co m m e nta r y

3. Rosen S, Peters CA, Chevalier RL et al. The kidney early event following ischemic injury. Am J Physiol dense intramembranous and mesangial
in congenital ureteropelvic junction obstruction: Renal Physiol 2010; 299: F479–F486.
a spectrum from normal to nephrectomy. J Urol 8. Heyman SN, Fuchs, Jaffe R et al. Renal
deposits were the hallmark of type II, also
2008; 179: 1257–1263. microcirculation and tissue damage during acute known as dense deposit disease (DDD).2
4. Girshovich A, Vinsonneau C, Perez J et al. Ureteral ureteral obstruction in the rat: the effect of saline In type III, deposits could be subendo-
obstruction promotes proliferation and differentia- infusion, indomethacin and radiocontrast.
tion of the renal urothelium into a bladder-like Kidney Int 1997; 51: 653–663.
thelial and subepithelial (Burkholder
phenotype. Kidney Int 2012; 82: 428–435. 9. Conte C, Riant E, Toutain C et al. FGF2 translatio- subtype) or produce complex intramem-
5. Wu XR, Lin JH, Walz T et al. Mammalian uroplakins. nally induced by hypoxia is involved in negative branous, subendothelial, and subepithelial
J Biol Chem 1994; 269: 13716–13724. and positive feedback loops with HIF-1 alpha.
6. Romih R, Korosec P, de Mello Jr W et al. PLoS One [online 2008; 3: e3078. formations with fraying of the lamina
Differentiation of epithelial cells in the urinary 10. Giaccia AJ, Simon MC, Johnson R. The biology densa (Strife and Anders subtype).3 Even
tract. Cell Tissue Res 2005; 320: 259–268. of hypoxia: the role of oxygen sensing in within the same biopsy or a given glomer-
7. Vinsonneau C, Girshovich A, Ben M’rad M et al. development, normal function and disease.
Intrarenal urothelium proliferation: an unexpected Genes Dev 2004; 18: 2183–2194. ulus, pathologists may observe overlap-
ping features between these subtypes,
defying easy classification. While hypo-
complementemia was a well-recognized
see original articles on pages 454 and 465 clinical feature of all these entities, with
the exception of C3 nephritic factor,
C3 glomerulopathy: what’s in causation was largely unknown.
Over the next two decades, increasing
a name? attention was paid to the composition of
the deposits detected by immunofluores-
Vivette D. D’Agati1 and Andrew S. Bomback2 cence. Type I was typically characterized
by deposits of immunoglobulin as well as
Whereas early classifications of membranoproliferative glomerulo- complement components; type II was
nephritis (MPGN) were based on morphologic features, the modern distinguished by deposits of C3 only;
approach is directed at immunofluorescence findings. Glomerular the type III Burkholder variant typically
deposits of C3 alone, without immunoglobulin, are the hallmark of had deposits of immunoglobulin and
complement; whereas the type III Strife
alternative complement pathway dysregulation through inherited or and Anders variant could manifest C3
acquired defects. These immunoglobulin-negative forms are referred either alone or in combination with
to as C3 glomerulopathy, which encompasses both dense deposit immunoglobulins. As long as the patho-
disease and C3 glomerulonephritis. Distinguishing C3 glomerulopathy genesis of these forms remained obscure,
from immunoglobulin-mediated MPGN is opening the way to better the classification was cumbersome to use
and perplexing to clinicians and patholo-
diagnostic, prognostic, and treatment algorithms.
gists alike.
Kidney International (2012) 82, 379–381. doi:10.1038/ki.2012.80 In the past decade, our understanding
of the role of complement in the patho-
genesis of MPGN has illuminated the field
Among glomerular diseases, none has is defined as mesangial interposition and and led to a paradigm shift in classifica-
undergone greater conceptual metamor- duplication of glomerular basement mem- tion (Figure 1). First, the observation that
phosis over the past five years than mem- branes, typically associated with periph- a number of patients with intramembra-
branoproliferative glomerulonephritis eral capillary wall immune deposits. Over nous dense deposits lack an MPGN pat-
(MPGN). In the 1970s, before there was the next two decades, many secondary tern altogether caused the designation
knowledge of pathogenesis, MPGN was forms with clear etiologic associations MPGN II to be discarded in favor of DDD.
