IntroductionC3 glomerulopathy (C3GP) is caused by persisting activity of the complement system and characterized by glomerular deposition of C3, demonstrated by immunohistochemistry. Also in most of the other forms of glomerulonephritis (GN) deposition of complement factors occurs. However, here the significance and predominant activation pathways are less clear.MethodsWe used proximity ligation assays (PLA) to visualize C3/C5 complement convertases in renal biopsies. Close proximity of C3b and Bb or C2 and C4b was interpreted as assembled alternative or classical/lectin C3/C5 convertases, respectively. For quantitative analysis we established a two-stage, deep learning, explainable artificial intelligence (xAI) based workflow for a) the detection of glomeruli and b) detection and quantification of PLA signals. Signal densities were calculated as numbers of signals per glomerular area [signals/sqmm]. We analyzed ten cases per disease group of C3GP, immune complex membranoproliferative GN (IC-MPGN), IgA nephropathy (IgAN), infection-related GN (IRGN), and membranous nephropathy (MN) in comparison with 10 cases of thin basement membrane disease as control group, in which no local complement activity is expected.ResultsIn C3GP and IRGN a clear predominance of the alternative convertase (mean 8410 and 14483 signals/sqmm) was detected as compared to the control group (mean 798 signals/sqmm) whereas IC-MPGN and MN cases showed higher densities of the classical/lectin convertase (3039 signals/sqmm and 5015 signals/sqmm) as compared to the control group (mean 176 signals/sqmm). Interestingly, cases with IgAN revealed increased densities for the alternative convertase (mean 2088 signals/sqmm) but only very slightly increased densities for the classical/lectin convertase (225 signals/sqmm). IgAN cases with mesangial and endocapillary hypercellularity had higher values for the alternative convertase than cases without hypercellularity.ConclusionsOur work shows the applicability of a combined xAI based convertase detecting analysis to quantify and type local complement activity in renal biopsies. The results reveal insights into the pathogenesis of different forms of glomerulonephritis. The clear predominance of C3bBb signals in IgAN and IRGN suggests a direct activation of the alternative pathway rather than an immune complex mediated activation via the classical pathway in these two entities. Moreover, this approach opens up the possibility to assess the local complement activity in individual patients also with regard to novel anti-complement agents being under clinical investigation.Conflict of interest Corporate sponsored research or other substantive relationships:Novartis, Bayer, Glaxo Smith Kline IntroductionC3 glomerulopathy (C3GP) is caused by persisting activity of the complement system and characterized by glomerular deposition of C3, demonstrated by immunohistochemistry. Also in most of the other forms of glomerulonephritis (GN) deposition of complement factors occurs. However, here the significance and predominant activation pathways are less clear. C3 glomerulopathy (C3GP) is caused by persisting activity of the complement system and characterized by glomerular deposition of C3, demonstrated by immunohistochemistry. Also in most of the other forms of glomerulonephritis (GN) deposition of complement factors occurs. However, here the significance and predominant activation pathways are less clear. MethodsWe used proximity ligation assays (PLA) to visualize C3/C5 complement convertases in renal biopsies. Close proximity of C3b and Bb or C2 and C4b was interpreted as assembled alternative or classical/lectin C3/C5 convertases, respectively. For quantitative analysis we established a two-stage, deep learning, explainable artificial intelligence (xAI) based workflow for a) the detection of glomeruli and b) detection and quantification of PLA signals. Signal densities were calculated as numbers of signals per glomerular area [signals/sqmm]. We analyzed ten cases per disease group of C3GP, immune complex membranoproliferative GN (IC-MPGN), IgA nephropathy (IgAN), infection-related GN (IRGN), and membranous nephropathy (MN) in comparison with 10 cases of thin basement membrane disease as control group, in which no local complement activity is expected. We used proximity ligation assays (PLA) to visualize C3/C5 complement convertases in renal biopsies. Close proximity of C3b and Bb or C2 and C4b was interpreted as assembled alternative or classical/lectin C3/C5 convertases, respectively. For quantitative analysis we established a two-stage, deep learning, explainable artificial intelligence (xAI) based workflow for a) the detection of glomeruli and b) detection and quantification of PLA signals. Signal densities were calculated as numbers of signals per glomerular area [signals/sqmm]. We analyzed ten cases per disease group of C3GP, immune complex membranoproliferative GN (IC-MPGN), IgA nephropathy (IgAN), infection-related GN (IRGN), and membranous nephropathy (MN) in comparison with 10 cases of thin basement membrane disease as control group, in which no local complement activity is expected. ResultsIn C3GP and IRGN a clear predominance of the alternative convertase (mean 8410 and 14483 signals/sqmm) was detected as compared to the control group (mean 798 signals/sqmm) whereas IC-MPGN and MN cases showed higher densities of the classical/lectin convertase (3039 signals/sqmm and 5015 signals/sqmm) as compared to the control group (mean 176 signals/sqmm). Interestingly, cases with IgAN revealed increased densities for the alternative convertase (mean 2088 signals/sqmm) but only very slightly increased densities for the classical/lectin convertase (225 signals/sqmm). IgAN cases with mesangial and endocapillary hypercellularity had higher values for the alternative convertase than cases without hypercellularity. In C3GP and IRGN a clear predominance of the alternative convertase (mean 8410 and 14483 signals/sqmm) was detected as compared to the control group (mean 798 signals/sqmm) whereas IC-MPGN and MN cases showed higher densities of the classical/lectin convertase (3039 signals/sqmm and 5015 signals/sqmm) as compared to the control group (mean 176 signals/sqmm). Interestingly, cases with IgAN revealed increased densities for the alternative convertase (mean 2088 signals/sqmm) but only very slightly increased densities for the classical/lectin convertase (225 signals/sqmm). IgAN cases with mesangial and endocapillary hypercellularity had higher values for the alternative convertase than cases without hypercellularity. ConclusionsOur work shows the applicability of a combined xAI based convertase detecting analysis to quantify and type local complement activity in renal biopsies. The results reveal insights into the pathogenesis of different forms of glomerulonephritis. The clear predominance of C3bBb signals in IgAN and IRGN suggests a direct activation of the alternative pathway rather than an immune complex mediated activation via the classical pathway in these two entities. Moreover, this approach opens up the possibility to assess the local complement activity in individual patients also with regard to novel anti-complement agents being under clinical investigation.Conflict of interest Corporate sponsored research or other substantive relationships:Novartis, Bayer, Glaxo Smith Kline Our work shows the applicability of a combined xAI based convertase detecting analysis to quantify and type local complement activity in renal biopsies. The results reveal insights into the pathogenesis of different forms of glomerulonephritis. The clear predominance of C3bBb signals in IgAN and IRGN suggests a direct activation of the alternative pathway rather than an immune complex mediated activation via the classical pathway in these two entities. Moreover, this approach opens up the possibility to assess the local complement activity in individual patients also with regard to novel anti-complement agents being under clinical investigation.