- Email: [email protected]
Macrophages in childhood IgA nephropathy
Kidney International, Vol. 48 (1995), PP. 527—535
Macrophages in childhood IgA nephropathy MIcHIo NAGATA, YUKO Axioic, YURI TSUNODA, YASUHIRO KOMATSU, HIR05HI KAWAGUCHI, YUTAKA YAMAGUCHI, and KATUMI ITO Department of Pediatric Nephrology, Tokyo Women's Medical College, Tokyo; Department of Pathology, Institute of Basic Medical Science, University of Tsukuba, Tsukuba; and The Jikei University School of Medicine, Tokyo, Japan
Macrophages in childhood Ifl nephropathy. The role of glomerular macrophages in IgA nephropathy in children was investigated using a new monoclonal antibody (KP1) as a probe. The average number of glomer-
ular macrophages per patient (ANM/P) was closely correlated with the degree of hematuria (P < 0.01) as well as with the degree of leukocyturia (P C 0.01) in the absence of any correlation with proteinuria, serum IgA levels or the interval between the detection of urine abnormalities and renal biopsy. ANM/F was significantly higher in patients diagnosed pathologically as having focal and diffuse proliferative glomerulonephritis than in patients with minor glomerular abnormalities or advanced sclero-
sis (P C
0.05).
Among various types of glomerular morphology in
individual patients, macrophages predominantly infiltrated glomeruli with cell-proliferative lesions despite an absence of any increase in glomeruli
with minor abnormalities or with sclerosis. Macrophages were mainly localized within the capillary lumen in association with endocapillary proliferative lesions (tuft necrosis), they accumulated in areas of mesangial proliferation, and they were attached to Bowman's capsule in segmental lesions. Macrophages were less evident in sclerosis. Furthermore, ultrastructural analysis revealed macrophages in the paramesangial areas in close proximity to lytic changes in the glomerular basement membrane
and effacement of epithelial foot processes. In addition, some cases in repeat biopsy shows prolonged or increased values of ANM/P after several years of interval in association with progression of proliferative lesions. These results suggest that macrophages infiltrate glomeruli during acute glomerular inflammation, and that they are involved in mesangial prolif-
eration or the development of extracapillary lesions in the absence of apparent clinical symptoms. Furthermore, recurrence or prolonged infiltration may promote progression of IgA nephropathy.
In cases of glomerulonephritis, the numbers of infiltrated glomerular macrophages have been shown to increase [1—11]. Recent studies with experimental models of proliferative glomerulonephritis suggest that secretion of cytokines from infiltrated macrophages promotes progression of the disorder [12—14]. Moreover, cytokines also appear to play a role in the progression of glomerulonephritis in humans [15]. These observations suggest that macrophages might participate in the progression of human glomerulonephritis. Indeed, several studies have shown that the number of infiltrated glomerular macrophages increases in cryoglobulinemia-associated nephropathy, in crescentic glomerulonephritis, in lupus nephritis and in acute post-streptococcal glomerulonephritis [5—7, 11]. However, it has not been fully understood
Received for publication April 25, 1994 and in revised form December 7, 1994 Accepted for publication January 3, 1995
© 1995 by the International Society of Nephrology
whether or in what way macrophages participate in slowly progressive glomerulonephritis.
IgA nephropathy is now the most common type of slowly progressive glomerulonephritis and it is the major cause of end-stage renal disease. Previous studies of the role of macrophages in IgA nephropathy was analyzed in adults and discussions
were based on the numbers of macrophages that had infiltrated glomeruli [2—4, 9, 16]. Hooke and colleagues and Ferrario and colleagues failed to find any increases in numbers of glomerular macrophages in IgA nephropathy [2, 3]. By contrast, three recent reports, all from Japan, have demonstrated increased numbers of glomerular maerophages in patients with mesangial hypercellu-
larities [4, 9, 16]. The conflicting data may be the result of differences between the markers of macrophages used or of differences in the backgrounds of patients (phase of the disease, ages of patients or racial differences). Patients in previous studies were very heterogeneous in terms of age and disease state. In most previous studies only one or a few glomeruli per patient was analyzed. Since glomerular lesions and the extent of infiltration of macrophages differ among glomeruli in a single patient with IgA nephropathy (Fig. 1 C, D in this study), a single or a few glomeruli
would seem to be an inadequate source of data on the average number of macrophages infiltrated in a patient. Moreover, because previous studies involved analyses of frozen sections by immunofluorescence or peroxidase-antiperoxidase techniques, localization of macrophages on the background of detailed glomerular structure was not possible.
The present study was designed to investigate the role of macrophages in glomerular morphology in cases of IgA nephropathy in children. Since biopsy materials in this study were obtained
relative soon after onset of the disease, they might possibly provide evidence of glomerular pathology that is related to the onset of the disease. Macrophages were detected with the monoclonal antibody KP1, which is suitable for immunostaining of paraffin-embedded sections [17]. This antibody, applied to formalin-fixed paraffin-embedded sections, allowed us to investigate the localization of maerophages on a background of tuft structure and
to quantify the extent of infiltration in each patient more objectively than has been possible by previously reported methods. Our results indicate that macrophages are associated with endocapillary proliferative lesions, mesangial areas involved in mesangial proliferation and paramesangial destructive lesions with resultant segmental lesions. The results further suggest that asymptomatic recurrence or prolonged infiltration might promote progression of IgA nephropathy in children.
527
528
Nagata et a!: Macrophages in IgA nephropathy
Methods Patients and materials
Forty-one pediatric cases of IgA nephropathy were analyzed. Patients with various degrees of glomerular damage were examined in this study. All cases were defined as IgA nephropathy from clinical and histological findings. Cases of secondary glomerulonephritis with deposition of IgA were not included in this study.
Renal tissues were obtained by needle or open biopsy, fixed in 10% buffered formaldehyde for 24 hours and then embedded in paraffin.
Immunohistochemistiy
Immunohistochemistry was performed by the streptavidin-biotin staining technique as previously described [18]. In brief, after deparaffinization through a graded alcohol series, sections (4 jtm)
were incubated with a 0.1% solution of protease in phosphatebuffered saline (PBS) for four minutes at 37°C in a water bath for enzymatic digestion. Thereafter, endogenous peroxidase activity was blocked by immersing the slides in 0.3% H202 in methanol
and SCL was defined as when more than 80% of the total glomeruli showed evidence of predominant sclerotic lesions rather than proliferation. The average number of macrophages per glomerulus for each patient (ANM/P) was calculated as the total number of macrophages within all glomeruli divided by the total number of glomeruli in a single section. Glomeruli of less than 130 m in diameter were excluded from the analysis. To clarify more precisely the relationship between the infiltration of macrophages and the glomerular morphology, the pathological diagnosis for individual glomeruli was defined as no or minor proliferation (MIN), moderate proliferation (MOD; <50% of entire glomerular area), severe proliferation (SEV; >50% of entire glomerular area), or sclerotic glomerulus (SCL; sclerosis more dominant than proliferation). Average numbers of macrophages for each type of classified glomerular lesion (ANM!GL) were calculated as the total number of immunopositive cells divided by the number of glomeruli in each classified glomerular lesion in a single section from each patient. The localization of KP1-positive cells within the glomeruli was defined as intracapillary or intra-mesangial. To evaluate the distribution of immunopositive cells within glomeruli, we calculated a mesangial localization index (Cap/Mes), which was defined as the ratio of
for 30 minutes. After washing in PBS, the sections were incubated for 15 minutes with normal rabbit serum. Next, the diluted (X200) KP1 monoclonal antibody (CD 68; DAKO, Glostrup, Denmark) immunopositive cells in the capillaries to those in the mesangium. was overlaid on the tissue sections which were then incubated for When immunopositive cells tended to be localized in the mesanone hour at room temperature. The specificity of this antibody has gial area, this ratio was low. Additionally, to determine whether been well described previously [17]. After washing of specimens in infiltration of macrophages occurred in a prolonged or recurrent PBS, biotinylated antibodies against mouse IgG (DAKO) were manner and whether aggravation of glomerular morphology was applied and slides were incubated for 30 minutes at room present in cases with intensive infiltration, 15 cases out of 41 cases temperature. After three washes in PBS, streptavidin-peroxidase were analyzed with repeat biopsy. Changes in pathological diag(DAKO) was applied to the sections for 30 minutes at room nosis and ANM/P in each patient were compared. temperature. After an additional three washes in PBS, color was developed by incubation with 3,3'-diaminobenzidine in PBS with Ultrastructural analysis 1% H202. Finally, sections were counterstained with periodic Small pieces of cortical tissue from each patient were fixed in acid-Schiff reagent. 3% glutaraldehyde and post-fixed in 1% osmium tetroxide. TisAs a negative and positive controls for immunostaining, tonsil- sues were then dehydrated by the standard procedure and emlar tissues from patients after tonsillectomy were fixed by the same bedded in epoxy resin. Ultrathin sections were mounted on grids procedures as applied to renal tissues and treated either with KP1 and stained with uranyl acetate-lead citrate and then they were
antibody, normal mouse serum or PBS alone. Macrophages in tonsillar tissues reacted with the KP1 monoclonal antibody (serving as a positive control) while tonsillar tissues incubated with either normal mouse serum or PBS alone were entirely negative
observed under a transmission electron microscope (H-800; Toshiba, Japan). Approximately 1.4 glomeruli per patient on
for immunostaining. (serving as negative controls; data not
Statistics
shown).
