An Approach to the Structure and Function of the Glomerular

Journal of the American Society of Nephrology S65

An Approach to the Structure and Function of theGlomerular Mesangium1

Harrison Latta2

H. Latta, Department of Pathology and LaboratoryMedicine, University of California Medical Center,Los Angeles, CA

(J. Am. Soc. Nephrol. 1992; 2:S65-S73)

ABSTRACTThe mesangium of the glomerulus is a connectivetissue tree arising at the vascular pole of the glomer-ulus and supporting the glomerular capillaries. It is

partly covered by a basement membrane that fol-

lows the epithehial cells from the peripheral glomer-ular capillary wall over the supporting tissue. Thecapillary endothehium does not normally have a sep-

arate basement membrane. The endothelium hasfenestrations that open directly into the mesangiumand allow blood plasma and tracers to flow into themesangium. The fenestrations partially restrict (orsieve) particles over 405 A in mean length from entry.Tracers move in intercellular channels and are fil-tered and concentrated by the basement mem-

brane at the sides ofthe mesangium or by mesangialmatrix filaments in the channels between cells. Theirregular distributions of flow, matrix, and concentra-lions of tracers may account for irregular lobular

reactions in glomerular disease. Two main pathwaysof flow seem to be ( 1) through the basement mem-brane and between the epithelial foot processes toform part of the glomerular filtrate and (2) into the

efferent capillaries through their mesangial fenestra-

tions. Intrinsic mesangial cells can now be regardedas myofibroblasts associated with the production ofthe connective tissue matrix. These cells hold thebasement membrane to maintain the shape of theglomerular capillaries, they swell readily, and theycan constrict like smooth muscle cells with appropri-ate stimulation. These reactions may enable them tocontrol the flow of blood through the capillary net-work in glomerular disease. Mesangial cells can take

This article was prepared for a lecture presented at the Third Porno Nephrologyconference in Fiuggi. Italy.on Septernber 27. 1991.

2coffe$pondence to Dr. H. Lafla, Departrnent of Pathology and Laboratory

Medicine, University of california Medical center. Los Angeles, CA 90024-1732.

1046-6673/02 1 O-0S65$03.00/0Journal of the Arnerican Society of Nephrology

copyright © 1992 by the Arnerican Society of Nephrology

up large amounts of foreign material within 24 h.Intrinsic mesangial cells and monocytes can in-crease in numbers in disease. Cells in the polar cush-ion (Polkissen) react differently than do mesangialcells.

Key Words: Mesangial filtration, flow, fenestrations, basement

membrane, endocytosis, mesangial cells, myofibroblast

T his discussion will review the light and electronmicroscopy studies that describe the structure

of the mesangial region, its cellular and extracellularcomponents, its relationships to adjacent structures,some of its functions under physiologic conditions,and some of its reactions in disease. Figure 1 is adiagram of the main features of the mesangial region.Others have reviewed more recent studies of themesangium that elucidate biosynthetic activities andreactions to various substances including enzymes,hormones, cytokines, immune complexes, and me-diatons of inflammation (1 ,2).

The mesangium was considered by Zimmermann

(3,4) to be a connective tissue tree that arises fromthe vascular pole of the glomerulus, extends into thelobules, and supports the glomerular capillaries. Hedescribed and illustrated it in 1929 (3), but he did notuse the term mesangium until 1933 (4). Zimmen-mann’s observations were detailed and remarkable.However, they were not accepted by some investiga-tons who used light microscopy and early electronmicroscopy techniques. In this period, the terms ap-plied to this region included intercapillary, axial, andcentrolobular. The development of improved electronmicroscopy techniques enabled confirmation of Zim-menmann’s observations (5). A number of studiesfrom this laboratory have dealt with the mesangium(5-15), and we have reviewed it briefly in 1973 (11)and 1985 (15). Other studies and reviews have pre-sented further information with electron microscopyand other techniques (1,16-21).

BASEMENT MEMBRANE

The basement membrane (basal lamina) of theglomerulus appears early in the embryologic devel-opment of the glomerulus in association with theepithelial cells that form the renal vesicle. It retainsthis association with epithelial cells as the glomerularcapillaries and the mesangium develop. In the matureanimal, the basement membrane comes to lie over

Structure and Function of Mesangium

s66 Volume 2 - Supplement 2’ 1992

Figure 1 . Diagram of the mesangiah region. The mesangium is shown at the center of capillaries forming a glomeruhar lobule.Portions of the mesangium are covered by the lamina densa (LD) and lamina rara externa (IRE), which follow the footprocesses of the epithehium (Ep) from the peripheral glomerular capillary wall over the mesangium. The capillary basementmembrane (BM) (basal hamina) is composed of the IRE, the ID, and the lamina rara interna (LRI). Where it joins themesangium, the IRI is continuous with the mesangial matrix (MM). The MM varies in density and is irregularly distributed inthe mesangium. It is often more dense where it anchors processes of mesangiah cells (M) to the ID of the infohded BM. TheMM is usually less dense beneath central portions of endothelial cells (En) lying over the mesangium. Endothelial cells donot normally have a separate basement membrane. The central portions of endothehial cells often have fenestrations (F)that are similar in size to but fewer in number than fenestrations in the peripheral capillary wall. The fenestrations allow bloodplasma and particulate tracers to directly enter into the mesangium and move in intercellular channels (IC) where there islittle or no matrix. One pathway of flow out of the mesangium seems to be through the LD and between foot processes overthe mesangium, contributing to the glomerular filtrate (arrow to the right). A second pathway of flow out of the mesangiumseems to be through the intercellular channels and endothelial fenestrations of efferent capillaries into their lumens (arrowto the left). A third but small pathway of flow seems to be through the hilus into the polar cushion (Polkissen) of the JGA.Adjacent foot processes come from different cells, hence the number covered by epithehial cytoplasm is always odd.Modified from Latta et a!. (5) with permission.

