Growth Characteristics of Cultured Human Proximal … Characteristics of Cultured Human Proximal Tubule Cells Exposed to ... 1,250; 750; 500; 250; 100; 50; 25; and, 5 (xg per ml. The
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
Growth Characteristics of Cultured Human Proximal Tubule Cells Exposed to Aminoglycoside Antibiotics*
MARYANN SENS, M .D ., P h .D ., DEBRA J. HAZEN-MARTIN, P h .D .,
JOHN G. BLACKBURN, P h .D ., GORDON R. HENNIGAR, M .D .,
and DONALD A. SENS, Ph.D .
Departments o f Pathology and Physiology Medical University o f South Carolina,
Charleston, SC 29425
ABSTRACT
It is well known that the nephrotoxic lesions that occur during aminoglycoside-induced nephrotoxicity are both dose- and tim e-dependent. It was the purpose of this study to determ ine if a cell culture model based on the hum an proximal tubule would exhibit similar dose- and tim e-depen- dent relationships when exposed to aminoglycosides of various nephrotoxic potential. For this determination, the hum an proximal tubule (HPT) cells w ere exposed to increasing concentrations of streptomycin, kanamycin, gentamicin, and neomycin and m onitored for cell growth and toxicity over an 18-day period of exposure. Both actively-dividing and resting cells w ere assessed with regard to aminoglycoside exposure. At high levels of aminoglycoside exposure, linear regression analysis disclosed that the rank order of toxicity of the aminoglycosides to be: neomycin > kanamycin > gentamicin > streptomycin. Both actively-dividing and resting cultures of HPT cells displayed both dose- and tim e-dependency with regard to toxicity and the ability of the cells to regenerate in the continued presence of aminoglycoside exposure. This pattern of dose- and tim e-dependency was unique for each aminoglycoside and varied depending on the replicative state of the cells.
W ith the exception of neomycin, clear evidence was obtained that toxicity and cell regeneration were occurring simultaneously throughout the tim e course of aminoglycoside exposure; the equilibrium betw een the two processes determ ining overall cell toxicity or regeneration. In addition, the HPT cells exposed to gentamicin displayed a unique pattern of toxicity and cell regeneration when compared to the o ther aminoglycosides tested, with gentamicin having an increased ability to stimulate cell proliferation. W hile the results obtained are in excellent agreem ent with that known from the clinical experience with the aminoglycosides, the dose- and tim e-dependency of the responses will require careful attention to growth state during em ploym ent in experimental protocols.
Send reprint requests to: Mary Ann Sens, M .D ., m ent of Pathology, Medical University of South Car-Ph.D ., Division of Cellular Biopathology, Depart- olina, 171 Ashley Avenue, Charleston, SC 29425.
The am inoglycoside an tib io tics are used ex tensively in th e tre a tm e n t of gram-negative infections, and their clinical use is associated with dose-limiting nephrotoxicity .2’1018,27 The renal pathogenicity associated with the aminoglycosides can be attribu ted to their selective accu m u la tio n in th e ren a l proxim al tubule of hum ans2,7,24,27 and experim ental anim als.9,21 That pathogenicity is due to accumulation is suggested since it is known that the aminoglycosides are not m e ta b o liz e d by th e k id n ey b u t are excreted unchanged in the urine (parenteral administration) or in the feces (oral administration ) . 2,20'21 The accumulation of the aminoglycosides within the proximal tubule results in the early ultrastruc- tural appearance of cytoplasmic myeloid bodies19 along with alterations in tubular function21 and release of tubular proteins into the u r in e . 1,24 O ne of the striking features of aminoglycoside nephrotoxicity is that the morphologic lesions are both dose- and tim e-dependent18,25 and run a course best characterized as one of proximal tubular damage with potential recovery.
Histologic examination of the kidneys of rats following aminoglycoside exposure reveals proximal tubule cell damage even at low m ultip les of the hum an dose . 15,16,19 This tubular degeneration is focal in nature and usually progresses to tu b u la r n ec ro sis , w hich may e ith e r resolve by com plete regeneration of the necrotic proxim al tubu le or resu lt in scarring and interstitial fibrosis. 15,16 Concurrent recovery from the nephrotoxicity induced by large doses of aminoglycosides in rats has been docum ented with active proximal tubule cell necrosis and regeneration occurring simultaneously in adjacent nephrons . 5,13,16,22 D espite the localization of pathologic damage to the
proximal tubule , the initial injury and m echanism (s) of ce llu la r toxicity and regeneration rem ain unknown.
W hile num erous cellular alterations have been noted to occur with aminoglycoside exposure , 15,18 no agreem ent has been reached on which are prim ary or secondary events in nephrotoxicity. This difficulty arises prim arily owing to the complexity of the in vivo environm ent w h e re fa c to rs su c h as g lo m e ru la r involvement, tubular obstruction, vascular compromise, electrolyte imbalances, and infections render the definition of p rim ary v e rsu s seco n d ary ev en ts in n e p h ro to x ic ity v e ry d ifficu lt. In an a t te m p t to l im it th e in te r fe re n c e s present in the in vivo setting, this labora to ry has d e v e lo p e d a ce ll c u ltu re model derived from the human proximal tu b u le 3,6 w hich, w hen exposed to the aminoglycoside antibiotic, streptomycin, dem onstrated growth responses and cellu la r changes sim ilar to those d o cu m ented to occur in vivo.26 In this report, the characterization of this model system is ex tended to d e te rm in e the growth response of the cells to several aminoglycoside antibiotics chosen to represent a spectrum of known nephrotoxicities. It will be shown that the effects on cell growth are dose- and tim e-dependent and that cellular toxicity correlates to the nephrotoxicities known from the in vivo experience w ith these antibiotics. Of particular in terest was the finding that gentamicin dem onstrated the ability to stimulate proximal tubule cell growth.