Evaluations
average were examined.
standard errors (SE). A Values are presented as means non-parametric test, with Spearman's test, was performed for
two-factor correlation. Additionally, one-way ANOVA was perPathological diagnosis in individual patients was made by two of formed for comparisons among several groups. A value of P of the authors (M.N. and Y.Y.) without knowledge of any clinical less than 0.05 was considered significant.
and pathological information about individual patients. Immunopositive cells were defined as cells that reacted with KP1 and Results contained nuclei. The disease state was defined as minor glomerA total of 665 glomeruli from 41 patients was analyzed (an ular abnormality (MGA), focal proliferative glomerulonephritis average of 16 1.3 per patient). The minimum number of (FGN), diffuse proliferative glomerulonephritis (DPGN), or advanced sclerosis (SCL), on the basis of the frequency of glomeruli glomeruli per patient was 8. that showed proliferative changes, in accordance with the criteria Clinical parameters, pathological diagnosis and ANMIP by Churg and Sobin [19]. In brief, MGA was defined as being present when less than 20% of the total glomeruli in a single Table 1 shows several clinical parameters and ANM/P for each section showed evidence of proliferative lesions; FGN was defined group of patients with a specific pathological diagnosis. The as when 20 to 80% of the total glomeruli showed evidence of degree of proteinuria tended to be higher in patients with more proliferative lesions; DPGN was defined as when more than 80% severe histologic damage (P < 0.05: MGA vs. FGN, DPGN vs. of the total glomeruli showed evidence of proliferative lesions; SCL; P < 0.01: MGA vs. DPGN, DPON vs. SCL). FGN and
529
Nagata et al: Macrophages in IgA nephropathy Table 1. Clinical parameters for each group of patients and the correlation between various parameters and ANM/P
Age year Male/female Detection of urine abnormalities to biopsy years
Ug/day
URBC (degree)
MGA
FUN
(N = 8) 1.1
(N = 15) 12.7 0.7
5/3 0.4 1.3
11/4 1,4 0.51
0.1
0.6 0.2"
12,1
1.4
0,04b,C 0,3a,b
2.3 5.7
0.3
DPUN
(N = 12.9
13)
0.9 Q,3C
1.5
2.4 8.2
Total
0.2 1.9
Correlation with ANM/P
(N = 41)
P
0.7
13 0.4
NS
NS
3.1
0.6
27/14 1.7 0.2
3.1 1.8
1.1 0.4 1.3
1.1 2.0 5.5
11.8
3/2
8/5 1.8
SCL (N = 5)
0.2 0.2 0.9
NS
NS <0.01 <0.01 NS NS
4.0 1.5 1.9 04b (/HPF) 276± 14 247± 28 243±20 279± 34 312±19 Serum Igamg/dl 4.7 0.6 6.9 O.9 0.4 0.3 1.4 0.3 5.6 l.3' ANM/P (Iglomendus) Data are shown as means SE. Two-group comparisons are: a MGA vs. FGN; b MGA vs. DPGN; C, MGA vs. SCL; ci and 0, FUN vs. SCL; C and E, DPGN vs. SCL. Lower case letters (a, b, c...) indicate P < 0.05 and upper case letters (A, B, C. ..) indicate P < 0.01. Abbreviations are: NS, not significant; NE, not examined.
DPGN were associated with significantly higher levels of hemaCap/Mes and glomerular lesions tuna compared to MGA (P < 0.05). Leukocyturia in FGN was Cap/Mes for individual patients was significantly lower in significantly higher than that in MGA (P < 0.05). ANM/P in cases glomeruli with MOD and SEV as compared to those with MIN (P of FGN and DPGN was significantly higher than in cases of MGA < 0.05, MIN vs. MOD; P < 0.01, MIN vs. SEV), as shown in (P < 0.05) and SCL (P < 0.05: FGN vs. SCL; P < 0.01: DPGN vs. Figure 3.
SCL). A non-parametric test of the correlation between each parameter and ANM/P, the degree of hematuria and the degree
Changes in ANMIP and morphology in cases of repeat biopsy of leukocyturia revealed significant correlations (P < 0.01 in each ANM/P and the pathological diagnosis in 15 cases in which case). However, the age and gender of the patient, the extent of repeat biopsy was performed are shown in Figure 4. In three cases proteinuria, the serum IgA levels and the time from detection of for which data are shown in this Figure, ANM/P was still high even
urine abnormality to renal biopsy were not correlated with two to four years after the first biopsy. In these cases, the ANM/P.
Morphological correlations Localization of macrophages and structural changes in glomeruli. As
shown in Figure 1, KP1-positive cells (macrophages) were clearly visible and the localization of macrophages within glomeruli was easily assessed. The extent of infiltration by macrophages of individ-
ual glomeruli in a single patient was sometime heterogeneous and depended on glomerular morphology, as shown schematically in Figure 1 C, D. Macrophages were associated with endocapilary proliferative lesions (occasionally with polymorphs), tuft necrosis, mesangial proliferation and extracapifiary cellular lesions. In some cases, macrophages represented a significant proportion of the total number of intraglomerular cells. In areas of proliferation of tuft segments, foamy changes in macrophages were occasionally seen. In addition, macrophages were found at the site of ruptured glomerular
histological diagnosis had progressed from FGN to DPGN. In addition, three patients with low values of ANM/P at the first biopsy showed increases in ANM/P at the second biopsy, after one to six years, with morphological progression. Ultrastructure
Special attention was paid to the structural relationship between macrophages and glomerular tuft segments in a study by transmission electron microscopy. In general, electron-dense deposits in GBM were not dominantly associated with macrophages.
Macrophages localized in the capillary lumen were frequently attached to endothelial cells. However, under these conditions, no structural changes in endothelial cells, GBM or podocytes were noted. By contrast, macrophages that had invaded the mesangial area were associated with considerable and destructive structural
changes. Furthermore, in podocytes, the effacement of foot
basement membrane (GBM). Such lesions were frequently con- processes and dilated cytoplasmic vacuoles were frequently apnected to cellular crescents. However, macrophages were not the parent (Fig. 5). Marked attenuation, undulation and thinning of main constituents of cellular crescents. Furthermore, glomeruli with the GBM accompanied extensive mesangial infiltration by macan accumulation of matrix or sclerotic changes tended to contain rophages (Fig. 5), and lytic changes in the GBM and the mesanfewer macrophages than glomeruli with endocapillary proliferation gial area were occasionally seen (Fig. 6). or mesangial proliferation. In areas of fibrocellular crescents, an Discussion accumulation of macrophages was evident of periglomerular sites. This study represents an effort to resolve the confusion that has ANMIGL and glomerular lesions in each patient arisen from previously published conflicting data about the role of As shown in Figure 2, glomeruli with MOD and SEV contained macrophages in IgA nephropathy [2—4, 9, 16]. Our study has two significantly more macrophages than those with MIN or SCL in each distinct advantages as compared to earlier studies. First, our study patient (P < 0.01 for all comparisons). Glomeruli with SEV con- is limited to pediatric patients. Since the onset of the disease can be estimated from the yearly screening of children's urine, and tained more macrophages than glomeruli with MOD (P < 0.05).
530
A t -
I
-
'',.
'4
1'
cl
I I
4
1
Sn
-
•1
_w
-A
t7i1.
,ap) tctt
Nagata et a!: Macrophages in IgA nephropathy
531
Nagata et al: Macrophages in IgA nephropathy
I
T
20 15
___
a 0
I')
ANM/GL
I
a10
5
1
*
3
4
5
6
7
Time, years
Fig. 2. ANMIGL and classified glomerular lesions. Abbreviations are: MIN, minor proliferation; MOD, moderate proliferation; SEV, severe proliferation; SCL, sclerosis; see text for detailed definitions. Values are < 0.01. mean SE. < 0.05,
I
2
I
2
Fig. 4. ANMIP and pathological diagnosis in eighteen cases of repeat biopsy. and SCL (•). FGN (0), DPGN
()
Second,
the precise localization of macrophages on the back-
ground of detailed glomerular structures was analyzed. The results from this study should contribute considerably to a better further understanding of the role of macrophages in IgA nephropathy. The average number of macrophages per patient (ANM/P) was significantly higher in patients with proliferative lesions (endocapillaiy and mesangial proliferation). Previous authors have classified adult patients by degree of mesangial proliferation and found that the number of macrophages per patient increased with the severity of mesangial proliferation [4, 9, 161. From such results, earlier authors concluded that macrophages are involved in mesangial proliferation. However, their results simply mean that more macrophages were evident in patients with mesangial proliferation and do not directly support their conclusion. To obtain more valid evidence, we classified glomeruli by degree of proliferation in individual patients and found that glomeruli with proliferative lesion contained more macrophages
text; MIN, minor proliferation; MOD, moderate proliferation; SEV, severe proliferation; SCL, sclerosis; see text for detailed definitions.
than those with minor changes in these individual patients. These results further suggest either that macrophages stimulate the formation of or actually form proliferative lesions. Since macrophages secrete several kinds of cytokine in this disease [20] and cytokines have been proven to stimulate proliferation of mesangial cells in vitro [21, 221, and furthermore, since mesangial macrophages seem to be better able to proliferate glomerular cells than macrophages in the
Values are mean
capillary lumen, the precise localization of macrophages within
0 MIN MOD SEV Fig. 3. Mesangial localization indices. Abbreviations are: Cap/Mes, see SE.
< oo5 J3 < 0.01.
