The Third Department of Internal Medicine, The Second Department of Internal Medicine, University of Tokyo, Tokyo, and Faculty of Pharmacology,University of Meijo, Aichi, Japan
Expression of a nonmuscie myosin heavy chain in glomerular cellsdifferentiates various types of glomerular disease in rats. To characterizethe phenotypic modulation of mesangial and glomerular epithelial cells,we investigated the expression of a nonmuscie type myosin heavy chain,SMemb, and a-smooth muscle actin (a-SM actin) in rat experimentalglomerular diseases, which included anti-Thy 1 nephritis, 5/6 nephrec-tomy, diabetes, and anti-glomerular basement membrane nephritis.SMemb was only slightly expressed in normal glomerular epithelial cellsbut not in mesangial cells. In the anti-Thy 1 nephritis rats, both SMemband a-SM actin were most conspicuously induced in mesangial cells.However, the expression profile was shifted from a-SM actin to SMembdominant pattern over the course of glomerulonephritis. The expressionof SMemb was also increased in epithelial cells in this model. In the otherthree models, glomerular cells did not express a-SM actin, but did so forSMemb. In the nephrectomized and the diabetic rats SMemb was newlyexpressed in mesangial cells at earlier stages, but at later stages wasremarkably enhanced in epithelial cells when severe glomerular hypertro-phy developed. In the anti-GBM nephritis rats, SMemb expression wasincreased in cpithelial cells. In all models examined, mesangial andepithelial expression of SMemb was confirmed by immunoelectron mi-croscopy, and enhanced expression of SMemb mRNA in glomeruli wasverified by RNase protection assay. We conclude from these results thatglomerular cells change their phenotypes differently depending on varioustypes of glomerular diseases. These phenotypic changes in glomcrular cellscan be revealed by the combined immunostaining for SMemb and a-SMactin. SMemb is especially useful to detect both mesangial and glomerularepithelial cell activation in these glomerular disease models. Understand-ing the functional difference and regulatory mechanisms of these cytoskel-etal proteins will provide insight into the pathogenesis and progression ofglomerular diseases.
There is increasing evidence that mesangial cells are activatedand show phenotypic modulation with enhanced expression ofa-smooth muscle actin (a-SM actin) in various types of glomeru-tar diseases [reviewed in 1—4]. The activated mesangial cells alsochange their character to the secretory phenotype and release amultitude of inflammatory mediators [5]. These pathologicalevents are considered to be associated with mesangial cell prolif-eration, which leads to glomerulosclerosis. Further studies by
Present address: The Second Department of Internal Medicine,University of Gunma, Maebashi 371, Japan.
Received for publication August 3, 1995and in revised form December 8, 1995Accepted for publication December 11, 1995
renal biopsy showed that mesangial cells also express nervegrowth factor receptor (NGF) in a variety of glomerular diseases[6]. However, glomerular expression of either a-SM actin andNGF was found to be limited to the mesangial cell. Recent studieshave shown that epithelial cells are also activated frequently andplay an important role in the progression of glomerular diseases[7—111, which indicates that sensitive markers for both mesangialcell and epithelial cell activation need to be developed.
We previously isolated and characterized eDNA clones forthree types of myosin heavy chain (MHC) isoforms: SMI, SM2,and SMemb. SM1 and SM2 are specific to smooth muscle cells,whereas SMemb is a nonmuscle type MHC. Although SMemb isubiquitously expressed in many types of nonmuscie cells, it is mostabundant in smooth muscles of embryonic aortas, and is down-regulated with vascular development [12—15]. Preliminary obser-vations in our laboratory have shown that SMemb is expressed inmany activated mesenchymal cells in vivo, including smoothmuscle cells and fibroblasts [16].
In the present study, to characterize glomerular cell activationin glomerulopathy, and to differentiate various types of glomeru-tar diseases, we examined the expression of both SMemb anda-SM actin in four glomerular disease models in the rat. Ourresults suggest that: (1) injured glonierular cells change theirphenotypes heterogeneously depending on the type or stage ofdiseases; (2) SMemb expression is closely correlated with mesan-gial and epithelial cell activation, whereas glomerular expressionof a-SM actin is limited to activated mesangial cells.
Methods
Anti-Thy 1 nephritis model
Mesangioproliferative glomerulonephritis was induced by intra-venous injection of anti-Thy 1.1 antiserum (22211D; 0.2 ml/100 gbody wt; Pharmingen, San Diego, CA, USA) into 55 femaleWistar rats weighing 160 to 200 g. Eleven rats were sacrificed fordetermination of SMemb expression in the kidney at days 1, 3, 5,8 or 21. As controls, the same number of female Wistar rats of thesame body wt were intravenously administered with physiologicalsaline.
Other disease models
5/6 Nephrectomy. Subtotal nephrectomy was achieved by rightuninephrectomy, and after two weeks of recovery, resection of the
1231
1232 Hiroi et al: Nonmuscie myosin heary chain in glomerular diseases
upper and lower thirds of the left kidney in 55 male Wistar ratsweighing 160 to 200 g was conducted [17]. Eleven rats weresacrificed for determination of SMemb expression at days 3, 10,and weeks 4, 8, or 16. As controls, the same number of maleWistar rats of the same body wt were sham-operated.
Diabetes. Diabetes mellitus was induced by intravenous injec-tion of streptozotocin (S-U 130; 65 mg/kg body wt; Sigma ChemicalCo., St. Louis, MO, USA) dissolved in citric acid buffer (pH 4.0)through tail veins in 44 male Wistar rats weighing 160 to 200 g[18]. Eleven rats were sacrificed for determination of SMembexpression at days 3, 10, or weeks 4 or 8. As controls, the samenumber of male Wistar rats of the same body wt were intrave-nously administered with physiological saline.
