MORPHOLOGIC EVALUATION AND INTEGRIN EXPRESSION PROFILE OF RENAL TUBULAR CELLS CULTURED FROM PERCUTANEOUS RENAL BIOPSY SPECIMEN

Abstract

Kidney biopsy is an indispensible procedure for making a pathologic diagnosis of renal diseases by fixing and staining the biopsy specimen. However, it is not a routine procedure to culture the cells from a renal biopsy specimen directly, or to utilize the cultured cells for any kind of diagnostic or functional evaluation.

In this study, primary culture of the renal tubular epithelial cells was tried from a piece of percutaneous kidney biopsy specimen. Successive passages of the cells were possible until fourth passage.

With these cells, morphologic characteristics of the cultured cells and integrin expression profiles were investigated.

On light and electron microscopy, these cells were characterized by the cobblestone-like growth, presence of microvilli and tight junction, and the preservation of polarity. Immunohistochemical studies demonstrated the epithelial nature of these cells and particularly their differentiation from renal tubular epithelial cells, of either proximal or distal nephronic segment. The integrin profile confirms the epithelial nature of the cell. We hope that our results facilitate the understanding of pathophysiology of renal tubular cells from the patient directly.

• Kidney biopsy
• Proximal tubular epithelial cell
• Cell culture
• Integrin

INTRODUCTION

Since the first closed needle renal biopsy by Ball in 1934 [[1]], the renal biopsy has been most instrumental in the development of our understanding of renal histopathologic abnormalities. The use of this technique not only improved our diagnosis but also has given new insights into the pathogenesis of human renal disease. However, our interpretation of biopsy specimens has been limited to histopathologic examination. So far it has not been successful to utilize the human renal biopsy specimen for culturing the cells.

Integrins are a family of cell surface receptor heterodimers composed of two noncovalently-linked protein chains having transmembrane portion across the cell wall [[2]]. There has been a rapid recent expansion of our knowledge relating to integrins, and these molecules are considered not only as glue that bind to ligand, but also as signaling molecules [[3]]. Integrin receptor subunits have been detected on cells in normal kidneys as well as in pathologic conditions [[4]]. These molecules are known to be localized to the specific cell types of the kidney [5-9].

In this study, we demonstrated that cells originating from renal tubular cells were successfully primary cultured from the biopsy specimen, were stably maintained until the fourth passages, and were used for evaluation of integrin subunit expression such as flow cytometry and northern blotting, as well as histochemical staining procedures.

MATERIALS AND METHODS

Primary Cell Culture from the Renal Biopsy Specimen

The percutaneous kidney biopsy specimen was obtained from a patient with minimal change nephrotic syndrome by gun-shot method. One tiny piece of each biopsy specimen was chopped with a surgical blade into 1–2 nm³ pieces, was planted on the T25 flask, and was left dry for 3 minutes. Then growth medium was added into the side wall of the flask gently in order not to detach the attached tissue. We used DMEM/F12 (Gibco BRL, Grand Island, NY) containing 10% FCS (Gibco BRL) as the first growth medium.

Culture Condition of Successive Passages

After the first trysinization (1/20 dilution of 0.5% trypsin-EDTA), only 0.5% EDTA (Versene 1:5000, Gibco BRL) without trypsin was used for passage, because the cells were detached and did not grow enough after repeated trypsinization. Cells were washed twice with PBS and plated into T25 flask. The medium was changed to DMEM/F12, supplemented with 7.5 pg/mL of thyroxine 3 (Sigma Chemical Co., St. Louise, MO), 12.5 ng/mL prostaglandin E1 (Sigma Chemical Co.), 50 ng/mL hydrocortisone (Sigma Chemical Co.), 5 μg/mL of insulin: transferrin: selenite (Sigma Chemical Co.) and 5 μg/mL of EGF (Sigma Chemical Co.). This medium was reported to culture the proximal tubular cell [11]. The morphologic evaluation, immunohistochemical study, flow cytometry, and Northern blotting was done at the fourth passages.