glomeruli seems to be crucial if we are eventually to characterize the role of macrophages in glomerular morphology in glomerulonephrisince biopsies were performed approximately 1.5 years after tis. The present study indicates that infiltration of glomerular macasymptomatic onset, the glomerular lesions in the majority of our rophages is closely associated with endocapillary proliferative lesions:
patients may represent pathology related to the onset of disease. tuft necrosis, and with mesangial proliferation. Moreover, the Fig. 1. Light micrographs of glomendi from patients with various types of glomerular morphology. Brown coloration indicates immunoreaction. (A) From a patient diagnosed with MGA (ANM/P = 1.2). No infiltration of macrophages (X400). (B) From a patient diagnosed with FGN (ANM/P 7.4). The localization of macrophages is clearly visible. There are two macrophages in the mesangial area (arrowhead) and three in the capillary (arrow) (X800). (C and D) From a single patient with DPGN (ANMIP = 6.3). The glomerulus with SEV (D) contains many more macrophages and polymorphonuclear cells than that with MIN (C). Proliferative lesions in (D) are mainly the result of endocapillary proliferation (X400). (E and F) From a single patient with FGN (ANM/P = 17.7). Macrophages have accumulated in association with a lytic and expanded mesangial area (E) and a small crescent (F). Note that macrophages are not components of the cellular crescent (X800). (G) From a patient diagnosed with FGN (ANM/P = 6.3). Segmental aggregation of macrophages with foamy changes (arrowhead) is seen in the segmental lesion with adhesions (X 400). (H) A glomerulus classified as SG from a patient diagnosed with SCL. (ANM = 0.4). No macrophages are evident in areas of sclerosis. (x400).
532
Nagata et al: Macrophages in IgA nephropathy
Fig. 5. Electron micrograph of a glomen4ar tuft from a patient with FGN (ANMIP = 17.7). Paramesangial infiltration of macrophages (M) and polymorphonuclear cells (PM) are visible. Note the various changes in the GBM and podocytes (PD). PE indicates parietal epithelial cells. See text for details. (x4,000).
mesangial localization index was lower in glomeruli with prolifera-
tion than in those with minor lesions or sclerosis. From these observations, we speculate that macrophages actively infiltrate gbmeruli as a mediator of the acute phases of glomerular inflammation (endocapillary proliferation) and that approximately half (Cap/Mes is approximately 1 in glomeruli with proliferative lesions) of the
to be essential for the process of progression of IgA nephropathy. Indeed, values of ANM/P were closely correlated with the degree
of hematuria, as well as the degree of leukocyturia, at renal biopsy. The results supports the previous notion that the signifi-
cant infiltration of polymorphonuclear cells were present in
association with gross hematuria or formation of crescents in IgA nephropathy [27, 28]. In our results, macrophages are frequently associated with endocapillary proliferative lesions and tuft necroproliferation of mesangial cells. And such mesangial proliferation is sis and, furthermore, analysis of serial section in our laboratory a major pathologic feature of IgA nephropathy [23, 24]. Mesangial proliferation followed by mesangial sclerosis is con- revealed that such lesions are very frequently apparent at the sites sidered to be a major morphological indicator of the course of IgA of formation of small cellular crescents in this disease [29]. nephropathy in children [23]. However, since mesangial sclerotic Moreover, our observations by electron microscopy confirm the lesions do not predominantly contain macrophages, as shown in light microscopic findings, that is, the close proximity of mesangial this study, the actual process of macrophage-promoted progres- macrophages to lytic mesangial changes and damaged gbomerular sion has not been well characterized. Basically, the pathology of basement membrane, whereas intracapillary macrophages are IgA nephropathy consists of intraglomerular lesions (mesangial associate a lesser extent with such changes. A recent report by proliferation and IgA deposits), while the progression of the Shigematsu, Kobayashi and Hiki also suggests that para-mesandisease largely depends on extracapillary lesions, namely, cres- gial destructive lesions are crucial for the extension of tuft lesions cents or adhesions [25, 261. Therefore, rupture of the GBM seems to extracapillary lesions [30]. Thus, it seems likely that inftltration
macrophages invade the mesangial area to form or to stimulate
Nagata et a!: Macrophages in IgA nephropathy
533
Fig. 6. Electron micrographs of a glomerular tuft from a patient with FGN (ANMIP = 7.4). A) Low-magnification view of tufts. Macrophages (M) are localized in the paramesangial area and in the capillary. B) High-magnification view of the paramesangial macrophage displayed in A). Around an expanded macrophage in the paramesangial area, discontinuous fragments of degenerated GBM (arrowheads) and the effacement of foot processes of podocytes are seen (small arrow). Fenestra of endothelial cell is absent. (X9,600).
of macrophages within the mesangial area can potentially destroy glomerular tufts with resultant hematuria, leukocyturia and formation of segmental lesions, namely, crescents or adhesions. On the other hand, KP1-positive cells were not the main constituent of
children's urine, we were able to estimate the timing of the onset of this disease in each of our patients. Since ANM/P was not correlated with the duration of urine abnormalities and, further, since cases with repeat biopsy revealed prolonged infiltration or increased values of
cellular crescents. Yoshioka and co-workers showed previously that
ANM/P after intervals of several years, we can conclude that
crescentic cells in IgA nephropathy were cytokeratin-positive but negative for several established markers of macrophages [31]. Our
infiltration of macrophages into glomeruli in IgA nephropathy occurs
recent study is consistent with their findings [32], suggesting a parietal epithelial origin. As Magil proposed, macrophages might participate
frequently observed endocapillary proliferative lesions in combination with sclerosis within a single glomerulus. Moreover, in three patients with prolonged high values of ANM/P and three cases of increased values of ANIVI/P at repeat biopsy, the progression of glomerular changes was clearly evident. Hence, we propose that acute inflammatory changes in the glomeruli, represented by the
in formation of crescents via stimulation of parietal epithelial proliferation in an acute phase of inflammation [33]. Further studies to identiti the stimulatory factors for such cell proliferation are warranted. From the clinical observation, we note that most of our cases were
at no fixed time during the course of the disease. In fact, we
infiltration of macrophages, may take place recurrently or in a
identified from urine abnormalities by a program for screening of
prolonged manner and may have a damaging effect on glomerular
urine from asymptomatic school children, and who did not have any clinical signs of acute exacerbation of nephritis. Therefore, it appears that hematuria and leukocyturia are good indicators of macrophage
morphology, without any clinical symptoms, in cases of IgA nephrop-
infiltration in IgA nephropathy in children in the absence of any clinical symptoms.
To obtain further insight into the contribution of macrophages
to the aggravation of abnormal glomerular morphology, it is necessary to determine whether infiltration of macrophages is a transient or a prolonged phenomenon and whether deleterious
athy in children. However, another four cases revealed the progression of glomerular morphology despite low values of ANMIP at first and second biopsy. Since infiltration of macrophages may not be persistent in every case of this disease, the timing of renal biopsy in these cases might not have correspond to a period of active infiltration. Alternatively, factors other than macrophages may be involved in the progression in some cases.
Furthermore, it still remains unknown whether macrophages morphological changes occur in cases with intense infiltration of participate in the chronic progression, namely, in the development macrophages. With the aid of the school program for analysis of of sclerosis. In this study, we found neither an increase in ANM/P
534
Nagata et a!: Macrophages in IgA nephropathy
in cases with advanced sclerosis nor an increase in ANM/GL in sclerotic glomeruli. Moreover, macrophages were not dominantly accumulated in the areas of sclerosis by morphological analysis. Our observations indicate that macrophages are less evident in areas of sclerosis, but they do not necessarily imply that macrophages do not contribute to sclerosis. Indeed, we found foamy changes in macrophages in areas of sclerosis and the aggregation
of macrophages in the periglomerular interstitium at sites of fibrocellular crescents The role of such macrophages in progression of sclerosis is unknown. One possible explanation for it is that macrophages stimulate phenotypic changes in mesangial cells. As
recently be shown, macrophage derived factors suggested to stimulate mesangial proliferation in IgA nephropathy [20]. In the experimental model of mesangial proliferative glomerulonephritis induced by anti-Thy 1 antibody, mesangial proliferation is associated with the expression of a-smooth muscle cell actin (pheno-
typic changes) and an increase in matrix production that leads later to sclerosis [34, 35]. Such phenotypic changes have also been
demonstrated in human glomerulonephritis [36, 37]. Hence, we can speculate that macrophages might participate the develop-
ment of sclerosis by the activation of phenotypic changes in mesangial cells and then the production of matrix by mesangial
cells might be accelerated after the disappearance of macrophages. Further study is clearly needed to determine the role of glomerular macrophages in phenotypic changes in mesangial cells in vivo and in sclerosis in this disease.
