(Figure 2: Nephrolithiasis)
We analysed urinary and plasma parameters of interest for nephrolithiasis in 66 astronauts that stayed on the International Space Station (ISS) for a period between 1 and 180 days. Values were analysed across time, including before, during and after flight. Our data showed an increased urinary excretion of parameters that are considered risk factors for nephrolithiasis (Figure 2A). In the figure, urinary excretion of calcium is expressed as fractional excretion (FECa). In the absence of plasma calcium abnormalities FECa should be <1%; in astronauts, FECa increases significantly during flight. This urinary abnormality normalises on return on Earth. A less significant increase was seen in oxalate, phosphate, uric acid urinary excretion during space flight, all rapidly decreasing at return. Urinary citrate did not significantly change and remained within normal range. There was also a decrease in urinary volume during spaceflight and an increase in urinary osmolality (Supplementary Figure X). Urinary electrolytes (chloride, sodium and potassium) remained stable during spaceflight compared to pre-flight with the exception of magnesium that, although within the normal range, significantly increased (Supplementary Figure X). The calculated values: estimated glomerular filtration rate (eGFR), aldosterone/renin ratio, FE for urinary electrolytes and water, transtubular potassium gradient (TTKG) and the ratio of tubular maximum reabsorption of phosphate (TmP) to GFR can be found in Supplementary Figure Y. Supplementary Figure Z shows plasma values from the 66 astronauts.
These data confirm an increased excretion of calcium and phosphate in the urine that normalises on return to Earth. Other parameters change towards a pro-lithogenic profile in space (urinary excretion of oxalate and uric acid and decreased urinary volume), however, citrate, a key player in lithogenesis, was not depleted. This suggests that there might be other mechanisms underlying the increased risk of nephrolithiasis in space.
To test this hypothesis, we analysed plasma and kidney tissue multi-omic data from human and mouse samples (Figure 2B) for disease ontologies related to kidney stone formation. The nephrolithiasis and related ontologies were heavily represented across different datasets, providing orthogonal confirmation of real changes in the gene products comprising this ontology. In contrast, hypercalciuria and hyperoxaluria ontologies were scantily represented.
The volcano plot displaying rodent kidney tissue phosphoproteomic data (Figure 2C) shows a decreased phosphorylation of SLC12A1, the furosemide-sensitive Na-K-Cl cotransporter expressed in the thick ascending limb of the loop of Henle. This transporter is regulated by the calcium-sensing receptor (CaSR) and genetic (i.e. Bartter syndrome) or pharmacological (e.g. furosemide) impairment of SLC12A1 leads to hypercalciuria.
The changes in faecal microbiome between space flight and ground controls (Figure 2D), highlighted in red are statistically significant. Overall, these microbiota changes would be expected to confer a protective profile for nephrolithiasis, implying that the increased urinary oxalate excretion is not diet or microbiota-dependent.
Pathway analysis of novel plasma metabolomic data from humans in spaceflight (Inspiration 4, Figure 2E) shows a striking over-representation of Arginine biosynthesis and Arginine-Proline metabolism; Hydroxy-L-Proline is a precursor to glyoxylate19, and acts as a substrate to produce glyoxylate in the mitochondria of renal proximal tubular cells20.Further, Arginine-Proline metabolism has been implicated in toxic nephrolithiasis21. Glycerophospholipids are the main lipid component of cell membranes and pathway enrichment is compatible with remodelling events (reviewed by Han 22). Citric Acid Cycle pathway enrichment is common to many metabolomic analyses of oxalosis; malate is a precursor to oxaloacetate and oxalate.
(Figure 3: Remodelling)
To investigate whether renal remodelling was occurring in spaceflight, we analysed Kyoto Encyclopaedia of Genes and Genomes (KEGG) pathway enrichment across 17 datasets comprising 2 human and 13 rodent spaceflight missions and 2 rodent simulated GCR experiments. The datasets were epigenomic, transcriptomic, proteomic and phosphoproteomic (Figure 3A). Enriched KEGG pathways that are repeatedly enriched across the datasets are compatible with remodelling events occurring in the kidney, these include focal adhesion, tight junction, gap junction, sphingolipids, actin and cell cycle pathways.