Anti-glomerular basement membrane (GBM) nephritis. Glomer-ulonephritis was induced by intravenous injection of anti-GBMantiserum [19] (1.5 mi/lOU g body wt) into 11 male Sprague-Dawley (SD) rats weighing 160 to 200 g. Eleven rats weresacrificed for determination of SMemb expression at day 7. Ascontrols, the same number of male SD rats of the same body wtwere intravenously administered with physiological saline.
All rats were fed standard rat pellets and had free access to tapwater. Before sacrifice, the systolic blood pressure was measuredby the tail cuff (Natume Co., Tokyo, Japan). The rats were thenhoused individually in metabolic cages (Natume Co.) for collec-tion of 24 hour-urine samples in order to measure daily proteinexcretion. Urinary protein was measured by the pyrogallol redmethod [20]. In the diabetic rats, the blood glucose and urinaryglucose excretion were also measured. Of 11 rats in each model,five rats were provided for conventional histological and immu-nohistochemical examinations. Immunoelectron microscopic lo-calization of SMemb expression was performed in one rat, andRNase protection assay in five rats.
Histological and immunohistochemical examination
The kidneys were removed and fixed in 95% ethanol with 1%acetic acid at 4°C for four hours. The specimens were thendehydrated and embedded in paraffin. One-j.tm thick serial sec-tions were obtained for conventional histological study (hematox-ylin and eosin or periodic acid-Schiff) or immunohistochemistiy.Four monoclonal antibodies were used for immunohistochemicalevaluation: mouse anti-rabbit SM2, mouse anti-rabbit SMemb[15, 161 (SM1 was not revealed in the ethanol fixed rat kidneys bythe anti-SM1 antibody), mouse anti-human a-SM actin (M851,DAKO A/S, Denmark), and mouse anti-human proliferating cellnuclear antigen/cyclin (PCNA) (M879, DAKO A/S). All primaryantibodies used in the present study have been shown to havecross reactivity to the rat antigens [2 1—23]. Immunohistochemicalstudy was performed by the labeled streptavidin biotin kit (DAKOCo.) [24]. Sections were preincubated with 3% hydrogen peroxideand 0.1% sodium azide to reduce non-specific reactions. Primaryantibodies against SM2, SMemb, a-SM actin or PCNA wereapplied at a dilution of 1:50 to 1:100 in Tris-buffered saline with
Table 1. The blood pressure and the urinary protein excretion
Blood pressuremm Hg
Urinary protein excretionmg/day
ATS-3 ControlATS
——
2 0211 42"
ATS-5 ControlATS
——
3 1
132 12°ATS-8 Control
ATS——
3 1
41±17°NX-4 SHAM
OP107±5106±4
7±289±8a
NX-8 SHAMOP
108 11128±8
12 3182±19a
NX-16 SHAMOP
104 6120 10
16 4237 25
DM-4 ControlDM
109 7106±6
5 167±1P
DM-8 Control
DM89 485±15
13 298±4k'
GBM Control
GBM——
11 3372±2a
Abbreviations are: ATS-3, 5, and 8, the anti-Thy 1 nephritis rats at days3, 5, and 8; NX-4, 8 and 16, the nephrectomized rats at weeks 4, 8, and 16;DM4 and 8, the diabetic rats at weeks 4 and 8; GBM, the anti-GBM rats.Values are mean SEM (N = 5).
P < 0.001 vs. controlbp < 0.01 vs. controlP < 0.1 vs. control
Tween 20 (TBST) and incubated for 20 minutes at room temper-ature. Sections were incubated with biotinylated rabbit anti-mouse immunoglobulin (E464, DAKO A/S) for 30 minutes andthen incubated with horseradish peroxidase-labeled streptavidinsolution for 10 minutes. The slides were rinsed in TBST after eachincubation step. Peroxidase was visualized by 3,3'-diaminobenzi-dine tetrahydrochloride (0.2 mg/mi; Sigma Chemical Co., St.Louis, MO, USA) as chromogen with 0.014% hydrogen peroxideas substrate. The slides were washed in tap water and counter-stained with hematoxylin, dehydrated and mounted. Negativecontrols were induced by omission of either of the primaryantibodies in the staining procedure.
Analysis of immunohistochemistiySMemb and a-SM actin expression score. Approximately 100
glomeruli for each kidney were graded according to degree ofSMemb and a-SM actin expression. Glomerular expression was
graded semiquantitatively according to the following proceduredescribed by Kimura et a! [25]: 0 = no expression; 1 = mildexpression (less than 30% of a glomerular area); 2 = moderateexpression (30% to 60% of a glomerular area); 3 = markedexpression (more than 60% of a glomerular area).
This grading was performed by one observer (J.H.) in a blindfashion with coded slides. A weighted composite expression scorewas then calculated for each kidney according to the following
Fig. 1. Immunohistochemical glomerular staining for SMemb (A-D) and a-SM actin (E-H) in the anti-Thy 1 nephritis rats. At day 5, SMemb was newly
expressed in mesangial cells (A). SMemb expression was greatest at day 8 (B), and decreased at day 21(C). In the control rats, SMemb was slightlyexpressed in glomerular epithelial cells (D). a-SM actin was strongly positive at day 5 (E), and expressed extensively at day 8 (F). a-SM actin expressiondisappeared earlier than SMemb at day 21(G). Normal glomeruli did not express a-SM actin, although a-SM actin expression could be observed in therenal arteries and arterioles (H) (original magnification: x260).
,C)
—
4'
w
- t
•f 44
s
It
.0
—
S
I &
' • !4
If"
'slp
iI_
S
t a
as
a a
S
—.
• •
I -
a-
p. £
4,.
t a.
- :_
:' 4ra1
t p
a.
I /
C,
- C
'.