Materials and Methods
T is s u e C u l t u r e
Sources of reagen ts and the p rocedures for the initiation and maintenance of stock c u ltu re s of hum an proxim al tubule (HPT) cells have been previously d e sc r ib e d . 6,26 Briefly, the H P T cells were grown in a serum -free formulation
268 SENS, HAZEN-MARTIN, BLACKBURN, HENNIGAR, AND SENS
consisting of a 1:1 mixture of Dulbecco’s M o d if ie d E a g le s (D M E ) a n d F -12 medium supplem ented with selenium (5 ng per ml), insulin (5 fig per ml), transferrin (5 (Jig per ml), hydrocortisone (36 ng p e r ml), triiodothyronine (4 pg per ml), and epiderm al growth factor (1 0 ng per ml). The cells were fed fresh growth medium every three days. For the measurem ent of cell growth and ultrastructure , cultures of HPT cells at passage 3 - 4 were subcultured into 12-well plates w ith each w ell con ta in ing two ml of growth m edium . C onfluent or resting culture experim ents were performed following a 1:1 subculture of the cells. In th e s e s tu d i e s , th e c e l l n u m b e r s rem ained constan t in the control cu ltu re s (no a m in o g ly co sid e ad d itio n ) th ro u g h o u t th e e x p e r im e n ta l tim e course. Subconfluent studies were performed following a 1:3 subculture such that the control wells dem onstrated an exponential growth phase during the initial four days of aminoglycoside exposure. Twenty-four hours following the ap p ro p ria te su b cu ltu re , the cu ltu res were fed fresh growth medium containing a given concentration of aminoglycoside. The aminoglycosides assessed were streptom ycin, kanamycin, gentam icin, and neom ycin, and all concentrations w ere tested in triplicate. The effect of each aminoglycoside concentration on cell growth was tested over an 18-day tim e course. The concentrations of each am inog lycoside te s te d w ere: 6,500; 5,000; 3,500; 2,500; 1,250; 750; 500; 250; 100; 50; 25; and, 5 (xg per ml. The cells were fed fresh growth medium containing th e ap p ro p ria te am inoglycosides every three days.
Throughout the experimental course, th e ce lls w e re m o n ito re d u s in g an inverted microscope.* Every third day, cell coun ts w ere o b ta ined using the
inverted microscope interfaced via video ca m era f to a co m p u te r-lin k ed im age analysis system. $ Random fields of cells were selected and the cells in the fields counted using the mouse attachm ent and digitizing tablet. For each concentration, a m inim um of 2 0 fields w ere selected and counted in each well at each tim e p o in t. All c o n c en tra tio n s w ere p e r form ed in triplicate. The criterion for acceptance of individual tim e points was the finding of no statistical difference in cell num bers within the triplicate samples at each concentration. All statistical analyses w ere perform ed utilizing the V ideoplan software and included the S tudent’s two-sided t-test as well as basic statistical m ethods such as mean, standard deviation, and standard variance. Trypan blue staining was utilized in p re liminary determ inations and at selected points of each tim e course to judge cell viability. Cells that were floating free in the growth m edium after concentration by centrifugation were dem onstrated to be over 99 percent non-viable. The HPT cells a tta c h e d to th e c u ltu re vesse l growth surface w ere shown to exclude trypan blue regardless of aminoglycoside concentration or time of exposure.
The results presented are representative of one isolate of hum an proxim al tubule cells. Two other independent isola tes w ere te s te d , and b o th y ie ld e d eq u iv a len t resu lts to th a t p re se n te d herein.
Results
C o n f l u e n t HPT C e l l s E x p o s e d t o A m in o g l y c o s id e A n t ib io t ic s
Confluent cultures of HPT cells were m onitored for cell growth and survival when exposed continuously over an 18- day period to a wide concentration range
* Zeiss IM35.t Hitachi GP5A4. t Zeiss Videoplan.
AMINOGLYCOSIDE EFFECTS ON CELL GROWTH 269
Dose Response of Human Proximal Tubule Cells to Aminoglycosides
TABLE I
P e r c e n t V i a b i l i t y *C o n c e n tr a t io n
♦Percent viability expressed as percent of control viability.
+TD 50 = dose producing 50 percent lethality.Values were calculated using linear regression analyses employing Systat software on IBM PS-2.The regression was linear at a = 0.005 with ANOVA.
of the aminoglycoside antibiotics streptom ycin, kanam ycin, gentam icin , and neomycin. The period of exposure and large concentration range (5 jjLg per ml to6,500 (JLg per ml) were chosen to afford the maximal opportunity to detect: (1 ) the rank o rder of toxicity of the four aminoglycosides; (2 ) the dose- and time- dependency of each aminoglycoside; (3) the ability of the cells to recover and regain proliferative capacity in the continued presence of antibiotic; and, (4) p o te n tia l d iffe ren ces b e tw e e n th ese closely related antibiotics.
The rank o rder of toxicity for HPT cells exposed to streptom ycin, kanamyc in , g en tam ic in , and neom ycin was d e te rm in e d follow ing seven days of exposure to 2.5, 3.5, 5.0, and 6.5 mg per ml of antibiotic. As shown in table I, the rank order of toxicity for the HPT cells, as determ ined through linear regression analysis, was neomycin > kanamycin > gentamicin > streptomycin. The highest concentration range of aminoglycoside exposure was chosen for analysis since, as will be subsequently dem onstrated, exposure of HPT cells to lower concentrations of these antibiotics resu lted in selected instances of cell renew al and resistance th a t w ere b o th dose- and tim e-dependent. As such, inclusion of
these lower concentrations of exposure into the calculations, while resulting in consistent T. D .50 values and rank order of toxicity, produced a significant deviation from linearity. This problem was avoided when employing the higher concen tra tio n s and sh o rt tim e p e rio d of exposure.