categorized on the basis of histologic and (such as related to cryoglobulinemia, dys- Second, pathologists began emphasizing
ultrastructural findings.1–3 The membran- proteinemia, autoimmune disease, and in their reports the presence of isolated
oproliferative pattern of glomerular injury infections) were differentiated from the deposits of C3 in examples of MPGN type
primary idiopathic form. But the patho- I and type III. A major breakthrough was
1Department of Pathology, College of Physicians logic classification of primary MPGN the discovery of genetic mutations or
and Surgeons, Columbia University, New York, New remained problematic because it was deficiencies in complement-regulatory
York, USA and 2Department of Medicine, College based largely on the ultrastructural proteins in patients with isolated C3
of Physicians and Surgeons, Columbia University, appearance and location of deposits, deposits.4,5 Thereafter, ‘C3 glomerulopa-
New York, New York, USA devoid of pathogenetic context. thy’ assumed center stage as a newly rec-
Correspondence: Vivette D. D’Agati, College
Subendothelial and mesangial deposits ognized subgroup encompassing DDD
of Physicians and Surgeons, Department of
Pathology, Columbia University, 630 W. 168th predominated in type I MPGN, where the and those examples of type I and type III
Street, New York, New York 10032, USA. membranoproliferative pattern was typi- (now termed ‘C3 glomerulonephritis’;
E-mail: vdd1@columbia.edu cally well developed.1 Highly electron- C3GN) in which immunofluorescence

Kidney International (2012) 82 379

com m enta r y

HISTORICAL
stration by Sethi et al.8 that the proteomic
MPGN type II, or dense
CLASSIFICATION MPGN type I
deposit disease (DDD)
MPGN type III profile of C3GN by mass spectrometry
is similar to that of DDD (with predom-
inance of C3 and distal complement path-
Mesangial prolifer-
ation with mesangial Diverse glomerular MPGN pattern, way components) further supports the
LIGHT interposition and histology with or usually with membranous emerging view that the presence of C3
MICROSCOPY GBM duplications without MPGN pattern features alone is the major hallmark of alternative
(MPGN pattern)
complement pathway dysregulation in C3
Mesangial and Mesangial and Mesangial, subendothelial, glomerulopathy, independent of histo-
ELECTRON
MICROSCOPY
subendothelial intramembranous highly subepithelial, and/or logic and ultrastructural features.
deposits electron-dense deposits intramembranous deposits However, the pitfall of lumping diseases
into a single descriptor or category is an
implied assumption of homogeneity.
As compared with atypical hemolytic ure-
C3 with IgG C3 with IgG
IMMUNO- and/or
C3 alone C3 alone C3 alone
and/or mic syndrome, which is also strongly
FLUORESCENCE IgM, C1 IgM, C1 associated with dysregulated alternative
complement pathway activity, C3 glomer-
C3GN DDD C3GN ulopathy manifests a much broader
MODERN
CATEGORIES spectrum of clinical presentation, histopa-
thology, genetic abnormalities, comple-
MPGN type I C3 glomerulopathy MPGN type Ill
ment activity, and prognosis. Servais et al.,
Figure 1 | The evolving classification of membranoproliferative glomerulonephritis (MPGN). looking only at the adult patients in their
Until recently, the classification of primary MPGN into types I, II, and III was based primarily on cohort, found worse 10–year renal sur-
histologic features (light microscopy) and the ultrastructural location and electron density of the
deposits (electron microscopy). With our increased understanding of the role of complement in vival in patients with DDD than in
the pathogenesis of these conditions, the immunofluorescence findings now play a crucial role in patients with MPGN type I and C3GN;7
categorizing MPGN as immunoglobulin-mediated vs. non-immunoglobulin-mediated disease; the Sethi et al., in their smaller cohort, also
latter grouping, which is distinguished by isolated C3 staining on immunofluorescence, has been report worse outcomes in DDD than in
termed ‘C3 glomerulopathy.’ C3 glomerulopathy encompasses C3 glomerulonephritis (C3GN) and
dense deposit disease (DDD). GBM, glomerular basement membrane; IgG, immunoglobulin G; C3GN and suggest that C3GN may be a
IgM, immunoglobulin M. less aggressive entity.8 One possible expla-
nation is the much higher rate of C3
nephritic factor (C3NeF) positivity in
reveals isolated deposits of C3, underscor- stabilize the C3 convertase of the alterna- DDD (86.4% in the French cohort, in
ing the pathogenetic importance of tive pathway (e.g., C3 nephritic factors) or accordance with the approximately 80%
dysregulation of the alternative comple- target the inhibitory complement factors rate described in other cohorts)9 than in
ment pathway. Thus, the modern approach (e.g., factor H autoantibodies). These C3GN (45.3% in the French cohort and
to classification distinguishes those forms abnormalities promote excessive activa- 50% in the American cohort).7,8 Yet even
