GDF15 Triggers Homeostatic Proliferation of Acid-Secreting Collecting Duct Cells
Jean Paul Duong Van Huyen*,
Lydie Cheval,
May Bloch-Faure,
Marie France Belair*,
Didier Heudes*,
Patrick Bruneval* and
Alain Doucet
* UPMC University of Paris 06, Unité Mixte de Recherche Scientifique (UMRS) 872, and INSERM, UMRS 872, and UPMC University of Paris 06, Unité Mixte de Recherche (UMR) 7134, Laboratoire de physiologie et génomique rénales, and CNRS, UMR 7134, Laboratoire de physiologie et génomique rénales, Paris, France
Correspondence: Dr. Alain Doucet, UMR 7134, Institut des Cordeliers, 15 rue de l'Ecole de Médecine, 75270 Paris cedex 6, France. Phone: 33-1-55-42-78-51; Fax: 33-1-46-33-41-72; E-mail: alain.doucet{at}bhdc.jussieu.fr
Received for publication July 19, 2007.
Accepted for publication May 26, 2008.
Although adult kidney cells are quiescent, enlargement of specificpopulations of epithelial cells occurs during repair and adaptiveprocesses. A prerequisite to the development of regenerativetherapeutics is to identify the mechanisms and factors thatcontrol the size of specific populations of renal cells. Unfortunately,in most cases, it is unknown whether the growth of cell populationsresults from transdifferentiation or proliferation and whetherproliferating cells derive from epithelial cells or from circulatingor resident progenitors. In this study, the mechanisms underlyingthe enlargement of the acid-secreting cell population in themouse kidney collecting duct in response to metabolic acidosiswas investigated. Acidosis led to two phases of proliferationthat preferentially affected the acid-secreting cells of theouter medullary collecting duct. All proliferating cells displayedpolarized expression of functional markers. The first phaseof proliferation, which started within 24 h and peaked at day3, was dependent on the overexpression of growth differentiationfactor 15 (GDF15) and cyclin D1 and was abolished when phosphatidylinositol-3kinase and mammalian target of rapamycin were inhibited. Duringthis phase, cells mostly divided along the tubular axis, contributingto tubule lengthening. The second phase of proliferation wasindependent of GDF15 but was associated with induction of cyclinD3. During this phase, cells divided transversely. In summary,acid-secreting cells proliferate as the collecting duct adaptsto metabolic acidosis, and GDF15 seems to be an important determinantof collecting duct lengthening.
Loss of kidney functions may be life-threatening and is alwayssocially and economically costly. Reparative therapies triggeringrenewal of kidney epithelial cells would greatly improve thecourse of these pathologies. Identification of the mechanismscontrolling proliferation of specific tubular renal cell populationsis a prerequisite for developing such therapies. Unfortunately,little is known about tubular cell proliferation in the maturekidney.
In contrast to the juvenile kidney, in which active cell proliferationpermits nephron growth in length,1 in adult kidney, tubularcells are quiescent under normal conditions.2 However, renalproximal tubule cell proliferation is triggered in postinjuryrepair after acute renal failure,3,4 ureteral obstruction, orfive-sixths nephrectomy.5 Hyperplasia of specific tubule cellpopulations also occurs in physiologic adaptive processes. Earlyreports suggested that the abundance of collecting duct intercalatedcells (IC), the acid-base transporting cells, increases homeostaticallyin response to changes in acid-base balance6 or to potassiumdepletion,7,8 but these results were not confirmed.9 More recently,the use of cell proliferation markers, such as incorporationof the bromodeoxyuridine (BrdU) into DNA or staining for proliferatingcellular nuclear antigen (PCNA), demonstrated proliferationof distal tubule cells in response to increased sodium delivery10,11and to potassium depletion.12 However, the issue of proliferationof collecting duct type A, acid secreting IC (AIC) during acidosishas not been re-addressed using this approach. Our first aimwas to reinvestigate whether metabolic acidosis increases thepopulation of AIC in the collecting duct and to determine thecontribution of cell proliferation versus transdifferentiationon this effect and the differentiation status of proliferatingcells.
The second aim was to evaluate the putative role of growth anddifferentiation factor 15 (GDF15) in the proliferation of AIC.GDF15, a member of the TGF-β superfamily, is synthesizedas a 40-kD propeptide and undergoes cleavage of its N-ter togenerate an active 30-kD disulfide-linked dimeric protein thatis secreted.13 GDF15 is detected mainly in liver and placentabut is induced in many tissues in response to various stresses.Large-scale analysis of gene expression in mouse outer medullarycollecting duct (OMCD) revealed tremendous overexpression ofGDF15 mRNA during metabolic acidosis14 and potassium depletion.12
Acidosis Induces Early Proliferation of IC
Combined immunofluorescence labeling of outer medullary sectionsof control mice kidney (Figure 1A) showed that all cells werestained either apically for aquaporin 2 (AQP2) or at their basalpole for kAE1, identifying them as principal cells (PC) andAIC, respectively. Electron micrographs (Figure 1C) confirmedthat OMCD cells display structural characteristics of eitherPC (spherical nucleus, basolateral membrane infoldings, smoothapical membrane) or AIC (light cytoplasm, apical microplicae,higher number of mitochondria). This confirms that type B IC(BIC) and non-A non-B IC are not detectable in mouse OMCD.15No cell was double-positive for AQP2 and kAE1. Analysis of 3-dacidotic mice (Figure 1, B and D) also demonstrated the exclusivepresence of PC and AIC and revealed an apparent increase inthe AIC proportion. Cell counting on kidney outer medulla sectionsstained for either AQP2 or kAE1 or on transmission electronmicroscopy (TEM) micrographs demonstrated increased proportionof AIC in 3-d acidotic mice (Table 1).