the glomerular capillary network and the mesangium glomeruli in the renal cortex of human fetuses underlike a folded sheet (11). Different stages of this proc- 35 wk of gestation. There is no separate basementess can be seen by light microscopy in developing membrane associated with glomerular endothelial

Latta

Journal of the American Society of Nephrology $67

cells either in the peripheral capillary wall or over

the mesangium in normal animals (5, 1 1), but onemay develop in glomerular disease as in human mes-

angiocapillary (membranoproliferative) glomerulo-nephnitis (20,22). Although the term, basement mem-

brane, seems well established in the literature onglomerular capillaries and refers to the lamina raraexterna, lamina densa, and lamina rara interna, his-

tologists now prefer the term basal lamina (23).

MESANGIAL MATRIX

The mesangial matrix was described by Zimmer-mann (4) as connective tissue with a fibrous networkthat is denser and stains more strongly with Azan

than the connective tissue outside the glomerulus.Other stains indicate that it has a large polysaccha-ride or proteoglycan content ( 1 1 ,24). Electron micros-copy studies showed a large component of filaments1 2 to 25 A wide (24) and a few banded collagen fibers(7). Banded collagen may become much more promi-

nent in glomerular disease (6). Fibronectin and lam-mm have been described (1). Ruthenium red staining

indicates a negative (polyanionic) charge on the mes-

angial fibnils (24), which are much more closelypacked than the relatively large, negatively chargedlattice-like network described by Kanwar and Far-quhar (25). Deposits in the mesangium representing

abnormal materials or increased amounts of normalcomponents are recognized in diabetic glomeruloscle-rosis, other types of glomerulosclerosis, amyloidosis,

focal and diffuse glomerulonephnitis, immune com-plex disease, etc. (22).

Mesangial cells and fibers seem to hold the glomer-

ular basement membrane against the blood pressurein the capillaries (1 1 , 1 5). This has been studied in

some detail by Kriz et at. (26). Loss of this holdingforce would allow the basement membrane to pop

out and form microaneurysms, which appear in mes-angiolysis with necrosis of mesangial cells, diabetes,and other glomerular diseases (11,20).

MESANGIAL PLASMA FLOW

The concept of plasma flow in the mesangiumseems to be a key to understanding the structure and

functions of the mesangium.

Entry into Mesangium

The rapid entry of small tracers into the mesan-gium within a few minutes after i.v. injection indi-cates a considerable flow of blood plasma into thisregion (5). Entry into the mesangium seems to occur

through fenestrations of the same size and shape asthose in the peripheral glomerular capillary wall (15).The mean width is 376 A in the rat. Analysis of themeasurements of the fenestrations and asymmetricthorium dioxide (ThO2) particles used as tracers in-

dicated that particles with a mean length of 405 A or

more were partially restricted from entry. This con-stitutes sieving, which partially excludes larger sub-stances. However, the presence of some large fenes-trations allowed the entry of a few thorium dioxideor carbon particles over 600 A in size. The entry of

tracers into the mesangium is favored by the absenceof a basement membrane under the endothehium over

the mesangium. It is also favored by the lesser densityof the mesangial matrix lying under the endothehium(15).

Slower entry into the mesangium of aggregates oftracers could occur in hours or days with envelop-ment by the endothelium. Envelopment was sug-

gested by clumps of Th02 particles in capillary lu-mens that appear to be incorporated into the mes-

angium by thin layers of endothelium moving overthem (8). This would be analogous to the movement

of the endothelium over the surface of a mural throm-bus. This mechanism could also explain incorpora-tion into the mesangium of masses of carbon (27),possibly fibrin, capillary thrombi, larger aggregatesof protein or immune complexes, etc.

Mesangial cells can project into the capillary lumen

(4,9, 1 0, 1 3, 1 6). Zimmermann (4) suggested that theymight do this for nourishment. No evidence has been

found by electron microscopy that these processestake up tracers, and our impression is that theseswollen processes are a response to ischemia or dis-ease conditions (9, 1 0, 13).

Substances may be carried into the mesangium bycells migrating from the capillaries. After the injec-tion of foreign organisms, which adhere to the gb-merular endothebium, monocytes and neutrophils

lodged in the glomeruli and began phagocytosis in 10mm (28,29). Monocytes containing ferritin began toinfiltrate the mesangium in 4 h (30). Such inflam-matory cells may contribute to the gbomerular hyper-cellularity apparent by light microscopy.

Other mechanisms of incorporation have been sug-gested (such as by passage along the peripheral gb-merular capillary wall, by transport across endothe-

hal cytoplasm, or by mesangial cells sweeping theperipheral capillary wall), but there is little evidencefor them (15,19,20).

