Portal de Periódicos da CAPES

Edmund Randerath (1899–1961): Experimental proof for the glomerular origin of proteinuria

1999; Elsevier BV; Volume: 56; Issue: 4 Linguagem: Inglês

10.1046/j.1523-1755.1999.00655.x

1523-1755

Rüdiger Waldherr, Eberhard Ritz,

Chronic Kidney Disease and Diabetes

Edmund Randerath (1899–1961): Experimental proof for the glomerular origin of proteinuria. A century ago, Edmund Randerath (1899–1961), who was one of the pioneers in nephrology that provided indirect experimental proof for the glomerular origin of proteinuria, was born. In the first decades of this century, the concept prevailed that “nephrosis” was a process of primary tubular cell degeneration. In contrast to prevailing opinion, he interpreted these changes to be the result of the uptake and storage of serum proteins after they had been filtered in the glomerulus. Edmund Randerath proved the glomerular origin of proteinuria by astute experiments in amphibia. In the salamander, an intraperitoneal injection of albumin provoked the supposedly “degenerative” changes of tubular epithelial cells in only those nephrons that drained the coelomic cavity and were devoid of glomeruli, but not in those nephrons that were closed and attached to glomeruli. This observation provided incontrovertible evidence that the presence of serum proteins in tubular fluid was a prerequisite for the development of the tubular epithelial cell changes typically seen in nephrotic patients. Edmund Randerath (1899–1961): Experimental proof for the glomerular origin of proteinuria. A century ago, Edmund Randerath (1899–1961), who was one of the pioneers in nephrology that provided indirect experimental proof for the glomerular origin of proteinuria, was born. In the first decades of this century, the concept prevailed that “nephrosis” was a process of primary tubular cell degeneration. In contrast to prevailing opinion, he interpreted these changes to be the result of the uptake and storage of serum proteins after they had been filtered in the glomerulus. Edmund Randerath proved the glomerular origin of proteinuria by astute experiments in amphibia. In the salamander, an intraperitoneal injection of albumin provoked the supposedly “degenerative” changes of tubular epithelial cells in only those nephrons that drained the coelomic cavity and were devoid of glomeruli, but not in those nephrons that were closed and attached to glomeruli. This observation provided incontrovertible evidence that the presence of serum proteins in tubular fluid was a prerequisite for the development of the tubular epithelial cell changes typically seen in nephrotic patients. Although an association between dropsy, renal disease, and albuminuria had been discussed by some predecessors, for example, William C. Wells and John Blackall [reviewed in[1.Cameron J.S.T. The nephrotic syndrome: A historical review.The Nephrotic Syndrome. edited by Cameron, JST, Glassock, RJ. Marcel Dekker, New York and Basel1988: 3-56Google Scholar]], it was the seminal observation of Richard Bright[2.Bright R. Reports of Medical Cases, Selected with a View of Illustrating the Symptoms and Cure of Diseases by a Reference to Morbid Anatomy. Longman, Rees, Orme, Brown, and Greene, London1827Google Scholar] which documented that the excretion of protein in the urine was linked to the presence of renal disease. Given today's insights into the physical and cellular basis of proteinuria[3.Bridges C.R. Myers B.D. Brenner B.M. Deen W.M. Glomerular charge alterations in human minimal change nephropathy.Kidney Int. 1982; 22: 677-684Abstract Full Text PDF PubMed Scopus (55) Google Scholar],[4.Hemesh O. Ross J.C. Deen W.M. Grant G.W. Myers B.D. Nature of the glomerular capillary injury in human membranous glomerulopathy.J Clin Invest. 1986; 77: 868-877Crossref PubMed Scopus (63) Google Scholar], it seems surprising that at the turn of the century a hot debate began about the precise site in the nephron where serum proteins entered the urinary space. In 1905, at the annual meeting of the German Society of Pathology in Meran, Müller proposed a scheme that distinguished between “nephritis” and “nephrosis”[5.Müller F. Morbus Brightii.Verh Dtsch Ges Pathol. 