not shown). However, when HMC-SNs of cells were used that had been treated with rh IL-1i, PDGF-BB, or LPS the result was an increase in the autocrinestimulated HMC growth by 320 ± 11%, 270 ± 9%, or 380 + 12%, respectively (Figure 1B). These autocrine effects of stimulated HMC-SN were time- and dose-dependent with respect to both the concentration of stimulus applied before the collection of the
Francki et al
AJP November 1995, Vol. 14 7, No. 5
A [3H]-thymidine incorporation % in of the control 40C
B 500 7o
*r0 300 0~ c c 0
0 E c
Autocrine Growth Regulation of Human Mesangial Cells 1377 AJP November 1995, Vol. 147, No. 5
SN and the doses of concentrated SN solution used (data not shown). A likely candidate that is possibly released from HMC by either IL-1f, or LPS is bFGF.9 25'26 Therefore we applied neutralizing polyclonal antibodies to human bFGF to either cells stimulated with recombinant cytokines or to HMCs stimulated with autocrine active SN (Figure 1B). In addition, neutralizing antibodies to PDGF and a combination of both were used to elucidate the factors which were responsible for the proliferative effect of HMC-SNs. Neutralizing anti-bFGF- or anti-PDGF antibodies alone partially reduced the activity of the SNs; however, the combination of both antibodies completely inhibited the proliferative effect. Subsequent experiments using the inflammatory mediators IL-1f3, LPS, and to a lesser extent PDGF-BB as stimuli for inducing autocrine activity demonstrated that the potent proliferative action of these SNs was always predominantly due to bFGF and only to a minor extent to PDGF released by HMC. A control rabbit IgG preparation had no inhibitory effect on the autocrine activity induced by all stimuli.
Stimulus-Dependent Expression of bFGF Protein Isoforms in Nuclei, Cytosol, and Membranes of HMC The previous data identified bFGF as one of the key factors regulating stimulated human mesangial cell autocrine growth behavior. To directly address the effects of cytokines, we determined the expression of HMC-derived bFGF protein isoforms in the different cellular compartments. In subsequent experiments we examined in detail the effects of IL-la, IL-1,B, IL-6, bFGF, PDGF, and LPS, respectively, on the distribution of the mesangial cell 18, 21.5, and 24 kd bFGF isoform in the nuclear cytosolic membrane and the non-concentrated SN fraction as compared with the respective medium control. The expression of the 18 kd isoform of bFGF in the membrane of HMC was increased after stimulation with IL-la (Figure 2A), bFGF, and PDGF-BB (not shown). Notably, membrane steady state levels of HMC-bFGF were
not stimulated by IL-1: (Figure 2B), IL-6 (not shown),
LPS (Figure 2C). Furthermore, we were able to show a nuclear localization of all three different bFGF isoforms in HMC, whether stimulated or unstimulated
(Figure 2A-C). The detection of bFGF in the SN fraction in this approach comparing the HMC compartments was negative for all stimuli (Figure 2 A-C). These results are due to the fact that each respective lane of the gel was loaded with 10 Ag protein of the unconcentrated SNs as well as for the other fractions throughout these tests, whereas in the experiments shown in Figure 3, 100 ,ug/ml of 100-fold concentrated SNs were loaded.
Secretion of bFGF in the HMC-SN after Stimulation with IL- 1 ,, PDGF, and LPS Figure 3 shows that a smaller isoform of bFGF (-16 kd), compared with recombinant bFGF (--18 kd) was released from HMC into SN already under resting, serum-free culture conditions. Stimulation with IL-1,B, PDGF, and LPS increased the detectable bFGF protein level in the SN. HMCs that had been stimulated with IL-la, IL-6, or bFGF did not release increased amounts as compared with the resting medium control. Confirming the antibody specificity and the sensitivity of these immunoblots recombinant bFGF used for stimulation was detected as a separate band in the respective SN (100-fold concentrated; Figure 3, lane 6, upper band).
