HUMAN RED BLOOD CELL HEMOLYSATE IS A POTENT MITOGEN FOR RENAL TUBULAR EPITHELIAL CELLS

Abstract

Conditions associated with intravascular hemolysis are often complicated by acute renal failure which is characterized by renal vasoconstriction and tubular injury. However, the effects of human red blood cell (RBC) hemolysate on renal tubular epithelial cell function have not been well characterized. We therefore measured the effect of distilled water-lysed human RBCs on cultured LLC-PK₁ cell function. We found that human RBC hemolysate produced a marked effect to promote LLC-PK₁ ³H-thymidine uptake that was several-fold higher than that produced by maximal concentrations of several known growth factors. Partial purification of human RBC hemolysate by sequential centrifugation and passage over a column that removes low molecular weight substances each significantly reduced, but did not totally eliminate, the effect of human RBC hemolysate to promote ³H-thymidine uptake. Exposure of LLC-PK₁ cells to horse myoglobin also stimulated ³H-thymidine incorporation in LLC-PK₁ cells. A reducing agent (ascorbic acid) decreased the effect of horse myoglobin and of human RBC hemolysate to promote LLC-PK₁ mitogenesis. Ascorbic acid also decreased the methemoglobin content of human RBC hemolysate. The effect of human RBC hemolysate to increase LLC-PK₁ incorporation of ³H-thymidine could also be significantly decreased by either of two inhib itors of tyrosine kinase. These results suggest that there are several components of human RBC hemolysate that promote LLC-PK₁ ³H-thymidine incorporation. One of these components appears to be methemoglobin that exerts its effect via a tyrosine kinase signal transduction pathway.

• Kidney cells
• Blood hemolysate
• Cell proliferation

INTRODUCTION

Clinical conditions associated with intravascular hemolysis are often complicated by the occurrence of acute renal failure [[1]]. Possible mechanisms contributing to hemolysis-associated acute renal failure include hypotension, intravascular volume depletion, renal vasoconstriction, and renal tubular epithelial cell damage resulting from stromal phospholipids and hemoglobin. Potential deleterious effects attributable to free hemoglobin include tubular obstruction due to heme pigment macromolecular precipitation and ischemic injury due to vasoconstriction [2-8]. For example, partially purified red blood cell (RBC) hemolysate has recently been found to induce vigorous constriction of rat afferent and efferent arterioles, to sensitize these arterioles to angiotensin II-induced vasoconstriction, and to promote calcium influx into cultured rat aortic smooth muscle cells [[8]].

The systemic and renal vascular effects of infused RBC hemolysate or of hemoglobin preparations renders determination of direct interactions between hemoglobin and the renal tubular epithelium difficult during in vivo studies [1-7]. Recent studies demonstrate the feasibility of studying the renal tubular epithelial effects of hemoglobin using cultured renal tubular cells [[9]]. The present studies were undertaken to test the hypothesis that human red blood cell hemolysate exerts direct effects on renal tubular epithelial mito- and motogenesis. To our surprise, we found that human RBC hemolysate results in striking proliferation of cultured LLC-PK₁ cells. This report describes these results as well as our investigations designed to clarify some of the mechanisms of this mitogenic effect.

METHODS

Materials

³H-thymidine (70–85 Cu/mmol) was obtained from Amersham (Arlington Heights, IL). Scintillation fluid (Scintisafe Econ2) was obtained from Fisher Science (St. Louis, MO). Culture media RPMI 1640, penicillin, streptomycin, N-2-hydroxyethyl piperazine-N′-2 ethanesulfonic acid (HEPES), platelet–derived growth factors AA and BB (PDGFAA and PDGFBB respectively), all cis and 13–trans retinoic acids (cis RA and trans RA respectively), purified human hemoglobin, bovine myoglobin, ascorbic acid, suramin, genistein, herbimycin A, and glutamine were obtained from Sigma (St. Louis, MO). Human recombinant growth factors including acidic and basic fibroblastic growth factor and fibroblastic growth factor seven (aFGF, bFGF and FGF-7 respectively) and epidermal and heparin-bound epidermal growth factor (EGF and HB-EGF respectively) were obtained from R & D Systems (Minneapolis, MN). Endothelins 1 and 2 (ET-1 and ET-2 respectively) were obtained from Calbiochem (La Jolla, CA).