Figure 1. Identification of AIC and PC in control and acidotic mice OMCD. (A and B) Immunofluorescence of kidney outer medulla sections shows PC with red apical AQP2 labeling and AIC with green basolateral kAE1 staining. In control (A) and 3-d acidotic mice (B), OMCD were continuously lined by either AQP2- or kAE1-labeled cells, without any double-negative or double-positive cell. (C and D) Electron micrographs (bars = 5 µm) show clear PC with smooth apical membrane and interdigitations of basolateral membrane and lighter AIC with numerous mitochondria, apical membrane microplicae, and smooth basolateral membrane. A slight increase in AIC abundance was apparent in 3-d acidotic mice (B and D) as compared with controls (A and C). Magnification, x2500 in C and D.
Table 1. Quantification of PC and AIC in control and acidotic mice OMCDa
Figure 2 shows co-stainings of outer medulla sections with threeproliferation markers and anti-kAE1 antibody. In control mice,no staining was observed for PCNA (Figure 2A), Ki-67, or BrdU(data not shown). In contrast, in 3-d acidotic mice, OMCD cellstaining was readily observed for PCNA (Figure 2, B and C),Ki-67 (Figure 2D), and BrdU (Figure 2, E and F). Most proliferatingcells were positive for kAE1 (AIC, arrows), but some were kAE1-negativePC (Figure 2C, inset, arrowhead). Occasionally, mitotic figureswere seen (Figure 2F). Similar conclusions were reached whenouter medulla sections were co-stained for AQP2 and either Ki-67(Figure 3, A through D), PCNA (Figure 3E), or BrdU (Figure 3F):No proliferating cells were detected in control mice (Figure 3A),and in 3-d acidotic mice, proliferation concerned preferentiallyAQP2-negative AIC (arrows) and occasionally PC (Figure 3D, arrowhead).Quantification of proliferation marker–positive cellsconfirmed that the proliferation rate of AIC exceeded that ofPC (Table 2), thereby accounting for the enlargement of AICpopulation in acidotic mice. Proliferation of both AIC and PCappeared as early as 24 h after the onset of metabolic acidosisand peaked after 3 d (Figure 4).
Figure 2. (A through F) Co-localization of kAE1 and proliferation markers in OMCD from control mice (A) and 3-d acidotic mice (B through F). IC are characterized by basolateral kAE1 labeling (in brown in A through D; in red in E and F), and cells in cell cycle are labeled by anti-PCNA antibody (A through C, red), anti–Ki-67 antibody (D, red), or anti-BrdU antibody (E and F, brown). PCNA-positive cells were present in 3-d acidotic (B, *) but not in control mice (A). Proliferation markers PCNA (B and C), Ki-67 (D), and BrdU (E and F) are mainly present in kAE1-positive AIC (arrows) and occasionally in kAE1-negative PC (C, arrowhead). Occasionally, mitotic cells are seen (F, arrows). Plasma acid-base parameters of mice are shown in Figure 7.
Figure 3. (A through F) Co-localization of AQP2 and proliferation markers in OMCD from control mice (A) and 3-d acidotic mice (B through F). PC show apical and cytoplasmic AQP2 labeling (A through D and F, red; E, brown), and collecting duct cells in cell cycle are labeled by anti–Ki-67 antibody (A through D, brown), anti-PCNA antibody (E, red), or anti-BrdU antibody (F, brown). Ki-67 labeling is increased in the collecting ducts from 3-d acidotic mice (B, *) whereas no Ki-67–positive cell is present in control mice (A). Proliferation markers Ki-67 (B through D), PCNA (E), and BrdU (F) are mainly present in AQP2-negative AIC (arrows) and occasionally in AQP2-positive PC (D, arrowhead).
Figure 4. Proliferating AIC and PC in control and acidotic mice OMCD. Outer medulla sections of kidneys from control (C) and 2- to 14-d acidotic mice (d2 through d14) were labeled with anti-AQP2 and anti–Ki-67 antibodies. In tubular sections with at least one AQP2-positive cell, Ki-67–positive and –negative cells with either positive (PC) or negative AQP2 staining (AIC) were counted. Two hundred cells were counted in each mouse, and the percentages of proliferating PC () and AIC () were calculated. Data are means ± SEM from four to 10 mice. Statistical significance between acidotic and control mice was assessed by ANOVA analysis followed by unpaired t test: *P < 0.05; ***P < 0.001.
Morphometric analysis of OMCD cross-sections on electron micrographs(Figure 5, A through C) revealed several acidosis-induced changes:(1) The cross-section epithelial surface area was slightly butsignificantly reduced in 3-d acidotic mice and returned to controllevel at day 14 (Figure 5D); (2) because the total number ofcells varied in parallel with the epithelium surface (Figure 5E),the mean cell surface area remained constant in the three groupsof mice (Figure 5F); and (3) changes in the total number ofcells were accounted for by a decreased number of PC at day3, which recovered at day 14, and an increased number of AICat day 14 (Figure 5E).
Figure 5. Morphometric analysis of OMCD in control and acidotic mice. (A through C) TEM micrographs of OMCD from control and acidotic mice (bars = 5 µm). In a control mouse (A), the duct lumen is lined by a normal-sized AIC and three PC. In a 3-d acidotic mouse (B), the apparent AIC over PC number ratio is increased. In a 14-d acidotic mouse (C), the AIC cross-sectional surface area is apparently increased and the tubular lumen is dramatically narrowed. AIC enlargement is associated with appearance of cell dimples at the luminal surface (arrowheads), which are not associated with either tight junction or border between two cells. Arrows, cell junctions; arrowheads, cell dimples. (D) Quantification of the cross-section epithelium surface area of OMCD from control mice (), 3-d acidotic mice (), and 14-d acidotic mice (). (E) Number of AIC () and PC () per OMCD cross-section from control and acidotic mice. (F) The cross-section surface area of cells (both AIC and PC) in the three groups of mice was calculated as the ratio of the epithelium surface area over the total number of cells in each OMCD section. Data are means ± SEM from three mice in each group. Statistical significance between acidotic and control mice was assessed by nonparametric Mann-Whitney test: *P < 0.05. Magnification, x2500.
