Salt intake and the dance of the macula densa cells

Structure and function are intimately intertwined, and, as the pioneering architect Louis Sullivan stated, “all things in nature have a shape, a form, an outward semblance that distinguishes them from each other, that can deliver the message of the building concept (function) without people even going into the building.” Physiology evolved from knowledge of structure, and such is the case with the masterful study by Gyarmati et al. (1) of the macula densa (MD) microanatomy published in the current issue of the American Journal of Physiology-Renal Physiology as “A new view of macula densa cell microanatomy.” The images of the MD cellular microstructure obtained by these investigators represent a meaningful advance in our understanding of the how these unique cells may function. The study brought together cutting-edge tools of genetic engineering with powerful imaging technologies, thereby providing a clearer contextual framework upon which to advance our understanding of MD functions. To visually observe the dance-like movements of MD cells in real time is to better understand how they communicate signals from the early distal tubule to the surrounding cells of the juxtaglomerular apparatus (JGA). To see these structures and how their movements are altered by a salt diet and the renin-angiotensin system provoke many new and interesting questions that must now be addressed.

The anatomic relation between glomerular afferent arterioles and tubules was first described by Golgi in 1893 and Goormaghtigh in 1945 (2), who first visualized the MD. He appreciated that these small clusters of cells were positioned to sense the contents of the early segments of the distal tubules and transmit this information to the vascular component of the glomerular hilus and the mesangium of the JGA. Since then, many outstanding physiologists have examined the merit of this hypothesis and identified key functional components of the molecular feedback pathways responsible for tubuloglomerular feedback (TGF) and autoregulation of glomerular filtration rate (GFR). Feedback gains and time responses of TGF have been characterized both in vitro and in vivo, which, together with arteriolar myogenic responses, cooperatively maintain the constancy of GFR. It is well recognized that these responses are essential for the kidney to carry out the homeostatic tasks required for survival (3).

Although the sensory and feedback responses of the MD upon the JGA have been well defined, clear visualization of MD cells in living systems has been quite limited due to their relative inaccessibility and sparsity (∼20–25 cells/nephron) (4). High-resolution static features of the MD have been obtained from studies of fixed tissue specimens using light and electron microscopy, and cytoplasmic processes of the basal side of human MD cells have been found (5). Single 5- to 8-µm-long primary cilia extending into the lumen from each MD cell was identified by presence of α-tubulin. These were shown to respond to physiological changes of tubular flow similar to other nephron segments (6). Small vesicles have also been observed on the apical side of MD cells that appear to be endocytic in nature. On the basal side of the MD, small dense granules have been detected in both the basal cytoplasm and cytoplasmic projections, which have been assumed to be related to the paracrine activity of MD cells. However, such images lacked clarity, and the nature of the molecules and function of these structures have remained unclear.

The study by Gyarmati et al. (1) used a novel MD_GFP mouse model to clearly identify MD cells within the kidney. Mice with MD-specific tamoxifen-inducible membrane-targeted enhanced green fluorescent protein (eGFP) were generated by intercrossing mice containing the MD cell-specific marker neuronal nitric oxide synthase (nNOS) driving CreER with the mTmG reporter. The colabeling of nNOS and estimated GFR provided clear identification of MD cells and, with adjustments of tamoxifen dosing, enabled visualization of fine morphological details of individual cells. Confocal/multiphoton laser scanning fluorescence microscopy and transmission electron microscopy were used to study kidneys in vivo and in freshly dissected microperfused JGA preparations. Live isolated MD cells obtained by fluorescence-activated cell sorting provided remarkable glimpses of the dynamic nature of MD cells. Optically cleared kidneys with multiphoton microscopy (MPM) optical slicing of fluorescence-immunolabeled whole mouse kidneys provided the first detailed three-dimensional features of the MD. Finally, double labeling of MD and juxtaglomerular (JG) renin cells on thick frozen sections provided novel views of the contextual relationship of these two critical structures. All in all, this study is a real tour de force.

Many novel insights emerged from these analyses. Novel images were obtained of an elaborate network of major and minor processes at the MD cell base with projections reaching toward the glomerulus and other MD cells. These were named “maculapodia” by the authors. The immunohistochemical analysis of these structures in fixed tissues showed them to be of differing lengths, with some of the major processes ranging from 1 to 34 µm and running in parallel with the tubular side of the basement membrane. Interestingly, female mice exhibited significantly longer major processes. The isolated MD cells enabled quantification of the numbers of immune-labeled MD minor cell processes in living cells. Importantly, mice fed a low-salt diet and treated with an angiotensin-converting enzyme inhibitor exhibited a significant increase in the numbers of minor cell processes. F-actin but not tubulin was found in these minor cell processes, as was found in human kidney sections obtained from the Human Protein Atlas. These observations are consistent with the presence of a vesicular transport system at the basal processes of MD cells in mice and humans.