The response of the confluent HPT cells to the aminoglycoside antibiotics was both dose- and tim e-dependent. As depicted in figure 1A, exposure of HPT cells to neomycin, the most nephrotoxic of the aminoglycosides, at a concentration of five (jig p e r ml resu lted in the death and detachm ent from the culture growth surface of 2 0 percent of the cells. This occurred w ithin the initial 10 days of exposure and, thereafter, no further d e c re a se s in c e ll n u m b e r o c c u rre d despite continued presence of the antibiotic. In creasin g th e co n cen tra tio n of neomycin resulted in a continued reduction in the percen tage of viable H PT cells within the initial 1 0 days of exposure; however, this reduction was clearly not in proportion to the increase in neomycin concentration. W hereas a five |xg p e r m l c o n c e n tra t io n o f n e o m y c in resulted in 20 percent cell death, 25 (JLg
per ml resulted in only an additional 13 percent reduction, and exposure to 250 jjLg per ml in only an additional 20 percent reduction. This slow increase in cellular toxicity with increasing neomycin concentration was ev iden t during the initial 1 0 days of exposure for all the concentrations tested. Additionally, regardless of concentration, the toxicity of neomycin occurred mainly within the initial 1 0 days of exposure and, thereafter, stabilized at a constant value despite continued exposure to the drug. Daily light microscopic examination of the cultures re v e a le d few d e ta c h e d ce lls in th e growth medium after 1 0 days of neomycin exposure, an observation suggesting that the rem aining cells were resistant to c o n tin u e d n e o m y c in ex p o su re and ,
270 SENS, HAZEN-MARTIN, BLACKBURN, HENNIGAR, AND SENS
ow ing to th e c o n s ta n t a tta c h e d cell n u m b er, no t u n dergo ing ap p rec iab le proliferation.
An identical analysis perform ed utilizing streptom ycin, the least nephrotoxic of th e am inoglycosides, disclosed this antibiotic to be far less toxic to the HPT cells (figure IB 26). Exposure of the HPT cells to concentrations of streptom ycin be tw een five and 50 ¡xg p e r ml w ere w ith o u t e ffec t on th e H P T ce lls as ju d g e d bo th by cell coun ts and ligh t microscopy. Concentrations of s trep to mycin betw een 100 and 500 (jug per ml elicited a small, bu t repeatable, gradual decrease in the num ber of attached HPT cells during the initial 1 0 days of exposure, an event followed by renew ed cell proliferation to control values in the contin u ed p resence of the drug. F u r th e r increasing the streptomycin concentration (750 to 3,500 |JLg per ml) resulted in im m ediate cell toxicity within the initial th ree days of exposure followed by rapid prolifertion of the cells to near-control leve ls of g row th . L igh t m icroscop ic examination of the cultures revealed the presence of detached, non-viable cells in the growth medium throughout the time course, suggesting that toxicity and p roliferation were occurring as simultaneous
FIGURE 1. Time course of confluen t hum an proximal tubule cells exposed to various concentrations of streptom ycin, neomycin, and kanamycin. The data points are expressed as the percent of control human proximal tubule cells, plated at the same time at the same density, and maintained in antibiotic-free media. They are derived from the mean of 20 cell count measurements, repeated in triplicate. The individual standard error of the mean (SEM) for each point was less than five percent of the mean (error bars not shown for graphical simplicity).
F ig u r e 1A. Confluent human proximal tubule cells exposed to neomycin. O = 5 (xg per ml; V = 25 (Jtg per ml; □ = 250 |xg per ml; A = 2500 p.g per ml; • = 6500 (jLg per ml.
F ig u r e IB . Confluent human proximal tubule cells exposed to streptomycin. O = 50 (xg per ml; V = 250 (xg per ml; □ = 750 |xg per ml; A = 2500 (xg per ml; 0 = 6500 |xg per ml.
F ig u r e 1C. Confluent human proximal tubule cells exposed to kanamycin. O = 50 ¡xg per ml; V = 500 |xg per ml; □ = 1250 ¡J.g per ml; A = 3500 |xg per ml; 0 = 6500 |j,g per ml.
AMINOGLYCOSIDE EFFECTS ON C ELL GROWTH 271
events. The HPT cells exposed to concentrations of streptom ycin greater than3,500 fxg per ml displayed similar patterns to that previously described, but dem onstrated only a marginal ability to proliferate during continued streptom ycin exposure and regain confluent cell density.
Kanamycin, an aminoglycoside antibiotic with clinical nephrotoxicity in term ediate to that of neomycin and streptom ycin, was also assessed w ith regard to HPT cell viability and growth (figure lc). Exposure of the HPT cells to kanamycin at concentrations betw een five and 1 0 0 (jug per ml resulted in a small, bu t repeatable, decrease in viable cells within the first seven to 10 days of exposure. This was followed by renew ed proliferation of the cells to control values during the rem aining days of the tim e course and occurred in the continuing presence of this antibiotic. Light microscopic examination during the period of cell renewal d isc losed the p resen ce of significant quantities of detached cells, indicating th a t p ro life ra tio n and tox icity w ere occurring simultaneously. Increasing the concentration of kanamycin (250 to 750 (J ig p e r ml) re su lte d in fu rth e r small decreases in viability during the initial 1 0 days of exposure and in the subsequen t ability of the cell population to undergo regeneration to control values. At a kanamycin concentration of 1,250 |xg pe r ml and above, findings similar to th e p re v io u s ones d e m o n s tra te d w ith fu rth e r gradual decreases in cell viability as the kanamycin concentration increased.