GDF15 and Acidosis-Induced Cell Proliferation
GDF15 mRNA was expressed at low level all along the nephronof control mice and was markedly induced in connecting tubulesand collecting ducts from 3-d acidotic mice, mostly in OMCD(>40-fold increase; Figure 6A). In OMCD, expression of GDF15mRNA increased already 24 h after the onset of acidosis, peakedat day 3, and decreased afterward (Figure 6B), a time coursesimilar to that of AIC proliferation (see Figure 4).
Figure 6. Expression of GDF15 mRNA in nephron segments of control and acidotic mice. (A) Expression profile of GDF15 mRNA along the nephron of control () and 3-d acidotic mice (). PCT, proximal convoluted tubule; PST, proximal straight tubule; cTAL, cortical thick ascending limb of Henle's loop; mTAL, outer medullary thick ascending limb of Henle's loop; DCT, distal convoluted tubule; CNT, connecting tubule; CCD, cortical collecting duct. (B) GDF15 mRNA expression in OMCD from control (C) and 1- to 14-d acidotic mice (d1 through d14). Data are means ± SEM from five to seven mice. Statistical significance between acidotic and control mice was assessed by ANOVA followed by unpaired t test: *P < 0.05; ***P < 0.001.
Compared with wild-type (WT) mice, proliferation of AIC wasreduced drastically in 3-d acidotic GDF15–/– miceand was increased at day 14 (Figure 7A). Similar changes wereobserved for PC (data not shown). GDF15–/– micedisplayed lower blood pH and bicarbonate concentration thanWT mice at days 1 to 3 but afterward compensated their acidosisback toward WT level, concomitantly with the marked increasein AIC proliferation, whereas WT mice did not (Figure 7, B andC).
Figure 7. Effect of acidosis on cell proliferation, and acid-base parameters in WT () and GDF15–/– mice (). (A) Percentage of proliferation of AIC in control (C) and 3- and 14-d acidotic mice (d3 and d14). GDF15 deficiency reduces and postpones acidosis-induced proliferation of AIC. (B and C) Blood pH and plasma bicarbonate concentration in control and 1- to 14-d acidotic mice. GDF15–/– mice display more intense acidosis than WT at the early stages of the treatment and recover afterward to the same level of acidosis. Data are mean ± SEM from five to seven mice. Statistical significance between WT and GDF15–/– mice was assessed by unpaired t test: **P < 0.005.
Under basal conditions, OMCD from GDF15–/– and WTmice displayed similar morphologic features except a narrowerlumen (in µm2± SE; WT 183 ± 8; GDF15–/–52 ± 12; n = 3 in each group; P < 0.05), and a greatercross-section surface area of AIC (in µm2± SE;WT 104 ± 0; GDF15–/– 126 ± 13; n =3 in each group; P < 0.05). In acidotic GDF15–/–mice, the cross-section epithelial surface area increased steadilyfrom day 3 to day 14 (Figure 8A), whereas the total number ofcells remained constant (Figure 8B). Therefore, the mean cross-sectioncell surface area increased in parallel with the epitheliumsurface area (Figure 8C). The constancy of the total numberof cells was accounted for by a steady increase in AIC and amirror decrease in PC (Figure 8B).
Figure 8. Morphometric analysis of OMCD in acidotic GDF15–/– mice. (A) Quantification of the cross-section epithelium surface area of OMCD from control mice (), 3-d acidotic mice (), and 14-d acidotic mice (). (B) Number of AIC (hatched bars) and PC (solid bars) per OMCD cross-section from control and acidotic mice. (C) The cross-section surface area of cells (both AIC and PC) in the three groups of mice was calculated as the ratio of the epithelium surface area over the total number of cells in each OMCD section. Data are means ± SEM from three mice in each group. Statistical significance between acidotic and control mice was assessed by nonparametric Mann-Whitney test: *P < 0.05.
Entry of quiescent cells into early G1 requires transcriptionalactivation of cyclins D.16 Cyclin D2 mRNA was not detected inOMCD of either control or acidotic mice. Cyclin D1 showed earlyoverexpression (days 1 to 2) in WT mice and returned towardcontrol level after. Overexpression of cyclin D1 mRNA was bluntedin GDF15–/– mice (Figure 9). Cyclin D3 displayeddelayed overexpression (day 14) in WT mice. In GDF15–/–mice, overexpression of cyclin D3 was evident earlier, as soonas day 3 (Figure 9). Expression of the V1d subunit of H-ATPase,considered as representative of the acid-base regulon,14 increasedgradually during acidosis in WT and GDF15–/– mice(Figure 9), indicating that cell proliferation and inductionof acid-base regulon are independent. p53, known to controlthe transcription of GDF15,17 showed early overexpression atday 1 in WT but not GDF15–/– mice (Figure 9).
Figure 9. Acidosis-induced gene expression pattern in OMCD from WT () and GDF15–/– mice (). mRNA expression of cyclins D1 and D3, V1d subunit of vacuolar H-ATPase (ATP1V1d), and p53 in OMCD from control (C) and 1- to 14-d acidotic mice (d1 through d14). Data are means ± SEM from five to seven mice. Statistical significance was assessed by ANOVA followed by unpaired t test: *P < 0.05, acidosis versus control; P < 0.05, GDF15–/–versus WT.