Cell and Culture Conditions

LLC-PK₁ cells were obtained from the American Type Tissue Collection and were used at passages 90–125. LLC-PK₁ cells were grown in RPMI 1640 medium supplemented with 20mM HEPES, 100 μg/mL streptomycin, 100 μg/mL penicillin, and 6% calf bovine serum. After initial seeding, LLC-PK₁ cells typically reached confluence in 48–72 hours.

Preparation of Hemolysate

Blood from 15 healthy normal human volunteers of both sexes not taking any medications was obtained by venipuncture, centrifuged and washed extensively with normal saline to remove plasma, white blood cells and platelets as previously described [[8]], [[10]]. Packed RBC (500ul) were added to 6.0 mL of sterile, distilled water to induce hemolysis. In most experiments this crude hemolysate was partially purified by centrifugation for one hour at 100,000×g to remove membranes and by passage over a Sephadex G-25M column (Amersham) to remove molecules of less than 1000 daltons as described previously [[8]], [[10]].

Growth Activity Assay

Confluent monolayers of 75-cm² flasks of LLC-PK₁ cells were treated with 0.05% trypsin-0.01% EDTA for 4 min at 37°C. The cells were rinsed from the flask with serum-containing media, centrifuged at 300–400 g for 5 min, resuspended in 2–3 mL of medium, and counted on a hemocytometer. LLC-PK₁ cells were seeded in 24-well plates at 2×10⁴ cells/well in RPMI media supplemented with 6% bovine calf serum, HEPES, and streptomycin/penicillin. The cells were allowed to adhere for 24 hours. The media then changed to serum-free RPMI media with HEPES and streptomycin/penicillin, which was maintained for 48 hours prior to study. Cells were then pulsed with 0.5 μCi/well ³H-thymidine for three hours at 37°C, and then harvested. The samples were harvested by rinsing the wells once with phosphate-buffered saline (PBS) and the cellular material was precipitated with perchloric acid and solubilized with 0.01N NaOH/1%SDS. Samples were transferred to vials, scintillation cocktail was added, and the radioactivity was determined by liquid scintillation counting (Beckman Instruments; Fullerton, CA).

Migration Assay

LLC-PK₁ cells were seeded in 35mm dishes at 7–10, × 10⁵ cells/plate in serum containing medium. The cells adhered overnight then the medium was changed to serum-free RPMI with 20mM HEPES for 20–24 hours. The medium was changed again with serum-free medium. Rectangular areas of denuded cells were created near the center of the monolayer using sterile needles as described previously [11-12]. Two vertical lines dissecting two horizontal lines (tic-tac-toe like) were made and the rectangular areas were measured (time 0) and labeled. The wounds were measured with an eyepiece reticule in an inverted microscope with phase optics at × 10–20 (Nikon IM 35). Compounds were added to the plates and the cells were incubated at 37°C. The denuded areas were measured again at 6 and 24 hours after injury. Each study condition was compared to a simultaneously measured control and the percentage of surface area healed was calculated. An average of 6–8 denuded areas were measured in each experiment for both control and treated cells. The individual measuring the denuded areas was unaware of the treatment status of the cells. We have previously demonstrated that these wounded cultures heal in the absence of cellular proliferation [[11]].

Statistical Analyses

All values are expressed as the means ± SE. Statistical analyses were done using paired and unpaired Student's t-tests and analysis of variance where appropriate. P < 0.05 was considered significant.