of MPGN with deposits of C3 only tion of the alternative complement path- this discrepancy is open to question, as
(known as C3 glomerulopathy and includ- way in the fluid phase, with deposition of Servais et al. report a fluctuation of the
ing both DDD and C3GN) from MPGN complement debris, including breakdown C3NeF activity in one-third of their
with deposits of immunoglobulin and products of C3b and components of the patients during follow-up and a normal
complement.6 terminal complement cascade, in the range of serum C3 levels in approximately
The reports from Servais et al.7 and glomerular capillary wall. A reclassifica- 40% of C3NeF-positive patients, suggest-
Sethi et al.8 in this issue of Kidney Interna- tion of MPGN into immunoglobulin- ing that the C3NeF-stabilized C3 conver-
tional highlight the strengths and weak- mediated disease (prompting a work-up tase may be subject to regulation by other
nesses of the term ‘C3 glomerulopathy.’ As for infectious, autoimmune, or dyspro- factors.7 The authors hypothesize that
validation of the classification, defective teinemia-associated etiologies) vs. com- complement-regulatory membrane cofac-
control of the alternative complement plement-mediated disease (prompting a tor protein (MCP) variants may explain
pathway was detectable in 71% of patients work-up of the alternative pathway) this heterogeneity of disease in C3 glomer-
in the French cohort7 (which included should lead to better diagnostic and treat- ulopathy.7 However, they are not able to
cases of C3GN, DDD, and MPGN type I) ment algorithms and, ideally, improved show a functional role of their proposed
and 100 % of those in the American outcomes in these diseases. For example, at-risk MCP haplotype, and the P-values
cohort8 (which included cases of C3GN 9 of the 12 patients with C3GN in the reported for the selected single-nucleotide
only). This defect can be due to mutations American cohort were initially misclassi- polymorphisms in the MCP gene are
in complement proteins (such as C3, fac- fied as having MPGN type I ( n = 4), not robust enough to denote significance,
tor B, factor H, and factor I) or due MPGN type III (n = 1), and postinfectious particularly without replication in an
to acquired autoantibodies that either glomerulonephritis (n = 4). The demon- independent cohort.7

380 Kidney International (2012) 82


One of the most intriguing findings to
emerge from the French cohort is the
identification of alternative complement
pathway dysregulation in more than half
of the 48 patients with MPGN type I.7 In
fact, a C3NeF was identified as frequently
in MPGN type I as in C3GN. The termi-
nology ‘C3 glomerulopathy’ is intended
for immunoglobulin-negative diseases,
to highlight the underlying alternative
pathway dysregulation as opposed to acti-
vation of the classical pathway by anti-
gen–antibody immune complexes. How,
then, should we classify the MPGN type I
patients described here with genetic
abnormalities in factor H (n = 5) or factor
I (n = 3) or with positive C3NeF activity
(n = 18)? One possible explanation is that
these patients were previously misclassi-
fied as MPGN type I, akin to some patients
in the American C3GN cohort, and that
the immunoglobulin staining was trace
and inconsequential; the lack of central
review of all biopsies in this study, as the
authors themselves acknowledge, is there-
fore a major limitation.7 Small amounts
of immunoglobulin may become trapped
in areas of sclerosis or accumulate in
podocyte protein resorption droplets,
confounding immunofluorescence inter-
pretation. However, if we are to accept that
these cases truly have immunoglobulin
deposits alongside serologic evidence of
alternative pathway dysregulation, should
we now consider these as yet another sub-
type of C3 glomerulopathy? These find-
ings suggest more commonality between
MPGN type I and C3 glomerulopathies
than previously recognized and raise the
question of whether patients with dys-
regulated alternative complement path-
way may not also develop disease-causing
immune deposits containing immu-
noglobulin and complement in the course
of antigen-driven immune responses.
Such observations are consistent with
the predisposition to immune complex-
mediated lupus nephritis in MRL-lpr mice
with genetic deletion of complement fac-
tor H10 and the increased susceptibility to
IgA nephropathy in people with muta-
Kidney International (2012) 82
tions in the complement factor H locus on
human chromosome 1q32.11 C3NeF has
also been identified in some patients with
postinfectious glomerulonephritis. 12
Because patients with complement-
regulatory disorders are not protected
from development of other immune-
mediated glomerulonephritides, future
studies will be needed to explore the
potential contribution of alternative path-
way complement dysregulation beyond
C3 glomerulopathy.
