Tertiary Lymphoid Structure–Like Lymphocyte Aggregate Generation in Human and Mouse Hypokalaemic Nephropathy: Research Letter

Introduction

The kidney handles a wide range of potassium (K⁺) load to keep homeostasis by modulating K⁺ excretion and sodium (Na⁺) reabsorption. If the kidney is exposed to an excessively low K⁺ milieu beyond physiological limits for an extended period, it develops aberrant kidney damage, known as hypokalaemic nephropathy (HN). To resolve the malnourished kidney microenvironment, HN is commonly treated with K⁺ supplementation. However, the long-term prognosis of HN has not been thoroughly explored. Furthermore, we recently found that HN kidneys exhibit intense inflammatory cell infiltration preceding cell death, suggesting that inflammation plays a primary role. ¹ In this process, inflammasome-associated proteins, nucleotide-binding and oligomerization domain-like receptor family pyrin domain-containing 3 (NLRP3) and apoptosis-associated speck-like protein-containing a caspase recruitment domain (ASC), play a pivotal role in the pro-inflammatory response. ¹ Whether or not, and how, this unique inflammatory response resolves after K⁺ supplementation has proved elusive.

Tertiary lymphoid structures (TLSs) are organized ectopic lymphocyte aggregates that develop de novo in nonlymphoid organs.^2,3 Heterogeneous cell populations, including T and B cells, dendritic cells, and fibroblasts, participate in the formation of TLSs under chronic inflammation. Tertiary lymphoid structure formation has been reported in several kidney diseases, including IgA nephropathy, anti-neutrophil cytoplasmic antibody-associated vasculitis, lupus nephritis, membranous nephropathy, transplanted kidney, and renal cell carcinoma. ³ Although still contentious, the presence of TLS correlates with non-cancer kidney disease progression. ³ It is not known whether TLSs are generated in the HN kidneys. We analysed inflammatory infiltrates in HN to expand perspectives for treating HN, with the help of a mouse model and observational data from human HN.

Methods

Detailed methods are described in Supplementary Methods.

Results

First, we analysed five human kidney biopsy samples of clinically diagnosed HN. ¹ Our patients represented diverse clinical manifestations and had heterogeneous backgrounds, including anorexia nervosa (K⁺ intake deficiency), Crohn disease, laxative abuse (extra-kidney K⁺ loss), Sjögren’s syndrome, and licorice overuse (kidney K⁺ loss) (Supplemental Table S1). Immunostaining for B and T lymphocytes (CD20 and CD3ε, respectively) revealed varying degrees of infiltration (Figure 1A-E). Hypokalaemic nephropathy kidneys with laxative abuse, Sjögren’s syndrome, and licorice overuse showed lymphocyte aggregations with segregated T and B cells, potentially forming TLSs (Figure 1C-E).

OPEN IN VIEWER

Figure 1.

Kidney biopsy and clinical course of human hypokalaemic nephropathy and a mouse model of recovery from hypokalaemic nephropathy. (A-E) Paraffin-embedded sections of biopsy samples of hypokalaemic nephropathy (HN) kidneys and the clinical course of patients with HN. Kidney sections were subjected to immunostaining for CD20 (green; B cells) and CD3ε (red; T cells). Sections were counterstained by DAPI (for nuclei, blue). Representative images of biopsy samples and clinical course of serum potassium (K⁺) levels and estimated glomerular filtration rate (eGFR) from patients with anorexia nervosa (A), Crohn disease and renal tubular acidosis (B), laxative abuse (C), Sjögren’s syndrome (D), and chronic licorice overuse (E) are shown. (F-I) Six-week-old C57BL6/J male mice were fed a low-K⁺ diet or standard chow for six weeks, followed by a recovery diet (standard chow) for six months. (F) Experimental design is shown. (G) Picrosirius red (PSR) staining of the kidney was observed under polarized microscopy. Collagen type I (red) area was quantified (n = 5 for each). (H) Representative immunostaining for B220 (green; B cells) and CD3ε (red; T cells) of aggregates and the number of tertiary lymphoid structures (TLSs) is quantified. (I) Correlation between the number of TLSs and the collagen type I area.

