Single-cell transcriptional phenotypes linked to anatomical localization of fructose transporters in rat proximal tubule segments

Abstract

Kidneys of healthy individuals filter 4 to 25 g of fructose (FRU) a day, equivalent to ~10% of the filtered glucose. FRU reabsorbed by proximal tubules (PTs) is mostly used in gluconeogenesis. Elevated dietary FRU alters hormonal signaling in PTs and increases oxidative stress, which overtime, leads to salt-sensitive hypertension, tubulointerstitial injury, and glomerular damage. Four apical transporters could transport FRU at physiological concentrations: SGLT5 (km0.62 mM), NAGLT1 (km4.5 mM), GLUT5 (km12.6 mM) and SGLT4 (undetermined km). We hypothesize that single-cell (sc) transcriptional phenotypes correspond to anatomical features and could inform FRU transport in PTs. We used scRNAseq rat kidney transcriptomes (3 males & 3 females) to measure the expression of FRU transporters in cell clusters corresponding to the S1, S2, and S3 PT segments. We integrated cell transcriptional phenotypes with spatial and structural features using quantitative proteomics and transcriptomics from microdissected PT segments. Clusters S1, S2, and S3 had 1281, 2123 and 1441 cells, respectively. More than 90% of cells in all clusters expressed KHK, the first enzyme in FRU metabolism (expression: S1<S2<S3), showing high correlation with segment-proteomics (PCC (Pearson correlation coefficient) = 0.87) and transcriptomics (PCC = 0.65). In contrast, there were striking imbalances in transporters’ distributions. High-affinity SGLT5 was expressed by 3%, 24%, 75% cells in clusters S1, S2, and S3, with 4, 43, 230 aggregated counts per 100 cells (nCounts), respectively. SGLT5 showed a log2FC > 0.8 in S3 as compared to S1-S2 (p-adj < 1x10-300). SGLT5, PCC was 0.95 for quantitative proteomics and 0.48 for sequencing. Low-affinity NAGLT1 was expressed by 70%, 81%, 47% cells in clusters S1, S2, and S3, with 206, 306, 121 nCounts, respectively. These values were highly correlated with RNAseq in microdissected tubules (PCC = 0.99), but NAGLT1 was absent in the proteomics dataset. GLUT5 was expressed by 9%, 10%, 38% cells in clusters S1, S2 and S3, with 11, 11, 63 nCounts, respectively, and PCC > 0.95 for both quantitative proteomics and sequencing. Finally, SGLT4 was expressed by 4%, 6%, 8% cell in clusters S1, S2 and S3, with 5, 6, 8 nCounts, respectively. Due to the low signal in microdissected segments, we could not conduct a correlation analysis for SGLT4. Our data shows that SGLT5 is most abundant in the straight portion of PTs, particularly S3. GLUT5 and SGLT4 follow a similar distribution. In contrast, low-affinity NAGLT1 is expressed in all segments, most predominantly S1. However, as NAGLT1 affinity for glucose (km3.7 mM) is near that for FRU, it likely reabsorbs glucose in early PT. In the straight portion FRU will be predominantly reabsorbed by the most abundant higher-affinity SGLT5. Our analysis shows a high correlation between scRNAseq clusters and spatial proteomics data and highlights the relevance of SGLT5 as the main FRU transporter in PTs BGT670107 This is the full abstract presented at the American Physiology Summit 2023 meeting and is only available in HTML format. There are no additional versions or additional content available for this abstract. Physiology was not involved in the peer review process.