In summary, in IgA nephropathy in children, macrophages infiltrate glomeruli during an active phase of glomerular inflammation (endocapillary proliferation) and they are involved in mesangial hypercellularity (either due to accumulation or by stimulating mesangial cell proliferation). We provide evidence
that macrophages also participate in the development of paramesangial destructive lesions, namely, in lysis or attenuation of GBM, as well as in the destruction of the integrity of podocytes. All these changes have the potential to lead to the development of extracapillary lesions, namely, the crescents
and adhesions that are the most reliable predictors of the
chronic progression of this disease. Furthermore, asymptomatic recurrent or prolonged infiltration of macrophages might take place and promote progression. Acknowledgments The authors thank Mr. Shigeru Horita, Mr. Hideki Nakayama and Miss Mayuko Kawashima with their skillful technical assistance. The
authors are grateful to Dr. Hiroshi Shiraga and Professor Teruo Watanabe (Tsukuba University) for critical discussions related to the manuscript. The contributions of anonymous reviewers are also gratefully acknowledged.
Reprint requests to Michio Nagata, M.D., Department of Pathology, Institute of Basic Science, University of Tsukuba, Tsukuba 305, Japan.
References 1. ATKINS RC, GIsGow EF, HOLDSWORTH SR, THOMPSON NM, HAN-
COCK WW: Tissue culture of isolated glomeruli from patients with
4. KOBAYASHI M, KOYAMA A, NARITA M, SHIGEMATSU H: Intraglomeru-
lar monocytes in human glomerulonephritis. Nephron 59:580—585, 1993 5. LAOHAPAND T, CArFELL V, G.uRiAJ. FRT: Monocyte infiltration in human glomerulonephritis: Alpha-1-antitiypsin as a marker for mononuclear phagocytes in renal biopsies. Clin Nephrol 19:309—316, 1983 6. MAGIL AB, WADSWORTH LD, LOEWEN M: Monocytes and human renal glomerular disease. A quantitative evaluation. Lab Invest 44:27— 31, 1981 7. MONGA G, MAZZUCCO G, CASTELLO G: Glomerular monocyte infiltration in human glomerulonephritis: Prevalence and correlation with clinical and morphological variables. Virchow Arch 405:483—496, 1985 8. SCHREINER GF, COTRAN RS, PIt.i)o V, UNANUE ER: A mononuclear
cell component in experimental immunologic glomerulonephritis. J Exp Med 147:369—384, 1978 9. TOMINO Y, OZAKI T, K0IDE H, YAGAME M, EGUCHI K, NoMoTo Y,
SARAI H: Glomerular T cell and monocyte populations in patients with IgA nephropathy. Jap J Nephrol 3 1:221—226, 1989 10. Li HL, HANCOCK WW, Hoia DH, DOWLING JP, ATKINS RC: Mono-
nuclear cell activation and decreased renal function in IgA nephropathy with crescents. Kidney mt 37:1552—1556, 1990 11. SHIGEMATSU H, SHISHIDO H, KUMARA K, TSUCI-IIDA H, SUZUKI H,
HIROSE K, ToJo S: Participation of monocytes in transient glomerular
hypercellularity in poststreptococcal glomerulonephritis. Virchow Arch 12:367—4370, 1973
12. BORDER WA, OKUDA S, LAGUINO LR, SPORN MB, RousIHTI E: Suppression of experimental glomerulonephritis by antiserum against transforming growth factor /31. Nature 346:371—374, 1990 13. TIPPING PG, LOWE MG, HOLDSWORTH SR: Glomerular interleukin-1
production is dependent on macrophage infiltration in anti-GBM glomerulonephritis. Kidney mt 39:103—112, 1991 14. HoRn Y, MURAGUCHI A, IwANo M, MATSUDA T, HIRAYAMA H, FuJii
Y, DOHI Y, ISHIKAWA H, OHMOTO Y, YosHIz.s.iu K, HIRANO T, KISHIMOTO T: Involvement of IL-6 in mesangial proliferative glomerulonephritis. J Immunol 143:3949—3955, 1989 15. Doin K, IwA.NO M, MURAGUCHI A, HoRn Y, HIRAYAMA T, OGAwA 5,
SHnIU H, Hnw'io T, KIHSIMOTO T, ISHIKAWA H: The prognostic significance of urinary interleukin 6 in IgA nephropathy. Clin Nephrol 35:1—5, 1991 16. ARIMA S, NAKAYAMA S, NAITO M, SATO M, TAKAHASFI! K: Signifi-
cance of mononuclear phagocytes in IgA nephropathy. Kidney mt 39:684—692, 1991 17. PULFORD KAF, RIGNEY EM, MICKLEM KJ, JONES M, STROSS W,
GATFER KC, MASON DY: KP1: A new monoclonal antibody that detects a monocyte/macrophage associated antigen in routinely processed tissue sections. J Clin Pat/so! 42:414—421, 1989 18. NAGATA M, YAMAGUCHI Y, ITO K: Loss of mitotic activity and expression of vimentin on glomerular epithelial cells in developing human kidneys. Anat Embryo! 187: 275—279, 1993
19. CI-IURG J, SOBIN LH: Renal disease, in Classification and Atlas of Glomerular Diseases (2nd ed), edited by CHURG J, Tokyo, IgakuShoin, 1982, pp 3—6
20. Yosioi K, TAKEMURA T, MURAI
in IgA nephritis. Kidney Int 44:825—833, 1993 21. MATSUMOTO K: Jnterleukin-1 production by monocytes from patients with glomerulonephritis after in vitro incubation with soluble immune complexes. C/in Nephrol 36:267—273, 1991 22. LOVET DH, RYAN JL, STERZEL RB: Stimulation of rat mesangial cell proliferation by macrophage interleukin 1. J Immunol 13 1:2830, 1983
23. YosHIIwA N, ImMA K, MAKiwt N, YOSHIARA 5, YOSHIYA K, MATSUO T, OKADA S: Mesangial changes in IgA nephropathy in children. Kidney mt 32:585—589, 1987 24. D'AMIco G, IMBASCIATI DI, BELGIOI0S0 GB, BERTOLI 5: Idiopathic
IgA mesangial nephropathy: Clinical and histological study of 374 patients. Medicine 64:49—60, 1985 25. YOSHIKAWA N, ITO H, NAKAMURA H: Prognostic indicators in child-
glomerulonephritis. Kidney mt 17:515—527, 1980 2. HooK DH, GEE DC, ATKINS RC: Leukocytes analysis using monoclonal antibodies in human glomerulonephritis. Kidney mt 31:964—972, 1987 3. FERRARIO F, CASTIGLIONE A, C0L&sANll G, BELGI0I0s0 GR, BER-
hood IgA nephropathy. Nephron 60:60—67, 1992 26. KATAFUCHI R, OH Y, HORI K, KOMOTA T, YANASE T, IKEDA K,
TOLl S, D'AMIGo G: The detection of monocytes in human glomerulonephritis. Kidney mt 28:513—519, 1985
Clin Nephrol 41:191—198, 1994 27. KINCAID-SMITH P, NICHOLLS K, BIRCHALL I: Polymorphs infiltrate
OMURA T, FUJIMI 5: An important role of glomerular segmental lesions on progression of IgA nephropathy: A multivariate analysis.
Nagata et al: Macrophages in IgA nephropathy glomeruli in mesangial IgA glomerulonephritis. Kidney mt 36:1108— 1111, 1989 28. BENNETF WM, KINCAID-SMITH Pr Macroscopic hematuria in mesang-
ial IgA nephropathy, Correlation with glomerular crescents and renal dysfunction. Kidney mt 23:393—400, 1983
29. AXI0KA Yr Structure of segmental tuft lesions in childhood IgA nephropahty. Three dimensional analysis indicates development of segmental lesions. Jap J Nephrol 36:973—981, 1994 (in Japanese) 30. SHIGEMATSU H, KOBAYASHI Y, HIKI Yr Para-mesangial destructive lesions in immunoglobulin A nephritis. Ultrastruct Pathol 14:129—134, 1990
31. Yosmoi K, TAJcEMURA T, AKANO N, MIYvoTo H, IsElu T, Mi Sr Cellular and non-cellular compositions of crescents in human glomerulonephritis. Kidney mt 32:284—291, 1987 32. NAGATA M, HORITA 5, ITo Kr Evidence for phenotypic changes of
glomerular epithelial cells in renal disease. (abstract) J Am Soc Nephrol 4:282, 1993 33. MAGIL AB: Histogenesis of glomerular crescents. Immunohistochem-
535
ical demonstration of cytokeratin in crescent cells. Am J Pathol 120:222—229, 1985 34. FLOEGE J, JOHNSON R I, GORDON K, IIDA H, PRITZL P, YosHIMuitA A,
CAMPBELL C, ALPERS CE, CousER WGr Increased synthesis of extracellular matrix in mesangial proliferative nephritis. Kidney mt 40:477—488, 1991 35. JOHNSON RI, IIDA H, ALPERS CE, MAJESICY MW, SCHWARTZ SM, PRITZL P, GORDON K, GowN AM: Expression of smooth muscle cell
phenotype by rat mesangial cells in immune complex nephritis. J Clin Invest 87:847—858, 1991
36. ALPERS CE, HUDKINS KL, GowN AM, JOHNSON RJ: Enhanced expression of muscle-specific actin in glomerulonephritis. Kidney mt 41:1134—1142, 1992
37. Siouio Y, SHIM0ICAMA T, SxMI T, HARADA A, YAMAGUCHI M, WATANABE T: Segmental glomerular necrosis as an active index of IgA nephropathy: A study with 100 serial sections of 128 cases. Acta Pathol Jpn 43:723—729, 1993
Macrophages in childhood IgA nephropathy MIcHIo NAGATA, YUKO Axioic, YURI TSUNODA, YASUHIRO KOMATSU, HIR05HI KAWAGUCHI, YUTAKA YAMAGUCHI, and KATUMI ITO Department of Pediatric Nephrology, Tokyo Women's Medical College, Tokyo; Department of Pathology, Institute of Basic Medical Science, University of Tsukuba, Tsukuba; and The Jikei University School of Medicine, Tokyo, Japan
Macrophages in childhood Ifl nephropathy. The role of glomerular macrophages in IgA nephropathy in children was investigated using a new monoclonal antibody (KP1) as a probe. The average number of glomer-
ular macrophages per patient (ANM/P) was closely correlated with the degree of hematuria (P < 0.01) as well as with the degree of leukocyturia (P C 0.01) in the absence of any correlation with proteinuria, serum IgA levels or the interval between the detection of urine abnormalities and renal biopsy. ANM/F was significantly higher in patients diagnosed pathologically as having focal and diffuse proliferative glomerulonephritis than in patients with minor glomerular abnormalities or advanced sclero-
sis (P C
0.05).