Flow Cytometry and Northern Blotting

3 × 10⁵ cells were pelleted and suspended in FACS buffer plus 50 ng/mL goat IgG (DAKO, Denmark) for 30 minutes on ice. Cells were harvested and then resuspended in primary antibody for 60 minutes on ice. Two FACS buffer washes were preceded by second antibody application for 30 minutes. The samples were mixed with 300 μg of 2% formaldehyde. Nytex filtered cells were employed for integrin analysis. The mouse anti-integrin subunit hybridoma antibodies are either commercially available or produced and kindly provided by Dr. E. Wayner and were used in two other experiments [[7]], [[11]]: α2:P1H6, α3:P4C2, α4:P1D6, β1:P4G11 except α5:G0H3 (AMAC), which was commercially available. Northern blotting was performed as previously described [[11]]. The cell lines used for comparing the integrin expression were human mesangial and human lymphocyte cells. The mesangial cells were also used in other experiments and were well characterized [[12]]. The peripheral lymphocytes were obtained from a volunteer's blood by Ficoll-Hypaqe gradients. The following c-DNA were used as probes: human α2, α3, α4 integrin subunit (Dr. M. Hemler, Dana Farber Institute), human α5, β1 integrin subunit (Dr. E. Ruoslahti).

Immunohistochemical Staining and Electron Microscopy

For immunohistochemical studies, seven markers that have been known to stain the specific segment of nephron were used (Tab. 1). The reagents were: Tetragonolobus purpureas (1:1000 Sigma), Phaseolus vulgaris erythroagglutinin (1:2000, Sigma), epithelial membrane antigen (1:50, M613, DAKO), low M.W. cytokeratin (1:80, M631, DAKO), Arachis hypogaea (1:50, Sigma), Tamm-Horsfall protein (1:10, L1032A, Cedar Lane), and vimentin (1:80, M7020, DAKO). The staining was performed both on the cultured cells and on the renal tissue section for companarison.

Table 1. Immunohistochemical Staining of Cultured Cells

Lectin or Antibody	Tissue Staining	Cultured Cell
Tetragonolobus purpureas	Proximal tubule	±
Phaseolus vulgaris erythroagglutinin	Proximal tubule	+
Epithelial membrane antigen	Distal tubule & collecting	±
Low M.W. cytokeratin	Distal tubule & collecting	±
Arachis hypogaea	Distal tubule & collecting	−
Tamm-Horsfall protein	Thick ascending limb	−
Vimentin	Mesenchyme	−

For transmission electron microscopy study, cultured cells were grown on a porous filter and Millicell-CM (millipore, Bedford), and were fixed in situ at day 5 for 30 minutes at room temperature in 2% glutaraldehyde-phosphate buffered saline solution. For scanning EM, glutaraldehyde fixed cells were washed in 0.2M cacedylate buffer and postfixed first in 1% osmium tetroxide with 0.1% cacodylate buffer for 10 minutes. They were treated with 1% osmium tetroxide for 10 min and 1% tannic acid for 10 minutes. These cells were subsequently cast by 4 nm gold/palladium evaporation and were observed.

RESULTS

Characterization of the Renal Cells Cultured from DMEM/F12 Medium

Outbudding of the epithelial cells was evident after the fifth day of implanting biopsy tissue to the T25 flask. There were two proliferating cell types. One type was made up of small cuboidal cells and the other consisted of large pleiotropic multinucleated cells that ceased to proliferate one week later. Trypsinization was performed two weeks after the budding and the cells were plated into T150 flasks. The number of passages, rather than the absolute cell number, was the limiting step. Only small cuboidal cells proliferated to form a uniform morphology after the second passage. After the third passage, the cells from the serum-free medium were characterized by a cobblestone appearance, that is with pseudocyst formation.

However after the fifth passage, these cells ceased to proliferate and showed vacuolization of the cytoplasm. Until then, the cells had a monotonous appearance when observed with a light microscope.

Flow Cytometry and Northern Blotting for Evaluation of Integrin Subunits

α2, α3, α4, α5 and β1 integrin subunit monoclonal antibodies were used for flow cytometry study. α2, α3 and β1 integrin subunits were expressed at the surfaces of these cells, while the expression of α4 and α5 subunit were absent when examined by flow cytometry.

When α4, β1, β4, β5, and β6 integrin subunit m-RNA were evaluated by Northern blotting, there were expected size signals with β1, β4, and β6 c-DNA probes, which suggest that these cells are epithelial cells in origin (Fig. 1).