-
: —
__
-
•1
•'•
- •
- 4'
a
.t-
• t.
p a-
1
6 '4
Hiroi et a!: Nonmuscie myosin heavy chain in g!omen4ar diseases 1233
1234 Hiroi et al. Nonmuscie myosin heavy chain in glomendar diseases
2a)0C.)Cl)
0Cl)Cl,ci)
0.8.0E
0.4U) 0.2
0
2.5
**10918
0
U...Z 5-1 **421J
01 3 5 8
Time, days
Fig. 2. Semiquantitative evaluation of immunohistochemistiy in the anti-Thy 1 nephritis rats. (A) SMemb expression scores in the anti-Thy 1nephritis rats. The scores were higher in the anti-Thy I nephritis rats(shaded column) (N = 5) than in the controls (unshaded column) (N = 5)at days 5, 8, and 21. (B) a-SM actin expression scores in the anti-Thy 1nephritis rats. The scores were higher in the anti-Thy 1 nephritis rats thanin the controls at all points. (C) The number of PCNA-positive cells perglomerular cross section in the anti-Thy 1 nephritis rats. PCNA-positivecells peaked at day 5, and then decreased. *p < 0.05, **p < 0.01.
formula: expression score = [1 X (number of grade 1 glomeruli)+ 2 x (number of grade 2 glomeruli) + 3 X (number of grade 3glomeruli)]/(number of glomeruli observed). At the light micro-scopic level, we regarded peripheral cells delineating the glomer-ular tuft as epithelial cells.
RNase protection assayRats were sacrificed under ether anesthesia and kidneys were
removed, and decapsulated. Glomeruli were isolated by differen-tial sieving at 4°C using PBS [26]. Total RNA was extracted byRNA Zol B (Bioteck Laboratories Inc., Houston, TX, USA)according to the acid guanidinium thiocyanate phenol-chloroformmethod [27]. The cRNA probe for rat SMemb mRNA (riboprobe1) was produced by polymerase chain reaction from the cDNA ofrat fibroblast, using primers designed for SMemb-specific se-quences which are conserved in rabbit and human. The amplifiedDNA was subcloned into pCR II (In vitrogen, San Diego, CA,USA) and sequenced by the Taq dye terminator cycle sequencingkit (Applied Biosystems Inc., Foster City, CA, USA). Homologyof amino acid sequence of rat SMemb was 80% and 84% withhuman and rabbit, respectively. In the assay for SMemb mRNA,we included a cRNA probe for human 18S ribosomal RNA (pT7RNA 18S, Ambion Inc., Austin, TX, USA) (riboprobe 2) to
* PCNA-positive cells in glomerular tufts. Proliferation was ex-pressed as the number of PCNA-positive cells per glomerularcross section. Quantitation of PCNA-positive cells was performedby counting all positive cells within the glomerular tufts anddividing by the number of glomeruli present in each section [1].
Determination of glomerular areaThe glomerular areas were measured by tracking the outline of
glomerular tufts with a computer-aided video manipulator(Hamamatsu-Photonics Co., Hamamatsu, Japan). For this pur-pose, 25 nonsclerotic glomeruli were chosen separately from outerand inner cortexes (total 50 glomeruli) for each rat.
Immunoelectron microscopy
B
1 3 5 8 21
ci)
00C,,
Cl)C/)a)0.
C.)
**
*
2
1.5
0.5
0
*
1 3 5 8 21
The kidneys were fixed by perfusing periodate-lysine-parafor-maldehyde (PLP) solution, then cut into three-mm thick slices
* * and immersed in PLP solution overnight at 4°C. After rinsing withphosphate buffered saline (PBS), the tissue was processed forelectron microscopic immunohistochemistry using the pre-em-bedding immunoperoxidase procedure. Fifty-pm sections werecut from the tissue slides on a vibratome (Microsclicer, DosakaCo., Osaka, Japan), washed with 50 mivi NH4C1 in PBS, andpermeabilized by incubation for one hour in PBS containing 1%ovalbumin, 0.2% gelatin, and 0.05% saponin (solution A). Thesections were incubated overnight at 4°C with primary antibodyagainst SMemb diluted 1:200 in PBS with 1% ovalbumin. As anegative control, sections were incubated in PBS with ovalbuminwithout antibody. After rinsing in solution A, the sections wereincubated for two hours with biotinylated rabbit antiserum againstmouse immunoglobulins diluted 1:200. They were rinsed in solu-tion A, and then incubated in horseradish peroxidase-labeledstreptavidin solution for one hour. The sections were rinsed andfixed in 1% glutaraldehyde in PBS. After they were washed in PBSand Tris buffer, peroxidase was detected by incubation in 0.1%diaminobenzidine in Tris buffer for five minutes followed by theaddition of hydrogen peroxide to a final concentration of 0.01%and subsequent incubation for 10 minutes. The sections werepost-fixed in 2% osmium tetroxide in 0.1 M sodium cacodylatebuffer containing 7.5% sucrose for one hour at 4°C, dehydrated in
* graded ethanol, and then embedded in epoxy resin. The ultrathinsections were stained with lead citrate and observed with the
21 transmission electron microscope.
t - •• —C"
A
—
4.trs
;-I . ;1Mgt ,, . ia_p
'/.J(•.('4.
I.. '•
— •
Al•
Hiroi et al: Nonmuscie myosin heavy chain in glomerular diseases 1235
Fig. 3. Immunoelectron micrographs of glomerulin the anti-Thy 1 nephritis rats. In the controls,SMemb reactive material (black dots) wasdemonstrated only in epithelial cells (A)(x2550). In the anti-Thy 1 nephritis rats,SMemb was positive for mesangial, epithelialand endothelial cells (B) (X3400).Abbreviations are: Me, mesangial cell; Ep,epithelial cell; En, endothelial cell.
standardize SMemb mRNA levels with pT7 mRNA levels in eachspecimen. The antisense RNA probe was synthesized using theRiboprobe Gemini System II kit (Promega Corp., Madison, WI,USA). Hybridization with RNA samples, digestion with RNase,and analysis on polyacrylamide gels were performed using theribonuclease protection assay RPA II kit (Ambion Inc.). The fullprotection of riboprobe 1 with SMemb mRNA is 200nt. The fullyprotected fragment of riboprobe 2 with pT7 mRNA is 8Ont.