RESULTS

Effect of Human RBC Lysate on LLC-PK₁ Cell ³H-thymidine Uptake and Wound Healing

The effect of 40 and 80μl of partially purified human RBC hemolysate, relative to 2% serum, are depicted in Figure 1. These hemolysate samples were obtained from at least 15 healthy donors of both sexes ranging in age from 21 through 57 years. In order to put these results in better overall perspective, Figure 2 demonstrates the effects of 80μl of human RBC hemolysate and the maximal effect of several other growth factors on LLC-PK₁ incorporation of ³H-thymidine. Also in Figure 2 is the effect of 80 l of partially purified human hemoglobin on LLC-PK₁ cell ³H-thymidine uptake. All tested factors significantly (P < 0.05) increased ³H-thymidine incorporation. However, the effect of human RBC hemolysate to enhance ³H-thymidine uptake was several-fold higher than maximal concentration of several known mitogens and of human hemoglobin.

Figure 1. Effect of human RBC lysate on LLC-PK₁ cell mitogenesis as assessed by ³H-thymidine incorporation. Each bar represents the mean ±SEM of 10 paired samples performed in triplicate relative to control values. All values (2% serum, 40 and 80 μl lysate) are significantly greater than control (P < 0.001), while the 40 and 80 μl RBC lysate samples are significantly greater than the 2% serum values (P < 0.01).

Figure 2. Relative effects of 80 μl human RBC lysate, 80 μl human hemoglobin and several other growth factors on LLC-PK₁ mitogenesis as assessed by ³H-thymidine incorporation. All bars represent the mean ± SEM of 4–10 paired samples done in triplicate relative to simultaneously studied controls. All growth factors were studied at 10⁻⁹M, which elicited a maximal effect on ³H-thymidine uptake except for cis-RA and Trans RA, which exerted maximal effects at 10⁻⁸M. All values are significantly greater than control (P < 0.05) while the value for lysed RBC is significantly greater (P < 0.01) than all other values.

We next examined if the effect of human RBC hemolysate to enhance ³H-thymidine uptake in LLC-PK₁ cells was additive to the effect of maximal concentrations of known mitogens for these cells. In these experiments, the effects of concentrations of FGF-7 and HB-EGF that we have found to induce maximal stimulation of LLC-PK₁ ³H-thymidine uptake (10⁻⁹M) were examined in the presence and absence of 80 l of human RBC hemolysate. An increase in cpm from 1,574 ± 209 at baseline to 2,399 ± 393 after exposure to FGF-7 was seen (P < 0.05). Red blood cell lysate increased thymidine uptake to 6,535 ± 1,120 cpm (P < .0001) in the absence and to 7,429 ± 1,592 cpm in the presence of 1.0 nM FGF-7 (n = five paired studies done in triplicate, P < 0.05 versus lysate alone). Similarly, HB-EGF increased basal ³H-thymidine uptake (baseline 1,562 ± 246 cpm to 2,481 ± 382 cpm (P < 0.05) in the absence of hemolysate. Hemolysate increased thymidine uptake to 6,920 ± 1,290 cpm (P < 0.0001) in the absence and significantly further to 7,681 ± 856 in the presence of HB-EGF (n = five paired studies done in triplicate, P < 0.05). These results demonstrate that the effect of human RBC hemolysate to stimulate LLC-PK₁ ³H-thymidine uptake is additive to the effect of maximum concentrations of two known mitogens for LLC-PK₁ cells.

To delineate if human RBC hemolysate affects another aspect of LLC-PK₁ cell function, we examined the effect of 40 μl human RBC hemolysate on the healing of small wounds made within confluent monolayers of relatively quiescent LLC-PK₁ cells. In these paired studies (n = 4), human RBC hemolysate increased the degree of 24-hour wound healing by 19.5 ± 5.7% (P < 0.025) over control.

Characterization of Human RBC Lysate Components Responsible for Promoting LLC-PK₁ ³H-thymidine Uptake

These studies were done to better characterize our human RBC lysate preparation and its components that potentially contribute to stimulation of LLC-PK₁ mitogenesis. Initially, we did six paired experiments in which we compared the effects of crude lysate (washed to remove white blood cells and platelets and lysed with distilled water), crude lysate that underwent centrifugation (to remove cell membranes and phospholipids) and crude lysate that was washed, centrifuged and passed over a Sephadex G-25M column (to remove small molecular weight substances). These results (Figure 3) demonstrate a small (17%) but significant (P < 0.05) effect of centrifugation and a larger (37%, P < 0.025) effect of passage over a G-25 column to decrease human RBC lysate effect to stimulate LLC-PK₁ incorporation of ³H-thymidine. These results are compatible with the notion that there is more than a single component of our human RBC lysate preparation that contributes to an effect to promote LLC-PK₁ ³H-thymidine incorporation.