Among various types of glomerular morphology in
individual patients, macrophages predominantly infiltrated glomeruli with cell-proliferative lesions despite an absence of any increase in glomeruli
with minor abnormalities or with sclerosis. Macrophages were mainly localized within the capillary lumen in association with endocapillary proliferative lesions (tuft necrosis), they accumulated in areas of mesangial proliferation, and they were attached to Bowman's capsule in segmental lesions. Macrophages were less evident in sclerosis. Furthermore, ultrastructural analysis revealed macrophages in the paramesangial areas in close proximity to lytic changes in the glomerular basement membrane
and effacement of epithelial foot processes. In addition, some cases in repeat biopsy shows prolonged or increased values of ANM/P after several years of interval in association with progression of proliferative lesions. These results suggest that macrophages infiltrate glomeruli during acute glomerular inflammation, and that they are involved in mesangial prolif-
eration or the development of extracapillary lesions in the absence of apparent clinical symptoms. Furthermore, recurrence or prolonged infiltration may promote progression of IgA nephropathy.
In cases of glomerulonephritis, the numbers of infiltrated glomerular macrophages have been shown to increase [1—11]. Recent studies with experimental models of proliferative glomerulonephritis suggest that secretion of cytokines from infiltrated macrophages promotes progression of the disorder [12—14]. Moreover, cytokines also appear to play a role in the progression of glomerulonephritis in humans [15]. These observations suggest that macrophages might participate in the progression of human glomerulonephritis. Indeed, several studies have shown that the number of infiltrated glomerular macrophages increases in cryoglobulinemia-associated nephropathy, in crescentic glomerulonephritis, in lupus nephritis and in acute post-streptococcal glomerulonephritis [5—7, 11]. However, it has not been fully understood
Received for publication April 25, 1994 and in revised form December 7, 1994 Accepted for publication January 3, 1995
© 1995 by the International Society of Nephrology
whether or in what way macrophages participate in slowly progressive glomerulonephritis.
IgA nephropathy is now the most common type of slowly progressive glomerulonephritis and it is the major cause of end-stage renal disease. Previous studies of the role of macrophages in IgA nephropathy was analyzed in adults and discussions
were based on the numbers of macrophages that had infiltrated glomeruli [2—4, 9, 16]. Hooke and colleagues and Ferrario and colleagues failed to find any increases in numbers of glomerular macrophages in IgA nephropathy [2, 3]. By contrast, three recent reports, all from Japan, have demonstrated increased numbers of glomerular maerophages in patients with mesangial hypercellu-
larities [4, 9, 16]. The conflicting data may be the result of differences between the markers of macrophages used or of differences in the backgrounds of patients (phase of the disease, ages of patients or racial differences). Patients in previous studies were very heterogeneous in terms of age and disease state. In most previous studies only one or a few glomeruli per patient was analyzed. Since glomerular lesions and the extent of infiltration of macrophages differ among glomeruli in a single patient with IgA nephropathy (Fig. 1 C, D in this study), a single or a few glomeruli
would seem to be an inadequate source of data on the average number of macrophages infiltrated in a patient. Moreover, because previous studies involved analyses of frozen sections by immunofluorescence or peroxidase-antiperoxidase techniques, localization of macrophages on the background of detailed glomerular structure was not possible.
The present study was designed to investigate the role of macrophages in glomerular morphology in cases of IgA nephropathy in children. Since biopsy materials in this study were obtained
relative soon after onset of the disease, they might possibly provide evidence of glomerular pathology that is related to the onset of the disease. Macrophages were detected with the monoclonal antibody KP1, which is suitable for immunostaining of paraffin-embedded sections [17]. This antibody, applied to formalin-fixed paraffin-embedded sections, allowed us to investigate the localization of maerophages on a background of tuft structure and
to quantify the extent of infiltration in each patient more objectively than has been possible by previously reported methods. Our results indicate that macrophages are associated with endocapillary proliferative lesions, mesangial areas involved in mesangial proliferation and paramesangial destructive lesions with resultant segmental lesions. The results further suggest that asymptomatic recurrence or prolonged infiltration might promote progression of IgA nephropathy in children.
527
528
Nagata et a!: Macrophages in IgA nephropathy
Methods Patients and materials
Forty-one pediatric cases of IgA nephropathy were analyzed. Patients with various degrees of glomerular damage were examined in this study. All cases were defined as IgA nephropathy from clinical and histological findings. Cases of secondary glomerulonephritis with deposition of IgA were not included in this study.
Renal tissues were obtained by needle or open biopsy, fixed in 10% buffered formaldehyde for 24 hours and then embedded in paraffin.
Immunohistochemistiy
Immunohistochemistry was performed by the streptavidin-biotin staining technique as previously described [18]. In brief, after deparaffinization through a graded alcohol series, sections (4 jtm)
were incubated with a 0.1% solution of protease in phosphatebuffered saline (PBS) for four minutes at 37°C in a water bath for enzymatic digestion. Thereafter, endogenous peroxidase activity was blocked by immersing the slides in 0.3% H202 in methanol
and SCL was defined as when more than 80% of the total glomeruli showed evidence of predominant sclerotic lesions rather than proliferation. The average number of macrophages per glomerulus for each patient (ANM/P) was calculated as the total number of macrophages within all glomeruli divided by the total number of glomeruli in a single section. Glomeruli of less than 130 m in diameter were excluded from the analysis. To clarify more precisely the relationship between the infiltration of macrophages and the glomerular morphology, the pathological diagnosis for individual glomeruli was defined as no or minor proliferation (MIN), moderate proliferation (MOD; <50% of entire glomerular area), severe proliferation (SEV; >50% of entire glomerular area), or sclerotic glomerulus (SCL; sclerosis more dominant than proliferation). Average numbers of macrophages for each type of classified glomerular lesion (ANM!GL) were calculated as the total number of immunopositive cells divided by the number of glomeruli in each classified glomerular lesion in a single section from each patient. The localization of KP1-positive cells within the glomeruli was defined as intracapillary or intra-mesangial. To evaluate the distribution of immunopositive cells within glomeruli, we calculated a mesangial localization index (Cap/Mes), which was defined as the ratio of
for 30 minutes. After washing in PBS, the sections were incubated for 15 minutes with normal rabbit serum. Next, the diluted (X200) KP1 monoclonal antibody (CD 68; DAKO, Glostrup, Denmark) immunopositive cells in the capillaries to those in the mesangium. was overlaid on the tissue sections which were then incubated for When immunopositive cells tended to be localized in the mesanone hour at room temperature. The specificity of this antibody has gial area, this ratio was low. Additionally, to determine whether been well described previously [17]. After washing of specimens in infiltration of macrophages occurred in a prolonged or recurrent PBS, biotinylated antibodies against mouse IgG (DAKO) were manner and whether aggravation of glomerular morphology was applied and slides were incubated for 30 minutes at room present in cases with intensive infiltration, 15 cases out of 41 cases temperature. After three washes in PBS, streptavidin-peroxidase were analyzed with repeat biopsy. Changes in pathological diag(DAKO) was applied to the sections for 30 minutes at room nosis and ANM/P in each patient were compared. temperature. After an additional three washes in PBS, color was developed by incubation with 3,3'-diaminobenzidine in PBS with Ultrastructural analysis 1% H202. Finally, sections were counterstained with periodic Small pieces of cortical tissue from each patient were fixed in acid-Schiff reagent. 3% glutaraldehyde and post-fixed in 1% osmium tetroxide. TisAs a negative and positive controls for immunostaining, tonsil- sues were then dehydrated by the standard procedure and emlar tissues from patients after tonsillectomy were fixed by the same bedded in epoxy resin. Ultrathin sections were mounted on grids procedures as applied to renal tissues and treated either with KP1 and stained with uranyl acetate-lead citrate and then they were
antibody, normal mouse serum or PBS alone. Macrophages in tonsillar tissues reacted with the KP1 monoclonal antibody (serving as a positive control) while tonsillar tissues incubated with either normal mouse serum or PBS alone were entirely negative
observed under a transmission electron microscope (H-800; Toshiba, Japan). Approximately 1.4 glomeruli per patient on
for immunostaining. (serving as negative controls; data not
Statistics
shown).