Because expression of cyclin D1 and G1 cell-cycle progressionare regulated by phosphatidylinositol-3 kinase (PI3K)/Akt/mammaliantarget of rapamycin (mTOR) signaling in cancer epithelial cells,18,19we investigated the involvement of this pathway in the earlyphase of acidosis-induced proliferation. Administration of thePI3K inhibitor LY294002 or the mTOR inhibitor rapamycin abolishedthe early AIC proliferation (day 2) and the overexpression ofcyclin D1 (Figure 10).
Figure 10. Effect of PI3K and mTOR inhibition on cyclin D1 expression and cell proliferation. (A) Expression of cyclin D1 mRNA in OMCD from control and 2-d acidotic mice treated or not (C) with PI3K inhibitor LY294002 (LY, 30 mg/kg once) or mTOR inhibitor rapamycin (Ra, 0.2 mg/kg per d). (B) Percentage of proliferation of AIC (AQP2-negative, Ki-67–positive cells) in the same groups of mice. Data are means ± SEM from three to six mice. Statistical significance was assessed by ANOVA followed by unpaired t test: *P < 0.05, acidotic versus control.
Acidosis Increases AIC Population via Proliferation of Fully Differentiated Cells
Early reports suggested that the AIC population grows duringmetabolic acidosis,6 but later this finding was not confirmed.9Meanwhile, Schwartz and Al Awqati20 developed the concept ofcell plasticity (transdifferentiation) to account for homeostaticadaptation of the rabbit cortical collecting duct to metabolicacidosis. They interpreted this adaptation to acidosis as representingconversion of BIC to AIC and demonstrated the role of hensinin that process.20,21 Our results (Table 2) demonstrate an enlargementof the AIC population in OMCD during acidosis that is accountedfor by preferential proliferation of AIC and not transdifferentiation:(1) We confirmed that mouse OMCD contains neither BIC nor non-Anon-B IC,15 the potential precursors of AIC in the plasticityprocess, and (2) proliferating IC display basolateral expressionof the AIC marker kAE1 (Figure 2). This conclusion does notpreclude the possibility that transdifferentiation may participatein adaptation of the cortical collecting duct, which containsboth BIC and non-A non-B IC.
It is commonly accepted that terminally differentiated epithelialcells can no longer enter cell cycle and divide. Accordingly,the turnover of permanently renewing epithelia, such as theintestine or the skin, proceeds through division of progenitorcells that differentiate afterward during their migration fromintestinal crypts to the apex of the villi or from the epidermisbasal layer to the skin surface. In epithelial cells with highproliferating capacity, such as hepatocytes during liver regeneration,reappearance of fetal markers22 suggests that they undergo partialdedifferentiation at the G0/G1 transition. Postinjury regenerationof renal tubules also relies on proliferation of partially dedifferentiatedepithelial cells and, possibly, of progenitors.3,23–25In contrast, our results show proliferation of cells displayingnormal polarized expression of specific functional markers.In some instances, mitotic cells with basolateral kAE1 labelingwere seen. Proliferation of fully differentiated cells in thenormal kidney was documented by another group.1,11 Teleologically,it is significant that in adaptive processes, cell proliferationdoes not alter the integrity of the epithelium barrier so asto maintain the tubular function. In contrast, when the integrityof the epithelium is compromised beforehand, as in postinjuryrepair, proliferation may rely on the recruitment of progenitorcells and/or on cell dedifferentiation.
AIC Proliferation Contributes to Acid-Base Homeostasis
Collecting ducts play a major role in acid-base homeostasisby controlling bicarbonate, proton, and ammonium excretion.Thus, loss-of-function mutations of proteins involved in collectingduct proton secretion are responsible for metabolic acidosis.26Reciprocally, diet-induced acidosis promotes homeostatic adaptationsof collecting ducts, including targeting of acid-base transportersfrom intracellular pools to the plasma membrane27 and transcriptionalinduction of genes involved in acid-base metabolism.14 Our resultsshow that AIC proliferation also participates in early homeostaticadaptation of the collecting ducts. As a matter of fact, reducedAIC proliferation observed in 1- to 3-d GDF15–/–acidotic mice was associated with more pronounced acidosis.Conversely, acid-base parameters in GDF15–/– micereturned to WT level at day 14, when proliferation was triggeredback (Figure 7, B and C).
Mechanisms Underlying Proliferation of AIC
In WT mice, the high proliferation rate of AIC observed at days2 and 3 (Figure 4) and the lack of variation of their cross-sectionnumber at day 3 (Figure 5E) indicate that AIC divided mainlyalong the tubular axis (longitudinal proliferation), which therebyshould lengthen OMCD. Given that during the same period PC proliferationrate was weaker (Figure 4), tubule lengthening required thelongitudinal migration of some PC, accounting for the reductionof their cross-section number (Figure 5E). Conversely, fromday 3 to day 14, proliferation of AIC decreased drastically(Figure 4), whereas their cross-section number increased (Figure 5E),indicating that dividing AIC were oriented mainly parallel tothe cross-section plane of OMCD (transverse proliferation).Transverse proliferation was also evidenced by the marked increasein the number of AIC doublets, without interposed PC in cross-sectionsfrom day 3 to day 14 acidotic mice (data not shown). Again,PC proliferation was slower during that period (Figure 4), suggestingthat increased cross-section number of PC (Figure 5E) mighthave been accounted for by elongation of PC along the tubuleaxis. This cell remodeling is compatible with the decreasedcross-section surface area of PC observed at day 14 (data notshown). Longitudinal AIC proliferation is concomitant with overexpressionof GDF15 (Figure 6B) and cyclin D1 (Figure 9), whereas transverseproliferation appears later and concomitantly with cyclin D3overexpression (Figure 9). Figure 11A schematizes the proposedcell proliferation, migration, and remodeling processes inducedby acidosis in WT mice. Longitudinal cell proliferation andsubsequent lengthening of OMCD in acidotic mice, in absenceof lengthening of neighboring thick ascending limbs (i.e., withoutincreasing the height of the outer medulla) requires that OMCDbecome sinuous in acidotic mice, which was commonly observedduring microdissection (Figure 11B).