Factors favoring mesangial localization have beendiscussed by Michael et at. (19), Sterzeb et at. (20),and Men#{232}et at. (1) and include high concentrationsin the blood, decreased activity of the systemic phag-

ocytic system, corticosteroids, barge size, positivecharge, and high avidity of immune complexes

(31,32).

Pathways of Flow and Filtration in theMesangium

If plasma flows into the mesangium rapidly, fluidmust flow out rapidly (5,11 ,15). Pathways of flow


$68 Volume 2’ Supplement 2’ 1992

have been suggested by the appearance and accu-mubation of tracers and immune complexes. Tracersentering in a few minutes move in channels betweenmesangial cells where the matrix is less concen-trated. With time, tracers accumulate in two placesin the mesangium. One is under the dense layer of

the basement membrane, which follows the epithe-hum over the mesangium. Tracers (hemoglobin, fer-ritin, dextrans, plasma proteins) and immune com-

plexes can concentrate in this paramesangiab area(see references 1 5, 33, and 34). This concentrationsuggests that these substances are being filtered outby the basement membrane and that the fluid withthem flows on through the basement membrane,between foot processes, and contributes to the gb-merular filtrate (34). This forms one pathway of flow

through the mesangium. The behavior of thesetracers suggests that the permeability of the base-ment membrane and foot processes over the mesan-gium is similar to that over the peripheral glomerular

capillary wall (34).A second pathway of flow through the mesangium

was suggested by the concentration of tracers onportions of the mesangiab matrix in intercellular

channels (8). This pathway would conduct flow fromthe afferent vessels in the gbomerular capillary net-work to the efferent capillaries. A hydrodynamic

analogy is suggested by the different directions andrates of flow a river can take passing through a deltaregion. Another consideration favoring flow throughthe mesangium from afferent to efferent capillariesis that, without such flow, plasma proteins filteredout at the basement membrane over the mesangiumwould become unduly concentrated. They would haveto be carried back into the capillary circulation forthe mesangium to maintain its normal state. Theaccumulation of tracers on portions of the mesangiabmatrix represents another type of filtration in addi-tion to filtration at the basement membrane men-tioned above. Concentration on the matrix would givetracers more time to be taken up by cells in themesangium. The irregular distribution of the mes-angial matrix and variations in flow account for thedifferent numbers of tracers found in different gb-merubi and in different lobules of the same glomeru-bus. It may also explain the more prominent changesin some lobules in focal gbomerubonephritis and other

gbomerular diseases. The flow of blood plasma

through the mesangium would also enable the mes-angiab cells to react to substances carried in theplasma, such as hormones, cytokines, immune com-plexes, etc.

A third pathway of flow to the hilus of the glomer-ulus and the juxtagbomerular apparatus (JGA) hasbeen suggested (8), but the small numbers of tracersfound in the JGA in 24 h (8) and the delayed timesof concentrations in the JGA (several hours for irondextran 1351 and ferritin [361 and 4 wk for carbon

[27]) indicate that flow in this pathway would have

to be quite small (1 5). Most of the hilus of a gbomer-ulus is occupied by the two arterioles, which leaves

the remaining area of contact between the mesan-gium and the JGA so small that it must act as abottleneck limiting the passage of tracers, whetherextraceblubar or intracellular. Moreover, the cells ofthe JGA are packed closely together, and where thereis interstitial matrix between them, it is quite dense,as if to partially seal off the JGA from extraneous

influences. This enables the components of the JGAto respond better to physiologic stimuli. A possible

alternative means by which tracers could enter theJGA is from the adjacent interstitial tissue and yes-sels, but the evidence for this is less impressive thanthat for slow passage from the mesangium.

Other fates for tracers in the mesangium have beensuggested (1 5, 1 9). These include intracellular andextracellular digestion, regurgitation into the blood,transportation by cell movement or cell-to-cell pas-

sage, passage through the basement membrane byliquefaction of its thixotropic gel structure, or disso-lution of immune complexes by excess antigen. In-digestible tracers may remain for months.

CELL TYPES IN THE MESANGIUM

It is generally agreed that the normal mesangiumcontains at beast two cell types, the intrinsic mesan-gial cell and the monocyte/macrophage. Each ofthese cell types performs multiple functions.

The Intrinsic Mesangial Cell

The Fibroblastic Nature of Intrinsic Mesangial

Cells. Zimmermann (4) considered mesangial cells tobe fibrobbasts because they bay in a fibrous matrixthat stained like connective tissue. Electron micros-copy support for the connective tissue nature of themesangium was obtained in this laboratory whenbanded collagen fibers were observed in experimen-tab (6) and human (37) gbomerular disease. Occasionalsmall bundles of banded collagen fibers were alsofound in the normal rat mesangium (7). Increaseddeposition of the mesangiab matrix (and sometimes

collagen) can occur in diabetes and different types ofgbomerulonephritis and gbomerulosclerosis. Amyloiddeposits occur in various connective tissue sites inthe body, including the mesangium (22).