HMC bFGFmRNA Expression: Induction of bFGF Messenger RNA by IL- 1 a, IL- 1 ,, and LPS Using the sensitive RT-PCR technique we established a experimental protocol to determine the changes of HMC bFGF mRNA expression following stimulation. The synthesis of bFGF mRNA in HMC was enhanced after stimulation with IL-1,B after 2 hours (Figure 4A), and with IL-la within 6 hours (Figure 4B) after stimulation. HMCs that were stimu-
Figure 1. (A) Effects of cytokines and growth factors on HMCproliferation. HMCs weregrown to subconfluence in multiwellplates, switched to resting medium, and after 48 hours stimulatedfor additional 12, 24, 36, and 48 hours with the respective mediators under serum-free conditions: rh IL-la (10 ng/ml), IL-1, (10 ng/ml), rh IL-6(10 ng/ml), rh bFGF (10 ng/ml), rh PDGF-AA, -AB, -BB (each 10 ng/ml), rh TGF-)3 (1 ng/ml), and rh EGF (10 ng/ml). [3Hlthymidine incorporation into DNA was allowed for the final 4 hours and radioactivity determined in a ,3-scintillation counter. Results shown represent the mean of five to eight experiments done with six replicates each. (B) bFGF- and PDGF-specific antibodies inhibit the autocrine growth induction ofserum-free cultured HMCs. 10% solutions of SNs harvestedfrom resting, cytokine, orLPS-stimulated HMC concentrated above a cut-off size of 10 kd as described. SN with (hatched bars) or without (dotted bars) a 30-minute preincubation with a polyclonal rabbit anti-human recombinant bFGF antiserum at an optimized neutralizing concentration (10 p.g/ml) were added to either the same HMCpreparation at a later passage or to primary HMC culturesfrom four other patients. [33Hlthymidine incorporation was measured after 24 hours. Results represent the mean of six replicates of one representative experiment out of a series of three. The distribution of bFGF protein isoforms in nuclear cytosol and membrane fractions of HMCs after stimulation with cytokines, growth factors and LPS.
Francki et al
AJP November 1995, Vol. 147, No. 5
45 31 -
Figure 3. The release of bFGF protein in HMC-SNs after stimulation with cytokines, growth factors, and LPS. bFGF protein released intto HMC-SNs. Th7e SNs were collected and concentrated as described, and 100 ,ggml protein were separated by SDS-PAGE and analyzed hy Western blotting using the neutralizing bFGF antiserum as described in Figure I C and in Materials and Metbods. The basal release of a 16 kd bFGF immunoreactivitV was enbanced by stimulation of HMC with IL-1f, PDGF-BB, and LPS. -
MW (kDa) 97.4 66.2-
C MW (kDa)
45 31 21.5-
Figure 2. A, B, and C shout the distribution of bFGFprotein isoforms in the nuclear and membrane fractions of HMCs induced by stimu-
latiotn uwith IL-la, IL-1(3, and LPS, respectively. For every stimulated condition the respective control with protein prepared from restinig HMC at the same tinie is depicted on the left part of the We,stern blots. 10 ,tg of HMC protein of every cellularfraction had been subjected to SDS-PAGF and after blotting bFGF protein detected with a rabbit polyclonal anti-human bFGFantiserum and visualized uith a biotinstreptavidin alkaline phosphatase staining sy.stem as described. Whereas membrane expression of mesangial 18 kd bFGF was increased by stimulation uwith IL- la,(A), IL-1(3 (B), and LPS(C), which enhanced the release of soluble bFGF into the SN (see Figure 3), had no cffcect on membrane expression. Thy nuclearfraction of HMC showed three different isofornis of bFGF already, under resting conditions (A-C).
lated with LPS showed, like IL-la, an elevated amount of bFGF mRNA after 6 hours (Figure 4C). After stimulation with IL-6, bFGF, and PDGF (not shown), no increased induction of bFGFmRNA was detectable. The results obtained in the agarose gel
electrophoresis of the RT-PCR cDNA products (Figure 4 A-C, upper part) were controlled by means of Southern blotting, followed by hybridization with a-[32P]ATP end-labeled human bFGF oligonucleotides and subsequent autoradiographies. These experiments confirmed that the results are due to the increase of the respective mRNAs and the higher amount of reverse-transcribed cDNA. The bands of the autoradiographies were scanned and are shown in Figure 4 A-C, lower part).