Screening tests of alternative comple-
ment pathway activity may help clear up
some of this confusion and provide much-
needed physiologic information supporting
the diagnosis of C3 glomerulopathy. Sethi
et al. used three different screening tests of
the alternative pathway in their cohort of
C3GN: a hemolytic assay, an alternative
pathway functional assay, and levels of
serum membrane attack complex.8 In such
a small cohort, in which most patients
showed remarkable stability of disease
despite varied treatment regimens, it is
impossible to discern whether these mark-
ers of alternative pathway activity are linked
to outcomes. Such testing would likely have
been far more informative in the larger
French cohort but was not performed.7
Instead, levels of C3 and C4 were used as
baseline markers of complement activity,
which by themselves are substandard in
these complex diseases; not surprisingly,
neither low C3 ( < 600 mg/l) nor very low
C3 ( < 200 mg/l) levels were predictive of
renal outcomes.7 Wider access to the types
of genetic and complement testing
described in both studies should allow
more etiology-specific diagnosis.
Another provocative finding is the iden-
tification of both C3NeF and a genetic
defect in the complement pathway in a
proportion of patients in the French
cohort: among 24 patients with an identi-
fied mutation in a complement gene, 13
also had detectable C3NeF.7 Why is there
coexistence of inherited and acquired
abnormalities of the alternative comple-
ment pathway in such cases? These num-
bers are difficult to ascribe to coincidence
co m m e nta r y
and raise the question of whether increased
activity of the alternative complement
pathway on a genetic basis might promote
autoimmune phenomena by unbridled
activation (and perhaps conformational
change) of the C3 convertase. These find-
ings suggest a two-hit disease model in
some people. Larger cohorts are needed to
evaluate the prevalence of this phenome-
non and the implications for evidence-
based treatment.
DISCLOSURE
The authors declared no competing interests.
REFERENCES
1. Levy M, Gubler MC, Sich M et al. Immunopathology
of membranoproliferative glomerulonephritis
with subendothelial deposits (Type 1 MPGN).
Clin Immunol Immunopathol 1978; 10: 477–492.
2. Habib R, Gubler M-C, Loirat C et al. Dense
deposit disease: a variant of membranopro-
liferative glomerulonephritis. Kidney Int 1975; 7:
204–215.
3. Strife CF, McEnery PT, McAdams AJ et al.
Membranoproliferative glomerulonephritis
with disruption of the glomerular basement
membrane. Clin Nephrol 1977; 7: 65–72.
4. Appel GB, Cook HT, Hageman G et al.
Membranoproliferative glomerulonephritis type
II (dense deposit disease): an update. J Am Soc
Nephrol 2005; 16: 1392–1403.
5. Servais A, Fremeaux-Bacchi V, Lequintrec M
et al. Primary glomerulonephritis with isolated
C3 deposits: a new entity which shares common
genetic risk factors with haemolytic uraemic
syndrome. J Med Genet 2007; 44: 193–199.
6. Sethi S, Fervenza FC. Membranoproliferative
glomerulonephritis: pathogenetic heterogeneity
and proposal for a new classification. Semin
Nephrol 2011; 31: 341–348.
7. Servais A, Noël L-H, Roumenina LT et al. Acquired
and genetic complement abnormalities play a
critical role in dense deposit disease and other C3
glomerulopathies. Kidney Int 2012; 82: 454–463.
8. Sethi S, Fervenza FC, Zhang Y et al. C3
glomerulonephritis: clinicopathologic findings,
complement abnormalities, glomerular proteomic
profile, treatment and follow-up. Kidney Int 2012;
82: 465–473.
9. Zhang Y, Meyer NC, Wang K et al. Causes of alter-
native pathway dysregulation in dense deposit
disease. Clin J Am Soc Nephrol 2012; 7: 265–274.
10. Bao L, Haas M, Quigg RJ. Complement factor H
deficiency accelerates development of lupus
nephritis. J Am Soc Nephrol 2011; 22: 285–295.
11. Gharavi AG, Kiryluk K, Choi M et al. Genome-wide
association study identifies susceptibility loci for
IgA nephropathy. Nat Genet 2011; 43: 321–327.
12. Fremeaux-Bacchi V, Weiss L, Demouchy C et al.
Hypocomplementaemia of poststreptococcal
acute glomerulonephritis is associated with C3
nephritic factor [C3NeF] IgG autoantibody activity.
Nephrol Dial Transplant 1994; 9: 1747–1750.
381