The Smooth Muscle Nature of Mesangial Cells.Our early electron microscopy studies showed bun-dles of microfibaments and attachment bodies inmesangiab cells like those in arterial smooth muscle

(8). Others demonstrated actomyosin activity in mes-angiab cells by immunofluorescence (38,39). Mesan-giab cells localize angiotensin II in situ (40) andcontract in tissue culture when stimulated with an-

Latta


giotensin (41). Mesangial cells swell easily and lift

the overlying endothebium or project through it intothe capillary lumen (9, 1 0, 1 3). Zimmermann de-scribed and illustrated this process by bight micros-copy in 1933 (4). Similar changes have been observedin the smooth muscle of the aortic wall (42). Contrac-

tion and swelling could control blood flow through

the gbomerular capillary network. Endothebial swell-ing may aid in this process. Considerable evidencesuggests a contribution of mesangial cells to the reg-ulation of glomerubar filtration (2 1 ,43). Swelling is

especially prominent after the cessation of blood flow(9) or after ischemia for 1 5 to 60 mm (1 3). Suchswelling could account for the loss of gbomerularperfusion and filtration in shock, toxemia of preg-nancy, or glomerulonephritis.

The concept of the myofibroblast developed fromstudies on fibrobbasts in wound healing (44). Darbyet at. (45) have described the development and regres-sion of smooth muscle actin in granulation tissuefibrobbasts. This is an example of the plasticity ofphenotype in the expression of different features inconnective tissue cells under various stimuli (46). Forexample, smooth muscle cells can acquire fibro-

blastic features ( 1 ) in the uterus (and possibly ac-count for the development of uterine fibromyomas)or (2) when migrating from the media to the intimain atherosclerosis. It has even been suggested thatthe macrophage can express muscle proteins (46).The fibrous histiocytoma may be derived from a mes-enchymab cell capable of mubtidirectional differentia-tion to express fibrous and histiocytic features (47).A contractile system would help mesangial cells holdthe “infolded” basement membrane against the bloodpressure in the capillaries (8,26).

Mesangiab cells have been considered to be pen-

cytes by several investigators. However, pericytes areusually described as having bong branching proc-esses that extend around capillary endothelium (23).Zimmermann (4), having earlier studied pericytesaround capillaries in other tissues, was particularlyinterested in searching for them in the kidney. In onepreparation of a cat kidney, he found five cells withsmall processes, apparently grasping gbomerular cap-illaries, and concluded that they were “perhaps truepericytes.” No one using electron microscopy seemsto have described appropriate pericyte processes out-side of the endothehium of the peripheral gbomerularcapillary wall in the normal animal. Therefore, itdoes not seem that normal mesangial cells ordinarilyfulfill the histologic criteria for pericytes. Mesangialprocesses may appear in the capillary walls of mes-angiocapillary gbomerubonephritis, but these wallsare greatly altered.

Endocytic Function of Mesangial Cells. In 1960,we (5) described the special features of the mesangiab

region and its cells with newer electron microscopytechniques. In 1961, before the concept of the mes-

angial region was generally accepted, Farquhar et at.

(48) reported that injected ferritin accumulated onthe luminal side of the gbomerular basement mem-brane, especially in axial regions, and was taken upby the deep cells, which were considered to be endo-thelial. In 1962, we (8) demonstrated that largeamounts of thorium dioxide particles were engulfedby centrobobular mesangial cells within 24 h. Adja-cent capillary endothelial cells showed little or nouptake. Later, in 1962, Farquhar and Palade (16)wrote an article agreeing that the axial region hadspecial characteristics and showing that the deepcells accumulated more of the tracers than the su-perficial endothebium. They concluded that the deepcells could be considered a distinctive “third” celltype.

A number of foreign substances have been foundin cells in the mesangium ( 1 , 1 9,20), but the evidenceneeds to be reevaluated in light of two questions:were the cells that were involved intrinsic mesangialcells or macrophages, and what type of endocytosiswas stimulated by each substance? The uptake ofmacromolecular substances into a membrane-him-ited organelle in a living cell is now called endocy-

tosis and is divided into three types: ( 1 ) phagocytosisof large particles, (2) macropinocytosis, and (3)

coated pit-mediated endocytosis, which accounts forthe majority of uptake of fluid-phase and receptor-bound materials by most cells, with the formation ofreceptosomes (or endosomes) (49). These matters willbe considered again below when the gbomerular mon-ocyte/macrophage is discussed.

The Multipotential Nature of Mesangial and

Other Connective Tissue Cells. While the conceptsof different functions for the mesangial cells havebeen developing, evidence has been accumulatingthat other connective cell types may show more thanone kind of activity. Myofibroblasts in wound healingand in uterine myofibromas and fibrous histiocyto-mas have been mentioned above. Smooth musclecells have been associated with the production ofcollagen and elastin in normal arteries (50). Smoothmuscle cells from arteries can ingest yeast cells andlatex spheres (5 1 ). They may acquire Fc and C3 recep-tors and the ability to ingest immunogbobulin G-coated red blood cells (52). In atherosclerosis, theymigrate into the intima and produce collagen (53).Considerable evidence has led to the concept of thearterial smooth muscle cell as a multifunctional mes-

enchymal cell (54). Another potential is shown byconnective tissue cells in muscle. When stimulatedby trauma (and bone morphogenetic protein), theycan produce cartilage and bone in a process cabledheterotropic bone formation (formerly myositis ossi-ficans) (55). Hence, mesangial cells seem to be pluri-potential and are bike other connective tissue cells inbeing able to exhibit different functions upon appro-priate stimulation.