Evaluations
average were examined.
standard errors (SE). A Values are presented as means non-parametric test, with Spearman's test, was performed for
two-factor correlation. Additionally, one-way ANOVA was perPathological diagnosis in individual patients was made by two of formed for comparisons among several groups. A value of P of the authors (M.N. and Y.Y.) without knowledge of any clinical less than 0.05 was considered significant.
and pathological information about individual patients. Immunopositive cells were defined as cells that reacted with KP1 and Results contained nuclei. The disease state was defined as minor glomerA total of 665 glomeruli from 41 patients was analyzed (an ular abnormality (MGA), focal proliferative glomerulonephritis average of 16 1.3 per patient). The minimum number of (FGN), diffuse proliferative glomerulonephritis (DPGN), or advanced sclerosis (SCL), on the basis of the frequency of glomeruli glomeruli per patient was 8. that showed proliferative changes, in accordance with the criteria Clinical parameters, pathological diagnosis and ANMIP by Churg and Sobin [19]. In brief, MGA was defined as being present when less than 20% of the total glomeruli in a single Table 1 shows several clinical parameters and ANM/P for each section showed evidence of proliferative lesions; FGN was defined group of patients with a specific pathological diagnosis. The as when 20 to 80% of the total glomeruli showed evidence of degree of proteinuria tended to be higher in patients with more proliferative lesions; DPGN was defined as when more than 80% severe histologic damage (P < 0.05: MGA vs. FGN, DPGN vs. of the total glomeruli showed evidence of proliferative lesions; SCL; P < 0.01: MGA vs. DPGN, DPON vs. SCL). FGN and
529
Nagata et al: Macrophages in IgA nephropathy Table 1. Clinical parameters for each group of patients and the correlation between various parameters and ANM/P
Age year Male/female Detection of urine abnormalities to biopsy years
Ug/day
URBC (degree)
MGA
FUN
(N = 8) 1.1
(N = 15) 12.7 0.7
5/3 0.4 1.3
11/4 1,4 0.51
0.1
0.6 0.2"
12,1
1.4
0,04b,C 0,3a,b
2.3 5.7
0.3
DPUN
(N = 12.9
13)
0.9 Q,3C
1.5
2.4 8.2
Total
0.2 1.9
Correlation with ANM/P
(N = 41)
P
0.7
13 0.4
NS
NS
3.1
0.6
27/14 1.7 0.2
3.1 1.8
1.1 0.4 1.3
1.1 2.0 5.5
11.8
3/2
8/5 1.8
SCL (N = 5)
0.2 0.2 0.9
NS
NS <0.01 <0.01 NS NS
4.0 1.5 1.9 04b (/HPF) 276± 14 247± 28 243±20 279± 34 312±19 Serum Igamg/dl 4.7 0.6 6.9 O.9 0.4 0.3 1.4 0.3 5.6 l.3' ANM/P (Iglomendus) Data are shown as means SE. Two-group comparisons are: a MGA vs. FGN; b MGA vs. DPGN; C, MGA vs. SCL; ci and 0, FUN vs. SCL; C and E, DPGN vs. SCL. Lower case letters (a, b, c...) indicate P < 0.05 and upper case letters (A, B, C. ..) indicate P < 0.01. Abbreviations are: NS, not significant; NE, not examined.
DPGN were associated with significantly higher levels of hemaCap/Mes and glomerular lesions tuna compared to MGA (P < 0.05). Leukocyturia in FGN was Cap/Mes for individual patients was significantly lower in significantly higher than that in MGA (P < 0.05). ANM/P in cases glomeruli with MOD and SEV as compared to those with MIN (P of FGN and DPGN was significantly higher than in cases of MGA < 0.05, MIN vs. MOD; P < 0.01, MIN vs. SEV), as shown in (P < 0.05) and SCL (P < 0.05: FGN vs. SCL; P < 0.01: DPGN vs. Figure 3.
SCL). A non-parametric test of the correlation between each parameter and ANM/P, the degree of hematuria and the degree
Changes in ANMIP and morphology in cases of repeat biopsy of leukocyturia revealed significant correlations (P < 0.01 in each ANM/P and the pathological diagnosis in 15 cases in which case). However, the age and gender of the patient, the extent of repeat biopsy was performed are shown in Figure 4. In three cases proteinuria, the serum IgA levels and the time from detection of for which data are shown in this Figure, ANM/P was still high even
urine abnormality to renal biopsy were not correlated with two to four years after the first biopsy. In these cases, the ANM/P.
Morphological correlations Localization of macrophages and structural changes in glomeruli. As
shown in Figure 1, KP1-positive cells (macrophages) were clearly visible and the localization of macrophages within glomeruli was easily assessed. The extent of infiltration by macrophages of individ-
ual glomeruli in a single patient was sometime heterogeneous and depended on glomerular morphology, as shown schematically in Figure 1 C, D. Macrophages were associated with endocapilary proliferative lesions (occasionally with polymorphs), tuft necrosis, mesangial proliferation and extracapifiary cellular lesions. In some cases, macrophages represented a significant proportion of the total number of intraglomerular cells. In areas of proliferation of tuft segments, foamy changes in macrophages were occasionally seen. In addition, macrophages were found at the site of ruptured glomerular
histological diagnosis had progressed from FGN to DPGN. In addition, three patients with low values of ANM/P at the first biopsy showed increases in ANM/P at the second biopsy, after one to six years, with morphological progression. Ultrastructure
Special attention was paid to the structural relationship between macrophages and glomerular tuft segments in a study by transmission electron microscopy. In general, electron-dense deposits in GBM were not dominantly associated with macrophages.
Macrophages localized in the capillary lumen were frequently attached to endothelial cells. However, under these conditions, no structural changes in endothelial cells, GBM or podocytes were noted. By contrast, macrophages that had invaded the mesangial area were associated with considerable and destructive structural
changes. Furthermore, in podocytes, the effacement of foot
basement membrane (GBM). Such lesions were frequently con- processes and dilated cytoplasmic vacuoles were frequently apnected to cellular crescents. However, macrophages were not the parent (Fig. 5). Marked attenuation, undulation and thinning of main constituents of cellular crescents. Furthermore, glomeruli with the GBM accompanied extensive mesangial infiltration by macan accumulation of matrix or sclerotic changes tended to contain rophages (Fig. 5), and lytic changes in the GBM and the mesanfewer macrophages than glomeruli with endocapillary proliferation gial area were occasionally seen (Fig. 6). or mesangial proliferation. In areas of fibrocellular crescents, an Discussion accumulation of macrophages was evident of periglomerular sites. This study represents an effort to resolve the confusion that has ANMIGL and glomerular lesions in each patient arisen from previously published conflicting data about the role of As shown in Figure 2, glomeruli with MOD and SEV contained macrophages in IgA nephropathy [2—4, 9, 16]. Our study has two significantly more macrophages than those with MIN or SCL in each distinct advantages as compared to earlier studies. First, our study patient (P < 0.01 for all comparisons). Glomeruli with SEV con- is limited to pediatric patients. Since the onset of the disease can be estimated from the yearly screening of children's urine, and tained more macrophages than glomeruli with MOD (P < 0.05).
530
A t -
I
-
'',.
'4
1'
cl
I I
4
1
Sn
-
•1
_w
-A
t7i1.
,ap) tctt
Nagata et a!: Macrophages in IgA nephropathy
531
Nagata et al: Macrophages in IgA nephropathy
I
T
20 15
___
a 0
I')
ANM/GL
I
a10
5
1
*
3
4
5
6
7
Time, years
Fig. 2. ANMIGL and classified glomerular lesions. Abbreviations are: MIN, minor proliferation; MOD, moderate proliferation; SEV, severe proliferation; SCL, sclerosis; see text for detailed definitions. Values are < 0.01. mean SE. < 0.05,
I
2
I
2
Fig. 4. ANMIP and pathological diagnosis in eighteen cases of repeat biopsy. and SCL (•). FGN (0), DPGN
()
Second,
the precise localization of macrophages on the back-
ground of detailed glomerular structures was analyzed. The results from this study should contribute considerably to a better further understanding of the role of macrophages in IgA nephropathy. The average number of macrophages per patient (ANM/P) was significantly higher in patients with proliferative lesions (endocapillaiy and mesangial proliferation). Previous authors have classified adult patients by degree of mesangial proliferation and found that the number of macrophages per patient increased with the severity of mesangial proliferation [4, 9, 161. From such results, earlier authors concluded that macrophages are involved in mesangial proliferation. However, their results simply mean that more macrophages were evident in patients with mesangial proliferation and do not directly support their conclusion. To obtain more valid evidence, we classified glomeruli by degree of proliferation in individual patients and found that glomeruli with proliferative lesion contained more macrophages
text; MIN, minor proliferation; MOD, moderate proliferation; SEV, severe proliferation; SCL, sclerosis; see text for detailed definitions.
than those with minor changes in these individual patients. These results further suggest either that macrophages stimulate the formation of or actually form proliferative lesions. Since macrophages secrete several kinds of cytokine in this disease [20] and cytokines have been proven to stimulate proliferation of mesangial cells in vitro [21, 221, and furthermore, since mesangial macrophages seem to be better able to proliferate glomerular cells than macrophages in the
Values are mean
capillary lumen, the precise localization of macrophages within
0 MIN MOD SEV Fig. 3. Mesangial localization indices. Abbreviations are: Cap/Mes, see SE.
< oo5 J3 < 0.01.