An identical assessment of the toxicity of gentam icin, an aminoglycoside with sim ilar clinical nephrotoxicity to kanam ycin, on the viability and grow th of H PT cells y ielded contrasting results when compared to the other aminoglycosides. E xposure o f co n flu en t H PT cells to five |xg per ml of gentamicin disclosed a small increase in the cell growth
w ithin th ree days of exposure (figure 2A). This was followed by a decrease in attached cells within the next four days of exposure and then renew ed cell proliferation over the rem aining seven days of exposure that exceeded that of HPT cells unexposed to gentamicin. This pattern of g ro w th s t im u la t io n , to x ic ity , an d regrowth (exceeding control values) was also dem onstrated for gentam icin concentrations betw een 25 and 100 (xg per m l, w ith th e only d ifference b e ing a gradual decrease in the m agnitude of the growth phases.
Increasing the gentamicin concentration to 250 jjug per ml (figure 2B) resulted in a small, gradual decrease in cell viability during the first 1 0 days of exposure and a gradual regrowth of the cell population to near control levels. F u r th e r increasing the gentamicin concentration to 500 jxg per ml resulted in a decrease in cell viability of approximately 2 0 percent within the initial 1 0 days of exposure; however, this was followed by a strik ing increase in cell p ro life ration which surpassed control levels (figure 2B). Furtherm ore, exposure of the HPT cells to h igher dosages of gentam icin (750 and 1,250 jjug per ml) resulted in no appreciable reduction in cell num ber, b u t ra ther a clear, striking increase in cell proliferation which exceeded control values by over 20 percen t (figure 2B). Light microscopic examination of these cu ltu res rev ea led very few d e tach ed cells at any point during the tim e course.
Additionally, this increase in cell proliferation as a result of gentamicin (750 (xg per ml for 15 days) presented light microscopically as areas of the monolayer containing tightly packed cells with some evidence of focal m ultilayer formation (figures 3A and 3B). O f additional in terest is the observation that the control HPT cells (figure 3B) appear with a flatte n e d m orphology, a m orpho log ica l characteristic routinely noted after these ce lls h av e b e e n m a in ta in e d fo r an
PERC
ENT
OF
CON
TRO
L
272 SENS, HAZEN-MARTIN, BLACKBURN, HENNIGAR, AND SENS
extended tim e period in a non-prolifera- tive state. This is in m arked contrast to the cells exposed to gentamicin (figure 3A), which possess a m ore epithelioid morphology which is routinely noted for d iv id in g o r re c e n tly c o n flu e n t H PT cells. F u rth e r increases in gentam icin concentration (2,500 and 3,500 |xg per ml) resu lted in initial cell toxicity, followed by appreciable regrow th of the monolayer, and subsequent total cell toxicity by the end of the tim e course (figure 2C). Concentrations of gentamicin in excess of 3,500 jJLg p e r ml resu lted in almost total cell death without evidence of any recovery period (figure 2C).
S u b c o n f l u e n t HPT C e l l s E x p o s e d t o A m in o g l y c o s id e A n t ib io t ic s
In order to com pare possible differe n c e s in a m in o g ly c o s id e to x ic ity betw een resting versus actively-dividing HPT cell cultures, the previous protocol was repeated using HPT cells plated at subconfluent density. To achieve active cell growth, the HPT cells w ere subcultu red at a ratio of 1:3, w hich allowed approxim ately four days of active division before the cells attained confluency.
F i g u r e 2. Time course of confluen t hum an proximal tubule cells exposed to various concentrations of gentamicin. The data points are expressed as the percent of control human proximal tubule cells, plated at the same tim e at the same density, and m ain tained in an tib io tic -free m edia. They are derived from the mean of 20 cell count m easurements, repeated in triplicate. The individual SEM for each point was less than five percent of the mean (error bars not shown for graphical simplicity).
F i g u r e 2A. Confluent human proximal tubule cells exposed to low dosages of gentamicin. 0 = 5 H -g per ml; V = 25 ( J ig per ml; □ = 50 (xg per ml; A = 100 (xg per ml.
F i g u r e 2B. Confluent human proximal tubule cells exposed to moderate dosages of gentamicin. O = 250 (jLg per ml; V = 500 (xg per ml; □ = 750 |xg per ml; A = 1250 |xg per ml.
F i g u r e 2C. Confluent human proximal tubule cells exposed to high dosages of gentamicin. O = 2500 (xg per ml; V = 3500 |xg per ml; □ = 5000 |JLg per ml; A = 6500 (Jig per ml.
AMINOGLYCOSIDE EFFECTS ON CELL GROWTH 273
F i g u r e s 3A and 3B. Confluent human proximal tubule cells exposed to: A = 750 |xg per ml of genta- micin for 15 days; B = control cells at 15 days which were not exposed to gentamicin. Cells were fixed and stained with toluidine blue and photographed using a Zeiss IM 35 microscope equipped with Hoffman optics. Magnification = 250 x .
A comparison of the toxicity of neomycin betw een actively-dividing (figure 4A) and stationary cultures (figure 1A) clearly d em o n stra ted tha t neom ycin was far m ore toxic to H PT cells undergo ing active cell division. W hereas exposure to neomycin at five |xg per ml elicited a 20 percent reduction for HPT cells at conf lu e n t density , th is reduc tion was in excess of 50 percent for actively-dividing cells. This increased toxicity of neomycin for actively-dividing HPT cells was true for all the concentrations of neomycin
tested. In addition, the toxicity of neomycin for rapidly dividing H PT cells occurred throughout the tim e course in contrast to resting cultures, which exhibited initial toxicity followed by a plateau phase.