Figure 3. Effect of human RBC lysate preparations on LLC-PK₁ mitogenesis as assessed by ³H-thymidine incorporation. Each bar represents the mean ± SEM of actual cpm for five samples studied under each (no addition/control, lysed only, lysed and centrifuged, and lysed, centrifuged and partially purified by G-25 column passage) condition. Centrifugation and centrifugation plus column passage each produced a modest but significant (P < 0.05) reduction in ³H-thymidine uptake relative to the lysed preparation that was neither centrifuged nor passed over the G-25 column.

Since human RBC hemolysate contains substantial free hemoglobin and since partially purified hemoglobin modestly stimulated LLC-PK₁ ³H-thymidine uptake (Figure 2), we assessed the effect of another iron containing porphyrin ring (myoglobin) on LLC-PK₁ ³H-thymidine uptake (upper panel, Figure 4). In initial studies, we found a dose-dependent effect of a crude horse myoglobin preparation to significantly promote LLC-PK₁ cell ³H-thymidine uptake (upper panel, Figure 4). Observations noted by Zager et al suggest that this effect of myoglobin can be attenuated by a reducing agent and may be due to metmyoglobin [[13]]. We therefore determined the effect of a reducing agent (ascorbic acid) on myoglobin-promoted ³H-thymidine uptake (lower panel, Figure 4). Ascorbic acid pretreatment resulted in a dose-dependent effect of myoglobin to inhibit LLC-PK₁ ³H-thymidine incorporation.

Figure 4. Effect of various concentration of either untreated (upper panel) or ascorbic acid (asc)-treated (lower panel) horse myoglobin on LLC-PK₁ proliferation as assessed by ³H-thymidine uptake. Each bar represents the mean ± SEM relative to control samples of four studies done in triplicate. In the absence of ascorbic acid pretreatment (upper panel), a concentration-dependent effect of myoglobin to increase ³H-thymidine uptake was seen with concentrations greater than 0.1 mg/mL producing a significant (P < 0.05) increase. Ascorbic acid pretreatment (lower panel) resulted in a concentration-dependent inhibition of ³H-thymidine incorporation significant (P < 0.05) at concentrations > 0.1 mg/mL.

To determine if a reducing agent affected human RBC hemolysate promotion of LLC-PK₁ ³H-thymidine uptake, the studies depicted in Figure 5 were done. In these five-paired studies, we found a significant reduction in human RBC lysate promotion of ³H-thymidine uptake by ascorbic acid pretreatment. However, in contrast to the studies with horse myoglobin, ascorbic acid treated lysate did not inhibit ³H-thymidine incorporation and continued to modestly but significantly (P < 0.05) stimulate ³H-thymidine uptake.

Figure 5. Effect of ascorbic acid (asc) pretreatment on human RBC lysate preparations to promote ³H-thymidine uptake in LLC-PK₁ cells. In these paired studies (n = 5), ³H-thymidine incorporation was measured under control (DMSO-treatment, first pair of bars>) conditions and after exposure to crude lysate (second pair of bars) or partially purified lysate (lysate that was centrifuged and passed over a G-25 sephadex column, third set of bars) in the absence (solid bars) or presence (hatched bars) of 10⁻⁴M ascorbic acid. The bars represent the mean ± SEM of the cpm of triplicate samples. Ascorbic acid pretreatment significantly (P < 0.001) reduced the effect of both lysate preparations to promote LLC-PK₁ ³H-thymidine uptake. Ascorbic acid pretreated lysate preparations however still modestly but significantly (P < 0.05) promoted ³H-thymidine uptake relative to control.