1905; 9: 64-99Google Scholar]. The former was thought to be of inflammatory and the latter of degenerative origin without considering the potential mechanisms involved. This distinction was carried one step further by Volhard and Fahr in their famous monograph “Die Brightsche Nierenkrankheit”[6.Volhard F. Fahr T.H. Die Brightsche Nierenkrankheit: Klinik, Pathologie und Atlas. Julius Springer, Berlin1914Crossref Google Scholar]. The observation of protein and lipid droplets in the proximal tubular epithelial cells of patients with nephrotic proteinuria, in the presence of morphologically intact glomeruli, led to the mistaken concept that proteinuria was the result of a primary degenerative process of tubular epithelial cells. The centenary of Edmund Randerath's birthday provides a welcome opportunity to describe the line of thoughts and the experiments that led to the concept of morphologically normal, but leaky glomeruli, which are the source of filtered serum protein, particularly in the condition that today we would call “minimal change nephropathy.” Edmund Randerath was born on March 18, 1899, in Düsseldorf Figure 1. After World War I in 1919, he began to study medicine in Marburg, subsequently in Düsseldorf and then in Cologne, where we wrote his doctoral thesis in 1923. In 1924, he joined the Department of Pathology of the Medical Academy in Düsseldorf under the directorship of Paul Hübschmann, an expert of world renown on the morphology of tuberculosis. The topic of the first publications of Edmund Randerath not surprisingly dealt with tuberculosis, particularly skeletal tuberculosis. In 1932, he wrote his professorial thesis on the topic of skeletal tuberculosis. Investigations to delineate the fate of thorium in the organism led him to study glomerular podocytes and capsular epithelial cells[7.Randerath E. Zur Frage des Glomerulothels.Beitr Pathol Anat. 1930; 85: 85-100Google Scholar],[8.Randerath E. Zur normalen und pathologischen Anatomie der Deckzellen des Nierenkörperchens.Z Zellforsch Mikroskop Anat. 1932; 15: 182-187Crossref Scopus (2) Google Scholar]. With the wisdom of hindsight, this was a fitting introduction into the topic of proteinuria, although his concept and terminology (“Glomerulothel”) did not stand the test of time. The renal changes in Bence Jones proteinuria raised his interest in the morphology of the kidney in proteinuric states[9.Randerath E. Pathologisch-anatomische und experimentelle Untersuchungen zur Frage der Nierenveränderungen bei Bence-Jonesscher Proteinurie.Z Klin Med. 1934; 127: 527-541Google Scholar]. These studies led him to question the then popular concepts concerning the site in the nephron that was responsible for protein excretion in nephrotic states[10.Randerath E. Über den Ort der Eiweiβausscheidung in der Niere bei nephrotischen Nierenkrankheiten, nebst Bemerkungen über den Begriff und die Einteilung der Nephrosen.Beitr Pathol Anat. 1935; 95: 403-430Google Scholar],[11.Randerath E. Pathologische Anatomie der Nephrosen.Zentralbl Inn Med. 1936; 57: 921-934Google Scholar]. To provide solid evidence for the concept of glomerular leakiness, previously proposed by others as the primary cause of proteinuria, he performed a series of studies (described in detail later in this article) that clearly documented that “hyalintropfige Eiweiβspeicherung” (hyaline droplet protein storage) and “trübe Schwellung” (hydropic swelling) of proximal tubular epithelial cells Figure 2 were caused by the uptake of serum proteins from tubular urine[12.Govaerts P. Cordier M.R. Contribution á l'étude clinique et anatomique de la néphrose lipoïdique.Bull Acad Royal Méd Belgique. 1928; 5: 510-548Google Scholar]. This line of research was interrupted by World War II, which Randerath spent as deputy director of the Pathological Institute of the Academy for Military Medicine in Berlin, Germany. His main publications at that time concerned trench nephritis[13.Randerath E. Die pathologische Anatomie der Kriegsnephritis.Arch Klin Med. 1948; 193: 119-156Google Scholar], tetanus, and tularemia. After the war, in 1947, Randerath became the head of the Department of Pathology in Göttingen. In 1949, he was appointed in Heidelberg, where he remained active as the head of the Institute of Pathology until his death in 1961. One of his major interests in this period concerned the development and specificity of intercapillary diabetic glomerulosclerosis[14.Randerath E. Zur Frage der intercapillären (diabetischen) Glomerulosklerose.Virchows Arch. 1952; 323: 483-523Crossref Scopus (1) Google Scholar].Figure 2Oil immersion, Altmann stain (A.G. = Altmann granules). Tubular degeneration with hyaline droplets (T) in the proximal tubular epithelial cells. Some of the degenerating cells still have intact nuclei. Others have lost their nuclei (with kind permission of Springer Verlag, Berlin, Germany)[5.Müller F. Morbus Brightii.Verh Dtsch Ges Pathol. 