Expression of bFGF Receptor Messenger RNA by HMC Using a similar RT-PCR technique we finally determined the expression of bFGF receptor mRNA by HMC. As mentioned earlier, the primers chosen identified the membrane coding sequence common for bFGF receptor isoforms.22 In comparison to simultaneously loaded tubulin RT-PCR products, bFGF receptor mRNA was significantly expressed (Figure 5).
Discussion This study was designed to elucidate the interactions of cytokines and growth factors, which regulate the proliferation of HMCs. Although recently performed investigations with rat models of experimental glomerulonephritis demonstrated that both bFGF in the earlier phase of postinjury mesangial cell proliferation9 and PDGF in the later phase27 are the key regulators of glomerular cell growth, these studies could not resolve the role of participating inflamma-
Autocrine Growth Regulation of Human Mesangial Cells 1379 AJP November 1995, Vol. 14 7, No. 5
A RT-PCR products
(expected size) 1 353 -, 1078
1353'.. 1078-. 872--
e bFGF-receptor K B-tubulin
310O 281 271
RT-PCR products (expected size)
MW (Bp) 1353
Figure 5. Expression of mRNA for bFGF receptors by HMCs. As described in Materials and Metbods an analytical RT-PCR (24 to 40 cycles) approach was used. Primers were chosen to detect a cDNA sequence codingfor a transmembrane region common to several bFGF receptor isoforms (see Materials and Methods). Unlike Figure 3 the respective expression of bFGF receptor mRNA (802 bp, lanes 8 to 12) was directly compared with the expression of -tubulin-mRNA (486 bp, lanes 3 to 7) in unstimulated HMC kept in resting medium alone.
tory mediators such as IL-1, IL-6,
LPS in detail.
The question that especially cannot be answered in
the complex in vivo environment is whether autocrine regulatory loops are involved in the reparative me-
sangial cell proliferation, which eventually may be disturbed and lead to chronic inflammatory dis-
31 0-" 281.
C MW (Bp)
(expected size) \
1 078 ---872 603 31 0|'I 2 hr 281
N 6 hrs,
1- 18 hrs..|
Figure 4. Upper parts of A, B, and C: Effect of rh IL-la, IL-1(3, and LPS, respectively, on the bFGF mRNA expressioi n of HMC as determined by RT-PCR 2, 6, and 16hours after addition of'te stimuli. 1 of total RNA extractedfrom stimulated HMC was reverrse-transcribed and subsequently amplified in the presence of specific bFGFprimers. In each case the respective stimulated condition was directly compared uwith both the bFGF mRNA template extracted fromicontrol cells in resting medium (A-C, lanes 3, 7, and 11, respectiviety), and the f-tubulin mRNA expression (A-C, lanes 2, 4, 6, 8, 10 , and 12, respectively), which was not altered by the stimuli used h Perein. As confirmed by Southern blots of these 2% agarose gels shown bvelou IL-1,8 after 2 hours (B, lane 5) and IL-la and LPS after 6 hoiurs (A and C, lane 9, respectively) led to an increased bFGF mRNA iexpression. Lower parts of A, B, and C: Scannings of the autoradio)graphies of the control Southern blots for the RT-PCR cDNA products as shown in the respective upper parts of this figure. The results obtc iined in the agarose gel electrophoresis of the RT-PCR-cDNA products w)ere controlled by means of Southern blotting followed by hybridizaitions with a P2_12-ATP
The data presented herein confirmed that, in addition to PDGF, bFGF is a very efficient growth factor for HMCs. Experiments with neutralizing bFGF and PDGF antibodies showed that the presence in HMCSNs accounted for the majority of the IL-1B-, PDGFBB-, or LPS-induced autogrowth activity of these cells. Furthermore, we were able to demonstrate that the proliferative effect of PDGF itself was partially mediated by bFGF, as the growth-inductive effect of PDGF was to some extent inhibitable by neutralizing anti-bFGF antibodies. Summarizing the results