Figure 11. Longitudinal and transverse proliferation of AIC. (A) Model picturing morphometric changes observed in OMCD between control (C) and day 3 and day 14 (d3 and d14) WT acidotic mice. For the sake of clarity and because the proliferation rate of AIC (dark) exceeded that of PC (light), only divisions of AIC are shown in this simplified model. From C to d3, longitudinal division (arrows) of A and D AIC lengthens the tubule, whereas migration of PC from one ring of cells to the next, newly formed one decreases the number of cells per ring. These changes account for the decreased cross-section number of PC and epithelia surface area and the maintenance of the cross-section surface area of individual cells. From d3 to d14, transverse division of A and E AIC (arrows) and remodeling of PC account for increased cross section number of AIC and decrease cross-section number of PC. (B) Light microscopy photographs of OMCD and mTAL microdissected from control and acidotic mice. OMCD from acidotic mice were sinuous, whereas mTAL remained rectilinear.
GDF15 and Longitudinal Proliferation of AIC
In acidotic GDF15–/– mice, the early proliferationof AIC was reduced (Figure 7A), but their cross-section numberwas increased (Figure 8B), indicating transverse proliferation.Early transverse proliferation of AIC was associated with overexpressionof cyclin D3, which was evident earlier in GDF15–/–mice than in WT mice (Figure 9).
On the basis of the marked induction of GDF15 expression inregenerating liver, Lee et al.28 first proposed that GDF15 mightcontrol cell proliferation, but they failed to demonstrate decreasedliver regeneration in GDF15–/– mice; however, theyevaluated hepatocyte proliferation 1 to 7 wk after partial hepatectomy,whereas they observed increased expression of GDF15 within hours.Given our results showing early effect of GDF15 in the kidney,it is likely that they missed the period when GDF15 controlsliver cell proliferation.
Altogether, our results suggest that GDF15 may control longitudinalproliferation of AIC and point to a role of cyclin D1 in thismechanism. GDF15 also prevented transverse proliferation, becauseoverexpression of cyclin D3 (Figure 9) and transverse proliferationwere triggered earlier in GDF15–/– mice than inWT mice. Thus, GDF15 seems more important to determine the orientationof the mitotic spindle along the OMCD axis than to induce growthitself.
GDF15-dependent proliferation of AIC and overexpression of cyclinD1 were prevented by inhibition of the PI3K/Akt/mTOR pathway(Figure 10), but whether this pathway is directly activatedby GDF15 remains unknown. Interestingly, Akt was recently reportedto play a major role in the orientation of the mitotic spindlein Drosophila embryo.29
Also unknown is the mechanism triggering GDF15 expression duringacidosis. GDF15 promoter contains a p53 response element,17and p53-dependent induction of GDF15 has been reported in severalsystems and conditions; however, early overexpression of p53observed in OMCD of acidotic mice is likely not responsiblefor induction of GDF15 because it was blunted in GDF15–/–mice (Figure 9). Overexpression of p53 might rather be a consequenceof GDF15-induced proliferation and serve as gatekeeper of celldivision.30 p53-independent induction of GDF15 has been reportedin kidney, lung, and liver.31,32 Alternately, acidosis mightinduce directly GDF15. As a matter of fact, metabolic acidosisand GDF15 overexpression are also associated in mice after potassiumdepletion12 and liver resection.28,33
Acidosis and OMCD Hypertrophy
Cell hypertrophy is defined as cell enlargement as a resultof increased protein/DNA ratio. It is either dependent or independenton cell cycle.16 Cell cycle–independent hypertrophy resultsfrom inhibition of protein degradation in quiescent (G0) cells.In contrast, cell cycle–dependent hypertrophy resultsfrom entry in the cell cycle and blockade before progressionthrough the S phase, before DNA replication.
Hypertrophy could not be evaluated directly, because this wouldrequire evaluating protein/DNA ratios in AIC and PC separately.Alternately, increased cell volume can be considered as presumptiveevidence of cell hypertrophy, but cell volume could not be determinedeither. As a last and questionable resource, we used the cross-sectionepithelium surface area/number of cells ratio as an index ofhypertrophy. This approximation entailed errors because cellremodeling may modify cross-section surface area but not cellvolume (as proposed for PC in Figure 11); however, this biasis circumvented in part when considering AIC and PC together,because axial elongation of one cell type is supposed to palliatetransverse invasion by the other cell type.
Provided these biases do not induce redhibitory errors, theconstancy of the mean cross-section cell surface area in controland acidotic WT mice (Figure 5F) suggests that OMCD do not hypertrophyduring acidosis. In contrast, in acidotic GDF15–/–mice, mean cross-section cell surface increased steadily (Figure 8C),suggesting cell hypertrophy. Thus, GDF15 seems to prevent acidosis-inducedcell hypertrophy in OMCD. An anti-hypertrophy role of GDF15was reported in the mouse heart, where overexpression of GDF15protected cardiomyocytes against pressure overload–inducedhypertrophy.34 Increased cross-section surface area of OMCDand of AIC and decreased lumen surface area observed in controlGDF15–/– mice suggest that GDF15 may also play itsanti-hypertrophy role during kidney development.