Among the most exciting observations were those obtained from live cells in which the highly dynamic nature of the MD maculopodia became evident. Time-lapse imaging of MD cells showed that the outgrowth or shorting of these cilia occurs with a few seconds in both the minor and major cell processes. Cytoplasmic vesicles were clearly visible within the minor processes that moved rapidly in both directions between the cell bodies and their tips. So too, remarkable images were obtained from the high-resolution three-dimensional multiphoton confocal optical slicing and reconstruction of MD cells. From these images, the major and minor processes can be clearly observed to be projecting into the extraglomerular mesangium at the glomerular vascular pole with the processes wrapping around individual cells of the afferent and efferent arterioles including JG renin cells. The functional implications of the maculapodia in these images become apparent as you see them positioned to secrete and uptake signaling molecules communicating to the JG apparatus and renin-secreting cells. To see this, “dance of the maculapodia” gives a fuller appreciation of how the physical forces and ionic contents of MD cells of the early distal tubule are able to transmit this information to surrounding elements of the JGA. It also stimulates one’s imagination to conceive of ways whereby we can now study the many yet unidentified intracellular molecular and biochemical pathways within the MD responsible for transducing from renal tubules to the surrounding JG structures. How are these pathways altered under normal physiological conditions, in response to dietary challenges, or during pregnancy? How are these responses altered in chronic diseases such as hypertension or eclampsia?

The release of ATP/adenosine is known to be a key effector of TGF regulation of afferent arteriolar resistance (3), but these responses can be modulated by nitric oxide, angiotensin, cytochrome P-450 metabolites, thromboxane, superoxide, kinins, and other molecules (3). It is unknown how many of these modulators of TGF are produced within the MD or from the surrounding cells. One of the novel and interesting cargos visualized in the vesicles of the maculapodia was pregnancy-associated plasma protein A2 (PAPPA2), which is secreted from the placental villi in high amounts in the first trimester of pregnancy (7). The present study identified the Pappa2 gene and protein in both freshly isolated and fixed MD cells of human and mouse kidneys. A high level of expression was found in a vesicular pattern within the cell body and the cell processes of the basal regions of the MD. PAPPA2 has been colocalized in Na⁺-K⁺-2Cl⁻ cotransporter-expressing cortical thick ascending limbs of Henle near the glomerulus (8). As with renin, PAPPA2 (mRNA and protein) is significantly reduced in the renal cortex of rats fed a high-salt diet and expressed at low levels in Dahl salt-sensitive rats. PAPPA2 is but one example of the many molecules that could now be studied to determine their release and paracrine mediators in the MD-JGA regulation of GFR and renal blood flow.

As was the case with the iconic picture of earthrise viewed from the moon by the crew of Apollo 11, the present study opens new vistas to better appreciate how these few specialized cells can cast such a large shadow upon the function of the kidney. Just seeing the maculapodia in real time reaching out to surrounding cells already extends our understanding of the function of the MD. The highly dynamic nature of MD cells to quickly form new connections with other JGA cells and the profound vesicular transport observed within these elements raise many interesting questions and possibilities for future studies. Exactly what is being produced and released from the macropodia and how are these functions regulated? Are the short and long filaments functionally distinct? Can the MD transition from one cell type to another? Do some release and others take up substances?

Regarding the MD-JGA feedback control of renin secretion, it was observed that the dense network minor processes of MD cells could extend their arms up to 30 µm in length toward individual JG renin cells of the afferent arterioles and efferent arterioles. The relative density of these extensions to the afferent versus efferent arterioles would have been interesting to determine. Those extending to efferent cells are intriguing and would be of considerable interest to further explore. Could there actually be fine regulation and coordination by the MD of the pre- and postglomerular vascular resistance? Could cell-specific chemoattractants exist that invite these filaments to extend to different cells types within the JGA? Within the MD, what mediates the vesicular transport of the cargo within the maculopodia from an organelle to specific sites at the cell membrane? What signals the release and/or uptake of these molecules? The present observations and the methods applied in these experiments provide opportunities to identify and study the physiological mechanisms related to any number of pathways yet undiscovered that may contribute to the normal or pathophysiological regulation of TGF and renal function.

Goormaghtigh (2) could not have imagined the technologies that would be required to obtain such images of these MD cells. Clarity of structure as achieved in the present study will surely motivate others to now address the complex ways in which MD cells exploit their unique microstructural forms to regulate kidney function. The timing is perfect given the current state of development of single-cell transcriptomics and proteomics technologies (9). As shown in the present study, the ability to now isolate the MD by fluorescence-activated cell sorting enables the application of these powerful technologies to dissect the complex regulatory and cell-cell communication networks that drive MD cell identification and function. I expect that the 21st century dance of the MD has just begun, and I look forward to an exciting and deeper understanding of the functions of these remarkable cells of the kidney.

GRANTS

This work was supported by National Heart, Lung, and Blood Institute Grants R01HL137748, P01HL116264, and R01HL151587.

DISCLOSURES

No conflicts of interest, financial or otherwise, are declared by the author.

AUTHOR CONTRIBUTIONS

A.W.C. drafted manuscript; edited and revised manuscript; and approved final version of manuscript.