Since our results and those of Zager et al. suggest a role for methemoglobin and/or metmyoglobin in the effects described in Figures 4 and 5, we measured methemoglobin in our human RBC lysate preparation. The whole blood samples studied (n = 3) contained < 0.5% of methemoglobin. The partially purified human RBC hemolysate preparation (n = 5) contained an average concentration of methemoglobin of 22.8 ± 0.7%. In three paired studies, ascorbic acid pretreatment lowered partially purified human RBC lysate methemoglobin from 20.7 ± 1.7 to 3.7 ± 0.8%. Since iron can also potentially exert effects on cell proliferation [[2]], [13-16], we measured iron in the crude hemolysate (n =) before (1,018 and 1,166 μg/dL) and after column passage (93 and 182 μg/dL).

Mechanism of Human RBC Hemolysate Promotion of LLC-PK₁ ³H-thymidine Incorporation

These studies were designed to clarify the pathway(s) whereby our human RBC hemolysate stimulates LLC-PK₁ cell thymidine uptake. In these paired studies (n = 5), we examined the effect of genistein, an inhibitor of tyrosine kinase, on the effects of various lysate preparations to stimulate LLC-PK₁ cell ³H-thymidine incorporation (Figure 6). As is apparent from Figure 6, genistein significantly (P < 0.05) decreased the effect of all lysate preparations to promote LLC-PK₁ ³H-thymidine uptake. This decrease was greatest from the crude lysate preparation (33% reduction) and least from the partially purified lysate preparation (22% reduction). To confirm these observations, paired experiments were done with a second, chemically dissimilar tyrosine kinase inhibitor (herbimycin A). In three paired studies done in triplicate, the ³H-thymidine cpms were significantly reduced by herbimycin A in the crude hemolysate (from 3,149 ± 296 to 1,916 ± 154), the centrifuged hemolysate (from 2,089 ± 202 to 1,154 ± 149) and in the centrifuged hemolysate passed over a G-25M column (from 1,078 ± 176 to 296 ± 20, all P < 0.05). We also examined the effect of genistein on the effect of nonascorbic acid-treated horse myoglobin preparation known to contain significant amounts of metmyoglobin [[13]] to increase LLC-PK₁ ³H-thymidine uptake. In these paired studies (n = 3 performed in triplicate, Figure 7), genistein significantly decreased the effect of 0.5, 1.0 and 2.0 mg/mL of horse myoglobin to increase LLC-PK₁ ³H-thymidine uptake (P < 0.01).

Figure 6. Effect of genistein on human RBC lysate preparations to promote ³H-thymidine uptake in LLC-PK₁ cells. In these paired studies (n = 5), ³H-thymidine incorporation was measured under control (DMSO) conditions and following exposure to various human RBC lysate preparations. The bars represent the mean ± SEM of cpm of triplicate samples in the absence (open bars) and presence (hatched bars) of 10⁻⁴M genistein. Genistein modestly but significantly (P < 0.05) reduced ³H-thymidine incorporation in all lysate preparations but not under control conditions.

Figure 7. Effect of genistein on horse myoglobin-promoted ³H-thymidine uptake in LLC-PK₁ cells. In these paired studies done in triplicate (n = 3), genistein (10⁻⁴M, solid bars) significantly decreased (P < 0.01) the effect of all concentrations of horse myoglobin to promote ³H-thymidine incorporation. The bars represent the mean ± SEM of cpm under control and after exposure to varying concentrations of myoglobin.

Suramin is a nonspecific inhibitor of the interaction of numerous polypeptide growth factors with their receptors. We therefore next measured the effect of suramin on basal and human RBC hemolysate-stimulated incorporation in LLC-PK₁cells. As is depicted in Figure 8, suramin did not attenuate the effects of human RBC hemolysate to promote LLC-PK₁ proliferation.