Figure 1. Northern hybridization of cultured cell with α4, β1, β4, β5 and β6 integrin subunit probes. When α4, β1, β4, β5, and β6 integrin subunit m-RNA were evaluated by Northern blotting, there were expected size signals with β1, β4 and β6 c-DNA probes, which suggests that these are epithelial in origin.

Ultrastructural and Immunohistochemical Characterization

On scanning and transmission electron microscopy, the cells were characteristically growing with the cobblestone appearance. Individual cells were rather primitive with short microvillus-like undulations along the upper cytoplasmic surface, and sparse intercellular dense bodies along lateral interdigitating membranes, remniscent of a primitive form of tight junction (Fig. 2).

Figure 2. Electron microscopic findings of cultured renal tubular cells. a) The cell surfaces were covered by many short microvilli. b) Occasionally, lateral digitations were located between adjacent cells.

Immunohistochemically, the cells were reactive for Tetragonolobus purpureas, Phaseolus vulgaris erythroagglutinin, epithelial membrane antigen, and low molecular weight cytokeratin with variable degrees. They were nonreactive for Arachis hypogaea, Tamm-Hosfall protein, and vimentin (Tab. 1).

DISCUSSION

Primary cultures derived from well-defined kidney segments have been used extensively to isolate the homogenous cell population using microdissection or magnetic iron oxide particles and differential sieving method [14-16]. But it was not possible to obtain enough cell numbers from the biopsy specimen with these methods. To overcome this problem, we used the selection media for proximal tubular cells. In our experience, it was the number of passages rather than the absolute cell numbers that determines the final cell number at the fourth passage. Thus we split the second passage cells directly into T150 flask and third passage into 3T150 flasks, in which the passaged cells proliferated quite vigorously.

The most success with serum-free media was in the field of tubular cell cultures. It is known that unlike glomerular cells, tubular cells can grow and differentiate in the total absence of serum, which further prevents the growth of fibroblasts. For this purpose, we used DMEM/F12 medium without serum after the second passage. Because we had now demonstrated that enough cell numbers can be obtained by this procedure, we pursued the characterization of the cells as the next step.

In our experiment, most of the cultured cells were stained with proximal tubular cell markers. Although the cultured cells used for experimental research were quite uniform in L.M., immunohistochemical staining showed that there were the cells that were weakly stained with distal tubular cell markers. Thus we speculated that the cultured cells were mixed tubular epithelial cells in origin. The cobblestone appearance of the cultured cells strongly suggested that these cells make dome formations and were derived from tubular epithelial cells.

Also, the profile of the integrin subunits supports the notion that the cultured cell are epithelial cells in origin and may function as a barrier of solutes, as β4 and β6 integrin subunits are a necessary component of cellular barrier, that is, tight junction formation.

The immunohistochemical staining was consistent with the SEM/TEM findings in that tubular epithelial cells are main components of this population. Although we speculate that these cultured tubular cells may still possess the characteristics of tubular epithelial cells such as polarity, microvillus formation, tight junction and specific integrin expression of epithelial cells, we should mention that these cultured cells are not fully differentiated. As shown in scanning EM, the observed microvilli are shallow and not well formed.

We speculate that our method of tubular cell culture may be quite useful for investigating tubular cell pathologies such as cystic disease of the kidney. It is known that tubular epithelial hyperplasia, increase in Na⁺−K⁺−ATPase activity in proximal tubule, and abnormal extracellular matrix production are responsible for the pathogenesis of polycystic kidney [17-19]. If our cell culture method can be used, a more direct approach to evaluate the tubular cell function can be investigated. We hope that the culture method of tubular epithelial cells from the patient's biopsy specimen will facilitate the elucidation of the mechanism of many diseases caused by abnormal tubular transport or cystic diseases of the kidney.

Figure 2c. Sparse intercellular dense bodies were observed along lateral interdigitating membranes, remniscent of a primitive form of tight junction.

ACKNOWLEDGMENTS

This work was supported by grants (#HMP-97-B-3-0030) from the 97 Good Health R&D project, Ministry of Health & Welfare, R.O.K., and Samsung grant #SBRI C-96-018.