Structure and Function of Mesanglum

$70 Volume 2’ Supplement 2’ 1992

The Monocyte/Macrophage in theMesangium

Studies have shown that normal rat mesangia con-tam phagocytic cells different from intrinsic mesan-gial cells and derived from the bone marrow (56).Phagocytosis was considered as the uptake of latexbeads, bacteria, aggregated gamma globulin, or other

large particles. There were 8 to 1 8 mesangial phago-

cytes per glomerulus, and all were found to express

the leukocyte common antigen. About half of thesebear Ia determinants. With an average of 28 1 mes-

angial cells in a rat glomerulus (57), the phagocyteswould represent between 3 and 7% (or 1 in 35 to 1 In

1 5) of the normal mesangial population. The kineticsof a slow traffic of monocytes from the bone marrowand the blood through the mesangium were consid-

ered (56).According to the above measurements, 93 to 97%

of the mesangial population is composed of intrinsic

mesangial cells. The amount of uptake of variousforeign tracers in the mesangium within a few mm-utes or hours of injection makes it seem likely that

the intrinsic mesangial cells are involved (8,19,20);however, which type of endocytosis they use needsto be determined. The type probably depends on thetracer.

Monocytes can localize in glomeruli within minutes(29) or hours (28) of an injection of pathogenic orga-nisms and can contribute to the hypercellularity ofthe developing focal gbomerulonephritis. With the

stimulus of foreign tracers, monocytes seem to mi-grate into the mesangium (30,36). After polyvinyl

alcohol injections, Ia-positive mesangial monocytes-macrophages showed no uptake at 24 h, whereasthere was some in stellate mesangial cells (58). Con-siderable amounts were found in the monocytes-macrophages after 3 days. The number of mono-cytes-macrophages also increased between 1 and 3days (36).

Hence, it seems that the intrinsic mesangial cellsaccount for much of the endocytosis in the mesan-gium within the first 24 h, and after that, endocytosis

by monocytes-macrophages becomes prominent. It

may be that after encountering some substances, theintrinsic mesangial cells stimulate the recruitmentof monocytes into the mesangium from the circulat-

ing blood. Macrophages can be prominent in glomer-ular disease and may serve as the principal effectorcell in some types of glomerulonephritis (2).

OTHER FUNCTIONS AND REACTIONS IN THEMESANGIUM

The proliferation of mesangial cells has been meas-ured in diabetic glomerulosclerosis (59), compensa-

tory hypertrophy (57), and immunologically mediatedglomerulonephritis (2). Lobular or diffuse prolifera-

tions of mesangial cells have been observed in other

types of glomerular disease involving glomeruloneph-ritis or glomerulosclerosis (22,60). Increased mesan-

gial matrix is evident in different types of glomeru-losclerosis. Deposits of amyloid, immune complexes,and other substances may occur. There is evidencefor the mediation of glomerular disease by neutro-phils, monocytes and macrophages, lymphocytes,

and platelets (2).

In some types of glomerulonephritis, particularlymembranoproliferative (or mesangiocapillary) gb-merulonephritis, endotheliab cells may be lifted awayfrom the capillary basement membrane and the mes-angiab cells may extend out in the capillary wall in aprocess cabled mesangial interposition (22,60). A sec-ond basement membrane forms beneath the dis-placed endothelium, producing the double-contoureffect seen by light microscopy.

THE RELATIONSHIP OF THE MESANGIUM TOTHE JGA

Granular epithebioid cells had been found in thewalls of afferent arterioles by Ruyter (6 1) and Ober-ling (62) when Goormaghtigh (63) and Zimmermann(4) described in more detail juxtaglomerular struc-tures containing modified smooth muscle cells. Zim-mermann cabled a group of cells arising from the

afferent arteriole outside of the vascular pole the

Potkissen, which means “polar cushion” in English.The cells in it made close contact with the macuba

densa of the distal tubule. The cells in the polarcushion were closely packed and were separated onlyby a thin fibrous network, which was much lessprominent than the more abundant and darker stain-ing connective tissue fibers in the adjacent mesan-

glum, which also contains fewer cells. The nuclei inthe polar cushion may be elongated and placed par-albel to each other. Striking differences at the tran-sition from the polar cushion to the mesangium werewell illustrated in several figures (4). Electron ml-

croscopy studies by us (8, 1 1 ,64-68) and others(69,70) have confirmed and extended the light mi-croscopy observations.

Although the structures of the polar cushion and

the mesangium appear quite different, the cells ofthe polar cushion and the mesangial cells seem to bemodified smooth muscle cells and functional similar-ities have long been postulated (8,69). However, anumber of studies have shown functional differ-

ences. After adrenalectomy, rat juxtagbomerular cellshad a striking increase in granularity, but no suchgranularity was found in mesangial cells (67). Afterthe injection of tracers, penetration and accumula-

tion in the mesangial region was rapid, but it wasdelayed and smaller in amount in the polar cushion

(8,27,35,36). The delayed appearance in the polar

Latta

Journal of the American Society of Nephrology $71

cushion suggested that tracers could pass into it from

the mesangium. In human (71) and experimental (72)unilateral renovascular hypertension, increases insize and granularity were found in the polar cushionsbut not in the mesangia. With excessive NaCl intakein rats, proliferation of agranular cells was found inpolar cushions but not in mesangia (73). Hence, afterconsideration of the anatomic and functional evi-dence presented above, it seems better to maintain a

distinction between the mesangium and the polarcushion.