glomeruli seems to be crucial if we are eventually to characterize the role of macrophages in glomerular morphology in glomerulonephrisince biopsies were performed approximately 1.5 years after tis. The present study indicates that infiltration of glomerular macasymptomatic onset, the glomerular lesions in the majority of our rophages is closely associated with endocapillary proliferative lesions:
patients may represent pathology related to the onset of disease. tuft necrosis, and with mesangial proliferation. Moreover, the Fig. 1. Light micrographs of glomendi from patients with various types of glomerular morphology. Brown coloration indicates immunoreaction. (A) From a patient diagnosed with MGA (ANM/P = 1.2). No infiltration of macrophages (X400). (B) From a patient diagnosed with FGN (ANM/P 7.4). The localization of macrophages is clearly visible. There are two macrophages in the mesangial area (arrowhead) and three in the capillary (arrow) (X800). (C and D) From a single patient with DPGN (ANMIP = 6.3). The glomerulus with SEV (D) contains many more macrophages and polymorphonuclear cells than that with MIN (C). Proliferative lesions in (D) are mainly the result of endocapillary proliferation (X400). (E and F) From a single patient with FGN (ANM/P = 17.7). Macrophages have accumulated in association with a lytic and expanded mesangial area (E) and a small crescent (F). Note that macrophages are not components of the cellular crescent (X800). (G) From a patient diagnosed with FGN (ANM/P = 6.3). Segmental aggregation of macrophages with foamy changes (arrowhead) is seen in the segmental lesion with adhesions (X 400). (H) A glomerulus classified as SG from a patient diagnosed with SCL. (ANM = 0.4). No macrophages are evident in areas of sclerosis. (x400).
532
Nagata et al: Macrophages in IgA nephropathy
Fig. 5. Electron micrograph of a glomen4ar tuft from a patient with FGN (ANMIP = 17.7). Paramesangial infiltration of macrophages (M) and polymorphonuclear cells (PM) are visible. Note the various changes in the GBM and podocytes (PD). PE indicates parietal epithelial cells. See text for details. (x4,000).
mesangial localization index was lower in glomeruli with prolifera-
tion than in those with minor lesions or sclerosis. From these observations, we speculate that macrophages actively infiltrate gbmeruli as a mediator of the acute phases of glomerular inflammation (endocapillary proliferation) and that approximately half (Cap/Mes is approximately 1 in glomeruli with proliferative lesions) of the
to be essential for the process of progression of IgA nephropathy. Indeed, values of ANM/P were closely correlated with the degree
of hematuria, as well as the degree of leukocyturia, at renal biopsy. The results supports the previous notion that the signifi-
cant infiltration of polymorphonuclear cells were present in
association with gross hematuria or formation of crescents in IgA nephropathy [27, 28]. In our results, macrophages are frequently associated with endocapillary proliferative lesions and tuft necroproliferation of mesangial cells. And such mesangial proliferation is sis and, furthermore, analysis of serial section in our laboratory a major pathologic feature of IgA nephropathy [23, 24]. Mesangial proliferation followed by mesangial sclerosis is con- revealed that such lesions are very frequently apparent at the sites sidered to be a major morphological indicator of the course of IgA of formation of small cellular crescents in this disease [29]. nephropathy in children [23]. However, since mesangial sclerotic Moreover, our observations by electron microscopy confirm the lesions do not predominantly contain macrophages, as shown in light microscopic findings, that is, the close proximity of mesangial this study, the actual process of macrophage-promoted progres- macrophages to lytic mesangial changes and damaged gbomerular sion has not been well characterized. Basically, the pathology of basement membrane, whereas intracapillary macrophages are IgA nephropathy consists of intraglomerular lesions (mesangial associate a lesser extent with such changes. A recent report by proliferation and IgA deposits), while the progression of the Shigematsu, Kobayashi and Hiki also suggests that para-mesandisease largely depends on extracapillary lesions, namely, cres- gial destructive lesions are crucial for the extension of tuft lesions cents or adhesions [25, 261. Therefore, rupture of the GBM seems to extracapillary lesions [30]. Thus, it seems likely that inftltration
macrophages invade the mesangial area to form or to stimulate
Nagata et a!: Macrophages in IgA nephropathy
533
Fig. 6. Electron micrographs of a glomerular tuft from a patient with FGN (ANMIP = 7.4). A) Low-magnification view of tufts. Macrophages (M) are localized in the paramesangial area and in the capillary. B) High-magnification view of the paramesangial macrophage displayed in A). Around an expanded macrophage in the paramesangial area, discontinuous fragments of degenerated GBM (arrowheads) and the effacement of foot processes of podocytes are seen (small arrow). Fenestra of endothelial cell is absent. (X9,600).
of macrophages within the mesangial area can potentially destroy glomerular tufts with resultant hematuria, leukocyturia and formation of segmental lesions, namely, crescents or adhesions. On the other hand, KP1-positive cells were not the main constituent of
children's urine, we were able to estimate the timing of the onset of this disease in each of our patients. Since ANM/P was not correlated with the duration of urine abnormalities and, further, since cases with repeat biopsy revealed prolonged infiltration or increased values of
cellular crescents. Yoshioka and co-workers showed previously that
ANM/P after intervals of several years, we can conclude that
crescentic cells in IgA nephropathy were cytokeratin-positive but negative for several established markers of macrophages [31]. Our
infiltration of macrophages into glomeruli in IgA nephropathy occurs
recent study is consistent with their findings [32], suggesting a parietal epithelial origin. As Magil proposed, macrophages might participate
frequently observed endocapillary proliferative lesions in combination with sclerosis within a single glomerulus. Moreover, in three patients with prolonged high values of ANM/P and three cases of increased values of ANIVI/P at repeat biopsy, the progression of glomerular changes was clearly evident. Hence, we propose that acute inflammatory changes in the glomeruli, represented by the
in formation of crescents via stimulation of parietal epithelial proliferation in an acute phase of inflammation [33]. Further studies to identiti the stimulatory factors for such cell proliferation are warranted. From the clinical observation, we note that most of our cases were
at no fixed time during the course of the disease. In fact, we
infiltration of macrophages, may take place recurrently or in a
identified from urine abnormalities by a program for screening of
prolonged manner and may have a damaging effect on glomerular
urine from asymptomatic school children, and who did not have any clinical signs of acute exacerbation of nephritis. Therefore, it appears that hematuria and leukocyturia are good indicators of macrophage
morphology, without any clinical symptoms, in cases of IgA nephrop-
infiltration in IgA nephropathy in children in the absence of any clinical symptoms.
To obtain further insight into the contribution of macrophages
to the aggravation of abnormal glomerular morphology, it is necessary to determine whether infiltration of macrophages is a transient or a prolonged phenomenon and whether deleterious
athy in children. However, another four cases revealed the progression of glomerular morphology despite low values of ANMIP at first and second biopsy. Since infiltration of macrophages may not be persistent in every case of this disease, the timing of renal biopsy in these cases might not have correspond to a period of active infiltration. Alternatively, factors other than macrophages may be involved in the progression in some cases.
Furthermore, it still remains unknown whether macrophages morphological changes occur in cases with intense infiltration of participate in the chronic progression, namely, in the development macrophages. With the aid of the school program for analysis of of sclerosis. In this study, we found neither an increase in ANM/P
534
Nagata et a!: Macrophages in IgA nephropathy
in cases with advanced sclerosis nor an increase in ANM/GL in sclerotic glomeruli. Moreover, macrophages were not dominantly accumulated in the areas of sclerosis by morphological analysis. Our observations indicate that macrophages are less evident in areas of sclerosis, but they do not necessarily imply that macrophages do not contribute to sclerosis. Indeed, we found foamy changes in macrophages in areas of sclerosis and the aggregation
of macrophages in the periglomerular interstitium at sites of fibrocellular crescents The role of such macrophages in progression of sclerosis is unknown. One possible explanation for it is that macrophages stimulate phenotypic changes in mesangial cells. As
recently be shown, macrophage derived factors suggested to stimulate mesangial proliferation in IgA nephropathy [20]. In the experimental model of mesangial proliferative glomerulonephritis induced by anti-Thy 1 antibody, mesangial proliferation is associated with the expression of a-smooth muscle cell actin (pheno-
typic changes) and an increase in matrix production that leads later to sclerosis [34, 35]. Such phenotypic changes have also been
demonstrated in human glomerulonephritis [36, 37]. Hence, we can speculate that macrophages might participate the develop-
ment of sclerosis by the activation of phenotypic changes in mesangial cells and then the production of matrix by mesangial
cells might be accelerated after the disappearance of macrophages. Further study is clearly needed to determine the role of glomerular macrophages in phenotypic changes in mesangial cells in vivo and in sclerosis in this disease.
