Human parietal epithelial cells (PECs) and proteinuria in lupus nephritis: a role for ClC-5, megalin, and cubilin?

In 2009, Ronconi et al. demonstrated the existence of human renal progenitors located in Bowman’s capsule [1]. This is a strategic location due to its close proximity to both tubular cells and podocytes. Many studies focused on the pathogenesis of hyperplastic lesions and the formation of crescents in glomerulopathies, providing evidence that renal progenitor cells located in Bowman’s capsule may differentiate into podocytes under certain clinical circumstances [2, 3].

Parietal epithelial cells (PECs) are a heterogeneous cell population lining the outer membrane of Bowman’s capsule characterized by specific molecular markers. Besides the classical flat PECs, a subpopulation of cuboidal (cPECs) and intermediate PECs (iPECs) has also been identified [4]. Mature PECs are recognizable by the expression of annexin A3 (ANXA3), LKIV69, Pax-2, Pax-8, and claudin-1. Podocyte-committed progenitors are detectable by the co-expression of the podocyte marker podocalyxin and the progenitor markers CD24 and CD133/2. Adult parietal epithelial pluripotent progenitors (APEMP) have been shown to express CD24 and CD133/2 but not podocyte or tubular markers [5]. Activated PECs play a key role in the pathogenesis of glomerulosclerosis and possess the capacity to proliferate, migrate and produce extracellular matrix. These cells are known to express the CD44 marker [2, 3].

The cellular protein uptake machinery composed of megalin, cubilin, and ClC-5 is normally present in proximal tubular cells (PTCs) but it is also present in human podocytes [6‐8]. It was also reported that mouse and rat PECs expressing megalin and cubilin were able to internalize albumin in normal and overload conditions [9, 10].

Lupus nephritis (LN) is a frequent complication of systemic lupus erythematosus, a systemic autoimmune disease [11] that has a major impact on the prognosis of the disease, resulting in a risk of end-stage renal disease. The pathogenesis of LN involves genetic, epigenetic, immunoregulatory, hormonal, and environmental phenomena. Lupus nephritis is associated with a wide spectrum of kidney lesions, mostly characterized by glomerular involvement. Mesangial cells and podocytes have been identified as direct targets of immune complexes through specific epitopes [12]. Proteinuria is an important marker of LN. Crosstalk between podocytes and PECs in some proliferative glomerular diseases is well known [3], while the role of PECs in the mechanism of protein uptake needs to be further explored.

In this study we aimed to investigate the role of ClC-5, megalin, and cubilin in PECs of kidney biopsies from patients with lupus nephritis as compared to control kidneys, and through the characterization of the different PEC subpopulations, to identify phenotypical changes of PECs occurring in the pathological milieu.

Histological analysis and clinical parameters of the LN group are reported in Table 1. Most of the patients (n = 15) had stage IV according to the ISN/RPS classification, 3 patients had stage II, 3 patients had stage III, and 2 patients had stage V. In LN renal biopsies we evaluated 158 glomeruli, 3.8% were sclerotic and another 3.8% showed cellular crescents and belonged to patients in stage IV (Table 1).

Inflammatory infiltrate was present in 5 patients out of 23: one patient belonged to class II, one to class III, and three to class IV. These small numbers did not allow us to verify a possible link between inflammation and the presence of hypertrophic PECs positive for ClC-5, megalin, and cubilin.

We did, however, observe, for the first time in both C and LN glomeruli, PECs expressing the protein uptake macromolecular machinery composed of ClC-5, megalin, and cubilin which were scattered around the glomerulus and did not significantly differ between C and LN. Semiquantitative analysis considering the number of glomeruli with all positive human PECs (flat and hypertrophic) vs number of total glomeruli in C and LN are reported in Table 2.

Different morphologies of ANXA3 positive cells were observed. Besides the classical flat PECs, we noticed in both control and proteinuric biopsies some cells showing a hypertrophic appearance with larger size and more represented cytoplasm than flat PECs mainly located at the tubular pole of the glomerulus (Figs. 1A, 2A). Hypertrophic PECs displayed positivity for megalin, cubilin, and ClC-5. Table 3 shows the percentage of hypertrophic PECs expressing the components of the macromolecular system in C and in biopsies of LN patients. The different expression between the number of glomeruli with positive hypertrophic PECs in C and LN was statistically different for all molecules (ClC-5 χ² 8.17 p = 0.0042; megalin χ² 17.4537 p = 0.0000; cubilin χ² 15.3634 p = 0.00002). These proteins were not always co-expressed by the same hypertrophic cells (Figs. 1B–D, 2B , C). In order to better characterize which type of PEC was able to express the components of the tubular uptake machinery, we tested different markers: CD24 for progenitor cells, CD44 for pro-sclerotic activation, podocalyxin for podocytes, LTA for PTC brush border, and Ki67 for proliferating cells. Furthermore, anti-HSA was used to detect albumin. Semiquantitative analysis performed on CD24 and CD44 positive PECs revealed, though only for CD24, a significant difference between C and LN (χ² 16.5968 p = 0.00004).