IL-ia, IL-1i3, and LPS induced synthesis of bFGF-mRNA; that the release of a translated 16 kd protein isoform into HMC-SNs was stimulated by IL-1i, LPS, and PDGF; and finally, that the bFGF receptor mRNA is constitutively expressed by HMC. Together these results establish secretion of bFGF as a potent autocrine growth-regulating mechanism in HMC. As a prerequisite for an autocrine growth factor, quentially
HMC synthesized bFGF. When cell-associated bFGF analyzed, several molecular species were de-
tected. The observation that low and high molecular weight isoforms of bFGF reported
existing has also been
identified three intracellular
autoradiographies to confirm that the results were due to the increase of the respective mRNAs and the higher amount of reversetranscribed cDNA and not to a false ethidium bromide staining. The bands of the autoradiographies were scanned and are shown in A-C. For conditions please refer to Materials and Methods and to the upper parts of this figure.
Francki et al
AJP November 1995, Vol. 14 7, No. 5
isoforms of bFGF in our Western blot analysis. These three isoforms (18, 21.5, and 24 kd) plus a fourth one (22 kd) have been described formerly.12'28 It is possible that in our detection system the 21.5 and the 22 kd isoforms were not distinguishable; therefore only three isoforms have been detected by the anti-bFGF antibody we used. In the cells bFGF was found to be localized in nuclei, cytosol, and membranes. The localization of bFGF in nuclei of HMC has also been reported for other growth factors and cytokines, notably EGF, nerve growth factor, IL-1, and acidic FGF in other cellular systems.29 In other cells it has been speculated that bFGF might directly regulate genes relating to cell division.29 IL-ia and to a minor extent PDGF led to an accumulation of bFGF (mainly the low molecular weight isoform) in the membranes of HMC, whereas IL-13, IL-6, bFGF, and LPS showed no significant effect on the intracellular distribution. These results may suggest that the membranes may function as storage depots for bFGF. bFGF protein could not be detected by Western blot analysis in native cell culture SNs. However, when the SNs were concentrated it became readily detectable also in growth arrested cells. As we have recently shown that these cells still exhibit low proliferation rates,20 we speculate that bFGF secretion contributes to this basal growth. The secreted bFGF was somewhat smaller (-16 kd) as compared with the small intracellular isoforms found in the nuclear and membrane fractions of HMC or the recombinant bFGF (-18 kd). Similar observations have been reported for other cells.30 That the secreted bFGF is bioactive with respect to growth induction is evident from the growth-inducing capacity of the concentrated HMC-SN, which was inhibitable by neutralizing anti-bFGF antibodies. Secretion of bFGF has also been found in aortic endothelial cells and smooth muscle cells.30 Most interestingly, the release of this specific bFGF isoform can be stimulated in HMC by various cytokines including IL-1, PDGF, or bacterial LPS. The inhibition of the growth-inducing capacity by specific antibodies suggested that the major growth factor secreted could be attributed to bFGF. These experiments also showed that growth induction by PDGF is partially mediated by bFGF, as well. A combination of anti-bFGF and anti-PDGF antibodies in all instances completely abrogated the growth-induction capacity that was induced by cytokines or LPS. As it has been shown that PDGF can also be secreted by mesangial cells, this strongly suggests that bFGF and PDGF are the only autocrine growth factors that are induced by cytokines or LPS. In contrast to rat MC, where PDGF appears to be the predominant