The different names that have been used in de-scribing different parts of the JGA have been con-fusing. The JGA is composed of the polar cushion,

the afferent and efferent arterioles, and the maculadensa (1 1 ,67,69,70). It is well innervated (67,74). Thepolar cushion (Potktssen, lads, juxtagbomerular cell

mass, extragbomerular mesangium) contains mostly

agranular cells (the pseudomeissnerian cells de-scribed by Goormaghtigh) and a few granular cells(68). Most of the granular cells (epithelioid or myo-epithelial cells) are found in the walls of the afferent

and efferent arterioles (68) and are thought to contain

renin (11,70).

ACKNOWLEDGMENTS

Original research from this laboratory was supported by grants from

the National Institutes of Health, United States Public Health Serv-

ice. I have been fortunate in having very capable colleagues and

research associates, whose names appear in each research paper

published. The diagram used for Figure 1 was redrawn by the very

capable medical artist, Gwynne Gloege. It was drawn in ink to avoid

halftone reproduction.

REFERENCES

1 . Men#{232}P, Slinonson MS. Dunn MJ: Physiologyof the mesangial cell. Physiol Rev 1 989;69:1347-1420.

2. Couser WG: Mediation of immune glomerularinjury. J Am Soc Nephrol 1 990; 1 : 1 3-�29.

3. Zlmmermann KW: Uber den Bau des Gbomeru-bus der menschlichen Niere. Z Mikrosk AnatForsch 1929; 18:520-552.

4. Zlmmermann KW: Uber den Bau des Glomeru-lus der Saugerniere. Weitere Mitteilungen. ZMikrosk Anat Forsch 1 933;32: 176-278.

5. Latta H, Maunsbach AB, Madden SC: The cen-trolobular region of the renal gbomerulus studiedby electron microscopy. J Ultrastruct Res 1960;4:455-472.

6. Bencosme SA, Stone RS, Latta H, Madden SC:Acute reactions with collagen production inrenal glomeruli of rats as studied electron micro-scopically. J Ultrastruct Res 1 959;3: 171-185.

7. Latta J, Cook ML: Collagen in normal rat gb-meruli. J Ultrastruct Res 1961 ;5:364-373.

8. Latta H, Maunsbach AB, Cook ML: Relations ofthe centrolobular region of the glomerulus to thejuxtagbomerular apparatus. J Ultrastruct Res1 962;6:562-578.

9. Cook ML, Osvaldo L, Jackson JD, Latta H:

Changes in renal gbomeruli during autolysis.Electron microscopic observations. Lab Invest1965; 14:623-634.

10. Johnston WI!, Latta H, Osvaldo L: Variationsin gbomerular ultrastructure in rat kidneysfixed by perfusion. J Ultrastruct Res 1973;45:149-167.

1 1 . Latta H. Ultrastructure of the gbomerulus andjuxtagbomerular apparatus. In: Orboff J, BerlinerRW, Geiger SR. eds. Handbook of Physiology.Renal Physiology. Washington, DC: AmericanPhysiological Society; 1973:1-29.

1 2. Schneeberger EF, Collins AB, Latta H, Mc-Cluskey RT: Diminished gbomerular accumula-tion of colloidal carbon in autobogous immunecomplex nephritis. Lab Invest 1 977;37:9- 19.

13. Johnston WI!, Latta H: Gbomerular mesangialand endothelial cell swelling following tempo-rary renal ischemia and its role in the no-reflowphenomenon. Am J Pathol 1977;89:153-166.

14. Latta, H: The multipotential nature of mesangialcells. Biol Cell 1 980;39:245-248.

1 5. Latta H, Fligiel 5: Mesangial fenestrations, siev-ing, filtration, and flow. Lab Invest 1985;52:591-598.

1 6. Farquhar M, Palade GE: Functional evidence forthe existence of a third cell type in the renalgbomerulus. Phagocytosis of filtration residuesby a distinctive “third” cell. J Cell Biol 1 962; 13:55-87.

17. Michielsen P, Creemers F. The structure andfunction of the gbomerular mesangium. In: Dab-ton AJ, Haguenau F, eds. Ultrastructure of theKidney, Vol. 2. New York: Academic Press; 1967:57-72.

18. Simon GT, Chatelanat F. Ultrastructure of thenormal and pathological gbomerulus. In: RouillerC, Muller AF, eds. The Kidney, Vol. 1 . New York:Academic Press; 1969:262-349.

1 9. Michael AF, Keane WF, Raij L, Vernier RL,Mauer SM: The gbomerular mesangium. KidneyInt 1980;17:141-154.

20. Sterzel RB, Lovett DH, Stein LW, KashgarianM: The mesangium and glomerubonephritis. KlinWochenschr 1 982;60: 1077-1094.