Hypertrophic PECs positive for megalin, cubilin, and ClC-5 at the tubular pole were negative for podocalyxin, CD44 (Fig. 1E and F), and HSA (Fig. 1H), with scattered positivity for the marker CD24 both in LN and in C biopsies (Figs. 1G and 2D). The degree of co-localization between megalin and CD24 signals was highly correlated both in LN and C biopsies (Rr 0.81 ± 0.07). Ki67 and LTA signals were absent (data not shown).

On the other hand, we observed that there was a distinct population of hypertrophic PECs showing co-localization for megalin, cubilin, and LTA (Rr 0.82 ± 0.06) (Fig. 3A and B). These cells were also positive for HSA (Fig. 3C) and negative for CD24 (Fig. 3D). When analyzing the regions with these hypertrophic PECs we observed that the HSA and LTA signals were scattered in a distributed manner. The two subpopulations of hypertrophic PECs described in our study are reported in Supplementary Fig. 1.

Recently, PECs have attracted a great deal of attention in an attempt to clarify their molecular phenotype and understand their function and responses in normal and in pathological conditions [4, 14].

The ability of PECs to internalize albumin is well-known in animal models [9]. However, the mechanism underlying this activity is yet to be clarified.

ClC-5, megalin, and cubilin are involved in protein uptake, and for the first time our study identified their presence also in the PECs of patients with lupus nephropathy and proteinura.

In humans, CD24 and CD133 are used to characterize cells with stem or progenitor properties [1]. The co-expression of CD24 and CD133 has been identified in a subset of PECs on the Bowman’s capsule of adult human kidneys [5]. We identified a CD24⁺/Podocalyxin⁻ subpopulation of PECs with hypertrophic appearance, expressing the protein uptake system components that we speculate to be APEMP, supporting the recently proposed hypothesis of the existence of tubular committed progenitor cells among PECs [3, 4].

The cells we identified could morphologically be iPECs sharing similarities with the tubular cell phenotype, without having reached the functional differentiation state, as suggested by the absence of the apical brush border.

We also demonstrated the presence of another population of CD24⁻/LTA⁺ PECs with the morphological appearance of cPECs at the tubular pole as well. They expressed ClC-5, megalin and cubilin with the ability to internalize albumin, suggesting a further step of commitment of these cells toward PTCs.

The different CD24 and LTA expression observed among hypertrophic PECs could likely be attributable to the different stages of PEC differentiation occurring between intermediate and cuboid PECs.

Many studies focused on the role of aPECs (CD44⁺/Ki67⁺) in the genesis of sclerotic lesions [3, 14], while less information is available regarding the role of PECs in the mechanism of protein uptake in normal kidney and in proteinuric nephropathies, a key step that may eventually lead to glomerulosclerosis.

The observation in our study of inactivated and non-proliferating hypertrophic PECs (CD44⁻/Ki67⁻) suggests that these cells are not involved in the pathogenesis of sclerotic lesions, at least at this stage of the nephropathy [10, 15].

The presence of hypertrophic CD24⁺ PECs, in a significantly greater number in lupus nephropathy biopsies than in controls, suggests that both proliferative glomerulonephritis and proteinuric milieu might induce an increment in these cells.

The observation that, even in controls, different subpopulations of PECs were present within the same glomerulus (with/without hypertrophy) highlights the great plasticity of PECs and suggests that these cells may change their morphology and protein expression in response to different stimuli (ischemia, proteinuria, hyperglycemia) [4].

In this study we have shown that two main sub-populations of hypertrophic PECs are present in human glomeruli, both of which express ClC-5, megalin, and cubilin. One subtype seems to show characteristics of APEMP, while the other is suggestive of the PTC phenotype.

The main limitation of our study is that we analyzed only one type of glomerular disease, lupus nephritis, both proliferative (Classes II, III and IV) and non-proliferative (Class V). Although the urinary hallmark is proteinuria, the pathophysiological pathways may differ. In this regard, we believe it would be interesting to evaluate glomerulonephritis of different etiologies, such as IgA and membranous nephropathies, to support the hypothesis that protein milieu can induce the increase in APEMPs and to demonstrate whether the phenotypic changes in PECs, described in our research, are independent or not of the immunologic pathogenesis of the underlying nephropathy.

However, our results obtained in LN biopsies may provide a starting point for further studies. Another limitation is that we did not perform characterization for progenitor cells using CD133/2, a well-known CD24 molecular partner. However, many studies report that CD133/2 and CD24 are co-expressed in the adult kidney [1, 5, 16, 17] allowing us to suppose that positivity for CD24 could identify progenitor-like cells.

In conclusion, our results describe two main sub-populations of hypertrophic PECs in human glomeruli in proteinuric proliferative and non-proliferative human lupus kidney biopsies, both of which express megalin, cubilin, and ClC-5, components of the macromolecular complex normally expressed by PTCs. The first subset seems to show characteristics of APEMP, the second is more in keeping with a PTC phenotype. In addition, our data suggest that the presence of hypertrophic PECs in the Bowman’s capsule could reflect a response to protein overload in LN, regardless of the proliferative setting. The link between proteinuria and renal disease progression remains very strong. The observation that PECs can modify their phenotype in order to counteract protein overload could be a first step toward identifying PEC phenotypes mainly involved in the formation of glomerulosclerotic lesions.