autocrine growth factor,27'31 our results thus establish a major role for bFGF in human mesangial cells. bFGF is a particular cytokine, because like IL-1,26 it has no signaling sequence and is probably not secreted in the usual way.25 It has therefore been postulated that bFGF is mostly released by sublethal or lethal injuries of the cells.9'10 Mechanisms of secretion of bFGF by nonlethal stimuli have been proposed,25'32 but so far no specific factors leading to secretion of bFGF from viable cells have been identified. In our experiments we only detected one, ie, the low molecular bFGF isoform (-16 kd) in the concentrated HMC-SNs, which suggests that bFGF release is not the result of mesangial cell death. In addition to inducing bFGF release, IL-1, PDGF, and LPS also enhanced the production of bFGF mRNA as shown by our RT-PCR findings and confirmed by the hybridization with radiolabeled oligonucleotides for bFGF. It should be noted that only those stimuli that led to the release of bFGF, namely IL-1,, PDGF, and LPS, induced the generation of autocrine growth. The differences in the activity of IL-la and IL-1,B in this respect deserve special attention. Although both cytokines enhanced transcription of bFGF, only IL-i1 led to a concomitant release of bFGF and subsequent autocrine growth promotion, whereas in the case of IL-la only an enrichment of bFGF in the membrane fraction was observed. Thus, apparently only secreted bFGF can act via its own receptors and in this way constitutes an autocrine growth-stimulating mechanism. Our data suggest that the proliferative effect of IL-1 sometimes described in other studies seems to be an indirect effect mediated by induction of bFGF mRNA and enhanced bFGF release from HMC resulting in growth of HMC. This suggests that in an inflammatory process or during a bacterial infection IL-1 and LPS, via bFGF, are able to induce HMC to proliferate. In a counterregulation of this proinflammatory process, TGF-f might be involved in downregulation of HMC growth,20 as indicated in our proliferation assays with HMC. However, for a more complete understanding of the cytokine network studied herein we have to consider additional data. 1) Previous studies of our group with HMC have identified a positive autocrine IL-1 loop, ie, IL-1 a and IL-1,B induced both IL-1 a and IL-i1 mRNA and protein in HMC (HH Radeke, CA Kelting, K lossifidou, K Resch, manuscript submitted for publication). 2) IL-ia and -1 induced IL-623'24 and PDGF (A Francki and HH Radeke, unpublished observations). 3) LPS induced PDGF protein release into HMC-SN (A Francki, P Uciechowski, and HH
Autocrine Growth Regulation of Human Mesangial Cells
AJP November 1995, Vol. 147, No. 5
Radeke, unpublished observations). Thus, HMCs once activated by an initial inflammatory stimulus may have the potential for a sustained proliferative response. Many human glomerular diseases are characterized by mesangial cell proliferation, which also contributes to changes in matrix turnover, accumulation of matrix, glomerulosclerosis, and subsequently the final loss of glomerular function.631 33 In chronic inflammatory glomerular diseases a self-limiting repair process after induced injury9 34 followed by reparative growth of mesangial cells after destruction is disturbed.6'8'31'33 A better understanding of the autocrine and/or paracrine growth mechanisms involved may ultimately result in novel therapeutic approaches to chronic inflammatory glomerular diseases.
For a fruitful cooperation and the generous supply of rh bFGF and bFGF antiserum we would like to thank the PBH (Hannover, Germany), especially C Behrens. For helpful hints and detailed discussion on preparing the manuscript we thank Dr. DH Lovett (University of California-San Francisco Department of Medicine, San Francisco, CA).