Statistical analysis
All values are expressed as the mean standard error of themean. For the blood pressure, the urinaly protein excretion, andthe blood glucose concentration, statistical significance was eval-uated using one-way analysis of variance with modified t-testsperformed using Bonferroni's method [28]. The semiquantitative
evaluation of the immunostaining was compared between groupswith non-parametric analysis based on the Kruskal-Wallis test.
Results
Anti-Thy 1 nephritisUrinaiy protein excretion and light microscopy. In control rats,
the urinary protein excretion did not change throughout theexperiment. In anti-Thy I nephritis rats, the urinary proteinexcretion was markedly increased at day 3, and then decreasedgradually from day 5 to day 21 (Table 1). At the light microscopy,the total glomerular cellularity was decreased with acute mesan-giolysis at day 3, and remarkably increased at day 5. At day 21, theglomerular cellularity was decreased again with recovery of nor-mal glomerular configuration. In a number of glomeruli, thecapillary tufts adhered to the Bowman's capsules.
1236 Hiroi et al: Nonmuscie myosin heavy chain in glomemlar diseases
Table 2. Light microscopic changes of the nephrectomized rats and thediabetic rats
Glomerular area,jn2, io-
Sclerosis indexa%
NX-4 SHAMOP
7.35 0.158.80 023b
0 018 i'
NX-8 SHAMOP
9.62 0.21c15.95 032b,c
0 040 2h,c
NX-16 SHAMOP
10.50 0.32e17.01 0.42'
553 4'°
DM-4 ControlDM
6.14 0.14799 019h
0 0b
DM-8 ControlDM
7.91 0.24"'11.44 024b,c
4 l20 i"'
(N = 5).Valucs are mean SEMa Percentage of glomerular cross sections with focal or global glomeru-
lar sclerosishp < 0.001 vs. controlP < 0.001, 4 vs. 8 weeks' P < 0.01, 4 vs. 8 weeks
°P < 0.05, 8 vs. 16 weeks
Immunohistochemistty. Representative glomeruli stained forSMemb or cs-SM actin are shown in Figure 1. In the anti-Thy 1nephritis rats at days I and 3, SMemb was slightly expressed inglomerular epithelial cells, but not in mesangial cells. At day 5,SMemb was newly expressed in mesangial cells, and then ex-pressed in diffuse and nodular patterns in mesangial areas at day8. SMemb expression was still positive at day 21. Collecting ductswere positive for SMemb in the disease models as well as in thecontrols. In the control rats, epithelial cells were slightly positivefor SMemb. SMemb expression scores in the anti-Thy 1 nephritisrats were higher than the controls at days 5, 8 and 21 (Fig. 2A).Glomerular expression of cs-SM actin was already observed at day1, and was greatest at day 8. However, cs-SM actin expression wasdecreased earlier than SMemb at day 21 (Fig. 2B). Normalglomeruli did not express cs-SM actin, although cs-SM actinexpression could be observed in the renal arteries and arterioles.PCNA-positive cells were markedly increased in the anti-Thy 1nephritis rats at day 5 (Fig. 2C).
Immunoelectron microscopy. In the control rats, we confirmedonly epithelial cells were positive for SMemb (Fig. 3A). In theanti-Thy 1 nephritis rats, SMemb-positive cells were identified asmesarigial, epithelial and endothelial cells (Fig. 3B).
mRNA expression of SMemb. SMemb mRNA were markedlyexpressed in glomeruli from the anti-Thy 1 nephritis rats at day 5,and decreased gradually at days 8 and 21 (ATS-5, -8, and -21 inFig. 6). In the control rats, SMemb mRNA was only slightlydetectable (Control in Fig. 6).
Other disease models
5/6 Nephrectomy. The blood pressure in the nephrectomizedrats was similar to that in the controls. The daily urinary proteinexcretion in the nephrectomized rats was remarkably augmentedat week 4, and increased at week 16 (Table 1). Glomerularhypertrophy and sclerosis were observed from week 4, andprogressed until week 16 (Table 2). In a number of glomeruli, theglomerular tufts adhered to the Bowman's capsules with orwithout segmental sclerosis. Representative glomeruli stained forSMemb are shown in Figure 4 A-D. In the nephrectomized rats,SMemb expression was comparable to the controls at day 3. Atday 10, SMemb was de novo expressed in mesangial cells, andmesangial expression of SMemb was markedly increased at week4. At weeks 8 and 16, SMemb was not expressed in mesangialcells, but epithelial expression of SMemb was increased. In thecontrol rats, SMemb was slightly expressed in epithelial cellsthroughout the experiment. SMemb expression scores werehigher in the nephrectomized rats than in the controls (NX4, -8,and -16 in Fig. 5). a-SM actin expression was not detected in anyglomeruli of the disease models or the controls. The number ofPCNA-positive cells in the nephrectomized rats was increased atday 10 (not shown). Immunoelectron microscopy confirmedSMemb expression in mesangial cells at day 10 and week 4, andenhanced expression in epithelial cells at weeks 8 and 16 (notshown). SMemb mRNA was abundant in glomeruli at week 4, andincreased with the progression of disease stages (NX-4, -8, and -16
in Fig. 6).Diabetes. The blood pressure in the diabetic rats was compara-
ble to that in the controls. The daily urinary protein excretion inthe diabetic rats was higher than the controls at weeks 4 and 8(Table 1). Blood glucose levels had already increased by week 4(426 27 mg/dl vs. 128 5 mg/dl in the control, P < 0.001).Extracellular matrix expansion, glomerular hypertrophy and scle-rosis were observed from week 4, and developed at week 8 (Table2). Representative glomeruli are shown in Figure 4 E-H. In thediabetic rats at days 3 and 10, SMemb expression was newlyobserved in mesangial cells. At weeks 4 and 8, SMemb expressionwas markedly increased in epithelial cells but not detected inmesangial cells. In the control rats, SMemb was expressed inepithelial cells but to a much lesser extent. SMemb expressionscores were higher in the diabetic rats than in the controlsthroughout the experiment (DM4 and DM8 in Fig. 5). a-SM actinwas not expressed in any glomeruli. The number of PCNA-positive cells in the diabetic rats was increased at day 3, and thendecreased (not shown). Immunoelectron microscopy confirmedthat SMemb expression in mesangial cells at days 3 and 10, andenhanced expression in epithelial cells at weeks 4 and 8 (notshown). SMemb mRNA was expressed at high levels at week 4,and increased at week 8 (DM-4 and DM-8 in Fig. 6).