As reported previously,26 a decrease in toxicity was observed for actively-dividing HPT cells exposed to streptom ycin (figures IB and 4B). This observation was most apparent for streptomycin concentrations early in the time course within the initial seven days of exposure. Thereafter, the HPT cells exposed to streptomycin displayed a decreased ability to regenerate to control values. An identical analysis of the HPT cells exposed to kanamycin dem onstrated, during the initial three days of exposure, a very m arginal trend for an increase in toxicity for actively-dividing cells compared to stationary cells (figures 1C and 4C). However, following this initial increase in toxic ity , th e s u b c o n f lu e n t H P T c e lls exposed to kanamycin dem onstrated no additional loss of viable cells and, at most dosages , d e m o n s tra te d a t re n d for ren ew ed cell p ro life ra tion to contro l levels. This is in contrast to stationary cultures, w here continued exposure to kanamycin resulted in further decreases in attached, viable cells. Light m icroscopic examination of the actively-divid- ing cultures exposed to kanamycin dem o n s tr a te d a p p re c ia b le n u m b e rs of detached, non-viable cells in the growth m edium th roughout the tim e course, providing evidence that both cell death and proliferation were occurring simultaneously.
Exposure of the actively-dividing HPT cells to gentam icin disclosed a unique pattern of cell response when compared both to that dem onstrated with the other aminoglycosides and to that found for confluent HPT cells exposed to gentam icin. Actively-dividing cells exposed to five |xg p e r ml of gentam icin dem on-
274 SENS, HAZEN-MARTIN, BLACKBURN, HENNIGAR, AND SENS
strated a large increase in cell proliferation during the first seven days of exposure, an event which was followed by m arked cell toxicity (figure 5A). This pattern of prom inent cell proliferation followed by m arked toxicity was dem onstrated for all gentamicin concentrations up to and including 500 |xg per ml (figures 5A and 5B). Light microscopy of the cells undergo ing m arked proliferation above control levels revealed focal areas of th e m ono layer c o n ta in in g tig h tly packed cells and evidence of multilayer formation in a m anner identical to that described ea rlie r for confluen t cells. Cells exposed to gentamicin at 1,250 |xg per ml displayed minimal cell toxicity, fo llow ed by m arg inal recovery , and finally m arked toxicity (figure 5B). Concentrations of gentam icin in excess of 1,250 per ml were toxic to the HPT cells w ithout evidence of cell recovery (figure 5B). Concentrations of gentam icin in excess of 1,250 |xg p e r ml were toxic to the HPT cells w ithout evidence of cell recovery (figure 5C). Light microscopic examination revealed that even during periods of significant cell recove ry , th e r e w e re s t i l l a p p re c ia b le
F igu re 4. Time course of subconfluent human proximal tubule cells exposed to various concentrations of streptom ycin, neomycin, and kanamycin. The data points are expressed as the percent of control human proximal tubule cells, plated at the same time at the same density, and maintained in antibi- otic-free media. They are derived from the mean of 20 cell count measurements, repeated in triplicate. The individual SEM for each point was less than five percent of the mean (error bars not shown for graphical simplicity).
F ig u re 4A. S ubcon fluen t hum an proxim al tubule cells exposed to neomycin. O = 5 (Xg per ml; V = 250 ( x g per ml; □ = 2500 | x g per ml; A = 6500 fig per ml.
F igure 4B. S ubcon fluen t hum an proxim al tubule cells exposed to streptomycin. O = 50 (xg per ml; V = 250 |xg per ml; □ = 1250 (xg per ml; A = 2500 (ig per ml; • = 6500 |xg per ml.
F igure 4C. S ubcon fluen t hum an proxim al tubule cells exposed to kanamycin. O = 50 (xg per ml; V = 500 (xg per ml; □ = 1250 |xg per ml; A = 5000 |xg per ml.
AMINOGLYCOSIDE EFFECTS ON C ELL GROWTH 275
num bers of detached, non-viable cells in the growth medium. W hen compared to co n flu en t cells, the actively-dividing cultures appeared to possess a greater capacity for proliferation in the presence of gentamicin at early stages of exposure, bu t also a m arked increase in toxicity at later times of continued exposure to the drug.
Discussion
The artificial environm ent provided by cell culture, through its elimination of the many complex factors present in the in vivo environm ent, could be a valuable adjunct in research efforts probing the cellular basis of aminoglycoside-induced nephrotoxicity. This concept was tested in the p resent study by determ ining the dose- and tim e-dependency of aminoglycoside exposure on the growth and viability of cultured human proximal tubule (HPT) cells. The HPT cell culture system, which proliferates in a serum-free growth media, has been dem onstrated to retain consistent properties from isolate to isolate through a combination of m orphological, u ltrastructu ra l, electrical, and enzyme histochemical techniques . 3,6 W hile clearly hom ogeneous in d e te r-
F i g u r e 5. Time course of subconfluent human proximal tubule cells exposed to various concentrations of gentamicin. The data points are expressed as the percent of control human proximal tubule cells, p lated at the same tim e at the same density, and m ain ta ined in an tib io tic-free m edia. They are derived from the man of 20 cell count m easurements, repeated in triplicate. The individual S E M for each point was less than five percent of the mean (error bars not shown for graphical simplicity).
F i g u r e 5A. S ubcon fluen t hum an proxim al tubule cells exposed to low dosages of gentamicin. O = 5 (xg per ml; V = 25 (xg per ml; □ = 50 (xg per ml.