Four in 5 patients, except for licorice overuse, were followed for a mean duration of 8.4 ± 3.4 years. All patients with HN were treated with K⁺ supplementation, and none progressed to kidney failure. A female patient with anorexia nervosa demonstrated that the change of estimated glomerular filtration rate (eGFR) correlated well with serum K⁺ levels (r = .7236, Figure 1A and Supplemental Figure S1A). On the contrary, a male patient with HN with Crohn disease who received effective supplementation in serum K⁺ levels did not improve kidney function dramatically for more than 12 years after an initial drop before supplementation (Figure 1B). A patient with HN who failed to achieve serum K⁺ levels above 3.5 mEq/L continued to progress kidney dysfunction (Figure 1C). Another patient with HN with excessive kidney K⁺ excretion complicated with Sjögren’s syndrome was treated with oral corticosteroid combined with K⁺ supplementation, aiming for the underlying massive tubulointerstitial nephritis (Figure 1D). Following a maintenance dose of prednisolone and proper K⁺ supplementation, serum K⁺ levels and kidney function continued to improve. Intriguingly, modulation of serum K⁺ levels did not show a positive correlation with eGFR in most HN patients, except one with anorexia nervosa (Supplemental Figure S1). Together, HN patients were treated with K⁺ replenishment as a mainstay strategy, but its effectiveness seemed inconsistent.

Next, we performed mouse experiments to assess the long-term kidney prognosis after restoring K⁺ intake following HN induction. Age-matched mice were compared at six months of recovery diet after feeding a standard chow or a low-K⁺ diet for six weeks (Figure 1F). Hypokalaemia induced by a six-week low-K⁺ diet is sufficient to cause kidney dysfunction, tubular necrosis, inflammation, and fibrosis. ¹ Six-month standard chow after feeding a low-K⁺ diet (L6w + Recovery) effectively restored serum K⁺ and blood urea nitrogen levels (Supplemental Figure S2A and B). Histological assessments, however, demonstrated different manifestations. Fibrosis induced by a low-K⁺ diet (L6w) tended to further progress six months after the recovery diet (L6w + Recovery, Figure 1G). Furthermore, six months of standard chow (L6w + Recovery) did not fully recover from tubular injury compared with the L6w group, albeit with significant improvement (Supplemental Figure S2C). Notably, focal interstitial cell aggregates, which were barely visible in kidneys just after a six-week low-K⁺ diet (L6w) or age-matched controls (C6w + Recovery), were documented in the perivascular space of the kidneys six months after recovery (L6w + Recovery, Supplemental Figure S2C). These aggregates were structured by B and T cells (B220 and CD3ε, respectively; Figure 1H), suggesting potential TLSs. The number of these lymphocyte aggregates forming in the HN kidney highly correlated with the degree of fibrosis (r = .9438, Figure 1I).

Discussion

We found for the first time that HN kidneys can develop TLS-like lymphocyte aggregates alongside worsening fibrosis, with a positive correlation (Figure 1I). Lymphocytes accumulate in injured kidneys, and stromal cell–derived homeostatic lymphoid chemokines, such as CCL19, CCL21, and CXCL13, contribute to the organization of lymphocyte aggregates into TLSs.^4,5 Stromal reprogramming supports dysregulated B-cell and T-cell activation, facilitating local antigen presentation and, potentially, autoantibody generation. Functional activation of adaptive immunity within TLSs is further characterized by increased proliferation of B and T cells.

The HN kidneys we presented here exhibited infiltrated T cells, which can play a crucial role in TLSs for local lymphoid neogenesis. In line with the effects of sodium and osmolytes on T cells in the kidney, ⁶ an abnormally altered K⁺ milieu in HN kidneys may manipulate T-cell functions, metabolism, and proliferation, possibly accelerating TLS formation. Indeed, the elevated extracellular K⁺ released by necrotic cells disturbs effector functions and Akt-mammalian target of rapamycin signalling in T cells and, alternatively, drives starvation response and stemness.^7,8 Furthermore, K⁺ deficiency triggers Akt signalling and tubular cell expansion in the kidney. ⁹ Therefore, it is conceivable that the microenvironmental K⁺ may regulate the functions and kinetics of TLS-forming lymphocytes.

Several limitations of this study should be acknowledged. First, the biopsy analysis was limited by a small sample size and the lack of age-matched normokalaemic controls. Second, the mice used in the experiments may have been too young to adequately study TLS, as ageing is a critical biological driver of TLS formation. ¹⁰ Finally, this study demonstrated only TLS-like lymphocyte aggregation without detailed characterization of functional TLS. Evidence of local antigen presentation and lymphocyte proliferation should be investigated in future studies with larger sample sizes and well-designed experiments.

Conclusions

We documented TLS-like lymphocyte aggregates in human HN kidneys and demonstrated the de novo formation of TLS-like aggregates and progressive fibrosis in the mouse HN kidneys after long-term K⁺ repletion. Our results suggest a potential role of K⁺ for the generation of TLSs in the injured kidney. Along with a fibrotic scar that persists after a lengthy recovery period, TLSs may also serve as an inflammatory scar of a malnourished microenvironment. Strategies targeting inflammation, including TLSs, need to be explored as additional treatments for HN.

Supplemental Material

Supplemental Material

Supplemental Material

Supplemental Material

Supplemental Material