Fig. 4. Immunohistochemical glomerular staining for SMemb in the nephrectomized model (.4-D) and the diabetic (E-H). In the nephrectomized rats,SMemb was expressed in mesangial cells at wcek 4 (A). At weeks 8 (B) and 16 (C), SMemb expression was enhanced in epithelial cells. In thesham-operatcd rats (D), SMemb was slightly expressed in epithelial cells. In diabetes, SMemb was expressed in mesangial cells at day 10 (E). Thelocation of SMemb expression was changed to epithelial cells at week 4 (F) and 8 (G). In the controls (H), SMemb was slightly expressed in epithelialcells (X260).
—
— - S
9'—4,
S. ••• CI pr, p •
ft en. .5.• Vb I - •—•1I —_, I
S
I
-I.¼
4,
-I.
•— 0 — _
— — Ia; • - 4 ' -.
1—
I
St
da .*I nt
'F— 5-
H
t.
•1t— . — -t
a- L;b'
I-
—b
—- •I'-. -•
I0
I
Hiroi et al: Nonmuscie myosin heavy chain in glomerular diseases 1237
1238 Hiroi et al: Nonmuscie myosin heavy chain in glomerular diseases
Anti-GBM nephritis. The urinaly protein excretion in the GBMnephritis rats was elevated at day 7 (Table 1). Mesangial expan-sion, tuft adhesions to the Bowman's capsule, and glomerularcrescent formation were observed. Substantial amounts ofSMemb was expressed in glomerular epithelial cells, parietalepithelial cells of the Bowman's capsule, and cellular crescents(not shown). SMemb expression scores were higher in the GBMnephritis rats than in the controls (GBM in Fig. 5). a-SM actinwas expressed in fibroblast-like cells surrounding the Bowman'scapsule, but not in glomerular cells. The number of PCNA-positive cells in the GBM nephritis rats was comparable with thecontrols (not shown). Immunoelectron microscopy revealed thatonly epithelial cells expressed SMemb (not shown). SMembmRNA was markedly increased in the GBM nephritis rats (GBMin Fig. 6).
In the four models, none of the glomeruli expressed 5M2, whilesmooth muscles of the renal arteries and arterioles were stronglypositive for SM2.
Discussion
We demonstrated in the present study that a nonmuscie myosinheavy chain, SMemb, is induced in both mesangial and glomerularepithelial cells with glomerular diseases. SMemb displays a dis-tinct expression pattern from a-SM actin depending on the stageor type of the glomerular disease, indicating usefulness of anonmuscle myosin isoform in diagnosis as well as pathophysiolog-ical understanding of glomerulopathy.
The expression pattern of SMemb and a-SM actin in glomer-ular cells is summarized in Table 3. In the control rats, epithelialcells slightly expressed SMemb. In the anti-Thy I nephritis rats,SMemb expression was enhanced in epithelial cells and newlyobserved in mesangial cells. SMemb appeared later than a-SMactin and continued to be expressed after day 21. In the ncphrec-tomized model and the diabetic rats, SMemb expression wasnewly observed in mesangial cells at earlier stages, but at laterstages, remarkably enhanced in epithelial cells. In the anti-GBM
nephritis rats, the expression of SMemb was increased in epithe-hal cells. a-SM actin expression was only detected in the anti-Thy1 nephritis rats.
SMemb is abundantly expressed in embryonic smooth muscles,in proliferating smooth muscle cells of experimental atheroscle-rotic or arteriosclerotic lesions, and in cultured smooth musclecells [15, 16]. SMemb is also expressed in other mesenchymalcells; cultured fibroblasts, myocardium with acute rejection incardiac transplantation, and hypervascular lesions of hepatic cellcarcinoma (unpublished observations). Because SMemb is thusconsidered to be expressed in activated mesenchymal cells ingeneral, we expected the expression of SMemb to be useful inidentifying activated glomerular cells in glomerulopathy. In thepresent study, we define the activated mesangial and epithelialcells as those cells that change the phenotypes and expressSMemb.
Single injection of anti-Thy 1.1 serum caused transient mesan-gial cell proliferation and massive proteinuria. SMemb expressionwas enhanced both in mesangial cells and epithehial cells, whichwas confirmed by immunoelectron microscopy and RNase pro-tection assay. In this model, the disease is characterized by directbinding of the antibody to mesangial cells, and mesangial cellswere considered to be selectively damaged [1, 29]. However, at theultrastructual level, Bagchus et al have reported that epithelialcells were also damaged with an extensive effacement of footprocesses [30]. Our present results also indicate that both mesan-gial and epithelial cells change their phenotypes in this model.