F ig u r e 5 B . S u b c o n f lu e n t h u m a n p ro x im a l tu b u le cells ex p o sed to m o d e ra te dosages o f g e n ta m icin . ▲ = 100 (xg p e r m l; O = 250 jxg p e r m l; V = 500 (Xg p e r m l; □ = 750 (xg p e r ml; A = 1250 (xg p e r m l.
F i g u r e 5C. S u b c o n f l u e n t h u m a n p r o x i m a l t u b u l e c e l l s e x p o s e d t o h i g h d o s a g e s o f g e n t a m i c i n .O = 2500 ( X g p e r m l ; □ = 5000 ( x g p e r m l ; A = 6500 ( x g p e r m l .
276 SENS, HAZEN-MARTIN, BLACKBURN, HENNIGAR, AND SENS
m ined structure and function, the culture itself is likely composed of all represen tative segm ents (S1; S2, S3) of the proximal tubule. As such, it is possible th a t cells from ce rta in segm ents are m ore susceptible to a given antibiotic than those of o ther segments.
A previous report from this laboratory has detailed that these cultures of human proximal tubule cells, when exposed to s trep tom ycin , d isp layed m any of the characteristics associated w ith in vivo am in o g ly co sid e-in d u ced n e p h ro to x icity. 26 These alterations included the formation of cytoplasmic m yeloid bodies, both cell toxicity and regeneration, and differing toxicity betw een stationary and ac tiv e ly -d iv id in g ce lls . T he p re se n t study was designed to extend the p re vious findings to other aminoglycosides of varying nephrotoxic potential, especially as it applies to the dose- and time- dependency of cell toxicity and regenera tio n . E m p lo y in g l in e a r re g re ss io n analysis of the HPT cells exposed to high concen tra tions o f am inoglycosides, a rank order of toxicity (neomycin > kana- mycin > gentam icin > streptom ycin) was determ ined which agreed well with th a t know n from in v ivo s tu d ie s in h u m an s and an im al m o d e ls . 2,18,20,21 Additionally, the total dose necessary to produce 50 percent lethality for the HPT cells exposed to neomycin and gentamicin was in close ag reem en t w ith that found by W illiam s and co -w orkers28 em ploying sim ilar cu ltu res of hum an proximal tubule cells. Thus, the HPT cells respond to aminoglycoside exposure with a rank order of toxicity similar to that known from in vivo experience with the antibiotics.
The response of the H PT cells to the aminoglycoside antibiotics was clearly dose- and tim e-d e p e n d en t. This is a w ell-docum ented finding w ith in vivo am inoglycoside nephrotoxicity , w here the m orphologic lesions produced are both dose- and tim e-dependent18,25 and
run a course characterized best as one of proximal tubular damage with potential recovery. In the HPT cell culture model, the growth response during exposure to streptom ycin, kanamycin, and gentam icin a t low and m o d era te doses was clearly one of simultaneous toxicity and reg en e ra tio n . This was convincingly dem onstrated for each aminoglycoside at several doses in areas of the time courses where cell growth was increasing while light m icroscopic observation dem onstra ted appreciable cells detached and floating free in the growth medium. The only exception to this was with neomycin , w h e re a p p re c ia b le tox icity was present even at the lowest dosage tested and visual observation did suggest some, bu t not appreciable, cell regeneration. This simultaneous occurrence of proliferation and toxicity correlates well to the in vivo circumstance where nephrotoxicity induced by large doses of aminoglycosides in rats dem onstrated active prox- im a l t u b u l e c e l l n e c r o s i s a n d regeneration as simultaneous events in ad jacen t n e p h ro n s . 5,13,16,22 Thus, the HPT cell culture model effectively m irrors the dose- and tim e-dependency of in vivo aminoglycoside-induced nephrotoxicity.
The ability of proximal tubule cells in vivo to undergo regeneration in the continued presence of gentamicin exposure8 has led to the hypothesis that regenerating rena l tu b u la r ep ith e liu m may be more resistant to the nephrotoxic insults of am inoglycosides than m ore m ature tubular epithelium . 12,16,22 In the present studies, both neomycin and streptom ycin elicited increased toxicity on HPT cells that w ere undergoing active division when compared to cells which were confluent and not undergoing appreciable cell division. However, this finding of increased toxicity for dividing cells was not noted for HPT cells exposed to kanam ycin or gen tam icin . For these am inoglycosides, the com petence for
SENS, HAZEN-MARTIN, BLACKBURN, HENNIGAR, AND SENS 277
cell renew al of the H PT cells was far m ore apparent for actively-dividing cells than for those in the resting stage of cell d iv ision . T his was v e ry s tr ik in g for actively-dividing H PT cells exposed to gentam icin, w here all bu t the highest levels of exposure y ielded significant increases in cell growth above control levels during the early periods of the tim e course. However, once cell proliferation reached control levels or above, gen tam ic in was th e n found to e lic it appreciable cell toxicity. The finding that a c tiv e ly -d iv id in g H P T c e lls , w h en exposed to gentam icin , dem onstra ted increased capability for cell growth correlates tem porally to the regeneration of proximal tubular elem ents during genta- micin-induced nephrotoxicity as noted in animals and man.
The present results also suggest that each individual aminoglycoside, as well as having a definable toxicity, may also have a definable capacity to stim ulate cell regeneration. This is suggested by the findings that exposure of HPT cells to neom ycin disclosed very little evidence for cell regeneration, while similar exposure to gentamicin elicited clear evidence for enhanced H PT cell growth. F u r th e rm o re , b o th kanam ycin and streptom ycin exhibited evidence for promoting cell regeneration in excess of that noted for neomycin, bu t at a much lower m a g n itu d e th an g e n ta m ic in . T h e se results suggest the possibility that the overall toxicity of a given aminoglycoside might be the average of both the toxic potential and the corresponding potential to alter cellular events in a m anner leading to a signal for cell renewal. It will be of in te re s t to expand th e p re se n t studies to include o ther clinically-useful am inoglycosides such as tobram ycin , amikacin, and netilmicin to determ ine if they have an intrinsic potential to stim ulate proximal tubule cell growth.