References 1. Striker LJ, Peten EP, Elliot SJ, Doi T, Striker GE: Mesangial cell turnover: effect of heparin and peptide growth factors. Lab Invest 1991, 64:446-456 2. Alpers CE, Hudkins KL, Gown AM, Johnson RJ: Enhanced expression of "muscle-specific" actin in glomerulonephritis. Kidney Int 1992, 41:1134-1142 3. Klahr S, Schreiner G, Ichikawa I: The progression of renal disease. N Engl J Med 1988, 318:1657-1666 4. Mene P, Simonson MS, Dunn MJ: Physiology of the mesangial cell. Physiol Rev 1989, 69:1347-1424 5. Floege J, Topley N, Resch K: Regulation of mesangial cell proliferation. Am J Kidney Dis 1991, 17:673-676 6. Radeke HH, Resch K: The inflammatory function of renal glomerular mesangial cells and their interaction with the cellular immune system. Clin Invest 1992, 70: 825-842 7. Sterzel RB, Schulze-Lohoff E, Marx M: Cytokines and mesangial cells. Kidney Int 1993, 43(suppl 39):S26-
S31 8. Floege J, Eng E, Young BA, Johnson RJ: Factors involved in the regulation of mesangial cell proliferation in vitro and in vivo. Kidney Int 1993, 43(suppl 39):S47-S54 9. Floege J, Eng E, Lindner V, Alpers CE, Young BA, Reidy MA, Johnson RJ: Rat glomerular mesangial cells synthesize basic fibroblast growth factor. Release, up-
regulated synthesis, and mitogenicity in mesangial proliferative glomerulonephritis. J Clin Invest 1992, 90: 2362-2369 D'Amore PA: Modes of FGF release in vivo and in vitro. Cancer Metastat Rev 1990, 9:227-238 Li Y, Koga M, Kasayama S, Matsumoto K, Arita N, Hayakawa T, Sato B: Identification and characterization of high molecular weight forms of basic fibroblast growth factor in human pituitary adenomas. J Clin Endocrinol Metab 1992, 75:1436-1441 Quarto N, Finger FP, Rifkin DB: The NH2-terminal extension of high molecular weight bFGF is a nuclear targeting signal. J Cell Physiol 1991, 147:311-318 Dono R, Zeller R: Cell-type-specific nuclear translocation of fibroblast growth factor-2 isoforms during chicken kidney and limb morphogenesis. Develop Biol 1994, 163:316-330 Gibran NS, Isik FF, Heimbach DM, Gordon D: Basic fibroblast growth factor in the early human burn wound. J Surg Res 1994, 56:226-234 Bouche G, Gas N, Prats H, Baldin V, Tauber JP, Teissie J, Amalric F: Basic fibroblast growth factor enters the nucleolus and stimulates the transcription of ribosomal genes in ABAE cells undergoing Go -> G1 transition. Proc Natl Acad Sci USA 1987, 84:6770-6774 Radeke HH, Meier B, Topley N, Floege J, Habermehl GG, Resch K: Interleukin 1-a, and tumor necrosis factor-a induce oxygen radical production in mesangial cells. Kidney Int 1990, 37:767-775 Radeke HH, Gessner JE, Uciechowski P, Magert H-J, Schmidt RE, Resch K: Intrinsic human glomerular mesangial cells can express receptors for IgG complexes (hFcyRIl-A) and the associated FcERI y-chain. J Immunol 1994, 153:1281-1292 Floege J, Topley N, Wessel K, Kaever V, Radeke HH, Hoppe J, Kishimoto T, Resch K: Monokines and platelet-derived growth factor modulate prostanoid production in growth arrested, human mesangial cells. Kidney Int 1990, 37:859-869 Radeke HH, Cross AR, Hancock JT, Jones OTG, Nakamura M, Kaever V, Resch K: Functional expression of NADPH oxidase components (a- and f3-subunits of cytochrome b558 and 45-kDa flavoprotein) by intrinsic human glomerular mesangial cells. J Biol Chem 1991, 266:21025-21029 