2 1 . Schlondorff D: The glomerular mesangiab cell:An expanding role for a specialized pericyte. FA-SEB J 1987;I:272-281.

22. Heptinstall RH: Pathology of the Kidney. 3rd Ed.Boston: Little Brown; 1983:1695.

23. Fawcett DW. A Texthook of Histology. 1 ith Ed.Philadelphia: WB Saunders; 1986:1017.

24. Latta H, Johnston WH, Stanley TM: Siabogly-coproteins and filtration barriers in the gbomer-ular capillary wall. J Ultrastruct Res 1975;51:354-376.

25. Kanwar YS, Farquliar MG: Anionic sites in thegbomerular basement membrane. In vivo and invitro localization to the laminae rarae by cationicprobes. J Cell Biol 1979;81:137-153.

26. Kriz W, Elger M, Lemley KV, Sakal T: Mesangialcell-glomerular basement membrane connec-tions counteract gbomerular capillary and mes-angial expansion. Am J Nephrol 1990;1 0(suppl):4- 13.

27. Elema JD, Hoyer JR, Vernier RI: The gbomer-ular mesangium: uptake and transport oT intra-venously injected colloidal carbon in rats. Kidneymt 1976;9:395-406.

28. Johnston WH, Latta H: Acute focal gbomerubo-


S72 Volume 2- Supplement 2’ 1992

nephritis in the rabbit induced by injection ofSaccharomyces cerevtsiae: An electron micro-scopic study. Lab Invest 1972;26:741-754.

29. Johnston WH, Latta H: Acute hematogenouspyebonephritis in the rabbit: an electron micro-scopic study of Escherichia coli localization andearly acute inflammation. Lab Invest 1978;38:439-446.

30. Cattell V, Gaskin de Urdaneta A, Arlidge 5,Collar JE, Roberts A, Smith J: Uptake andclearance of ferritin by the gbomerular mesan-gium. I. Phagocytosis by mesangial cells andblood monocytes. Lab Invest 1982;47:296-303.

3 1 . Germuth FG Jr, Rodriguez E, Lorelle CA,Trump El, Milano L, Wise 0: Passive immunecomplex gbomerubonephritis in mice: Models forvarious lesions found in human disease. I. Highavidity complexes and mesangiopathic glomer-ulonephritis. Lab Invest 1979;4l :360-365.

32. Iskandar 55, Jenette JC: Influence of antibodyavidity on gbomerular immune complex bocaliza-tion. Am J Pathol 1983;l 12:155-159.

33. Latta H: The gbomerular capillary wall. J. Ultra-struct Res 1970;32:526-544.

34. Latta H: Filtration barriers in the gbomerularcapillary wall. In: Maunsbach AB, Olsen TS,Christensen El, eds. Functional Ultrastructureof the Kidney. London: Academic Press; 1980:1-29.

35. Leiper JM, Thomson D, MacDonald MK: Uptakeand transport of Imposil by the gbomerular mes-an�ium in the mouse. Lab Invest 1977;37:52�-533.

36. Sterzel RB, Perfetto M, Biemesderfer D, Kash-garian M: Disposal of ferritin in the gbomerularmesangium of rats. Kidney mt l983;23:480-488.

37. Jackson JD, Smith FG, Litman NN, Yuile CL,Latta H: The Fanconi syndrome with cystinosis.Electron microscopy of renal biopsy specimensfrom five patients. Am J Med 1962;33:893-910.

38. Becker CG: Demonstration of actinomyosin inmesangial cells of the renal gbomerulus. Am JPathol l972;66:97-1 10.

39. Scheinman JI, Fish AJ, Matas AJ, Michael AF:The immunohistopathobogy of gbomerular anti-gens. II. The gbomerular basement membrane,actinomyosin, and fibrobbast surface antigens innormal, diseased, and transplanted human kid-neys. Am J Pathol 1978;90:71-84.

40. Osborne MJ, Droz B, Meyer P. Morel F: Angio-tensin II: Renal localization in gbomerular mes-angial cells by autoradiography. Kidney Int1 975;8:245-254.

41. Ausiello DA, Kreisberg JI, Roy C, KarnovskyMJ: Contraction of cultured rat gbomerular cellsof apparent mesangiab origin after stimulationwith angiotensin II and arginine vasopressin. JClin Invest 1980;65:754-5S0.

42. Stetz EM, Majno G, Joris I: Cellular pathologyof the rat aorta. Virchows Arch A Pathol AnatHistob 1 979;383: 135-148.

43. Pfeilschifter J: Cross-talk between transmem-brane signalling systems: A prerequisite for thedelicate regulation of glomerular haemody-namics by mesangiab cells. Eur J Clin Invest1989; 19:347-361.

44. Majno G: The story of the myofibroblasts. Am JSurg Pathol 1979;3:535-542.

45. Darby I, Skalli 0, Gabbiani G: a-Smooth muscleactin is transiently expressed by myofibroblastsduring experimental wound healing. Lab Invest1990;63:2 1-29.

46. Gown AM: The mysteries of the myofibroblast(partially) unmasked. Lab Invest 1 990;63: 1-3.

47. Nemes Z, Thom#{225}zy V: Factor XIIIa and the clas-sic histiocytic markers in malignant fibrous his-tiocytoma. A comparative immunohistochemicalstudy. Hum Pathol 1988;19:822-829.