The dose- and tim e-dependency of the HPT cells to aminoglycoside exposure is
consistent with these cells presenting as a valid in v itro m odel system for the s tu d y o f a m in o g ly c o s id e - in d u c e d nephrotoxicity. Furtherm ore, the results em ploying gentam icin also suggested th a t th e H P T c e ll c u l tu r e m o d e l responds to aminoglycoside exposure at concentrations equivalent to that know to occur in th e clin ical se tting . I t is known that a serum concentration of 1 2 |xg pe r ml is the limit suggested during clinical treatm ent in the hospital setting. As such, subconfluent HPT cells w ere dem onstrated to respond to five p-g per ml of gentam icin with over 40 percen t cell death following 1 0 days of exposure. W hile n o n -d iv id in g H PT cells w ere d em onstra ted to be m ore resis tan t to g e n ta m ic in e x p o s u re , to x ic ity s ti l l occurred w ithin a range that m ight be present within the proximal tubule. This is due to the fact that the serum level represents only a m inimal estim ate of the amount localized within the proximal tubule.
It is known that the aminoglycosides concen tra te w ithin the kidney cortex, and in aminoglycoside treated rats, cortical gentamicin concentrations have been shown to range from 900 to 2,230 |xg per ml w et w eight . 18,23 Assuming that the lysosomal volume represents five to 1 0 percent of the cell volume and given that most of the gentamicin is localized to the lysosom e , 13 it can be ca lcu la ted th a t in tra ly sosom al co n c en tra tio n s cou ld reach 18 to 80 m M . 5 Thus, even at the increased dosages needed to effect confluent HPT cells, it would appear that this cell culture model responds appropriately in term s of the aminoglycoside concentration needed to elicit cellular alterations.
W hile it is encouraging that this cell culture model appears valid for the study of the cellular basis of aminoglycoside- induced nephrotoxicity, the dose- and tim e-dependency of both toxicity and c e l l r e g e n e r a t i o n w ill r e n d e r i t
278 AMINOGLYCOSIDE EFFECTS ON CELL GROWTH
extremely difficult to employ. The rationale for this statem ent is that during an individual tim e course and dose, the HPT cells can undergo both toxicity and regeneration , usually as sim ultaneous events. This indicates that at any given time point, depending on the extent of the two processes, one could define cellu la r a lte ra tio n s o c c u rr in g ow ing to regeneration of the cells and not necessarily owing to toxicity of the aminoglyc o sid es , o r v ice v e rsa . W h ile b o th h u m a n 4,11 a n d a n im a l14,17 c u l tu r e s derived from the proximal tubule have been used to study the alterations of cellular metabolism that occur owing to the am inoglycosides, the resu lts have not been addressed as regards the growth state of the cells. O ne particular difficu lty arises w ith the use of defin ing appropriate cells for controls. If aminoglycoside exposure elicits both cell toxicity and renew al, th en confluen t cells a fte r various days of am inoglycoside exposure in culture are representative of new ly rep lica ted cells, w hile contro l cells would be quiescent during this time frame. W hile the use of proximal tubule- derived cell cultures will allow studies to be perform ed free of in vivo environmental influences, results will have to be ca re fu lly c o n tro lle d as re g a rd s th e growth state of the cultures.
Acknowledgments
This research was su p p o rted bv N IH grant DK35311.
References
1. B e c k , P. R . , T h o m p s o n , R . B . , and C h a u d - H U R A , A. K . R . : Aminoglycoside antibiotics and renal function: C hanges in urinary glutamyl transferase excretion. J. Clin. Pathol. 30:432- 437, 1977.
2. B e n n e t t , W. M.: Aminoglycoside nephrotoxicity. Nephron 35:7 3 - 77, 1983.
3. B l a c k b u r n , J . G ., H a z e n - M a r t i n , D . J . , D e t r i s a c , D . J . , a n d S e n s , D . A . : E l e c t r o p h y s i o l o g y a n d u l t r a s t r u c t u r e o f c u l t u r e d h u m a n p r o x i m a l t u b u l e c e l l s . K i d n e y I n t e r n . 33:508- 516, 1988.
4. C h a t t e r j e e , S . , T r i f i l l i s , A. L., and R e g e c , A. L .: Biochemical and morphological effects of gentamicin in human proximal tubular cells: effects on lipid and lipoprotein metabolism. Bio- chem. Cell Biol. 65:1049-1056, 1987.
5. C u p p a g e , F. E ., S e t t e r , K . , S u l l i v a n , L. P., R e i t z e s , E. S . , and M e l n y k o v y c h , A. O .: Gentamicin nephrotoxicity. II. Physiological, biochemical and morphological effects of p rolonged administration to rats. Virchows Arch. B 24:121-138, 1977.
6 . D e t r i s a c , C. J . , S e n s , M. A., G a r v i n , A. J., S p i c e r , S . S . , and S e n s , D . A.: Tissue culture of human kidney epithelial cells of proximal tubule origin. Kidney Intern. 25:383-390, 1984.
7. E d w a r d s , C. Q . , S m i t h , C. R . , B a u g h m a n , K. L . , R o g e r s , J. F., and L i e t m a n , P. S . : Concen tra tions of gen tam icin and am ikacin in hum an k idneys. A ntim icrob. A gents C he- mother. 9:925-927, 1976.