Figure 8. Effect of suramin on human RBC lysate promotion of ³H-thymidine uptake in LLC-PK₁ cells. The bars represent the mean ± SEM of cpm of four paired studies in which ³H-thymidine uptake was measured in the absence (open bars) or presence (solid bars) of 40 μl of human RBC lysate at increasing concentrations of suramin. Suramin did not affect lysate promotion of ³H-thymidine incorporation.

DISCUSSION

The major finding of the present experiments is that a human RBC hemolysate preparation exerts a potent effect to promote LLC-PK₁ cell ³H-thymidine uptake. The stimulation of LLC-PK₁ ³H-thymidine incorporation is several-fold greater than the maximal effect exerted by several known mitogens and exceeds the effect of any growth-promoting substance (including serum) heretofore studied in our laboratory. In view of the potential for growth-promoting substances to enhance the renal tubular repair process, we undertook several studies to clarify the substance(s) present in human RBC lysate that promotes ³H-thymidine incorporation and to delineate the mechanism of this effect.

Our results suggest that more than a single substance contributes to the promotion of LLC-PK₁ ³H-thymidine uptake by human RBC lysate. Specifically, paired studies demonstrate that both centrifugation and passage over a column each modestly but significantly reduces the effect of lysate to stimulate ³H-thymidine uptake. Centrifugation and column passage would be expected to remove cell membranes and low molecular substances. Despite such treatment, however, nearly 50% of the mitogenic effect of human RBC lysate continued to be present. This is perhaps not surprising since lysis of intact red blood cells produces not only phospholipid-rich cell membranes, but also large quantities of free hemoglobin. Free hemoglobin is unstable and degraded initially to at least three components including hemoglobin-heme, globin, and iron [[18]]. Of potential relevance to the present studies, subsequent proteolytic degradation of globin from hemoglobin has recently been found to result in production of several small- and medium-molecular weight growth-promoting peptides that have sequence homologies with known growth factors [19-21].

Our results provide evidence that methemoglobin is one of the predominant substances that contribute to this residual mitogenic effect. Direct measurements demonstrate that methemoglobin is present in modest but significant amounts in our lysate preparation. Moreover, a preparation known to contain metmyoglobin [[13]] exerts a comparable effect to promote ³H-thymidine uptake. Finally, a reducing agent (ascorbic acid) greatly attenuates the effect of human RBC lysate to stimulate LLC-PK₁ ³H-thymidine incorporation while concomitantly reducing methemoglobin content. We are unaware of published data demonstrating an effect of methemoglobin to directly stimulate mitogenesis. However an indirect effect of methemoglobin to enhance rabbit aortic smooth muscle cell ³H-thymidine uptake via modification of low-density lipoprotein has recently been reported [[22]]. Given the potent mitogenic effect that we observed, further studies delineating the precise structural/molecular element(s) of the methemoglobin moiety involved in stimulating LLC-PK₁ mitogenesis are indicated.

Another contribution of the present study is that one or more of the mitogenic substances present within the human RBC lysate preparation appears to promote LLC-PK₁ mitogenesis acting via a tyrosine kinase signaling pathway. Two chemically dissimilar inhibitors of tyrosine kinase significantly attenuated the effects of human RBC lysate to stimulate LLC-PK₁ ³H-thymidine uptake. While our studies support a tyrosine kinase pathway as a contributor to the observer mitogenesis, our results also suggest that a pathway independent of a tyrosine kinase signaling system may contribute since enhanced ³H-thymidine uptake continued to be observed even in the presence of high concentrations of the inhibitors used.

In summary, the present studies demonstrate that a human RBC hemolysate preparation exerts a potent effect to promote ³H-thymidine incorporation in cultured LLC-PK₁ cells. There appears to be more than a single factor in this hemolysate preparation that contributes to the enhanced ³H-thymidine uptake. One component of the human RBC hemolysate preparation that promotes LLC-PK₁ ³H-thymidine incorporation appears to be methemoglobin that acts, at least in part, via a tyrosine kinase signal transduction pathway.

ACKNOWLEDGMENTS

This work was supported by research funds from the Department of Veterans Affairs. Mary Miller provided expert secretarial assistance.