48. Farquhar MG, Wissig SL, Palade GE: Glomeru-bar permeability. I. Ferritin transfer across thenormal glomerular capillary wall. J Exp Med1961;1 1�3:47-56.

49. Pastan I, Willingham MC. The pathway of en-docytosis. In: Pastan I, Willingham MC, eds. En-docytosis. New York: Plenum; 1985:1-44.

50. Burke JM, Ross R: Synthesis of connective tis-sue macromolecules by smooth muscle. Int RevConnect Tissue Res 1979;8:l 19-157.

5 1 . Garfield RE, Chacko S, Blose 5: Phagocytosisby muscle cells. Lab Invest 1975;33:418-427.

52. Fowler 5, Shio H, Haley NJ: Characterization oflipid-laden aortic cells from cholesterol-fed rab-bit. IV. Investigation of macrophage-like prop-erties of aortic cell populations. Lab Invest 1979;41:372-378.

53. Ross R. The pathogenesis of atherosclerosis. In:Braunwald E, ed. Heart Disease. A Textbook ofCardiovascular Medicine. 3rd ed. Philadelphia:WB Saunders; 1988:1135-1152.

54. Campbell GR, Campbell JH, Manderson JA,Horngan 5, Rennick RE: Arterial smooth mus-cle. A mubtifunctional mesenchymab cell. ArchPathol Lab Med 1988;1 12:977-986.

55. Urist MR: Bone morphogenetic protein, bone re-generation, heterotropic ossification and thebone-bone marrow consortium. In: Peck WA, ed.Bone and Mineral Research. Vol. 6. Amsterdam:Elsevier; 1989:57-112.

56. Schreiner GF, Unanue ER: Origin of the ratmesangiab phagocyte and its expression of theleukocyte common antigen. Lab Invest 1984;51:515-523.

57. Olivetti G, Anversa P, Meissari M, Loud A:Morphometry of the renal corpuscle during post-natal growth and compensatory hypertrophy.Kidney Int 1980; 17:438-454.

58. Seiler MW, Hoyer JR. Sterzel RB: Role of mac-rophages in the gbomerular mesangiab uptake ofpolyvinyl alcohol in rats. Lab Invest 1983;49:26-37.

59. lidaka K, McCoy J, Kimmelstiel P: The glomer-ular mesangium. A quantitative analysis. LabInvest 1968; 19:573-579.

60. Spargo BH, Seymour AE, Ord#{244}#{241}ezNG: RenalBiopsy Pathology with Diagnostic and Therapeu-tic Implications. New York: Wiley; 1980:469.

61. Ruyter JHC: Uber elnen merkwUrdigen Absch-nitt der Vasa afferentia in der M#{227}useniere. ZZellforsch Mikroskop Anat 1 925;2:242-248.

62. Oberling C: L’existence d’une housse neuro-musculaire au niveau des art#{232}res gbom#{233}rulairesde l’homme. Compt Rend 1927;184:1200-1202.

63. Goormaghtigh N: Les segments neuro-myo-ar-teriels juxtagbomerulaires du rein. Arch Blob(Liege) 1932;43:575-591.

64. Latta H, Maunsbach AB: The juxtagbomerularapparatus as studied electron microscopically. J

Latta


Ultrastruct Res 1 962;6:547-56 1.65. Barajas L, Latta H: A three-dimensional study

of the juxtagbomerular apparatus in the rat.Light and electron microscopic observations.Lab Invest 1963;12:257-269.

66. Barajas L, Latta H: The juxtaglomerular appa-ratus in adrenalectomized rats. Light and elec-tron microscopic observations. Lab Invest 1963;12:1046-1059.

67. Barajas L, Latta H: Structure of the juxtaglo-merular apparatus. Circ Res 1967;20-21(suppl II):15-28.

68. Latta H, Johnston WH: Granular and agranularcell counts in the juxtagbomerular apparatusesof rats with unilateral renovascular hyperten-sion. Lab Invest 1978;39:219-224.

69. Hyatt PY: The juxtagbomerular apparatus. In:Dalton AJ, Haguenau F, eds. Ultrastructure ofthe Kidney. New York: Academic Press; 1967:101-141.

70. Rouiller C, Orci L. The structure of the juxtagbo-merular complex. In: Rouiller C, Muller AF, eds.The Kidney. Vol. 4. New York: Academic Press;1971:1-80.

71. Barajas L, Lupu AN, Kaufman JJ, Latta H,Maxwell MH: The value of the renal biopsy inunilateral renovascular hypertension. Nephron1 967;4:23 1-237.

72. Latta H, White FD, Osvaldo L, Johnston WH:Unilateral renovascular hypertension in rats.Measurements of medullary granules, juxtaglo-merular granularity and cellularity, and areas ofadrenal zones. Lab Invest 1975;33:379-390.

73. Latta H, Lee TC: The effects of excessive sodiumchloride on the juxtaglomerular apparatus andblood pressure of uninephrectomized rats. LabInvest 1 983;49:99- 106.

74. Barajas L: The innervation of the juxtaglomer-ular apparatus. Lab Invest 1964; 13:916-930.

Documents

An Approach to the Structure and Function of the Glomerular