Proteomic analysis of urine from the Russian Roscosmos human spaceflight mission (Figure 3B) show ECM-turnover related proteins that were enriched 7 days after spaceflight compared to baseline. These include Fibronectin type 3, Apolipoprotein H (which binds to cell surface phospholipids), Alpha2-HS-Glycoprotein, the basement membrane protein Nidogen1, APP secretase (involved in intracellular protein turnover, Proactivator Polypeptide (involved in the lysosomal degradation of sphingolipids and CDH13 (calcium -dependent cell adhesion protein). Angiotensinogen is also enriched, which probably reflects RAAS activation which may be involved in ECM-turnover and putative remodelling.
3D imaging of optically cleared rodent tissue from RR-10 spaceflight mission (Figure 3C). The sample is stained for the canonical distal convoluted tubule (DCT) marker SLC12A3 (Blue) and the collecting duct marker AQP2 (green). Automated morphometry shows a significant change in the intensity and distribution of DCT tubule/area ratio and nuclear density between baseline and spaceflight tissue.
(Figure 4: Disease Ontology Pathways)
To understand the common processes occurring across the multiple human and rodent, SF and GCR exposed proteomic and transcriptomic datasets, we plotted enriched kidney relevant disease ontologies that were reproduced in our diverse cohort of datasets. Many of the same ontologies are represented in both the SF and simulated GCR experiments. Ontologies that are particularly enriched include amyloidosis, nephrotic syndrome and membranous glomerulonephritis, reflecting the presence of acute phase and inflammatory gene products in addition to the profibrotic and cell death gene products represented in the CKD5, renal fibrosis and renal insufficiency ontologies. The heavily enriched peripheral vascular disease and endothelial damage ontologies reflect the known endothelial and microvascular damage marker gene products enriched across the multiple GCR and SF datasets.
(Figure 5: GCR induced damage)
We then examined tissue that was only exposed to simulated GCR, in order to understand the effect of simulated GCR alone, rather than SF exposed animals, which will always have a component of GCR exposure in addition to MG, even inlow earth orbit. To do this we utilised animals that had acute exposure to GCR followed by rapid sacrifice (BNL experiments) and also acute exposure to simulated GCR and sacrifice at 6 months post exposure (NSRL-22A experiment).
Firstly, we looked for functional evidence of glomerular, proximal tubular and distal tubular function. Urinary biochemical analysis (Figure 5A) showed significantly greater urinary protein excretion in GCR vs sham animals, consistent with a glomerular lesion. There was no significant difference in urinary glucose, which is a highly specific, but not very sensitive marker of proximal tubular dysfunction in non-diabetic animals. Finally, there was significantly greater magnesuria in GCR exposed animals, consistent with downregulated magnesium reabsorption in the loop of Henle or the distal convoluted tubule. Next we performed fluorescence in situ hybridisation (RNAScope) to ascertain the presence and density of pathogenic miRNA species previously implicated in tissue damage in SF23 (Figure 5B). Kidney tissue was algorithmically segmented into relevant regions (cortex, outer stripe of the outer medulla (OSOM), inner stripe of the outer medulla (ISOM), inner medulla (IM)) and the staining density was measured using QPath. The pathogenic mi-RNA 125A was significantly increased in the ISOM, particularly in the vascular compartment (arrows) which corresponds to the thick ascending limb of the loop of Henle, the thin descending limb of the loop and the medullary collecting duct. This mi-RNA species is implicated in vascular damage and the ISOM is also the site of the vascular bundle24, a complex structure which is the bottleneck of all blood flow to and from the inner medulla. Further, serial tissue sections of stained kidney tissue (hematoxylin and eosin, Masson’s trichrome stain) were examined by a blinded expert histopathologist. There was a thrombotic microangiopathy seen in GCR exposed tissue from two animals (Figure 5C).
To further understand the effect of simulated GCR on specific cell types in the kidney, we performed 2D spatial transcriptomics (Slide-seq25) on GCR exposed renal tissue (BNL-1 experiment). Cell types were inferred using a published renal cell atlas26, which were mapped onto individual data points which accurately reproduced anatomy (Figure 5D). Transcriptional changes in specific cell types were then examined. Immunoglobulin components were heavily overexpressed by many cell sub-populations (proximal convoluted tubular cells, proximal straight tubular cells, loop of Henle and distal convoluted tubular cells), presumably reflecting increased B-cell activity in the interstitium. SGK1 transcriptional activity was decreased in PCT and PST cells. SGK1 is a master regulator of a number of epithelial membrane transporters, including canonical segment specific transporters. Heavy metal scavenging ligands, such as metallothionein-2, were transcriptionally upregulated in the PST and LoH; possibly related to local oxidative damage; with simulated GCR this may be due to metallic HZE particles.