1905; 9: 64-99Google Scholar].View Large Image Figure ViewerDownload (PPT) Nephrosis was defined by Müller as a parenchymatous disorder of the kidney “which is either purely degenerative, or in which an inflammatory origin is not established beyond doubt”[5.Müller F. Morbus Brightii.Verh Dtsch Ges Pathol. 1905; 9: 64-99Google Scholar]. He had no definite opinion where proteins entered the urinary space and considered glomeruli as well as tubuli, including collecting ducts, as the sites of entry. Müller realized that in most cases, inflammatory renal diseases could not by clinical examination be distinguished from noninflammatory (degenerative) conditions. In 1913, Munk described a renal disease that he considered a distinctive entity, characterized by edema and marked albuminuria[15.Munk F. Klinische Diagnostik der degenerativen Nierenerkrankungen.Z Klin Med. 1913; 78: 1-52Google Scholar]. In the cases he described, it was associated with syphilis. In the urine of these patients, great amounts of birefringent lipid material were noted (lipoiduria), which was the focus of his interest (not albuminuria). He considered this finding as proof for the degenerative nature of this condition, which he accordingly designated “Lipoidnephrose” (lipoid nephrosis). Subsequently, in 1914, in their classic monograph, Volhard and Fahr further defined “nephrosis” as a degenerative disorder of the tubules, the origin of which was ascribed to poisoning by chemical substances, damage from bacterial products, or metabolic disturbances, respectively[6.Volhard F. Fahr T.H. Die Brightsche Nierenkrankheit: Klinik, Pathologie und Atlas. Julius Springer, Berlin1914Crossref Google Scholar]. As the histological hallmark of this condition, they pointed to the “trübe Schwellung” of tubular epithelial cells. Volhard and Fahr distinguished between cases of known origin, for example, amyloid nephrosis and “simple” or nonspecific nephrosis. In the evolution of this disorder, they distinguished four stages: (a) “trübe Schwellung,” (b) degenerative epithelial cell changes (hyaline and lipid droplets), (c) inflammatory reaction of the interstitium, and (d) interstitial fibrosis (scarring). In 1917, Epstein recognized changes in the colloid properties of plasma proteins in patients with “nephrosis”[16.Epstein A.A. The nature and treatment of chronic parenchymatous nephritis (nephrosis).JAMA. 1917; 69: 444-460Crossref Scopus (15) Google Scholar]. He proposed that nephrosis was primarily a disorder of protein metabolism. One has to realize that in those days the microscopic techniques were not sufficiently advanced to recognize subtle glomerular lesions. Nevertheless, in 1918, Munk noted some cases with chronic “nephrosis” in which he observed in the glomeruli lipid degeneration of capsular epithelial cells with consecutive compression of glomerular tufts[17.Munk F. Pathologie und Klinik der Nephrosen, Nephritiden und Schrumpfnieren. Urban & Schwarzenberg, Berlin, Wien1918Google Scholar]. In 1925, Fahr confirmed swelling and fatty changes of parietal and visceral glomerular epithelial cells, and furthermore described hyalinization of some capillary loops in the kidneys of patients with nephrosis in the advanced stage of “Schrumpfnieren” (contracted kidneys)[18.Fahr T.H. Henke-Lubarsch: Handbuch der Speziellen Pathologischen Anatomie VI/1. Springer, Berlin1925Google Scholar]. Bell used a modification of Mallory's aniline blue connective tissue stain, which permitted a clearer delineation of the endothelial and epithelial cells of glomeruli[19.Bell E.T. Lipoid nephrosis.Am J Pathol. 1929; 5: 587-622PubMed Google Scholar]. In cases of “nephrosis” in which glomeruli appeared normal by hematoxylin and eosin stain, he was able to demonstrate “a varying increase in the number and size of glomerular endothelial cells and an uneven thickening of the basement membrane,” even if the lesions were not of uniform type. These observations led to the suspicion that subtle alterations of the glomerular capillary wall existed that permitted transit of serum proteins into Bowman's space, although this hypothesis did not go undisputed[20.Kantrowitz A.R. Klemperer P. Über Lipoidnephrose.Virchows Arch. 1931; 280: 554-564Crossref Scopus (1) Google Scholar], and the concept of a primary tubular degeneration was sustained by Ehrich[21.Ehrich W. Über Nephrosen mit besonderer Berücksichtigung des nephrotischen Einschlags.Virchows Arch. 1933; 287: 333-347Crossref Google Scholar] and Terbrüggen[22.Terbrüggen A. Cytologische Untersuchungen zur Frage der Nierenfunktion unter normalen und abgeänderten Verhältnissen.Virchows Arch. 1933; 290: 574-647Crossref Scopus (1) Google Scholar]. Bell concluded that “lipoid nephrosis is to be regarded as a form of glomerulonephritis in which the glomeruli are damaged, but their capillaries are only partially obstructed; thus, they continue to function and tubular atrophy does not occur”[19.Bell E.T. Lipoid nephrosis.Am J Pathol. 