In kidney, both cell cycle–dependent and –independentmechanisms of hypertrophy have been described.35 Because OMCDhypertrophy in GDF15–/– mice is concomitant withcell proliferation, it is likely to be dependent on cell cycle.
In summary, this study demonstrated that acidosis induces homeostaticproliferation of acid-secreting cells in the mouse collectingduct according to a two-phase process: Early longitudinal proliferationis associated with overexpression of GDF15 and cyclin D1 andwith activity of the PI3K/Akt/mTOR pathway, whereas late transverseproliferation is independent of GDF15 and depends on cyclinD3. We propose that GDF15 is an important determinant of collectingduct lengthening, through the control of the orientation ofthe mitotic spindle, which may also play a role during development.
Animals
Reproduction couples of GDF15 gene targeted mice (GDF15–/–)in the strain background C57BL/6/129/SvJ were provided by Dr.Se-Jin Lee (John Hopkins University, Baltimore, MD). Animalswere backcrossed with C57BL/6J mice (Charles River, L'Arbresle,France) and genotyped by single PCR. Control mice were eitherWT littermates of GDF15 null mice or C57BL/6 mice. There wasno difference in any parameter analyzed between the two typesof controls. Animals were maintained on a standard diet (A04;SAFE, Epinay, France). Metabolic acidosis was induced by feedingthe mice an NH4Cl-supplemented diet (5 g of powdered A04 diet,0.2 g of agar agar, and 6 ml of 0.7 M NH4Cl) for 1 to 14 d.Acidosis was verified before the mice were killed by measurementof blood acid-base parameters (AVL Compact 1; AVL Instruments,Eragny, France). Some mice were administered an injection ofBrdU (0.1 mg/g body wt in PBS, intraperitoneally) 18 and 4 hbefore being killed. In vivo inhibition of PI3K and mTOR wasachieved as described previously with LY29400236 and rapamycin,37respectively. LY294002 (Calbiochem, Nottingham, UK) was administeredonce (30 mg/kg in DMSO 50%, intraperitoneally) at the onsetof NH4Cl feeding, and animals were killed 2 d later (day 2).Rapamycin (MPBio, Illkirch, France) was injected twice (0.2mg/kg per d in 3% DMSO, intraperitoneally), at the onset ofthe treatment and 1 d latter, and animals were killed on day2. Nonacidotic control mice were treated with LY294002 or rapamycinusing the same protocols. All animal experiments were performedin accordance with the French legislation.
Immunohistochemistry
Kidneys were fixed by in situ perfusion, removed, sliced intofour sections, and fixed in formalin for an additional 18 hbefore paraffin embedding. Double immunohistochemistry was performedon 5-µm-thick kidney sections with pairs of the followingantibodies: Monoclonal anti-AQP2 antibody (1:800012), polyclonalanti-kAE1 antibody (1:400; provided by Dr. Alper), monoclonalanti–Ki-67 antibody (clone TEC-3, 1:25; DAKO, Trappes,France), monoclonal anti-PCNA antibody (PC10 clone, 1:200; DAKO),and monoclonal anti-BrdU antibody (clone BU-33, 1:200; Sigma,St. Quentin Fallavier, France), using either a three-step streptavidin-biotinmethod or ENVISION amplification kit (DAKO), with previous antigenunmasking procedure.12,38 Secondary anti-rabbit (1:400), anti-mouse(1:200), and anti-rat (1:300) biotinylated antibodies and streptavidin-peroxidaseamplification ELITE and ABC-phosphatase kits were from Vectorlaboratories (Burlingame, CA). Diaminobenzidine (DAKO) and Fast-Red(DAKO) were used as chromogens. Secondary anti-mouse Cy3 antibody(1:800; Amersham, Les Ulis, France) and streptavidin-Cy2 (1:400;Amersham) were used in double-immunofluorescence studies. Anaverage of 200 to 400 cells per mouse were counted in a blindedmanner.
TEM and Morphometry Analysis
Kidneys were fixed in situ with 2% glutaraldehyde in PBS, followedby immersion fixation for 24 h of renal tissue slices. Blocksof tissue dissected around the inner stripe of the outer medullawere submitted to standard technique of TEM. On toluidine blue–stainedsemithin sections, only transverse sections of inner stripecollecting ducts were selected and studied on thin sections.Ten collecting ducts per mouse (n = 3 in control and acidoticgroups) were photographed at x2500 magnification so as to includea complete collecting duct section in a picture. For calibration,a reference grid was photographed at the same magnification.
Using ImageJ program and a macro written by one of us (D.H.,available upon request), image analysis allowed measurementof the following parameters in collecting duct TEM photographs:Numbers of PC and IC, surface areas of individual PC and ICand of the tubular lumen, and number of cells per section.
Microdissection and Reverse Transcription–PCR
The various segments of nephron were dissected from collagenase-treatedkidney slices under RNase-free conditions, as described previously,39and tubular length was determined by image analysis (Visilog,Noesis, France).
Total RNA were extracted from pools of 40 to 60 nephron segmentsusing an adaptation of the procedure of Chomczynski and Sacchi.40Reverse transcription was performed using first-strand cDNAsynthesis kit for reverse transcription–PCR (Roche Diagnostics,Meylan, France), according to the manufacturer's protocol.
Real-time PCR was performed using a cDNA quantity correspondingto 0.1 mm of nephron on a LightCycler (Roche Diagnostics) withLightCycler 480 SYBR Green I Master qPCR kit (Roche Diagnostics)according to the manufacturer's protocol, except that the reactionvolume was reduced to 10 µl. We verified that no amplificationproduct was produced when reverse transcriptase was omitted.Results (arbitrary units per millimeter of tubule length) areexpressed as means ± SEM from several mice. Specificprimers (Supplemental Table) were designed using ProbeDesign(Roche Diagnostics).