In the anti-Thy 1 nephritis rats, both SMemb and a-SM actinwere expressed in mesangial cells, but in a different way. Theexpression of SMemb became significant later than day 5, andremained positive after day 21 (Fig. 2A). On the other hand,a-SM actin started to be expressed at day 1 with mesangial cellproliferation, increased until day 8, and then disappeared afterday 21 (Fig. 2B). Along with the progression and regression of thedisease, the mesangial cell revealed transitional expression froma-SM actin to SMemb dominant expression pattern. In thismodel, there was a clear dissociation between the course of acutenephritis and SMemb expression pattern. At day 21, glomerularcells revealed a persistent increase in SMemb expression whenglomerular damage (that is, proteinuria or mesangial cell prolif-eration) was recovering. The healing process of transient injuryinvolves a well-coordinated cascade of cellular and molecularevents, consisting of three stages: an inflammatory stage, prolif-erative stage, and remodeling stage [31]. Thus, in this model, theincreased expression of SMemb in glomeruli may be associatedwith the tissue remodeling after proliferation.
These features of SMemb expression pattern were also ob-served in the early stage of the nephrectomized model and thediabetic. In the nephrectomized rats, PCNA-positive cells wereincreased in glomeruli at day 10, and SMemb was newly expressedin mesangial cells at day 10 and week 4. In the diabetic rats, thenumber of PCNA-positive cells was greatest at day 3, and afterproliferation, SMemb was markedly expressed in mesangial cellsat day 10. Thus, we assume that mesangial expression of SMembin the early stage of these models also reflects an overall repara-tive process.
In the nephrectomized rats, the location of SMemb expressionwas changed from mesangial cells to epithelial cells at week 8,which was supported by immunoelectron microscopy. At week 8,the expression of SMemb was increased in epithelial cells along
**3
1)
2.5
O.50
Rat model
Fig. 5. Semiquantitative evaluation of immunohistochemistiy for SMemb inthe nephrectomized rats, the diabetic rats, and the anti-GBM rats. SMembexpression scores in all models (shaded column) (N = 5) were higher thanin the control groups (unshaded column) (N = 5). Abbreviations are thesame as in Table I. *p < 0.05, **p < 001.
NX4 NX8 NX16 DM4 DM8 GBM
$ I a S
0 Ii S
S
S
S
Hiroi et al. Nonmuscie myosin heavy chain in glomen4ar diseases 1239
267nt
05—2 0oo u-
(0—
>< >< ><z z zL cc/) CI)Q)I.- I- I—< < <
SMemb200nt
Ribosomal RNA8Ont
Fig. 6. RNase protection assay of total glomerular RNA forSMernb gene expression. RNA probe (left), RNA samples from the control rat, cultured ratfibroblast, the nephrectomized rat at weeks 4, 8, 16, the diabetic rat at weeks 4, 8, the anti-Thy 1 nephritis rat at days 5, 8, 21, and the anti-GBM nephritisrat. The 267nt band indicates RNA probe, and the 200nt band indicates SMemb mRNA. The human ribosomal RNA represented by 8Ont bands wasdetermined as an internal control. All lanes contain 5 ig of RNA.
Table 3. The expression of SMemb and a-SM actin in mesangial andglomerular epithelial cells in various glomerular diseases
SMemb a-SM
Mesangial
actin
EpithelialMesangial Epithelialcell cell cell cell
Normal — — —
Anti-Thy 1 nephritis + + + + —
5/6 Nephrectomy +—— — —
— —
— + — —
with severe glomerular hypertrophy. In the previous reports usingelectron microscopy, epithelial cell injury was developed withglomerular hypertrophy, foot process effacement, and detachmentfrom GBM [7, 11]. Thus, the expression of SMemb at week 8 mayindicate progressive epithelial cell injury caused by mechanicalstress.
By RNase protection assay in the nephrectomized rats, SMembmRNA showed the greatest increase at week 16, when SMembexpression score was decreasing by immunohistochemistry. Atweek 16, glomerular hypertrophy progressed and thus epithelialexpression of SMemb continued to increase in nonsclerotic gb-meruli. Accordingly, SMemb mRNA was greatest at week 16.However, at the light microscopy, SMemb expression score wasevaluated by counting all glomeruli, including sclerosing glomeruliwhere SMemb was not expressed. Thus, SMemb expression scorewas decreased when glomerular sclerosis progressed at week 16.
In diabetes, the expression pattern of SMemb was similar to thenephrectomized rats, but epithelial expression was enhanced
earlier than the nephrectomized rats. The glomerular epithelialcells are known to be injured at the onset of metabolic glomerulardiseases including diabetes; ballooning of epithelial cell or flat-tening of foot processes was observed by transmission electronmicroscopy [32]. Therefore, we theorize that SMemb is also usefulto detect epithelial cell injury induced by metabolic disorders.
In the anti-GBM nephritis rats, SMemb expression was en-hanced in gbomerular epithelial cells and parietal epithelial cellsof the Bowman's capsule, which was confirmed by immunoelec-tron microscopy. The previous study using transmission electronmicroscopy described epithelial cell injury at the ultrastructurallevel in this model as the complete detachment of the epithelialcell from GBM, and prominent development of stress fibers infoot processes of parietal epithelial cells connecting to the de-nuded GBM [33]. The expression of SMemb in this model mayindicate the epithelial cell response to the immune-inducedgbomerular damage.
In the present study, a-SM actin was not expressed in anyglomeruli of nephrectomized model or the diabetic. However,recent studies proposed that a-SM actin is newly expressed inmesangial cells of these two models [34, 35]. The reason for lackof a-SM actin expression in the present study is not clear, but thedifference of fixation or antibody may have caused this dissocia-tion. Our results suggest that glomerular expression of a-SM actinmay be only limited to mesangial cell activation.