1929; 5: 587-622PubMed Google Scholar]. This anatomical observation provided support for the hypothesis, which was suggested by Govaerts and Cordier[12.Govaerts P. Cordier M.R. Contribution á l'étude clinique et anatomique de la néphrose lipoïdique.Bull Acad Royal Méd Belgique. 1928; 5: 510-548Google Scholar] on the basis of clinical observations in four cases, that leaky glomerular capillaries rather than degenerated tubules must account for proteinuria. They felt it was inconceivable that such massive amounts of protein are secreted by tubular epithelial cells, and favored capillary damage by toxic substances. At the same time in a scholarly review, Leiter considered two possibilities for the genesis of proteinuria, that is, altered glomerular permeability for native plasma proteins or a primary alteration of the properties of plasma proteins as a result of a metabolic disturbance[23.Leiter L. Nephrosis.Medicine (Baltimore). 1931; 10: 135-242Crossref Scopus (9) Google Scholar], echoing the concepts of Epstein[16.Epstein A.A. The nature and treatment of chronic parenchymatous nephritis (nephrosis).JAMA. 1917; 69: 444-460Crossref Scopus (15) Google Scholar]. Leiter explicitly stated that glomerular versus tubular excretion of proteins was a matter of view, “insusceptible at present of direct proof.” It is against this background that Edmund Randerath entered the scene and provided definite proof for the concept proposed by Govaerts and Cordier and Bell[12.Govaerts P. Cordier M.R. Contribution á l'étude clinique et anatomique de la néphrose lipoïdique.Bull Acad Royal Méd Belgique. 1928; 5: 510-548Google Scholar],[19.Bell E.T. Lipoid nephrosis.Am J Pathol. 1929; 5: 587-622PubMed Google Scholar]. It is interesting that Randerath did not refer to Bell in his early articles, and it is likely that he was unaware of his publication at that time. Edmund Randerath started with the tenet that the most likely explanation of proteinuria was leakiness of glomerular capillaries. In order to provide proof for this hypothesis, he made use of one anatomical peculiarity of the salamander kidney, that is, the existence of two populations of nephrons Figure 3. One set of nephrons resembles those of higher vertebrates in which glomerulus and tubulus form a closed unit. These nephrons are located in the dorsal and caudal portion of the kidney. In the ventral and cranial location, there are nephrons that are additionally connected with the coelomic cavity through a funnel-shaped channel so that the tubule is in open connection with the coelomic cavity. The channel is coated by ciliated epithelial cells. Randerath had the intuition to notice that the presence of “open nephrons” provided an opportunity to test the previously mentioned hypothesis: an injection of proteins into the coelomic cavity should permit the delivery of proteins to the “open nephrons” without intervention of the glomeruli, which would enable one to see whether or not the presence of proteins in tubular fluid led to the supposed “degenerative” alterations of the structure of the tubular epithelial cells. In previous studies Lambert[24.Lambert P. Sur les potentialités de résorption du tube contourné chez les urodèles.Comptes Rendues Sociéte Biologie (Paris). 1932; 110: 114-118Google Scholar] as well as Gerard and Cordier[25.Gerard P. Cordier R. Sur l'interprétation des altérations morphologiques caractéristiques observées dans le rein au cours de la néphrose lipoïdique.Arch Intern Med Exp. 1933; 8: 225-232Google Scholar] investigated the storage of colored colloids and cholesterol in the salamander kidney. Based on these observations, in an elegant series of experiments Randerath injected 0.5 ml of bovine serum, human serum, serum of patients with renal disease, purified human serum albumin, or purified human globulin into the coelomic cavity of salamanders[26.Randerath E. Die Entwicklung der Lehre von den Nephrosen in der pathologischen Anatomie.Erg Pathol. 