Vogetseder A, Karadeniz A, Kaissling B, Le Hir M: Tubular cell proliferation in the healthy rat kidney.
Histochem Cell Biol 124
: 97
–104, 2005[CrossRef][Medline]
Witzgall R, Brown D, Schwarz C, Bonventre JV: Localization of proliferating cell nuclear antigen, vimentin, c-Fos, and clusterin in the postischemic kidney: Evidence for a heterogenous genetic response among nephron segments, and a large pool of mitotically active and dedifferentiated cells.
J Clin Invest 93
: 2175
–2188, 1994[Medline]
Bonventre JV: Dedifferentiation and proliferation of surviving epithelial cells in acute renal failure.
J Am Soc Nephrol 14
[Suppl 1]: S55
–S61, 2003[Abstract/Free Full Text]
Megyesi J, Safirstein RL, Price PM: Induction of p21WAF1/CIP1/SDI1 in kidney tubule cells affects the course of cisplatin-induced acute renal failure.
J Clin Invest 101
: 777
–782, 1998[Medline]
Park SK, Kang SK, Lee DY, Kang MJ, Kim SH, Koh GY: Temporal expressions of cyclins and cyclin dependent kinases during renal development and compensatory growth.
Kidney Int 51
: 762
–769, 1997[Medline]
Hagege J, Gabe M, Richet G: Scanning of the apical pole of distal tubular cells under differing acid-base conditions.
Kidney Int 5
: 137
–146, 1974[Medline]
Oliver J, Macdowell M, Welt LG, Holliday MA, Hollander W Jr, Winters RW, Williams TF, Segar WE: The renal lesions of electrolyte imbalance: I. The structural alterations in potassium-depleted rats.
J Exp Med 106
: 563
–574, 1957[Abstract/Free Full Text]
Ordonez NG, Spargo BH: The morphologic relationship of light and dark cells of the collecting tubule in potassium-depleted rats.
Am J Pathol 84
: 317
–326, 1976[Abstract]
Hansen GP, Tisher CC, Robinson RR: Response of the collecting duct to disturbances of acid-base and potassium balance.
Kidney Int 17
: 326
–337, 1980[Medline]
Loffing J, Le Hir M, Kaissling B: Modulation of salt transport rate affects DNA synthesis in vivo in rat renal tubules.
Kidney Int 47
: 1615
–1623, 1995[Medline]
Wehrli P, Loffing-Cueni D, Kaissling B, Loffing J: Replication of segment-specific and intercalated cells in the mouse renal collecting system.
Histochem Cell Biol 127
: 389
–398, 2007[CrossRef][Medline]
Cheval L, Duong Van Huyen JP, Bruneval P, Verbavatz JM, Elalouf JM, Doucet A: Plasticity of mouse renal collecting duct in response to potassium depletion.
Physiol Genomics 19
: 61
–73, 2004[Abstract/Free Full Text]
Bootcov MR, Bauskin AR, Valenzuela SM, Moore AG, Bansal M, He XY, Zhang HP, Donnellan M, Mahler S, Pryor K, Walsh BJ, Nicholson RC, Fairlie WD, Por SB, Robbins JM, Breit SN: MIC-1, a novel macrophage inhibitory cytokine, is a divergent member of the TGF-beta superfamily.
Proc Natl Acad Sci U S A 94
: 11514
–11519, 1997[Abstract/Free Full Text]
Teng-umnuay P, Verlander JW, Yuan W, Tisher CC, Madsen KM: Identification of distinct subpopulations of intercalated cells in the mouse collecting duct.
J Am Soc Nephrol 7
: 260
–274, 1996[Abstract]
Shankland SJ, Wolf G: Cell cycle regulatory proteins in renal disease: Role in hypertrophy, proliferation, and apoptosis.
Am J Physiol Renal Physiol 278
: F515
–F529, 2000[Abstract/Free Full Text]
Osada M, Park HL, Park MJ, Liu JW, Wu G, Trink B, Sidransky D: A p53-type response element in the GDF15 promoter confers high specificity for p53 activation.
Biochem Biophys Res Commun 354
: 913
–918, 2007[CrossRef][Medline]
Gao N, Flynn DC, Zhang Z, Zhong XS, Walker V, Liu KJ, Shi X, Jiang BH: G1 cell cycle progression and the expression of G1 cyclins are regulated by PI3K/AKT/mTOR/p70S6K1 signaling in human ovarian cancer cells.
Am J Physiol Cell Physiol 287
: C281
–C291, 2004[Abstract/Free Full Text]
Grewe M, Gansauge F, Schmid RM, Adler G, Seufferlein T: Regulation of cell growth and cyclin D1 expression by the constitutively active FRAP-p70s6K pathway in human pancreatic cancer cells.
Cancer Res 59
: 3581
–3587, 1999[Abstract/Free Full Text]
Schwartz GJ, Tsuruoka S, Vijayakumar S, Petrovic S, Mian A, Al-Awqati Q: Acid incubation reverses the polarity of intercalated cell transporters, an effect mediated by hensin.
J Clin Invest 109
: 89
–99, 2002[CrossRef][Medline]
Bonney RJ, Hopkins HA, Walker PR, Potter VR: Glycolytic isoenzymes and glycogen metabolism in regenerating liver from rats on controlled feeding schedules.
Biochem J 136
: 115
–124, 1973[Medline]
Ichimura T, Hung CC, Yang SA, Stevens JL, Bonventre JV: Kidney injury molecule-1: A tissue and urinary biomarker for nephrotoxicant-induced renal injury.