On the other hand, SMemb expression was not limited tomesangial cells, but observed in both mesangial and epithelialcells in the present study. It has been reported that epithelial cellschange the phenotypes and extensively express SMemb in manydisease models including puromycin nephrosis rats [36]. In varioustypes of gbomerulopathy, the epithelial cell has been reported to
1240 Hiroi et at: Nonmuscie myosin heavy chain in glomerular diseases
be damaged with effacement of foot processes [9] and with thedenudation of GBM, leading to high permeability to macromol-ecules and eventually to glomeruloscierosis [37, 381. Thus, by theinimunostaining for SMemb, we found that epithelial cell activa-tion is a frequent pathological event, and may be closely related tothe progression of glomerular diseases.
To detect epithelial cell injury, several markers such as SPARC(osteonectin) or desmin have been proposed [391. However, thesemarkers have some limitation because enhanced expression ofSPARC was observed only in the complement-mediated glomer-ular diseases, and desmin was not expressed in the anti-Thy 1nephritis model. In contrast, we demonstrated that SMemb wasexpressed in any form of epithelial cell injury, including mechan-ical injury by glomerular hypertrophy, metabolic injury, andimmune-induced injury, showing its usefulness to sensitively de-tect various glomerular diseases.
In conclusion, glomerular cells (that is, mesangial and glomer-ular epithelial cells) change their phenotypes differently over thecourse of glomerular diseases. These phenotypic changes could berevealed by the combined staining for SMemb and a-SM actin.Further studies on the functional significance of the nonmusciemyosiri expression in glomerular cells and the molecular mecha-nisms controlling the nonmuscle myosin gene expression willgreatly enhance the understanding of the pathophysiology ofglomerular diseases.
Acknowledgments
This study was partly supported by Grant-in-Aid for Scientific Research(C) 05670949, the Ministry for Education, Science and Culture (K.Kimura). We are grateful to Drs. Masatoshi Yamamoto, Isao Ebihara,Noboru Watanabe, Toru Minamino, Shigeyoshi Ohba, Naoe Suzuki, andNaobumi Mise for their valuable support, and Dr. Yukio Hiroi for helpingwith RNase protection assay. We thank Mrs. Sanae Ogawa, Mr. Nobuyo-shi Machiyama and Mr. Hiroshi Miyazawa for their technical assistance.
Reprint requests to Kenjiro Kimura, The Second Department of InternalMedicine, University of Tokyo, 7-3-1 Hongo, Bunkyo-ku, Tokyo 113, Japan.
References
1. JOHNSON Ri, LIDA H, ALPERS CE, MAJESKY MW, SCHWARTZ SM,PRITZL F, GORDON K, GOWN AM: Expression of smooth muscle cellphenotype by rat mesangial cells in immune complex nephritis. J ClinInvest 87:847—858, 1991
2. FLOEGE J, JOHNSON RJ, COUSER WG: Mesangial cells in the patho-genesis of progressive glomerular disease in animal models. Clin Invest70:857—864, 1992
3. ALPERS CE, HUDKINS KL, GROWN AM, JOHNSON RJ: Enhancedexpression of 'muscle-specific" actin in glomerulonephritis. Kidney mt41:1134—1142, 1992
4. FLOEGE J, ALPERS CE, BURNS MW, PRITZL P, GORDON K, COUSERWG, JOHNSON RJ: Glomerular cells, extracellular matrix accumula-tion, and the development of glomerulosclerosis in the remnantkidney model. Lab Invest 66:485—497, 1992
5. REDEKE HH, RESCH K: The inflammatory function of renal glomer-ular mesangial cells and their interaction with the cellular immunesystem. Clin Invest 70:825—842, 1992
6. ALPERS CE, HUDKINS KL, FERGUSON M, JOHNSON RJ, SCHATTEMANGC, BOTHWELL M: Nerve growth factor receptor expression in fetal,mature and diseased human kidneys. Lab Invest 69:703—713, 1993
7. SCHWARTZ MM, BIDANI AK, LEWIS EJ: Glomerular epithelial cellfunction and pathology following extreme ablation of renal mass. AmJPathol 126:315—324, 1987
8. YAOITA E, KAWASAKI K, YAMAMOTO T, KTHARA I: Variable expres-sion of desmin in rat glomerular epithelial cells. Am J Pat/so! 136:899—908, 1990
9. LACHAPPELLE M, BENDAYAN M: Contractile proteins in podocytes:Immunocytochemical localization of actin and a-actinin in normal andnephrotic rat kidney. Virchow Arch 60:105—111, 1991
10. NAGATA M, SCHARER K, KRIZ W: Glomerular damage after unine-phrectomy in young rats. I. Hypertrophy and distortion of capillaryarchitecture. Kidney mt 42:136—147, 1992
11. SCHWARTZ MM, BIDANI AK: Role of glomerular epithelial cell injuryin the pathogenesis of glomerular scarring in the rat remnant kidneymodel. Am J Pathol 142:209—219, 1993
12. NAGAI R, LARSON DM, PERIASAMY M: Characterization of a mam-malian smooth muscle myosin heavy chain cDNA and its expression invarious smooth muscle types. Proc Natl Acad Sci USA 85:1047—1051,1988