1937; 32: 91-140Google Scholar]. It is a testimony to the academic habits of that era that the results were published in detail by the student, Arno Hein, who did the experiments[27.Hein A. Über die Entstehung und Bedeutung der hyalinen Tropfen in den Hauptstücken der Niere auf Grund von Experimenten an Salamandra Maculosa.Virchows Arch. 1938; 301: 339-356Crossref Scopus (3) Google Scholar]. The results were clear cut and left no doubt. As shown in Figure 4, after six daily injections of 0.5 ml of human serum albumin into the coelomic cavity, the epithelial cells of proximal tubuli showed storage in the form of fine granules, predominantly in the supranuclear zone. The extent of storage depended on the quantity of protein injected since the last injection and on the duration of the intracoelomic injection treatment. Of note, storage was less when serum of nephrotic patients was injected; this was interpreted as evidence of depletion of some protein(s) in the patients' sera. Confirming previous results of Lambert[28.Lambert P. Sur l'existence d'un gradient de perméabilité dans les néphrons ouverts des urodèles.Comptes Rendues Sociéte Biologie (Paris). 1933; 114: 1370-1372Google Scholar], he noted that the site of storage within the tubule was a function of the molecular weight of the injected proteins. Small molecules were preferentially stored in more proximal, larger molecules in more distal locations, adhering to the following ranking order: ovalbumin, serum albumin, and serum globulin. When injection was interrupted, the droplets disappeared within a few days. These observations led to the conclusion that droplet formation was not a sign of primary tubular degeneration, but rather proof of the vitality of epithelial cells, as they were obviously able to digest and remove stored droplets. It is of note that lower molecular weight material yielded droplet formation not only in open but also in closed nephrons. This finding led to the conclusion that low molecular weight proteins are filtered by glomeruli of closed nephrons. In independent experiments, he could show that the low molecular weight proteins were absorbed from the coelomic cavity into the blood. The results of these elegant experiments were interpreted as definite proof for the “filtration-reabsorption” hypothesis: “the appearance of hyaline droplets reflects the presence, within the tubular lumen, of proteinaceous colloids, susceptible of resorption and storage.” Because this occurs in only closed nephrons, altered permeability of the glomerular capillaries must be the necessary precondition for the appearance of hyaline droplets in closed nephrons, if higher molecular weight proteins are involved. In such kidneys, however, hyaline droplets in the proximal tubules can appear even without an increase in the permeability of the glomerular capillary wall, if proteins with a molecular weight below 70,000 are involved, for example, polypeptides[27.Hein A. Über die Entstehung und Bedeutung der hyalinen Tropfen in den Hauptstücken der Niere auf Grund von Experimenten an Salamandra Maculosa.Virchows Arch. 1938; 301: 339-356Crossref Scopus (3) Google Scholar]. These findings laid to rest the mistaken hypothesis that protein droplets are the “morphological sign of tubular protein secretion,” as “the appearance of proteinaceous droplets in tubular epithelial cells can only be interpreted as a sign of reabsorption and (transient) storage.” The idea that droplets are dynamic structures reflecting the balance between reabsorption and dissolution was further supported by the studies of Randerath and his pupil Kleier, who noted the rapid disappearance of the droplets within an average of 8 to 10 days[29.Kleier A. Experimentelle Untersuchungen über den Abbau der hyalinen Tropfen nach Eiweiβspeicherung in der Niere von Salamandra maculosa.Beitr Pathol. 1939; 103: 559-567Google Scholar]. It is a reflection of the limited exchange of scientific information in those unhappy days that knowledge of these experiments and their implications for the understanding of human disease apparently did not spread abroad, as reflected by the absence of any reference to Randerath's work in the authoritative textbook of Fishberg[30.Fishberg A.M. Hypertension and Nephritis. Lea & Febiger, Philadelphia1939Google Scholar]. The dissemination of information on Randerath's experiments was certainly further impeded by the advent of the Second World War. We noted that his work was quoted in some German texts, but not abroad. When scientific exchange was resumed after the war, the concept of the glomerular origin of proteinuria had been largely accepted, even in the absence of such detailed proof as provided by the experiments of Edmund Randerath and others. In that respect, Edmund Randerath was a tragic figure and did not receive the credit for his experiments and ideas that he undoubtedly deserved. To quote Sir William Osler[31.Sir William Osler Aphorisms. Edited by BEAN WB, Springfield, Charles C. Thomas. 1961: 112Google Scholar]: “In science the credit goes to the man who convinces the world, not to the man to whom the idea first occurs.”