Am J Physiol Renal Physiol 286
: F552
–F563, 2004[Abstract/Free Full Text]
Imgrund M, Grone E, Grone HJ, Kretzler M, Holzman L, Schlondorff D, Rothenpieler UW: Re-expression of the developmental gene Pax-2 during experimental acute tubular necrosis in mice 1.
Kidney Int 56
: 1423
–1431, 1999[CrossRef][Medline]
Oliver JA, Maarouf O, Cheema FH, Martens TP, Al-Awqati Q: The renal papilla is a niche for adult kidney stem cells.
J Clin Invest 114
: 795
–804, 2004[CrossRef][Medline]
Al-Awqati Q: Plasticity in epithelial polarity of renal intercalated cells: Targeting of the H(+)-ATPase and band 3.
Am J Physiol 270
: C1571
–C1580, 1996[Medline]
Hsiao EC, Koniaris LG, Zimmers-Koniaris T, Sebald SM, Huynh TV, Lee SJ: Characterization of growth-differentiation factor 15, a transforming growth factor beta superfamily member induced following liver injury.
Mol Cell Biol 20
: 3742
–3751, 2000[Abstract/Free Full Text]
Buttrick GJ, Beaumont LM, Leitch J, Yau C, Hughes JR, Wakefield JG: Akt regulates centrosome migration and spindle orientation in the early Drosophila melanogaster embryo.
J Cell Biol 180
: 537
–548, 2008[Abstract/Free Full Text]
Levine AJ: p53, the cellular gatekeeper for growth and division.
Cell 88
: 323
–331, 1997[CrossRef][Medline]
Zimmers TA, Jin X, Hsiao EC, McGrath SA, Esquela AF, Koniaris LG: Growth differentiation factor-15/macrophage inhibitory cytokine-1 induction after kidney and lung injury.
Shock 23
: 543
–548, 2005[Medline]
Zimmers TA, Jin X, Hsiao EC, Perez EA, Pierce RH, Chavin KD, Koniaris LG: Growth differentiation factor-15: Induction in liver injury through p53 and tumor necrosis factor-independent mechanisms.
J Surg Res 130
: 45
–51, 2006[CrossRef][Medline]
Cucchetti A, Siniscalchi A, Ercolani G, Vivarelli M, Cescon M, Grazi GL, Faenza S, Pinna AD: Modification of acid-base balance in cirrhotic patients undergoing liver resection for hepatocellular carcinoma.
Ann Surg 245
: 902
–908, 2007[CrossRef][Medline]
Xu J, Kimball TR, Lorenz JN, Brown DA, Bauskin AR, Klevitsky R, Hewett TE, Breit SN, Molkentin JD: GDF15/MIC-1 functions as a protective and antihypertrophic factor released from the myocardium in association with SMAD protein activation.
Circ Res 98
: 342
–350, 2006[Abstract/Free Full Text]
Preisig P: What makes cells grow larger and how do they do it? Renal hypertrophy revisited.
Exp Nephrol 7
: 273
–283, 1999[CrossRef][Medline]
Sourbier C, Lindner V, Lang H, Agouni A, Schordan E, Danilin S, Rothhut S, Jacqmin D, Helwig JJ, Massfelder T: The phosphoinositide 3-kinase/Akt pathway: A new target in human renal cell carcinoma therapy.
Cancer Res 66
: 5130
–5142, 2006[Abstract/Free Full Text]
Zahr E, Molano RD, Pileggi A, Ichii H, Jose SS, Bocca N, An W, Gonzalez-Quintana J, Fraker C, Ricordi C, Inverardi L: Rapamycin impairs in vivo proliferation of islet beta-cells.
Transplantation 84
: 1576
–1583, 2007[Medline]
Bariety J, Bruneval P, Hill G, Irinopoulou T, Mandet C, Meyrier A: Posttransplantation relapse of FSGS is characterized by glomerular epithelial cell transdifferentiation.
J Am Soc Nephrol 12
: 261
–274, 2001[Abstract/Free Full Text]
Deschenes G, Doucet A: Collecting duct (Na+/K+)-ATPase activity is correlated with urinary sodium excretion in rat nephrotic syndromes.
J Am Soc Nephrol 11
: 604
–615, 2000[Abstract/Free Full Text]
Marsy S, Elalouf JM, Doucet A: Quantitative RT-PCR analysis of mRNAs encoding a colonic putative H, K-ATPase alpha subunit along the rat nephron: Effect of K+ depletion.
Pflugers Arch 432
: 494
–500, 1996[CrossRef][Medline]
This article has been cited by other articles:
J. A. Oliver, A. Klinakis, F. H. Cheema, J. Friedlander, R. V. Sampogna, T. P. Martens, C. Liu, A. Efstratiadis, and Q. Al-Awqati Proliferation and Migration of Label-Retaining Cells of the Kidney Papilla
J. Am. Soc. Nephrol.,
November 1, 2009;
20(11):
2315 - 2327.
[Abstract][Full Text][PDF]
L. Lind, L. Wallentin, T. Kempf, H. Tapken, A. Quint, B. Lindahl, S. Olofsson, P. Venge, A. Larsson, J. Hulthe, et al. Growth-differentiation factor-15 is an independent marker of cardiovascular dysfunction and disease in the elderly: results from the Prospective Investigation of the Vasculature in Uppsala Seniors (PIVUS) Study
Eur. Heart J.,
October 1, 2009;
30(19):
2346 - 2353.
[Abstract][Full Text][PDF]
Top
Abstract
Introduction
) and AIC (
) were calculated. Data are means ± SEM from four to 10 mice. Statistical significance between acidotic and control mice was assessed by ANOVA analysis followed by unpaired t test: *P < 0.05; ***P < 0.001.
) and PC (
P < 0.05, GDF15–/– versus WT.