Floege J, Topley N, Hoppe J, Barrett TB, Resch K: Mitogenic effect of platelet-derived growth factor in human glomerular mesangial cells: modulation and/or suppression by inflammatory cytokines. Clin Exp Immunol 1991, 86:334-341 Murphy PR, Sato Y, Knee RS: Phosphorothioate antisense oligonucleotides against basic fibroblast growth factor inhibit anchorage-dependent and anchorageindependent growth of a malignant glioblastoma cell line. Mol Endocrinol 1992, 6:877-884 Isacchi A, Bergonzoni L, Sarmientos P: Complete sequence of a human receptor for acidic and basic fibroblast growth factors. Nucleic Acids Res 1990, 18:1906
Francki et al
AJP November 1995, Vol. 14 7, No. 5
23. Abbott F, Ryan JJ, Ceska M, Matsushima K, Sarraf CE, Rees AJ: Interleukin-1lB stimulates human mesangial cells to synthesize and release interleukins-6 and -8. Kidney Int 1991, 40:597-605 24. Zoja C, Ming Wang J, Bettoni S, Sironi M, Renzi D, Chiaffarino F, Abboud HE, Van Damme J, Mantovani A, Remuzzi G, Rambaldi A: lnterleukin-1l3 and tumor necrosis factor-a induce gene expression and production of leukocyte chemotactic factors, colony-stimulating factors, and interleukin-6 in human mesangial cells. Am J Pathol 1991, 138:991-1003 25. Mignatti P, Morimoto T, Rifkin DB: Basic fibroblast growth factor, a protein devoid of secretory signal sequence, is released by cells via a pathway independent of the endoplasmic reticulum-Golgi complex. J Cell Physiol 1992, 151:81-93 26. Stevenson FT, Torrano F, Locksley RM, Lovett DH: Interleukin 1: The patterns of translation and intracellular distribution support alternative secretory mechanisms. J Cell Physiol 1992, 152:223-231 27. Johnson RJ, Raines EW, Floege J, Yoshimura A, Pritzl P, Alpers C, Ross R: Inhibition of mesangial cell proliferation and matrix expansion in glomerulonephritis in the rat by antibody to platelet-derived growth factor. J Exp Med 1992, 175:1413-1416 28. Klagsbrun M, Smith S, Sullivan R, Shing Y, Davidson S, Smith JA, Sasse J: Multiple forms of basic fibroblast
growth factor: Amino-terminal cleavages by tumor celland brain cell-derived acid proteinases. Proc Natl Acad Sci USA 1987, 84:1839-1843 Baldin V, Roman AM, Bosc Bierne I, Amalric F, Bouche G: Translocation of bFGF to the nucleus is Gl phase cell cycle specific in bovine aortic endothelial cells. EMBO J 1990, 9:1511-1517 Speir E, Sasse J, Shrivastav S, Casscells W: Cultureinduced increase in acidic and basic fibroblast growth factor activities and their association with the nuclei of vascular endothelial and smooth muscle cells. J Cell Physiol 1991, 147:362-373 Floege J, Burns MW, Alpers CE, Yoshimura A, Pritzl P, Gordon K, Seifert RA, Bowen-Pope DF, Couser WG, Johnson RJ: Glomerular cell proliferation and PDGF expression precede glomerulosclerosis in the remnant kidney model. Kidney Int 1992, 41:297-309 Vlodavsky I, Bar Shavit R, Ishai Michaeli R, Bashkin P, Fuks Z: Extracellular sequestration and release of fibroblast growth factor: a regulatory mechanism? Trends Biochem Sci 1991, 16:268-271 Kashgarian M, Sterzel RB: The pathobiology of the mesangium. Kidney Int 1992, 41:524-529 Werber Hi, Emancipator SN, Tykocinski ML, Sedor JR: The interleukin 1 gene is expressed by rat glomerular mesangial cells and is augmented in immune complex glomerulonephritis. J Immunol 1987, 138:3207-3212