8. E l l i o t t , W . C., H o u g h t o n , D. C., G i l b e r t , D. N . , B a i n e s - H u n t e r , J., and B e n n e t t , W. M.: Gentamicin nephrotoxicity. 1. Degree and perm anence of acquired insensitivity. J. Lab. Clin. Med. 100:501-512, 1982.
9. F a b r e , J., F il l a s t r e , J. P., M o r i n , J. P., and R uD H A R D T , M .: Nephrotoxicity of gentamicin. Action on subcellular organelles and pharmacokinetics in the kidney. Contrib. Nephrol. 10:53 - 62, 1978.
10. F r a n c k e , E . a n d N e u , H . C . : N e p h r o t o x i c i t y o f a m i n o g l y c o s i d e s . I n f e c t . S u r g . March:205-214, 1983.
11. G h o s h , P . a n d C h a t t e r j e e , S . : E f f e c t s o f g e n t a m i c i n o n s p h i n g o m y e l i n a s e a c t i v i t y i n c u l t u r e d h u m a n r e n a l p r o x i m a l t u b u l e c e l l s . J. B i o l . C h e m . 262:12550-12556, 1987.
12. G i l b e r t , D. N., H o u g h t o n , D. C., B e n n e t t , W. M., P l a m p , C. E ., R e g e r , K., and P o r t e r , G . A.: Reversibility of gentamicin nephrotoxicity in rats: Recovery during continuous drug adm inistration. Proc. Soc. Exp. Biol. Med.i 60:99-103, 1979.
13. G i l b e r t , D. N . , P l a m p , C., S t a r r , P . , B e n n e t t , W . M., H o u g h t o n , D. C., and P o r t e r , G . : Comparative nephrotoxicity of gentamicin and tobramycin in rats. Antimicrob. Ag. Che- mother. 13:34-40, 1978.
14. H o r i , R., Y a m a m o t o , K . , S a i t o , J., K o h n o , M., and I n u i , K -I.: Effect of aminoglycoside antibiotics on cellular functions of kidney epithelial cell line (LLC-PKl): A model system for aminoglycoside nephrotoxicity. J. Pharmacol. Exp. Theraput. 230:742- 748, 1984.
15. H o t t e n d o r f , G . H . and G o r d o n , L. L.: Comparative low-dose nephrotoxicities of gentamicin, tobramycin and amikacin. Antimicrob. Ag. Chemother. /S :176 -181, 1980.
16. H o u g h t o n , D . C . , H a r t n e t t , M., C a m p b e l l - B o s w e l l , M., P o r t e r , G . , a n d B e n n e t t , W . : A l i g h t a n d e l e c t r o n m i c r o s c o p i c a n a l y s i s o f g e n t a m i c i n n e p h r o t o x i c i t y i n r a t s . A m J. P a t h o l . 82:589-612, 1976.
17. I n u i , K - I . , S a i t o , H . , I w a t a , T . , and H o r i , R . : Aminoglycoside-induced alterations in apical
279
membranes of kidney epithelial cell line (LLC- PK1). Am. J. Physiol.'254:C251-C257, 1988.
18. K a l o y a n i d e s , G. J. and P a s t o r i z a - M u n o z , E.: Aminoglycoside nephrotoxicity. Kidney Intern. iS:571-582, 1980.
19. K o s e k , J. C . , M a z z e , R . I., and C o u s i n s , M . J.: Nephrotoxicity of gentam icin. Lab. Invest. 30:48-57, 1974.
20. L u f t , F. C . , B l o c h , R., S l o a n , R. S . , Y u m , M . N . , C o s t e l l o , R., and M a x w e l l , D . R.: Comparative nephrotoxicity of aminoglycoside antibiotics in rats. J. Infect. Dis. 138-,541-545, 1978.
21. L u f t , F . C., P a t e l , V . , Y u m , M. N . , P a t e l , B . , and K l e i t , S. A.: Experimental aminoglycoside nephrotoxicity. J. L a b . Clin. Med. 86:213-220, 1975.
22. L u f t , F. C., R a n k i n , L . I., S l o a n , R. S . , and Y u m , M. N.: Recovery from aminoglycoside nephrotoxicity with continued drug administration. Antimicrob. Ag. Chemother. 14:284-287, 1978.
23. M o r i n , J. P., V i o t t e , G., V a n d e w a l l e , A., V a n H o o f , F . , T u l k e n s , P., a n d F i l l a s t r e ,
J. P.: Gentamicin-induced nephrotoxicity: A cell biology approach. Kidney Int. 18:583-590, 1980.
24. S c h e n t a g , J. J., C u m b o , T. J ., J u s k o , W. J., and P l a n t , M. E.: Gentamicin tissue accumulation and nephrotoxic reactions. J. Am. Med. Assoc. 240:2067-2069, 1978.
25. S e a l e , T. W . and R e n n e r t , O. M.: Mechanisms of antibiotic-induced nephrotoxicity. Ann. Clin. Lab. Sci. J2 :l-1 0 , 1982.
26. S e n s , M . A., H e n n i g a r , G. R . , H a z e n - M a r - t i n , D. J., B l a c k b u r n , J. G., and S e n s , D. A.: C ultu red hum an proximal tubule cells as a model for aminoglycoside nephrotoxicity. Ann. Clin. Lab. Sci. 78:204-214, 1988.
27. W h e l t o n , A .: Therapeutic initiatives for the avoidance of aminoglycoside toxicity. J. Clin. Pharmacol. 25:67-81, 1985.
28. W i l l i a m s , P. D., L a s k a , D. A., and H o t t e n - d o r f , G. H . : Comparative toxicity of aminoglycoside antibiotics in cell cultures derived from hum an and p ig k idney. In V itro Toxicol. 1:23—31, 1986/1987.