In summary, in IgA nephropathy in children, macrophages infiltrate glomeruli during an active phase of glomerular inflammation (endocapillary proliferation) and they are involved in mesangial hypercellularity (either due to accumulation or by stimulating mesangial cell proliferation). We provide evidence
that macrophages also participate in the development of paramesangial destructive lesions, namely, in lysis or attenuation of GBM, as well as in the destruction of the integrity of podocytes. All these changes have the potential to lead to the development of extracapillary lesions, namely, the crescents
and adhesions that are the most reliable predictors of the
chronic progression of this disease. Furthermore, asymptomatic recurrent or prolonged infiltration of macrophages might take place and promote progression. Acknowledgments The authors thank Mr. Shigeru Horita, Mr. Hideki Nakayama and Miss Mayuko Kawashima with their skillful technical assistance. The
authors are grateful to Dr. Hiroshi Shiraga and Professor Teruo Watanabe (Tsukuba University) for critical discussions related to the manuscript. The contributions of anonymous reviewers are also gratefully acknowledged.
Reprint requests to Michio Nagata, M.D., Department of Pathology, Institute of Basic Science, University of Tsukuba, Tsukuba 305, Japan.
References 1. ATKINS RC, GIsGow EF, HOLDSWORTH SR, THOMPSON NM, HAN-
COCK WW: Tissue culture of isolated glomeruli from patients with
4. KOBAYASHI M, KOYAMA A, NARITA M, SHIGEMATSU H: Intraglomeru-
lar monocytes in human glomerulonephritis. Nephron 59:580—585, 1993 5. LAOHAPAND T, CArFELL V, G.uRiAJ. FRT: Monocyte infiltration in human glomerulonephritis: Alpha-1-antitiypsin as a marker for mononuclear phagocytes in renal biopsies. Clin Nephrol 19:309—316, 1983 6. MAGIL AB, WADSWORTH LD, LOEWEN M: Monocytes and human renal glomerular disease. A quantitative evaluation. Lab Invest 44:27— 31, 1981 7. MONGA G, MAZZUCCO G, CASTELLO G: Glomerular monocyte infiltration in human glomerulonephritis: Prevalence and correlation with clinical and morphological variables. Virchow Arch 405:483—496, 1985 8. SCHREINER GF, COTRAN RS, PIt.i)o V, UNANUE ER: A mononuclear
cell component in experimental immunologic glomerulonephritis. J Exp Med 147:369—384, 1978 9. TOMINO Y, OZAKI T, K0IDE H, YAGAME M, EGUCHI K, NoMoTo Y,
SARAI H: Glomerular T cell and monocyte populations in patients with IgA nephropathy. Jap J Nephrol 3 1:221—226, 1989 10. Li HL, HANCOCK WW, Hoia DH, DOWLING JP, ATKINS RC: Mono-
nuclear cell activation and decreased renal function in IgA nephropathy with crescents. Kidney mt 37:1552—1556, 1990 11. SHIGEMATSU H, SHISHIDO H, KUMARA K, TSUCI-IIDA H, SUZUKI H,
HIROSE K, ToJo S: Participation of monocytes in transient glomerular
hypercellularity in poststreptococcal glomerulonephritis. Virchow Arch 12:367—4370, 1973
12. BORDER WA, OKUDA S, LAGUINO LR, SPORN MB, RousIHTI E: Suppression of experimental glomerulonephritis by antiserum against transforming growth factor /31. Nature 346:371—374, 1990 13. TIPPING PG, LOWE MG, HOLDSWORTH SR: Glomerular interleukin-1
production is dependent on macrophage infiltration in anti-GBM glomerulonephritis. Kidney mt 39:103—112, 1991 14. HoRn Y, MURAGUCHI A, IwANo M, MATSUDA T, HIRAYAMA H, FuJii
Y, DOHI Y, ISHIKAWA H, OHMOTO Y, YosHIz.s.iu K, HIRANO T, KISHIMOTO T: Involvement of IL-6 in mesangial proliferative glomerulonephritis. J Immunol 143:3949—3955, 1989 15. Doin K, IwA.NO M, MURAGUCHI A, HoRn Y, HIRAYAMA T, OGAwA 5,
SHnIU H, Hnw'io T, KIHSIMOTO T, ISHIKAWA H: The prognostic significance of urinary interleukin 6 in IgA nephropathy. Clin Nephrol 35:1—5, 1991 16. ARIMA S, NAKAYAMA S, NAITO M, SATO M, TAKAHASFI! K: Signifi-
cance of mononuclear phagocytes in IgA nephropathy. Kidney mt 39:684—692, 1991 17. PULFORD KAF, RIGNEY EM, MICKLEM KJ, JONES M, STROSS W,
GATFER KC, MASON DY: KP1: A new monoclonal antibody that detects a monocyte/macrophage associated antigen in routinely processed tissue sections. J Clin Pat/so! 42:414—421, 1989 18. NAGATA M, YAMAGUCHI Y, ITO K: Loss of mitotic activity and expression of vimentin on glomerular epithelial cells in developing human kidneys. Anat Embryo! 187: 275—279, 1993
19. CI-IURG J, SOBIN LH: Renal disease, in Classification and Atlas of Glomerular Diseases (2nd ed), edited by CHURG J, Tokyo, IgakuShoin, 1982, pp 3—6
20. Yosioi K, TAKEMURA T, MURAI
in IgA nephritis. Kidney Int 44:825—833, 1993 21. MATSUMOTO K: Jnterleukin-1 production by monocytes from patients with glomerulonephritis after in vitro incubation with soluble immune complexes. C/in Nephrol 36:267—273, 1991 22. LOVET DH, RYAN JL, STERZEL RB: Stimulation of rat mesangial cell proliferation by macrophage interleukin 1. J Immunol 13 1:2830, 1983
23. YosHIIwA N, ImMA K, MAKiwt N, YOSHIARA 5, YOSHIYA K, MATSUO T, OKADA S: Mesangial changes in IgA nephropathy in children. Kidney mt 32:585—589, 1987 24. D'AMIco G, IMBASCIATI DI, BELGIOI0S0 GB, BERTOLI 5: Idiopathic
IgA mesangial nephropathy: Clinical and histological study of 374 patients. Medicine 64:49—60, 1985 25. YOSHIKAWA N, ITO H, NAKAMURA H: Prognostic indicators in child-
glomerulonephritis. Kidney mt 17:515—527, 1980 2. HooK DH, GEE DC, ATKINS RC: Leukocytes analysis using monoclonal antibodies in human glomerulonephritis. Kidney mt 31:964—972, 1987 3. FERRARIO F, CASTIGLIONE A, C0L&sANll G, BELGI0I0s0 GR, BER-
hood IgA nephropathy. Nephron 60:60—67, 1992 26. KATAFUCHI R, OH Y, HORI K, KOMOTA T, YANASE T, IKEDA K,
TOLl S, D'AMIGo G: The detection of monocytes in human glomerulonephritis. Kidney mt 28:513—519, 1985
Clin Nephrol 41:191—198, 1994 27. KINCAID-SMITH P, NICHOLLS K, BIRCHALL I: Polymorphs infiltrate
OMURA T, FUJIMI 5: An important role of glomerular segmental lesions on progression of IgA nephropathy: A multivariate analysis.
Nagata et al: Macrophages in IgA nephropathy glomeruli in mesangial IgA glomerulonephritis. Kidney mt 36:1108— 1111, 1989 28. BENNETF WM, KINCAID-SMITH Pr Macroscopic hematuria in mesang-
ial IgA nephropathy, Correlation with glomerular crescents and renal dysfunction. Kidney mt 23:393—400, 1983
29. AXI0KA Yr Structure of segmental tuft lesions in childhood IgA nephropahty. Three dimensional analysis indicates development of segmental lesions. Jap J Nephrol 36:973—981, 1994 (in Japanese) 30. SHIGEMATSU H, KOBAYASHI Y, HIKI Yr Para-mesangial destructive lesions in immunoglobulin A nephritis. Ultrastruct Pathol 14:129—134, 1990
31. Yosmoi K, TAJcEMURA T, AKANO N, MIYvoTo H, IsElu T, Mi Sr Cellular and non-cellular compositions of crescents in human glomerulonephritis. Kidney mt 32:284—291, 1987 32. NAGATA M, HORITA 5, ITo Kr Evidence for phenotypic changes of
glomerular epithelial cells in renal disease. (abstract) J Am Soc Nephrol 4:282, 1993 33. MAGIL AB: Histogenesis of glomerular crescents. Immunohistochem-
535
ical demonstration of cytokeratin in crescent cells. Am J Pathol 120:222—229, 1985 34. FLOEGE J, JOHNSON R I, GORDON K, IIDA H, PRITZL P, YosHIMuitA A,
CAMPBELL C, ALPERS CE, CousER WGr Increased synthesis of extracellular matrix in mesangial proliferative nephritis. Kidney mt 40:477—488, 1991 35. JOHNSON RI, IIDA H, ALPERS CE, MAJESICY MW, SCHWARTZ SM, PRITZL P, GORDON K, GowN AM: Expression of smooth muscle cell
phenotype by rat mesangial cells in immune complex nephritis. J Clin Invest 87:847—858, 1991
36. ALPERS CE, HUDKINS KL, GowN AM, JOHNSON RJ: Enhanced expression of muscle-specific actin in glomerulonephritis. Kidney mt 41:1134—1142, 1992
37. Siouio Y, SHIM0ICAMA T, SxMI T, HARADA A, YAMAGUCHI M, WATANABE T: Segmental glomerular necrosis as an active index of IgA nephropathy: A study with 100 serial sections of 128 cases. Acta Pathol Jpn 43:723—729, 1993