13. NAGAI R, KuR0-O M, BABIJ P, PERIASAMY M: Identification of twotypes of smooth muscle myosin heavy chain isoforms by eDNA cloningand immunoblot analysis. J Biol Chem 264:9734—9737, 1989
14. KuR0-O M, NAGAI R, KATOH H, T5uCHIM0CHI H, YAZAKI Y,OHKUBO A, TAKAKU F: Developmentally regulated expression ofvascular smooth muscle myosin heavy chain isoforms. J Biol Chem164:18272—18275, 1989
15. KURO-O M, NAGAL R, NAKAHARA K, KATOI-I H, TSAI RC, T5UCHIM0-CHI H, YAZAKI Y, Oj-ucuao A, TAKAKU F: cDNA cloning of a myosinheavy chain isoform in embryonic smooth muscle and its expressionduring vascular development and in arteriosclerosis. J Biol Chem266:3758—3773, 1991
16. AIKAWA M, SIVAN PN, KURO-O M, K1MURA K, NAKAHARA K,TAKEWAKI 5, UEDA M, YAMAGUCHI H, YAZAKI Y, PERIASAMY M,NAGAI R: Human smooth muscle myosin heavy chain isoforms asmolecular markers for vascular development and atherosclerosis. CircRes 73:1000—1012, 1993
17. HOSTETrER TH, OLSEN JL, RENNKE HG, VENKATACHALAM MA,BRENNER BM: Hyperfiltration in remnant nephrons: A potentiallyadverse response to renal ablation. Am J Physiol 24 1:85—93, 1981
18. JENSEN PK, CHRISTIANSEN JS, STEVEN K, PARVING HH: Renal func-tion in streptozotocin-diabetic rats. Diabetologia 21:409—414, 1981
19. ITO H, YAMADA H, OKAMOT0 K, SUZUKI Y: Crescentic type nephritisinduced by anti-glomerular basement membrane (GBM) serum ne-phritis. Jpn J Pharmacol 33:1145—1154, 1983
20. WATANABE N, KAMEI S, OHKUBO A, YAMANAKA M, OHSAWA S,MAKINO K, TOKUDA K: Urinary protein as measured with a pyrogallolred-molybdate complex, manually and in a Hiachi 726 automatedanalyzer. Clin Chem 32:155—1554, 1986
21. KIMURA K, NAGAI R, SASAKI T, AIKAWA M, KuRo-O M, KOBAYASHIN, SHIRATO I, INAGAMI T, OsHI M, SUZUKI N, OBA S, MI5E N, TwoA, HIRATA Y, GoTo A, YAZAKI Y, OMATA M: Diversity and variabilityof smooth muscle phenotypes of renal arterioles as revealed by myosinisoform expression. Kidney mt 48:372—382, 1995
22. KIMURA K, SUZUKI N, OHBA S, NAGAI R, HIR0I J, Miss N, ToJo A,NAGAOKA A, HIRATA Y, GoTo A, YAzAIG Y, OMATA M: Hypertensiveglomerular damage as revealed by the expression of es-smooth muscleactin and non-muscle myosin heavy chain isoform, SMemb. Kidney list(in press)
23. WASEEM NH, LANE DP: Monoclonal antibody analysis of the prolif-erating cell nuclear antigen (PCNA). Structural conservation and thedetection of a nucleolar form. J Cell Sci 96:121—129, 1990
24. GI0RN0 R: A comparison of two immunoperoxidase staining methodsbased on the avidin-biotin interaction. Diagn Immunol 2:161, 1984
25. KIMURA K, Two A, MATSUOKA H, SUGIMOTO T: Renal arteriolardiameters in spontaneously hypertensive rats. Hypertension 18:101—110, 1991
26. JOHNsON RJ, ALPERS CE, PRIZL P, SCHULZE M, PRUNCHNO C,COUSER WG: Platelets mediate neutrophil-dependent immune com-plex nephritis in the rat. J Clin Invest 82:1225—1235, 1988
27. CHOMEZYNSKI P, SACCNI N: Single step method of RNA isolation byacid guanidium thiocyanate-phenol chloroform extraction. Anal Bio-chem 162:156—159, 1987
28. WALLENSTEIN 5, ZUCKER CI, FLEISS JL: Some statistical methodsuseful in circulation research. Circ Res 47:1—9, 1987
Hiroi et a1: Nonmuscie myosin heavy chain in glomerular diseases 1241
29. YAMAMOTO T, YAMAMOTO K, KAWAZAKI K, YAOITA E, SHIMIZU F,KIHARA I: Immunoelectron microscopic demonstration of Thy-iantigen on the surface of mesangial cells in the rat glomerulus.Nephron 43:293—298, 1986
32. FURNESS PN, TURNER SN, APPLEBY P, TURNER DR: A morphologicalstudy of experimental proteinuria using a novel form of surfacefixation. J Pathol 157:37—45, 1989
33. Srnaro I, Sci T, KJMURA K, T0MIN0 Y, K0IDE H: Parietalepithelial cells (PECs) of the Bowman's capsule (BC) actively con-tribute to the glomerular tuft adhesion. (abstract) JAm Soc Nephrol4:633, 1993
glomeruloselerosis in the remnant kidney model. Kidney mt 41:297—309, 1992
35. YOUNG BA, JOHNSON RJ, ALPERS CE, ENG E, GORDON K, FLOEGE J,COUSER WG: Cellular events in the evolution of experimental diabeticnephropathy. Kidney mt 47:935—944, 1995
36. NAKAMURA T, KIMURA K, EBIHARA I, TAKAHASHI T, T0MIN0 Y,AIKAWA M, YAZAKI Y, NAGAI R, K0IDE H: Glomerular expression ofsmooth muscle myosin heavy-chain isoforms in aminonucleosidenephrosis in rats. Cli Sci 89:45—52, 1995
37. RENNKE HG: How does glomerular epithelial cell injury contribute toprogressive glomerular damage? Kidney mt 45(Suppl 45):S58-.563,1994
38. CAMUSS! G, MARIANO F, BIANCONE L, MONTRUCCFJIO G, VERCEL-LONE A: Effect of cytokines on the cytoskeleton of resident glomerularcells. Kidney mt 43(Suppl 39):S32—S36, 1993
39. FLOEOE J, ALPERS CE, SAGE H, PRITZL P, GORDON K, JOHNSON RJ,COUSER WG: Markers of complement-dependent and complement-independent glomerular visceral epithelial cell injury in vivo. LabInvest 67:486—497, 1992
